**Nonlinear Epilepsy Forewarning by Support Vector Machines**

W.S. Ashbee, L.M. Hively and J.T. McDonald

Additional information is available at the end of the chapter

http://dx.doi.org/10.5772/57438

#### **1. Introduction**

Epilepsy is a neurological disorder that changes the observable behavior of an individual to the point of inducing complete loss of consciousness. Pharmaceutical drugs may reduce or eliminate the problems of epilepsy, but not all people respond to pharmaceuticals favorably, and some may find the side effects undesirable. EEG-based epilepsy prediction may offer an acceptable alternative or complementary treatment to pharmaceuticals. Invasive, intra-cranial EEG provides signals that are directly from the brain, without the muscular activity that infests non-invasive, scalp EEG. However, intra-cranial EEG requires surgery, which increases risk and cost of health care, while reducing the number of people able to receive medical attention. Algorithms to predict the seizure event—the ictal state—may lead to new treatments for chronic epilepsy. Finding solutions that involve non-invasive procedures may result in treatments for the largest section of the population.

#### **2. Background**

Epilepsy prediction is greater than 1 minute of forewarning before there is any visible indication that a seizure will occur. The physician does not label the pre-ictal periods that precede the seizure—states that may indicate a seizure is near. Event characterization only labels the start time of the seizure. Consequently, labeled data for the pre-ictal state is nonexistent, but is necessary to train a Support Vector Machine (SVM). Other researchers address this problem by assuming that the pre-ictal phase occurs immediately prior to a seizure [1]; see Figure 1 for an example.

© 2014 The Author(s). Licensee InTech. This chapter is distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/3.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

complex system like the brain is very difficult. Indeed, Stacey *et al.* [2] find that no algorithm

Nonlinear Epilepsy Forewarning by Support Vector Machines

http://dx.doi.org/10.5772/57438

31

One must also choose whether to use monopolar (single channel) or bipolar EEG (difference between two monopolar channels). Mirowski et al. assert that epilepsy can be predicted more effectively with bipolar features [3] because of changes in the brain's ability to synchronize regions during a seizure. Mirowski et al. consider their pre-ictal period to be 2 hours. They assert, "[most] current seizure prediction approaches can be summarized into (1) extracting features from EEG and (2) classifying them (and hence the patient's state) into pre-ictal or interictal". They go on to enumerate more specifically on the bipolar feature set in Figure 2.

Figure 2 enumerates a feature set from all unique channel pairs [3]. After the enumeration, they use a grid search to find appropriate parameters with their SVM with a Gaussian kernel. Mirowski et al. use intra-cranial EEG data and obtain 100% accuracy for patient-specific machine learning models. However, no single model provides 100% accuracy for all patients [3], so they choose from among a variety of algorithms to achieve high accuracy on a patient

By contrast, the present work uses non-invasive, scalp EEG. Moreover, the present work uses a SVM to extract seizure forewarning from the entire patient population. The goal is high accuracy. The long-term objective (not addressed in the present work) is lower health-care cost

Previous work by Hively *et al.* has used bipolar, scalp EEG and found the best seizure forewarning by using electrodes in the right frontal lobe [4, 5] in the 10-20 system; see Figure 3. The present work uses this same bipolar channel. Additionally, scalp EEG from one bipolar channel facilitates a simple, ambulatory device with two electrodes, which is far more manageable than an EEG headset with many channels. Our hypothesis is that the right-frontal region acts as a filter for pre-ictal condition change—a phase transition in the brain dynamics [6] that can be induced by noise [7]. Pittau *et al.* [8] reviewed the recent technical literature on sound-induced (musicogenic) seizures, which activate the fronto-temporo-occipital area.

by using one algorithm for all patients to analyze scalp EEG on a smartphone.

provides better-than-chance prediction of seizures in statistical tests to date.

**Figure 2.** Approach to using bivariate features in epilepsy prediction [3].

specific basis.

**Figure 1.** EEG with a seizure (ictal) event and potential class labels [1].

The labeling scheme of Figure 1 results in better than random predictions [1] under the assumption that the pre-ictal region immediately precedes the seizure and may be exploited for epilepsy prediction. This SVM approach provides the most obvious way to label the training and testing data without any extra information being available about the EEG.

Assuming pre-ictal dynamics occur within an hour of the seizure has the added benefit of being more likely to satisfy caregivers' requests to have forewarning within an hour of the seizure event. Netoff et al. achieve a specificity of 77% in classifying the pre-ictal region with no false positives with the above approach [1]. This level of accuracy is not high enough for a marketable prediction algorithm, but suggests that indicators of a seizure occur within an hour prior to a typical seizure. Netoff et al. use a "5 minute prediction horizon" where they label the pre-ictal region. They classify preictal as being within 5 minutes of the seizure and calculate specificities according to that labeling scheme. They assert that the short time frame makes the computational difficulty of the algorithm much more manageable than algorithms that have fewer restrictions on where the pre-ictal region is. They have a second stage of processing as well in which they look for 3 out of 5 pre-ictal indicators in a concentrated bundle in order to achieve prediction [1].

The assumption of pre-ictal indications near the event seems sound because a seizure resem‐ bles a dynamical phase transition. More specifically, the brain activity changes from some "normal" phase of brain activity into hyper-synchronous activity. The present work assumes that the brain dynamics within an hour of the seizure are approaching a phase transition, corresponding to measureable change in the scalp EEG. A simple example of a phase transition is liquid water becoming steam due to changes in pressure and temperature. However, scalp EEG exhibits nonlinear, chaotic features that are extremely difficult to predict over long periods and are extremely sensitive to initial conditions. Consequently, seizure prediction in a very complex system like the brain is very difficult. Indeed, Stacey *et al.* [2] find that no algorithm provides better-than-chance prediction of seizures in statistical tests to date.

One must also choose whether to use monopolar (single channel) or bipolar EEG (difference between two monopolar channels). Mirowski et al. assert that epilepsy can be predicted more effectively with bipolar features [3] because of changes in the brain's ability to synchronize regions during a seizure. Mirowski et al. consider their pre-ictal period to be 2 hours. They assert, "[most] current seizure prediction approaches can be summarized into (1) extracting features from EEG and (2) classifying them (and hence the patient's state) into pre-ictal or interictal". They go on to enumerate more specifically on the bipolar feature set in Figure 2.


#### **Figure 2.** Approach to using bivariate features in epilepsy prediction [3].

**Figure 1.** EEG with a seizure (ictal) event and potential class labels [1].

achieve prediction [1].

30 Epilepsy Topics

The labeling scheme of Figure 1 results in better than random predictions [1] under the assumption that the pre-ictal region immediately precedes the seizure and may be exploited for epilepsy prediction. This SVM approach provides the most obvious way to label the training and testing data without any extra information being available about the EEG.

Assuming pre-ictal dynamics occur within an hour of the seizure has the added benefit of being more likely to satisfy caregivers' requests to have forewarning within an hour of the seizure event. Netoff et al. achieve a specificity of 77% in classifying the pre-ictal region with no false positives with the above approach [1]. This level of accuracy is not high enough for a marketable prediction algorithm, but suggests that indicators of a seizure occur within an hour prior to a typical seizure. Netoff et al. use a "5 minute prediction horizon" where they label the pre-ictal region. They classify preictal as being within 5 minutes of the seizure and calculate specificities according to that labeling scheme. They assert that the short time frame makes the computational difficulty of the algorithm much more manageable than algorithms that have fewer restrictions on where the pre-ictal region is. They have a second stage of processing as well in which they look for 3 out of 5 pre-ictal indicators in a concentrated bundle in order to

The assumption of pre-ictal indications near the event seems sound because a seizure resem‐ bles a dynamical phase transition. More specifically, the brain activity changes from some "normal" phase of brain activity into hyper-synchronous activity. The present work assumes that the brain dynamics within an hour of the seizure are approaching a phase transition, corresponding to measureable change in the scalp EEG. A simple example of a phase transition is liquid water becoming steam due to changes in pressure and temperature. However, scalp EEG exhibits nonlinear, chaotic features that are extremely difficult to predict over long periods and are extremely sensitive to initial conditions. Consequently, seizure prediction in a very

Figure 2 enumerates a feature set from all unique channel pairs [3]. After the enumeration, they use a grid search to find appropriate parameters with their SVM with a Gaussian kernel. Mirowski et al. use intra-cranial EEG data and obtain 100% accuracy for patient-specific machine learning models. However, no single model provides 100% accuracy for all patients [3], so they choose from among a variety of algorithms to achieve high accuracy on a patient specific basis.

By contrast, the present work uses non-invasive, scalp EEG. Moreover, the present work uses a SVM to extract seizure forewarning from the entire patient population. The goal is high accuracy. The long-term objective (not addressed in the present work) is lower health-care cost by using one algorithm for all patients to analyze scalp EEG on a smartphone.

Previous work by Hively *et al.* has used bipolar, scalp EEG and found the best seizure forewarning by using electrodes in the right frontal lobe [4, 5] in the 10-20 system; see Figure 3. The present work uses this same bipolar channel. Additionally, scalp EEG from one bipolar channel facilitates a simple, ambulatory device with two electrodes, which is far more manageable than an EEG headset with many channels. Our hypothesis is that the right-frontal region acts as a filter for pre-ictal condition change—a phase transition in the brain dynamics [6] that can be induced by noise [7]. Pittau *et al.* [8] reviewed the recent technical literature on sound-induced (musicogenic) seizures, which activate the fronto-temporo-occipital area. Conversely, soothing music (e.g., Mozart's double piano sonata K.448) decreases the intensity and frequency of epileptic seizures [9].

protocols from 41 temporal-lobe-epilepsy patients (ages from 4 to 57 years; 36 datasets from females, and 24 datasets from males). The datasets range in length from 1.4 to 8.2 hours (average = 4.4 hours). Data characterization included patient activity. Forty datasets had

A patented zero-phase, quadratic filter enables analysis of scalp EEG by removing electrical activity from eye blinks and other muscular artifacts, which otherwise obscure the event forewarning. This filter retains the nonlinear amplitude and phase information [14]. The filter

points sampled of the eye blink filter; the value, N, represents the number of points in a cutset, which is represented by a graph; the value, g, is the artifact filtered set of points with eye blinks removed; the value, e, is the set of raw EEG data points; and the value, f, is the set of artifact

A trade-off is required between coarseness in the data to exclude noise, and precision in the

(*gn*) in the first base case cutset. Uniform symbols are generated by the form in Eq. (1).

gi - gn

Here, INT converts a decimal number to the closest lower integer. Takens' theorem [15] gives a smooth, non-intersecting dynamical reconstruction in a sufficiently high dimensional space by a time-delay embedding. The symbolized data from Eq. (1) are converted into unique

the underlying dynamics. The time-delay lag is *L*, which must not be too small (making *si*

ity). The embedding dimension is *d*, which must be sufficiently large to capture the dynamics,

nodes and links give a formal, diagrammatic construction, called a "graph." This form gives topologically-invariant measures that are independent of any unique labeling of individual nodes and links [16]. Figure 4 depicts the algorithmic steps to: extract the analysis window from the stream of EEG data; remove the artifacts from scalp EEG; symbolize the artifactfiltered data; and construct the graph nodes and links [4]. The parameter space in Figure 4

, that are uniformly distributed between the maximum (*gx*) and minimum


gx - gn <sup>≤</sup> <sup>S</sup> – <sup>1</sup> (1)

:

, si+L , . . . , si+(d–1)L (2)

and *si+L* independent by long-time unpredictabil‐

→ *yi*+*<sup>M</sup>*, forms state-to-state links. The


. The value, w, is a parameter that specifies the width in

. Essentially no low-frequency artifacts occur in

Nonlinear Epilepsy Forewarning by Support Vector Machines

http://dx.doi.org/10.5772/57438

33

) are symbolized into

and

seizures, and twenty had no event [13].

uses a moving window of 2*w* + 1 points of *ei*

filter points used to subtract out eye blinks.

the artifact-filtered signal, *gi*

*S* discrete values, *si*

Takens' theorem allows the *yi*

*si+L* indistinguishable) or too large (making *si*

The states from Eq. (2) are nodes. The process flow, *yi*

but not too large to avoid over-fitting.

sense, yielding *N* – 2*w* points of artifact data, *fi*

 = *ei* - *fi*

data to accurately follow the dynamics. Thus, the artifact-filtered data (*gi*

0 ≤si = INT S

dynamical states by the Takens' time-delay-embedding vector, *yi*

yi = si

enumerates the parameters used to generate the phase space graphs.

**Figure 3.** system (EEG) [10].

#### **3. Phase-space analysis**

We use one bipolar channel of *scalp* EEG (F8 – FP2) in the 10-20 system, as a measure of the noisy dynamics in cortical neurons over an area of roughly 6 cm2 . Our earlier work obtained channel-consistent forewarning across nineteen EEG channels [11]. The garbage-in-garbageout syndrome is avoided by rejecting data of inadequate quality [12].

These data were uniformly sampled in time, *ti* , at 250 Hz, giving *N* time-serial points in an analysis window (cutset), *ei* = *e*(*ti* ). Data acquisition was under standard human-studies protocols from 41 temporal-lobe-epilepsy patients (ages from 4 to 57 years; 36 datasets from females, and 24 datasets from males). The datasets range in length from 1.4 to 8.2 hours (average = 4.4 hours). Data characterization included patient activity. Forty datasets had seizures, and twenty had no event [13].

Conversely, soothing music (e.g., Mozart's double piano sonata K.448) decreases the intensity

We use one bipolar channel of *scalp* EEG (F8 – FP2) in the 10-20 system, as a measure of the

channel-consistent forewarning across nineteen EEG channels [11]. The garbage-in-garbage-

. Our earlier work obtained

, at 250 Hz, giving *N* time-serial points in an

). Data acquisition was under standard human-studies

noisy dynamics in cortical neurons over an area of roughly 6 cm2

= *e*(*ti*

These data were uniformly sampled in time, *ti*

out syndrome is avoided by rejecting data of inadequate quality [12].

and frequency of epileptic seizures [9].

32 Epilepsy Topics

**Figure 3.** system (EEG) [10].

**3. Phase-space analysis**

analysis window (cutset), *ei*

A patented zero-phase, quadratic filter enables analysis of scalp EEG by removing electrical activity from eye blinks and other muscular artifacts, which otherwise obscure the event forewarning. This filter retains the nonlinear amplitude and phase information [14]. The filter uses a moving window of 2*w* + 1 points of *ei* -data that are fitted to a parabola in a least-squares sense, yielding *N* – 2*w* points of artifact data, *fi* . Essentially no low-frequency artifacts occur in the artifact-filtered signal, *gi* = *ei* - *fi* . The value, w, is a parameter that specifies the width in points sampled of the eye blink filter; the value, N, represents the number of points in a cutset, which is represented by a graph; the value, g, is the artifact filtered set of points with eye blinks removed; the value, e, is the set of raw EEG data points; and the value, f, is the set of artifact filter points used to subtract out eye blinks.

A trade-off is required between coarseness in the data to exclude noise, and precision in the data to accurately follow the dynamics. Thus, the artifact-filtered data (*gi* ) are symbolized into *S* discrete values, *si* , that are uniformly distributed between the maximum (*gx*) and minimum (*gn*) in the first base case cutset. Uniform symbols are generated by the form in Eq. (1).

$$\text{I}0 \le \mathbf{s}\_{\mathbf{i}} = \text{INT} \left[ \mathbf{S} \frac{\mathbf{g}\_{\mathbf{i}} \cdot \mathbf{g}\_{\mathbf{n}}}{\mathbf{g}\_{\mathbf{x}} \cdot \mathbf{g}\_{\mathbf{n}}} \right] \le \mathbf{S} - 1 \tag{1}$$

Here, INT converts a decimal number to the closest lower integer. Takens' theorem [15] gives a smooth, non-intersecting dynamical reconstruction in a sufficiently high dimensional space by a time-delay embedding. The symbolized data from Eq. (1) are converted into unique dynamical states by the Takens' time-delay-embedding vector, *yi* :

$$\mathbf{y}\_{i} = \begin{bmatrix} \mathbf{s}\_{i'} & \mathbf{s}\_{i+L'} & \dots & \mathbf{s}\_{i+(d-1)L} \end{bmatrix} \tag{2}$$

Takens' theorem allows the *yi* -states to capture the topology (connectivity and directivity) of the underlying dynamics. The time-delay lag is *L*, which must not be too small (making *si* and *si+L* indistinguishable) or too large (making *si* and *si+L* independent by long-time unpredictabil‐ ity). The embedding dimension is *d*, which must be sufficiently large to capture the dynamics, but not too large to avoid over-fitting.

The states from Eq. (2) are nodes. The process flow, *yi* → *yi*+*<sup>M</sup>*, forms state-to-state links. The nodes and links give a formal, diagrammatic construction, called a "graph." This form gives topologically-invariant measures that are independent of any unique labeling of individual nodes and links [16]. Figure 4 depicts the algorithmic steps to: extract the analysis window from the stream of EEG data; remove the artifacts from scalp EEG; symbolize the artifactfiltered data; and construct the graph nodes and links [4]. The parameter space in Figure 4 enumerates the parameters used to generate the phase space graphs.

instances (*K*) above a threshold (*UT*) for each of *J* features, *U* (*V*

scalp EEG are captured by changes in the graph measures.

**Figure 5.** Normalized Graph dissimilarity measures, based on [4].

allows regions of the feature space to be found that forewarn for many patients. Because the graphs are diffeomorphic to the underlying dynamics (from Takens' theorem), changes in the

The present work uses a SVM approach to obtain forewarning from the normalized dissimi‐ larity measures—namely, we find nonlinear regions in the feature space using a SVM. Figure 5 shows the calculation of the dissimilarity measures [4]. The frequency of nodes and links is not used because Takens' theorem guarantees topology, but not density—meaning Takens'

The dissimilarity measures in Figure 5 capture topology changes between two graphs. While node and link differences are basic graph measures, they quantify the hypothesis in a very simple and general way. Less commonality of nodes and links between two graphs produces larger dissimilarity measures, which are used to capture changes in topology. Topology change is a necessary, but not sufficient condition for a phase transition [17]. Our results show that changes in topology over extended periods indicate a higher likelihood of observing a phase transition as an indicator of an impending seizure. Additionally, the four graph dissimilarity measures from nodes and links rely on two concepts from set theory and Venn diagrams. Node dissimilarity and link dissimilarity are broken into two measures of dissimilarity each. Comparing two graphs (A and B) results in differences in nodes, as well as links. The dissim‐ ilarity measures are used as SVM features (for a total of 4 features in the Stage-1 SVM described below), and include the nodes in graph A that are absent from graph B, links in graph A that

theorem doesn't guarantee useful information in the repetition of nodes or links.

\_ ) =


Nonlinear Epilepsy Forewarning by Support Vector Machines

*<sup>σ</sup>* . This normalization

http://dx.doi.org/10.5772/57438

35

**Figure 4.** Nonlinear phase space construction [4].

The value, B, is the number of base cases, which establishes a normal range of activity for the patient. The value, N, is the number of sampled points that are in a cutset and graph, The value, w, as mentioned previously is the half-width of the eye-blink filter in sampled point units. The value, S, is the number of bins that the EEG is discretized into in order to create the base-s number represented by the vector y(i) in Figure 4. The value, d, is number of numerals in the base-s number or elements in the d dimensional vector, y(i). L is a time delay embedding that specifies the interval between points sampled in order to create a node. M is a second time delay embedding that specifies the interval between two connected nodes. The parameters mentioned are all used to generate the phase space graphs illustrated in Figure 4.

The dissimilarity measures involve counting unique nodes and links (those not in common between the two graphs): (1) nodes in graph A but not in B; (2) nodes in B but not in A; (3) links in A but not in B; and (4) links in B but not in A. Nodes and links in common between graphs do not indicate change and are not useful. These measures sum the absolute value of differences, which is better than traditional measures that use a difference of averages. Each measure is normalized to the number of nodes (links) in A (for A not in B) or in B (for B not in A). This feature vector, *V*, is used to classify the EEG as pre-ictal or inter-ictal. The analysis obtains a vector of mean dissimilarities, *V*, and matching standard deviations,σ, by comparison among the *B*(*B*–1)/2 combinations of the *B* base-case graphs, as shown in Fig. 4. Subsequent test-case graphs are then compared to each of the *B* base-case graphs to get an average dissimilarity vector, *v*. Our previous approach to obtain forewarning was several successive instances (*K*) above a threshold (*UT*) for each of *J* features, *U* (*V* \_ ) = |*v* – *V* \_ | *<sup>σ</sup>* . This normalization allows regions of the feature space to be found that forewarn for many patients. Because the graphs are diffeomorphic to the underlying dynamics (from Takens' theorem), changes in the scalp EEG are captured by changes in the graph measures.

The present work uses a SVM approach to obtain forewarning from the normalized dissimi‐ larity measures—namely, we find nonlinear regions in the feature space using a SVM. Figure 5 shows the calculation of the dissimilarity measures [4]. The frequency of nodes and links is not used because Takens' theorem guarantees topology, but not density—meaning Takens' theorem doesn't guarantee useful information in the repetition of nodes or links.

**Figure 5.** Normalized Graph dissimilarity measures, based on [4].

**Figure 4.** Nonlinear phase space construction [4].

34 Epilepsy Topics

The value, B, is the number of base cases, which establishes a normal range of activity for the patient. The value, N, is the number of sampled points that are in a cutset and graph, The value, w, as mentioned previously is the half-width of the eye-blink filter in sampled point units. The value, S, is the number of bins that the EEG is discretized into in order to create the base-s number represented by the vector y(i) in Figure 4. The value, d, is number of numerals in the base-s number or elements in the d dimensional vector, y(i). L is a time delay embedding that specifies the interval between points sampled in order to create a node. M is a second time delay embedding that specifies the interval between two connected nodes. The parameters

The dissimilarity measures involve counting unique nodes and links (those not in common between the two graphs): (1) nodes in graph A but not in B; (2) nodes in B but not in A; (3) links in A but not in B; and (4) links in B but not in A. Nodes and links in common between graphs do not indicate change and are not useful. These measures sum the absolute value of differences, which is better than traditional measures that use a difference of averages. Each measure is normalized to the number of nodes (links) in A (for A not in B) or in B (for B not in A). This feature vector, *V*, is used to classify the EEG as pre-ictal or inter-ictal. The analysis obtains a vector of mean dissimilarities, *V*, and matching standard deviations,σ, by comparison among the *B*(*B*–1)/2 combinations of the *B* base-case graphs, as shown in Fig. 4. Subsequent test-case graphs are then compared to each of the *B* base-case graphs to get an average dissimilarity vector, *v*. Our previous approach to obtain forewarning was several successive

mentioned are all used to generate the phase space graphs illustrated in Figure 4.

The dissimilarity measures in Figure 5 capture topology changes between two graphs. While node and link differences are basic graph measures, they quantify the hypothesis in a very simple and general way. Less commonality of nodes and links between two graphs produces larger dissimilarity measures, which are used to capture changes in topology. Topology change is a necessary, but not sufficient condition for a phase transition [17]. Our results show that changes in topology over extended periods indicate a higher likelihood of observing a phase transition as an indicator of an impending seizure. Additionally, the four graph dissimilarity measures from nodes and links rely on two concepts from set theory and Venn diagrams. Node dissimilarity and link dissimilarity are broken into two measures of dissimilarity each. Comparing two graphs (A and B) results in differences in nodes, as well as links. The dissim‐ ilarity measures are used as SVM features (for a total of 4 features in the Stage-1 SVM described below), and include the nodes in graph A that are absent from graph B, links in graph A that are absent from B, nodes in graph B has that are absent from graph A, and links in graph B that are absent from graph A. All four dissimilarity measures are normalized and vary with cutset. Figure 6 shows these dissimilarity measures varying with time and how each cutset results in features and labels ("+" for pre-ictal, and "–" for inter-ictal).

kernel function evaluates to 0 as the distance between the two points becomes very large. The region where the kernel function evaluates to zero is parameterized by the value of gamma (*γ*), which is inversely proportional to the width of a multi-dimensional Gaussian function. SVMs with RBF kernels transform the decision boundary from a hyper-plane into much more amorphous decision boundary in the feature space. Figure 7 shows the difference in bounda‐ ries found by linear and RBF kernels for a representative SVM example in two dimensions.

Nonlinear Epilepsy Forewarning by Support Vector Machines

http://dx.doi.org/10.5772/57438

37

Each point in Figure 7 is equivalent to one instance of a class. Positive class values are denoted by positive signs and negative class values are denoted by negative signs. The Cartesian dimensions are the feature values—such as a dissimilarity measure. More than two features (two dimensions) can be used with a SVM, but it is more difficult to visualize when more than 3 dimensions (features) are involved. The main requirement of a RBF kernel is that the training set has a representative sample of the data that will be observed in the future and enough features to make distinctions between classes. Additionally, the range and scale of each feature has a large effect on the value of *γ*, the results, and the accuracy of predictions. Given a SVM model, future points are likely to be labeled as the class, to which they are most similar in the feature space. Similarity is defined by the kernel function as closeness between points of one class in the feature space. In essence, points from the training set are stored along with a weight associated with that point. A weighted sum of inner products is computed to evaluate how similar a new point is to all of the training data. The SVM training phase minimizes a cost function of several parameters, one of which is a vector, *θ*, of weights. Eq. (3) [18] gives the

) + (1 - y(i)

)cost0

(θTf (i) )) + 1

<sup>2</sup> <sup>θ</sup> <sup>2</sup> (3)

**Figure 7.** Types of boundaries found with two main SVM kernel types.

SVM cost function to be minimized with a kernel.

min <sup>θ</sup> C∑ i=1 m (y(i) cost1 (θTf (i)

**Figure 6.** Cutset labels (+/-) and features (four dissimilarity measures).

Analysis of graph dissimilarity measures by a SVM allows quantification of the change in topology over time by determining how dissimilar the graphs must be to predict an epileptic event. The details of the forewarning algorithm are in Section 5—with a brief overview of Support Vector Machines in Section 4.

#### **4. SVM with RBF kernels**

SVMs are one of the most commonly used supervised learning tools. The SVM approach was originally designed as a two-class (binary) classifier, but has been expanded to single and multiple classes. A SVM without a kernel function performs linear classification by finding a hyper-plane in the feature space that best maximizes a margin of separation between two classes with a given list of features.

SVM kernels define the similarity between two points in the feature space. For example, with a radial-basis-function (RBF) kernel, two points are said to be similar when they are proximate to one another in the feature space. The RBF kernel function for two points in a feature space evaluates to 1 when the distance between the two points approaches zero. The RBF, Gaussian kernel function evaluates to 0 as the distance between the two points becomes very large. The region where the kernel function evaluates to zero is parameterized by the value of gamma (*γ*), which is inversely proportional to the width of a multi-dimensional Gaussian function. SVMs with RBF kernels transform the decision boundary from a hyper-plane into much more amorphous decision boundary in the feature space. Figure 7 shows the difference in bounda‐ ries found by linear and RBF kernels for a representative SVM example in two dimensions.

**Figure 7.** Types of boundaries found with two main SVM kernel types.

are absent from B, nodes in graph B has that are absent from graph A, and links in graph B that are absent from graph A. All four dissimilarity measures are normalized and vary with cutset. Figure 6 shows these dissimilarity measures varying with time and how each cutset

Analysis of graph dissimilarity measures by a SVM allows quantification of the change in topology over time by determining how dissimilar the graphs must be to predict an epileptic event. The details of the forewarning algorithm are in Section 5—with a brief overview of

SVMs are one of the most commonly used supervised learning tools. The SVM approach was originally designed as a two-class (binary) classifier, but has been expanded to single and multiple classes. A SVM without a kernel function performs linear classification by finding a hyper-plane in the feature space that best maximizes a margin of separation between two

SVM kernels define the similarity between two points in the feature space. For example, with a radial-basis-function (RBF) kernel, two points are said to be similar when they are proximate to one another in the feature space. The RBF kernel function for two points in a feature space evaluates to 1 when the distance between the two points approaches zero. The RBF, Gaussian

results in features and labels ("+" for pre-ictal, and "–" for inter-ictal).

**Figure 6.** Cutset labels (+/-) and features (four dissimilarity measures).

Support Vector Machines in Section 4.

**4. SVM with RBF kernels**

36 Epilepsy Topics

classes with a given list of features.

Each point in Figure 7 is equivalent to one instance of a class. Positive class values are denoted by positive signs and negative class values are denoted by negative signs. The Cartesian dimensions are the feature values—such as a dissimilarity measure. More than two features (two dimensions) can be used with a SVM, but it is more difficult to visualize when more than 3 dimensions (features) are involved. The main requirement of a RBF kernel is that the training set has a representative sample of the data that will be observed in the future and enough features to make distinctions between classes. Additionally, the range and scale of each feature has a large effect on the value of *γ*, the results, and the accuracy of predictions. Given a SVM model, future points are likely to be labeled as the class, to which they are most similar in the feature space. Similarity is defined by the kernel function as closeness between points of one class in the feature space. In essence, points from the training set are stored along with a weight associated with that point. A weighted sum of inner products is computed to evaluate how similar a new point is to all of the training data. The SVM training phase minimizes a cost function of several parameters, one of which is a vector, *θ*, of weights. Eq. (3) [18] gives the SVM cost function to be minimized with a kernel.

$$\mathbf{f}^{\text{min}}\_{\theta} \mathbf{C} \sum\_{i=1}^{m} \left( \mathbf{y}^{\text{(i)}} \text{cost}\_{1} \{ \mathbf{\Theta}^{\text{T}} \mathbf{f}^{\text{(i)}} \} + \{ \mathbf{1} \cdot \mathbf{y}^{\text{(i)}} \} \text{cost}\_{0} \{ \mathbf{\Theta}^{\text{T}} \mathbf{f}^{\text{(i)}} \} \right) + \frac{1}{2} \| \mathbf{\Theta} \|^{2} \tag{3}$$

Here, *y* (*i*) is the class label of a training point i (positive or negative one); *f* (*i*) is the feature vector of the point (*i*), which compares a single point with the other points in the training set. Cost is an input parameter to the SVM and is a penalty for being in one class or the other.

Once the vector *θ* is found through the minimization, it can be used to determine the classes of new points. The method of determining the label of new points is shown in Eq. (4) [18].

$$\text{class label of new point} = \begin{cases} \text{Predict class 1 } \text{when } \Theta^{\text{T}} \mathbf{f}^{\text{(i)}} \mathbf{\hat{}} \mathbf{0} \\ \text{Predict class - 1 } \text{when } \Theta^{\text{T}} \mathbf{f}^{\text{(i)}} \mathbf{\hat{}} \mathbf{0} \end{cases} \tag{4}$$

Once the vector *θ* is obtained from training, it can be used to determine whether a new point in the feature space is of one class or another as given by Eq. (4). For each new point of index *i* in a test set, a vector f (i) is computed. The vector f (i) is a function of the kernel and the points in the training set. Each row in the vector f (i) is the value of the kernel function when the test set point and the training set points are inputs. The Gaussian kernel function in Eq. (5) will evaluate to 1 for points that are close together in the feature space and to zero when they are not. Eq. (5) gives the kernel function comparing a test set point *xi* to a landmark training set point, *l* (*k* ) [18].

$$\|f\_k^{(i)} = \exp\left( -\gamma \|\|\mathbf{x}\_i - l^{(k)}\|\|^2 \right) \tag{5}$$

is enough forewarning to stop or mitigate an event. Patients and caregivers [20] suggested 1-6 hours for safety, planning the day, and "driving myself to the hospital." Non-parent caregivers preferred 25 minutes to 1 hour for travel to the patient's location. Others gave 3-5 minutes, because longer forewarning was seen as more stressful to the patient. These requirements as well as previous research indicating that these constraints are a reasonable request—led to the labeling scheme used for pre-ictal indications for a SVM. For epileptic event data sets, the pre-ictal region is labeled as being anywhere from 3.3 minutes to 70 minutes before the seizure. Each epileptic patient is labeled as being pre-ictal for the same length of time prior to the seizure. Each plus and minus sign in Table 1 represents a 3.3 minute window (consistent with a cutset length of 49716 points, sampled at 250 Hz). The number of pluses is determined by a parameter (p) that is varied during cross validation. Figure 6 shows how the signs in Table 1

Nonlinear Epilepsy Forewarning by Support Vector Machines

http://dx.doi.org/10.5772/57438

39

relate to graphs, features, cutsets, and dissimilarity measures.

Stage algorithm.

**Patient data type 3.3 minute labeling 70 minute labeling**

Value of p 1 21

**Table 1.** Effect of variable p (number of + values) on Stage-1 pre-indication labeling of training data.

Event data set -----------------------------------+ --------------------------+++++++++++++++++++++ Non-event data set ------------------------------------- --------------------------------------------------------------

The input labeling (e.g., Table 1) assumes only approximate correctness and uses class weights to vary the correctness likelihood. The SVM methodology is implemented in three Stages with 10-fold cross validation. Stage-1 constructs a classifier that can label the pre-ictal state indica‐ tors. Stage-2 determines how long a patient must exhibit pre-ictal indicators in order to be certain of a seizure. Stages 1 and 2 establish cross validation accuracy and error. The SVM forewarning algorithm and previous voting method algorithm [4] both imply that patients must be in abnormal states a higher portion of the time before they are likely to have a seizure. Datasets without seizures can have infrequent abnormal states as well. Stage-3 obtains two models that can be used for seizure prediction in an ambulatory setting. Cross validation results in k different classifiers that leave out disjoint sets of data to establish an off-trainingset error (OTS error) to avoid overconfidence in accuracy. However, k slightly handicapped classifiers result in either less accuracy than is possible or more complexity in creating an ensemble. Stage-3 avoids this unnecessary choice by performing cross validation to create a final SVM model that includes all of the available data. Accuracy and error rates are statistically stronger, when they are reported from cross validation. The statistical claims are less robust, when one trains and tests on the same data. SVM with a RBF kernel is particularly susceptible to over-fitting—implying the need for cross validation. Figure 8 shows an outline of this three-

Without a kernel, the vector *θ* has n+1 dimensions for n features. With a kernel, the cardinality of *θ* and f (i) is m+1 for m training points. Often, this comparison to known points in the training set with a kernel function is referred to as mapping the feature space into a higher dimensional space. Indeed, the dimensionality of *f <sup>i</sup>* and *θ* increases from approximately the number of features without a kernel to roughly the number of training points with a kernel. The kernel results in linearly separable classes in the higher dimensional space, where the classes are not linearly separable in the feature space. The above description applies to binary classification with a RBF kernel and is intended to give the reader intuition on the types of boundaries that are found in the feature space as illustrated in Figure 7.

LIBSVM makes the details of the linear algebra of the training and testing transparent to the user and is easily used with the intuitions given in this Section [19]. The effect of a RBF kernel is that points proximate to one another will be labeled as belonging to the same class. Multiclass classification is treated as several binary classifications and is beyond the scope of the present work.

#### **5. SVM forewarning algorithm using graph dissimilarity features**

Labeling training data as pre-ictal or inter-ictal requires assumptions that must be sound. Current epilepsy prediction algorithms offer guidance about acceptable assumptions. The goal is enough forewarning to stop or mitigate an event. Patients and caregivers [20] suggested 1-6 hours for safety, planning the day, and "driving myself to the hospital." Non-parent caregivers preferred 25 minutes to 1 hour for travel to the patient's location. Others gave 3-5 minutes, because longer forewarning was seen as more stressful to the patient. These requirements as well as previous research indicating that these constraints are a reasonable request—led to the labeling scheme used for pre-ictal indications for a SVM. For epileptic event data sets, the pre-ictal region is labeled as being anywhere from 3.3 minutes to 70 minutes before the seizure. Each epileptic patient is labeled as being pre-ictal for the same length of time prior to the seizure. Each plus and minus sign in Table 1 represents a 3.3 minute window (consistent with a cutset length of 49716 points, sampled at 250 Hz). The number of pluses is determined by a parameter (p) that is varied during cross validation. Figure 6 shows how the signs in Table 1 relate to graphs, features, cutsets, and dissimilarity measures.

Here, *y* (*i*)

38 Epilepsy Topics

*i* in a test set, a vector f

point, *l*

of *θ* and f

present work.

(*k* ) [18].

(i)

(i)

in the training set. Each row in the vector f

space. Indeed, the dimensionality of *f <sup>i</sup>*

are found in the feature space as illustrated in Figure 7.

is the class label of a training point i (positive or negative one); *f* (*i*)

vector of the point (*i*), which compares a single point with the other points in the training set. Cost is an input parameter to the SVM and is a penalty for being in one class or the other.

Once the vector *θ* is found through the minimization, it can be used to determine the classes of new points. The method of determining the label of new points is shown in Eq. (4) [18].

Once the vector *θ* is obtained from training, it can be used to determine whether a new point in the feature space is of one class or another as given by Eq. (4). For each new point of index

set point and the training set points are inputs. The Gaussian kernel function in Eq. (5) will evaluate to 1 for points that are close together in the feature space and to zero when they are not. Eq. (5) gives the kernel function comparing a test set point *xi* to a landmark training set

Without a kernel, the vector *θ* has n+1 dimensions for n features. With a kernel, the cardinality

set with a kernel function is referred to as mapping the feature space into a higher dimensional

features without a kernel to roughly the number of training points with a kernel. The kernel results in linearly separable classes in the higher dimensional space, where the classes are not linearly separable in the feature space. The above description applies to binary classification with a RBF kernel and is intended to give the reader intuition on the types of boundaries that

LIBSVM makes the details of the linear algebra of the training and testing transparent to the user and is easily used with the intuitions given in this Section [19]. The effect of a RBF kernel is that points proximate to one another will be labeled as belonging to the same class. Multiclass classification is treated as several binary classifications and is beyond the scope of the

Labeling training data as pre-ictal or inter-ictal requires assumptions that must be sound. Current epilepsy prediction algorithms offer guidance about acceptable assumptions. The goal

**5. SVM forewarning algorithm using graph dissimilarity features**

(i)

=exp (-*γ xi* - *l*

*Predict class* - 1 *when* θTf

(i)

(*k* ) 2

is m+1 for m training points. Often, this comparison to known points in the training

(i) >0

> (i) <0

is a function of the kernel and the points

) (5)

is the value of the kernel function when the test

and *θ* increases from approximately the number of

class label of new point= {*Predict class* <sup>1</sup> *when* <sup>θ</sup>Tf

is computed. The vector f

*f k* (*i*) is the feature

(4)


**Table 1.** Effect of variable p (number of + values) on Stage-1 pre-indication labeling of training data.

The input labeling (e.g., Table 1) assumes only approximate correctness and uses class weights to vary the correctness likelihood. The SVM methodology is implemented in three Stages with 10-fold cross validation. Stage-1 constructs a classifier that can label the pre-ictal state indica‐ tors. Stage-2 determines how long a patient must exhibit pre-ictal indicators in order to be certain of a seizure. Stages 1 and 2 establish cross validation accuracy and error. The SVM forewarning algorithm and previous voting method algorithm [4] both imply that patients must be in abnormal states a higher portion of the time before they are likely to have a seizure. Datasets without seizures can have infrequent abnormal states as well. Stage-3 obtains two models that can be used for seizure prediction in an ambulatory setting. Cross validation results in k different classifiers that leave out disjoint sets of data to establish an off-trainingset error (OTS error) to avoid overconfidence in accuracy. However, k slightly handicapped classifiers result in either less accuracy than is possible or more complexity in creating an ensemble. Stage-3 avoids this unnecessary choice by performing cross validation to create a final SVM model that includes all of the available data. Accuracy and error rates are statistically stronger, when they are reported from cross validation. The statistical claims are less robust, when one trains and tests on the same data. SVM with a RBF kernel is particularly susceptible to over-fitting—implying the need for cross validation. Figure 8 shows an outline of this three-Stage algorithm.


there may be similar points in the feature space outside the hour time-frame being labeled as pre-ictal and inter-ictal, the variable for class weights are varied via a Monte-Carlo search over the parameter space during cross validation to determine how to tie-break. Other parameters are also varied randomly during the Monte-Carlo search, as shown in Table 2. To compensate for labeling uncertainty, the cost sensitive SVM adjusts a weight on the class labels to indicate how certain the labeling scheme is for the pre-ictal and the inter-ictal classes. The labeling scheme combined with the features, training set, class weights, and gamma creates regions in the feature space that will be associated with one class or another. Additionally, we use stratified cross validation—maintaining a ratio of 4 event patients and 2 non-event patients in each strata of cross validation. Cross validation is performed on 90% of the patients—54 patients in each training set and 6 patients in each test set—having varying numbers of cutsets due to having varying length observations. This process is repeated 10 times with disjoint sets


Cross validation average prediction distance)

Train on 9/10ths of the 60 stage 2 rows and then predict on 1/10th to obtain accuracies (repeat 10 times to get

**Stage 1 Feature**

**4**


**Stage 1 Feature** 

Non-event patient

41

http://dx.doi.org/10.5772/57438

Event prediction Non-event prediction

• Cutsets equate to Stage 1 training rows • Cutset labels from predictions when lain out temporally yield strings with distinct patterns

Nonlinear Epilepsy Forewarning by Support Vector Machines


**3**

**Stage 1 Feature** 

+1.0 -1.63347602 2.48750567 -0.61728311 2.61945629 +1.0 1.00212920 -0.71793008 1.17944241 -0.61776310

**2**


Predicting with stage 1 models on 6 patients 10 times produce 60 Stage 2 rows

Train on 9/10ths of the 60 patients' cutsets and then predict on 1/10th to produce the following (repeat 10

Successive, contiguous occurrences of pre-ictal indicators trigger an alert (prediction of an event). Non-event datasets have more inter-ictal indicators labeled with negative symbols, while event patients have fairly dense pre-ictal indicators—labeled with plus symbols. A single

The accuracy of the assumptions is reflected in the success rate of the predictions during cross validation. The parameters that appear to be uncertain are left as search variables, recognizing that more free parameters create a more computationally complex search. Too many variables result in a computational explosion in the CPU time to explore the search space. Each point in the parameter space corresponds deterministically with a cross-validation error rate. Assump‐ tions about the certainty of the training or testing example's labels are represented by class

pre-ictal indicator is usually not enough to make accurate predictions.

of patients in each test set.

**Stage 1 Label Stage 1 Feature 1**

times)

**(max contiguous +)**

**Figure 9.** Stage-1 and 2 flow (each + or – represents a 3.3 minute cutset).

**Stage 2 label Stage 2 Feature**

+1.0 5.0 -1.0 1.0

Figure 9 shows how Stages 1 and 2 flow together. Stage-3 involves training the RBF Model on all of the Stage-1 cutsets (4244 rows, instead of approximately 90% of it) and training the linear model on **Figure 8.** Steps in the three stages for cross validation and final model construction.

the all of the Stage-2 results (60 rows, instead of 90% of it). Then, one predicts on the training data to verify that the model is working as expected to produceܦௗ௦ . Figure 9 shows how Stages 1 and 2 flow together. Stage-3 involves training the RBF Model on all of the Stage-1 cutsets (4244 rows, instead of approximately 90% of it) and training the linear model on the all of the Stage-2 results (60 rows, instead of 90% of it). Then, one predicts on the training data to verify that the model is working as expected to produce *Dfinal models* .

Figure 8. Steps in the three stages for cross validation and final model construction.

Figure 9 shows that event datasets are labeled in Stage-1 as pre-ictal (+) in a window of p cutsets prior to the seizure and inter-ictal (-) outside of this window. All cutsets in non-seizure datasets are labeled as inter-ictal (-). A cost sensitive SVM is used to account for the uncertainty in the pre-ictal and inter-ictal labeling. The motive for this labeling scheme is the caregiver's desire to have forewarning within an hour of the event. Indicators are assumed to be near the event, and the time window is varied by the parameter (p) that is tested during cross validation. The assumption behind the design choice is that the pre-ictal state is a rare occurrence. Because


**Figure 9.** Stage-1 and 2 flow (each + or – represents a 3.3 minute cutset).

I. Stage-1

40 Epilepsy Topics

II. Stage-2

III. Stage-3

RBF kernel;

2 non-event patients;

kernel (see Table 5);

ி ଶ ቁ ଶ ቀிே ସ ቁ ଶ

forewarning (see Table 5);

optimistic over-fit);

**Figure 8.** Steps in the three stages for cross validation and final model construction.

verify that the model is working as expected to produceܦௗ௦ .

f. Repeat (2b) – (2e) 10 times;

from (2b);

e. Get ܦ ൌ ටቀ

a. Label each patient's feature set from (1g)

i. Event data sets are labeled as +1; ii. Non-event data sets are labeled as -1;

b. Labels

a. Obtain 4 dissimilarity measures from phase space analysis (for each cutset in a patient data set). These dissimilarity measures become the features of the

> i. Labels for +1: p cutsets immediately before an event; ii. Labels for -1: all other cutsets. See Table 1;

c. Divide datasets into 10 sets of patients: each set contains 4 event patients and

d. Train RBF Model on 9 sets of patients – obtain SVM model (see Figure 9);

b. Train on 9 sets of max contiguous pre-ictal indicators from (1g): linear

c. Predict on 1 set of max contiguous indicators from (1g): via linear model

(We use stratified cross validation);

d. Get false positive rate (FP) and false negative rate (FN) from (2c);

g. Get the average over ܦ) average cross validation OTS error rate);

d. Retrain linear model (similar to Stage-2, but with all 60 patients' max

successive indicators) to get the number of successive occurrences to trigger

e. Use data from (1b) to retrain RBF model on all 60 patients' cutsets (a total of

f. Use RBF Model (3e) result to predict +/- on data from (1b); see Table 7; g. Use linear model from (3d) to do event forewarning on all 60 patients (see Table 4-5, Table 7, Figure 9, Figure 10) and get ܦ ௗ௦ (represents an

a. Use D(average)=(Average prediction distance) from (2g); b. If D(average) < 0.7 create models to use in future via 3c-g; c. Use results (see Table 4-5) from all 60 data sets from (1g);

4244 cutsets) to obtain the RBF model (see Figure 10);

Figure 8. Steps in the three stages for cross validation and final model construction. Figure 9 shows how Stages 1 and 2 flow together. Stage-3 involves training the RBF Model on all of the Stage-1 cutsets (4244 rows, instead of approximately 90% of it) and training the linear model on the all of the Stage-2 results (60 rows, instead of 90% of it). Then, one predicts on the training data to

Figure 9 shows how Stages 1 and 2 flow together. Stage-3 involves training the RBF Model on all of the Stage-1 cutsets (4244 rows, instead of approximately 90% of it) and training the linear model on the all of the Stage-2 results (60 rows, instead of 90% of it). Then, one predicts on the

Figure 9 shows that event datasets are labeled in Stage-1 as pre-ictal (+) in a window of p cutsets prior to the seizure and inter-ictal (-) outside of this window. All cutsets in non-seizure datasets are labeled as inter-ictal (-). A cost sensitive SVM is used to account for the uncertainty in the pre-ictal and inter-ictal labeling. The motive for this labeling scheme is the caregiver's desire to have forewarning within an hour of the event. Indicators are assumed to be near the event, and the time window is varied by the parameter (p) that is tested during cross validation. The assumption behind the design choice is that the pre-ictal state is a rare occurrence. Because

training data to verify that the model is working as expected to produce *Dfinal models* .

e. Predict on the remaining, 10th set with RBF Model from (1d) ; f. Scan the results from (1e) for max # of contiguous +1. See Table 4-5. g. Repeat 1d-1f (and save results): 10 predicted sets. See Table 4-5.

> there may be similar points in the feature space outside the hour time-frame being labeled as pre-ictal and inter-ictal, the variable for class weights are varied via a Monte-Carlo search over the parameter space during cross validation to determine how to tie-break. Other parameters are also varied randomly during the Monte-Carlo search, as shown in Table 2. To compensate for labeling uncertainty, the cost sensitive SVM adjusts a weight on the class labels to indicate how certain the labeling scheme is for the pre-ictal and the inter-ictal classes. The labeling scheme combined with the features, training set, class weights, and gamma creates regions in the feature space that will be associated with one class or another. Additionally, we use stratified cross validation—maintaining a ratio of 4 event patients and 2 non-event patients in each strata of cross validation. Cross validation is performed on 90% of the patients—54 patients in each training set and 6 patients in each test set—having varying numbers of cutsets due to having varying length observations. This process is repeated 10 times with disjoint sets of patients in each test set.

> Successive, contiguous occurrences of pre-ictal indicators trigger an alert (prediction of an event). Non-event datasets have more inter-ictal indicators labeled with negative symbols, while event patients have fairly dense pre-ictal indicators—labeled with plus symbols. A single pre-ictal indicator is usually not enough to make accurate predictions.

> The accuracy of the assumptions is reflected in the success rate of the predictions during cross validation. The parameters that appear to be uncertain are left as search variables, recognizing that more free parameters create a more computationally complex search. Too many variables result in a computational explosion in the CPU time to explore the search space. Each point in the parameter space corresponds deterministically with a cross-validation error rate. Assump‐ tions about the certainty of the training or testing example's labels are represented by class

weights in a cost sensitive SVM. Table 2 shows the variables for the SVMs that were searched during the research for this paper.

forewarning of seizure events. A Monte-Carlo search is used over variables in Table 2 because the prediction distance has very irregular, fractal behavior—with sparse parameters generat‐ ing good predictions and the gradients of the parameter space being highly irregular with

The algorithm for forewarning is done in two Stages of processing after producing diffeomor‐ phic graphs and their dissimilarity measures for each graph. Stage-1 uses a cost sensitive SVM type (CSVC) from LIBSVM [19] with a RBF kernel. For each iteration of the cross validation, the algorithm labels event data sets as having pre-ictal indicators within a one-hour window prior to the seizure; see Table 1. The length of this window (p cutsets) is part of the Monte Carlo search. All other values—non-event data and event data far away from the seizure (outside the variable window)—are labeled as inter-ictal indicators. The analysis trains on k-1 sets, then predicts on a single, left-out set; this analysis is repeated k times in a k-fold cross validation (with k being 10). Table 4 shows a small sample of Stage-1 predictions for two patient outputs, but in actuality there are 60 sets of predictions—like Table 7 in the results section.

Event Data set --------------------+--------------------+-----++-++----------+++E 53 3 Non Event Data set -------------------------------------------------------NE no prediction 0

**Label meaning Label Single Feature (Maximum Successive occurrences of +)**

The analysis labels the Stage-2 training and testing values as either event or non-event data sets. In general, there are fewer pre-ictal indicators (+ in Table 4) in the non-event data sets on successful cross validation runs. One determines the cross validation average prediction

After making the six predictions on six patients for one of ten cross validation runs, one creates a new set of cross validation folds (representing sets of patients) for the Stage-2 analysis out of the Stage-1's predictions on the omitted set (similar to the middle column of Table 4). For the Stage-2 single feature, one scans each of the predictions made by the Stage-1 for the maximum number of contiguous occurrences of pre-ictal indicators. Specifically, the third column of Table 4 is obtained from scanning the second column of Table 4. Stage-2 of the algorithm has training and testing files that follow the format of the second and third columns

**Table 4.** Representative predictions from Stage-1 (E=event; NE=no event).

Event Data set +1.0 3 Non Event Data set -1.0 0

**Table 5.** Training/Testing set format for Stage-2.

**Example predictions from Stage-1 Maximum**

Nonlinear Epilepsy Forewarning by Support Vector Machines

http://dx.doi.org/10.5772/57438

43

**Successive occurrences of +**

many local maxima and minima [4].

**Common prediction output**

of Table 5.


**Table 2.** Parameters for the SVM portion of analysis.

Figure 4 and Table 3 lists other variables that might be searched in addition to those of the SVM. Those parameters were fixed in the present analysis because Takens' theorem is sufficiently powerful to show significant changes in topology with many sets of parameters when those topological changes are normalized properly. The dissimilarity measures for a patient reflect relative differences in graph topology. The parameter values for the phase-space graph generation are shown in the Table 3. Recall that these parameters were described near and are contained in Figure 4. Throughout the entire analysis, the parameters to generate the phase space graphs were kept fixed. These values were found in our prior work [4] to give good forewarning using ensemble voting methods mentioned in the background section.


**Table 3.** Parameters used in generation of phase space graphs [4].

Statistical validation of forewarning requires measures of success. One measure is the number of true positives (TP) for known event datasets (Ev), to yield the true positive rate (sensitivity) of TP/Ev. A second measure is the number of true negatives (TN) for known non-event datasets (NEv). The true negative rate is TN/NEv (specificity). The goal is a sensitivity and specificity of unity. Consequently, minimizing the distance from ideal (*D* = prediction distance) is an appropriate objective function for any event type:

$$\mathbf{D} = \sqrt{\mathbf{1} - \left(\frac{\mathbf{T}\mathbf{P}}{\mathbf{E}\mathbf{v}}\right)\mathbf{I}^2 + \left[\mathbf{1} - \left(\frac{\mathbf{T}\mathbf{N}}{\mathbf{N}\mathbf{E}\mathbf{v}}\right)\mathbf{I}^2\right]}\tag{6}$$

Eq. (6) is the objective function to be minimized for the OTS error rate and over-fit error rate. The OTS error rate is found from 10-fold cross validation from Stages 1 and 2. The over-fit error rate verifies that the final models in Stage-3 can correctly predict the training examples. Excessive false positives (inverse of a true negative) will cause real alarms to be ignored and needlessly expend caregiver resources. False negatives (inverse of a true positive) provide no forewarning of seizure events. A Monte-Carlo search is used over variables in Table 2 because the prediction distance has very irregular, fractal behavior—with sparse parameters generat‐ ing good predictions and the gradients of the parameter space being highly irregular with many local maxima and minima [4].

The algorithm for forewarning is done in two Stages of processing after producing diffeomor‐ phic graphs and their dissimilarity measures for each graph. Stage-1 uses a cost sensitive SVM type (CSVC) from LIBSVM [19] with a RBF kernel. For each iteration of the cross validation, the algorithm labels event data sets as having pre-ictal indicators within a one-hour window prior to the seizure; see Table 1. The length of this window (p cutsets) is part of the Monte Carlo search. All other values—non-event data and event data far away from the seizure (outside the variable window)—are labeled as inter-ictal indicators. The analysis trains on k-1 sets, then predicts on a single, left-out set; this analysis is repeated k times in a k-fold cross validation (with k being 10). Table 4 shows a small sample of Stage-1 predictions for two patient outputs, but in actuality there are 60 sets of predictions—like Table 7 in the results section.


**Table 4.** Representative predictions from Stage-1 (E=event; NE=no event).

weights in a cost sensitive SVM. Table 2 shows the variables for the SVMs that were searched

γ *crbf weigh t***<sup>+</sup>** *weigh t***-** *clinear* **p**

Figure 4 and Table 3 lists other variables that might be searched in addition to those of the SVM. Those parameters were fixed in the present analysis because Takens' theorem is sufficiently powerful to show significant changes in topology with many sets of parameters when those topological changes are normalized properly. The dissimilarity measures for a patient reflect relative differences in graph topology. The parameter values for the phase-space graph generation are shown in the Table 3. Recall that these parameters were described near and are contained in Figure 4. Throughout the entire analysis, the parameters to generate the phase space graphs were kept fixed. These values were found in our prior work [4] to give good forewarning using ensemble voting methods mentioned in the background section.

**B D L M N S W** 12 7 56 77 49716 3 29

Statistical validation of forewarning requires measures of success. One measure is the number of true positives (TP) for known event datasets (Ev), to yield the true positive rate (sensitivity) of TP/Ev. A second measure is the number of true negatives (TN) for known non-event datasets (NEv). The true negative rate is TN/NEv (specificity). The goal is a sensitivity and specificity of unity. Consequently, minimizing the distance from ideal (*D* = prediction distance) is an

<sup>+</sup> <sup>1</sup> –( TN

Eq. (6) is the objective function to be minimized for the OTS error rate and over-fit error rate. The OTS error rate is found from 10-fold cross validation from Stages 1 and 2. The over-fit error rate verifies that the final models in Stage-3 can correctly predict the training examples. Excessive false positives (inverse of a true negative) will cause real alarms to be ignored and needlessly expend caregiver resources. False negatives (inverse of a true positive) provide no

NEv ) <sup>2</sup> (6)

Weighs how powerfully the – class influences the decision boundary

Adjusts the pliability of the Stage-2 decision boundary given additional points Number of cutsets prior to the event that are labeled as being in the positive class.

Weighs how powerfully the + class influences the decision boundary

during the research for this paper.

Adjusts the pliability of the Stage-1

points

**Table 2.** Parameters for the SVM portion of analysis.

**Table 3.** Parameters used in generation of phase space graphs [4].

appropriate objective function for any event type:

<sup>D</sup> <sup>=</sup> <sup>1</sup> – ( TP

Ev ) <sup>2</sup>

decision boundary given additional

Sets the radius of the contribution of a single point to the decision boundary

42 Epilepsy Topics

After making the six predictions on six patients for one of ten cross validation runs, one creates a new set of cross validation folds (representing sets of patients) for the Stage-2 analysis out of the Stage-1's predictions on the omitted set (similar to the middle column of Table 4). For the Stage-2 single feature, one scans each of the predictions made by the Stage-1 for the maximum number of contiguous occurrences of pre-ictal indicators. Specifically, the third column of Table 4 is obtained from scanning the second column of Table 4. Stage-2 of the algorithm has training and testing files that follow the format of the second and third columns of Table 5.


**Table 5.** Training/Testing set format for Stage-2.

The analysis labels the Stage-2 training and testing values as either event or non-event data sets. In general, there are fewer pre-ictal indicators (+ in Table 4) in the non-event data sets on successful cross validation runs. One determines the cross validation average prediction distance by training on values of maximum contiguous, pre-ictal indicators from k-1 subsets at the Stage-2, making k predictions on the omitted subset, and then taking the average of the prediction distances.

One then retrains the Stage-2 model on all of the predictions from Stage-1. The retraining process results in two SVM models—one for Stage-1 and one for Stage-2. With these models,

Table 6 shows representative SVM parameter values (from the Monte Carlo search) and results. *NOCC* is the number of contiguous, pre-ictal indicators (+ labels) that must be present before forewarning occurs and is found by training the Stage-2 SVM model (and Stage 3g). A value of Nocc of 2 implies that dynamics over a 6.6 minute period must be observed to be abnormal for prediction. Nocc of 1 implies that dynamics over a 3.3 minute period must be observed to be abnormal to have forewarning. D(AVG) is the average OTS error rate during cross valida‐ tion runs; D(AVG)<0.7 indicates that the algorithm is performing more accurately than random guessing and simple heuristics. D(final) is a value verifying that the final models in Stage-3 have the capability of accurately classifying the training data. D(final) represents an over-fit, but is valuable in narrowing the Monte-Carlo search and determining whether the algorithm

Table 6 shows that the best cross validation accuracy with an average prediction distance of 0.287 and a final model prediction distance of 0.056. Table 6 shows additional representative cross validation averages—D(Avg)—and final model prediction distances—D(final). Hun‐ dreds of runs resulted in cross validation prediction distances of <0.5. The best cross validation accuracy of.287 achieved thus far also has a fairly decent over-fit error rate of 0.056—which

> 20 )2 <sup>+</sup> ( <sup>1</sup> 40 )2

positive and one false negative in the final model's ability to predict the original training data. Although, due to the fact that the number of events and non-events was unbalanced, the probability of having a false negative was twice as likely as false positive, which is still

**Ex#** γ *crbf weigh t***<sup>+</sup>** *weigh t***- NOCC** *clinear* **D(AVG) D(final)**

 3.929 9.732 19.160 24.860 1 3.723 0.287 0.056 21 2.640 87.825 32.867 81.832 1 2.945 0.342 0.025 22 2.860 10.022 85.200 50.843 2 8.033 0.374 0.075 18 0.964 85.719 71.740 33.237 2 9.259 0.396 0.125 19 7.709 2.587 28.705 26.903 2 2.347 0.413 0.050 20 2.207 75.920 52.493 36.102 2 8.994 0.438 0.125 18 6.783 18.035 49.974 30.441 2 6.985 0.456 0 21

≈.056, corresponding to one false

Nonlinear Epilepsy Forewarning by Support Vector Machines

http://dx.doi.org/10.5772/57438

45

**size of + label window**

one can make predictions on new data.

has a value of the objective function, *<sup>D</sup>* <sup>=</sup> ( <sup>1</sup>

**Table 6.** Summary of typical best results to date.

has any merit.

desirable.

Stage-3 obtains the cross-validation prediction distance via two models that predict on the basis of all of the data available instead of 90% of it. Specifically, Stage-3 takes the 4244 cutsets labeled in Stage-1 and the sixty predictions from Stage-2 to create two SVM models for predicting on future patients. Stage-3 involves retraining both the RBF model (from Stage-1) and linear model (from Stage-2). Figure 10 illustrates the flow in Stage-3 after optimal param‐ eters have been discovered via cross validation. Once the two models are obtained, they are used to predict on the original data sets to verify the model's validity.



**Figure 10.** Stage-3 process (builds on Stages 1 and 2).

#### **6. Representative results**

From Eq. (6), the scenario of a classifier never getting the answer correct is *D* = (1)2 + (1)2 ≈1.41. A random number generator that is guessing each class with equiprob‐ ability will have a prediction distance over time, *<sup>D</sup>* <sup>=</sup> ( <sup>1</sup> 2 )2 <sup>+</sup> ( <sup>1</sup> 2 )2 ≈0.7. A perfectly ideal classifier will have an average prediction distance (OTS error rate) of 0 during cross validation. A Monte-Carlo search over the parameters for the SVM attempts to find the set of parameters and corresponding SVM models that minimize the prediction distance at Stage-2 for classifi‐ cation of each dataset as an event or a non-event. Cross validation on Stage-2 leads to k prediction distances that are averaged. The average is minimized and corresponding "ideal" parameters for the SVM are found. Once the Monte-Carlo search has found parameters associated with an acceptable OTS error rate, one retrains the Stage-1 model on all of the data. One then retrains the Stage-2 model on all of the predictions from Stage-1. The retraining process results in two SVM models—one for Stage-1 and one for Stage-2. With these models, one can make predictions on new data.

Table 6 shows representative SVM parameter values (from the Monte Carlo search) and results. *NOCC* is the number of contiguous, pre-ictal indicators (+ labels) that must be present before forewarning occurs and is found by training the Stage-2 SVM model (and Stage 3g). A value of Nocc of 2 implies that dynamics over a 6.6 minute period must be observed to be abnormal for prediction. Nocc of 1 implies that dynamics over a 3.3 minute period must be observed to be abnormal to have forewarning. D(AVG) is the average OTS error rate during cross valida‐ tion runs; D(AVG)<0.7 indicates that the algorithm is performing more accurately than random guessing and simple heuristics. D(final) is a value verifying that the final models in Stage-3 have the capability of accurately classifying the training data. D(final) represents an over-fit, but is valuable in narrowing the Monte-Carlo search and determining whether the algorithm has any merit.

Table 6 shows that the best cross validation accuracy with an average prediction distance of 0.287 and a final model prediction distance of 0.056. Table 6 shows additional representative cross validation averages—D(Avg)—and final model prediction distances—D(final). Hun‐ dreds of runs resulted in cross validation prediction distances of <0.5. The best cross validation accuracy of.287 achieved thus far also has a fairly decent over-fit error rate of 0.056—which has a value of the objective function, *<sup>D</sup>* <sup>=</sup> ( <sup>1</sup> 20 )2 <sup>+</sup> ( <sup>1</sup> 40 )2 ≈.056, corresponding to one false positive and one false negative in the final model's ability to predict the original training data. Although, due to the fact that the number of events and non-events was unbalanced, the probability of having a false negative was twice as likely as false positive, which is still desirable.


**Table 6.** Summary of typical best results to date.

distance by training on values of maximum contiguous, pre-ictal indicators from k-1 subsets at the Stage-2, making k predictions on the omitted subset, and then taking the average of the

Stage-3 obtains the cross-validation prediction distance via two models that predict on the basis of all of the data available instead of 90% of it. Specifically, Stage-3 takes the 4244 cutsets labeled in Stage-1 and the sixty predictions from Stage-2 to create two SVM models for predicting on future patients. Stage-3 involves retraining both the RBF model (from Stage-1) and linear model (from Stage-2). Figure 10 illustrates the flow in Stage-3 after optimal param‐ eters have been discovered via cross validation. Once the two models are obtained, they are

From Eq. (6), the scenario of a classifier never getting the answer correct is *D* = (1)2 + (1)2 ≈1.41. A random number generator that is guessing each class with equiprob‐

classifier will have an average prediction distance (OTS error rate) of 0 during cross validation. A Monte-Carlo search over the parameters for the SVM attempts to find the set of parameters and corresponding SVM models that minimize the prediction distance at Stage-2 for classifi‐ cation of each dataset as an event or a non-event. Cross validation on Stage-2 leads to k prediction distances that are averaged. The average is minimized and corresponding "ideal" parameters for the SVM are found. Once the Monte-Carlo search has found parameters associated with an acceptable OTS error rate, one retrains the Stage-1 model on all of the data.

2 )2 <sup>+</sup> ( <sup>1</sup> 2 )2

≈0.7. A perfectly ideal

used to predict on the original data sets to verify the model's validity.

**Figure 10.** Stage-3 process (builds on Stages 1 and 2).

ability will have a prediction distance over time, *<sup>D</sup>* <sup>=</sup> ( <sup>1</sup>

**6. Representative results**

prediction distances.

44 Epilepsy Topics

Example 7 in Table 6 has an average cross validation accuracy of.456 with D(final)=0 (perfect prediction). Cross validation average accuracy or error rate is the more valid statistical claim. The best cross validation accuracy represented in Table 6 is in the same realm of accuracy as Netoff et al.'s intracranial methodology. Recall that Netoff et al. claim a specificity of 77.8% with no false positives, which is approximately a prediction distance of approximately.22 (*DNetoff* <sup>≈</sup> (1 - .78)2 <sup>+</sup> (0)2 <sup>≈</sup>.22) [1]. Our best cross validation accuracy is of the same order of magnitude. Given that we are using scalp EEG, and Netoff et al. are using intra-cranial EEG, this is a statistically significant result that cannot be said to be an over fit. Mirowski et al. claim 100% accuracy, which would be a prediction distance of zero (*DMirowski* =0). However, Mirowski et al. are making patient specific machine learning models that are tailored to individual patients. We are creating a novel algorithm that can be applied to a group of patients with noninvasive EEG. Very little research is being done to advance non-invasive EEG prediction algorithms in this way. Furthermore, comparing cross validation accuracies and error rates dispels any arguments of overconfidence due to over-fitting.

Patient ------------------------------------NE no prediction 1 ------------------------------------NE no prediction 2 ------------------------------------------NE no prediction 3 -------------------------------------------------------------------NE no prediction 4 -+--------------+--------------------------------------------------------------------+------------------------------NE no prediction 5 ----------------------------------------------------------------------------------------------------------NE no prediction 6 ---------------------------------NE no prediction 7 ------------------------------------NE no prediction 8 ------------------------------------NE no prediction 9 ---------------------------------NE no prediction 10 -------------------------NE no prediction 11 ------------------------------------+------------------------NE no prediction 12 -----------------------------------------+----------------+--------------------------+-----------NE no prediction 13 -------+-----------+----+----NE no prediction 14 -----------------------------------------------------------------------------------------------------------------------------------NE no prediction 15 -------------------------NE no prediction 16 ----------------------+---------------------------------+----------------------------------------NE no prediction 17 ------------------+------------------------------------------------------------------------------------------------NE no prediction 18 ---------------+-------------------------------------------------NE no prediction 19 ----------+----------------------+-----------------------------------------------------+----------------NE no prediction 20 ---------------------------+-----------------------------------------------------------NE no prediction 21 --+----------------+-----------------------------+++++++++++++++++++++E 62 22 ------------------------------------------+++---++--+-+++-+--++E 62 23 ----------------------------+------------------------------+++++++++++++++++++++E 62 24 -----------------------------+-++++++++++-++-+++++E 56 25 -------------------------+----------------+--+---++-++++++++++-++E 46 26 ---------------+---------------+-+-+-++-++++++++++++++++E 56 27 -------------------------------------+-++--++++++-++--+-++E 56 28 ------------------------+-------+----------+-----------------++--++++-+-+++-+-++++E 62 29 ----------------------+-------+--------+----++-+++++++++++E 39 30 ---------------------------------+--+-+++++++++++++++++++E 56 31 --------------------------+-----------------------+-----------------+++++++-++++++++-+-++E 62 32 ---------------+---+-+-----------------------+---------------------+++++++++++---+++++++E 62 33 ---------------------------+-+-+++++++++-++++-++-+--+E 66 34 --------------------------------+++++++++++++--++++++E 62 35 -------------------------------------------------------------------------------------------------++++--+-++-+++E 39 36 -------------+---------------------------------+++-++++++-----+--+--E 62 37 --------------------------------------------------------------------------------+++++++++++++++++++--E 62 38 ---------------------------+-----------------++---------------+-----+++++-------------------------++++++++-++++++++++-+E 238 39 ----------------------------------------+-------------------------------++++-+++++++++++++++-E 62 40 ------------++----------------+-----------------+---+-------------++++++++++++++++++++-E 241 41 ----------------------------+++-+++++++++-+++-+++E 62 42 ------------------------------------++++++-+++++++-+---++++E 69 43 -----------------+--------+++++++++++++++++++++E 62 44 --------------------------------------+------------------------------------------++-++++++++-+++++++++E 62 45 ---------------------------------------------++-++-+++++++-+--++++E 62 46 --------------------------+---+-++-+-++-+-+-++-+++-E 56 47 --------------------+++----+----+++++++++++++++-+--+++++++++E 125 48 -------------------------------------------------------+-------+-----------------------+-----+--------++----++----++-----E 56 49 -----------------------------------------------------+++++++++++++++++++++E 62 50 -------------++--+++--+-+-----++++++++-++-+-++++++++E 122 51 ----------------------------+---------------------------+++---++++-+++-+--++E 59 52 ---------------------+----------+---------+--+--+-++-++-++-++++++----E 56 53 --------++-------------------++-++++++-++++++++++-E 132 54 --------------------------+-------------------------+-++++--+-----+-+++-++++++++++++++++++E 112 55 ----------------+--------------------------------+++++-++++++++++++++-E 62 56 -------------+-----+-----++----++++-----+++++++++-+-+++++++++E 112 57 ------------------------------------------------------------------------+++++++++++++++++++++E 62 58 --------------------------------------------++++++++++++++++++++-E 62 59 -------------------------++-------+++++++++++++++++--+++E 96 60 ----------------------------------------------------------------------+++++++++++++++++++++E 62

Nonlinear Epilepsy Forewarning by Support Vector Machines

http://dx.doi.org/10.5772/57438

47

**Table 7.** Stage-3g predictions on all 60 patients for Example 7 in Table 6 (E=event; NE=no event; number following

"E"=forewarning time in minutes).

Figure 11 shows a plot of forewarning times (typically less than 1.5h) for the final-model of Example 7. The number of successive contiguous indicators to trigger forewarning was found to be 2 successive + values. Table 7 shows the Stage-1 predictions (Stage-3g of Figure 8 for all 60 patients) to produce the distribution of forewarning times in Figure 11. See Table 7 for the cutset indications (+ or -) that correspond to the parameters in Example 7 from Table 6.

**Figure 11.** Distribution of forewarning times for Example #7 in Table 6.

#### Nonlinear Epilepsy Forewarning by Support Vector Machines http://dx.doi.org/10.5772/57438 47


Example 7 in Table 6 has an average cross validation accuracy of.456 with D(final)=0 (perfect prediction). Cross validation average accuracy or error rate is the more valid statistical claim. The best cross validation accuracy represented in Table 6 is in the same realm of accuracy as Netoff et al.'s intracranial methodology. Recall that Netoff et al. claim a specificity of 77.8% with no false positives, which is approximately a prediction distance of approximately.22 (*DNetoff* <sup>≈</sup> (1 - .78)2 <sup>+</sup> (0)2 <sup>≈</sup>.22) [1]. Our best cross validation accuracy is of the same order of magnitude. Given that we are using scalp EEG, and Netoff et al. are using intra-cranial EEG, this is a statistically significant result that cannot be said to be an over fit. Mirowski et al. claim 100% accuracy, which would be a prediction distance of zero (*DMirowski* =0). However, Mirowski et al. are making patient specific machine learning models that are tailored to individual patients. We are creating a novel algorithm that can be applied to a group of patients with noninvasive EEG. Very little research is being done to advance non-invasive EEG prediction algorithms in this way. Furthermore, comparing cross validation accuracies and error rates

Figure 11 shows a plot of forewarning times (typically less than 1.5h) for the final-model of Example 7. The number of successive contiguous indicators to trigger forewarning was found to be 2 successive + values. Table 7 shows the Stage-1 predictions (Stage-3g of Figure 8 for all 60 patients) to produce the distribution of forewarning times in Figure 11. See Table 7 for the cutset indications (+ or -) that correspond to the parameters in Example 7 from Table 6.

dispels any arguments of overconfidence due to over-fitting.

46 Epilepsy Topics

**Figure 11.** Distribution of forewarning times for Example #7 in Table 6.

**Table 7.** Stage-3g predictions on all 60 patients for Example 7 in Table 6 (E=event; NE=no event; number following "E"=forewarning time in minutes).

In Figure 11, the solid black line is the occurrence frequency (arbitrary units) in half-hour bins. The blue line is the cumulative distribution of forewarning versus time. The red H-bar with the star in the middle indicates the mean value of the forewarning times (approximately 1 hour) and the sample standard deviation. The result in Example 7 of Table 6 is better than random guessing or biased heuristics with D(final)=0, despite poorer cross validation accuracy than other examples. This example shows most of the forewarning times of ≤1.5 hours with a statistically significant accuracy. One can visually make a prediction by scanning Table 7 from left to right and looking for 2 contiguous plus values. When 2 values are found, the seizure is highly likely to occur. One may be tempted to reduce the forewarn time by increasing the value of positive values that trigger a forewarning, but that would likely result in bad OTS error, which is why it is not the value found by the second and third stage SVMs.

present approach uses retrospective analysis of archival data on a desktop computer. Realworld forewarning requires analyst-independent, prospective analysis of real-time data on a portable device. (6) The results give forewarning times of 4 hours or less. A time-to-event estimate is needed. (7) All EEG involved temporal lobe epilepsy; other kinds of epilepsy need to be included. (8) A prospective analysis of long-term continuous data is the acid test for any predictive approach. Prospective data were unavailable for the present analysis. Clearly, much

Nonlinear Epilepsy Forewarning by Support Vector Machines

http://dx.doi.org/10.5772/57438

49

The present work uses Support Vector Machine analysis to extend earlier work by Hively *et al.* [4] for forewarning of epileptic seizures. The previous work obtained a prediction distance of 0.0559 and a maximum forewarning time of 5.1 hours. The present analysis divides the continuous data stream from one bipolar channel of scalp EEG into contiguous, non-overlap‐ ping windows; removes the muscular artifacts with a novel zero-phase quadratic filter; converts the artifact-filtered data into discrete symbols; applies the time-delay-embedding (Takens') theorem to create unique phase-space states that capture the brain dynamics; forms a graph from the nodes (phase-space states) and links (dynamical state-to-state transitions); extracts dissimilarity measures by pair-wise comparison of graphs (e.g., nodes in graph A that are not in B); uses these dissimilarity measures as features for a novel Support Vector Machine to classify the data as forewarning of a seizure event or not (i.e., not characteristic of the baseline, but characteristic of data near the event). The present work obtains a prediction distance as small as zero (sensitivity =1, specificity =1) with most of the forewarning times ≤1.5

hours. The best off-training-set error rate was.287 using 10-fold cross validation.

Our non-invasive (scalp) EEG analysis resulted in cross validation error rates comparable to other invasive EEG approaches. Additional accuracy could be obtained by applying this methodology to specific patients on a per patient basis for custom EEG models if the data were available. Modifications could conceivably be made to the algorithm to improve the compu‐ tational feasibility of per patient machine learning models. A research team could allow patients to start with group-based models that are less accurate while the patients collect and upload ambulatory data from their devices as they use them in real-world settings. Further‐ more, businesses could be compensated for creating patient specific models from patients' ambulatory data. Additionally, our algorithms have other applications as well, such as failure

This manuscript has been authored by UT-Battelle, LLC, under contract DE-AC05-00OR22725 with the U.S. Department of Energy. The United States Government retains and the publisher, by accepting the article for publication, acknowledges that the United States Government

work remains to address these issues.

forewarning in machines [21] and bridges [22].

**Acknowledgements**

**8. Conclusions**

#### **7. Discussion**

Ideally, we would like to achieve an average cross validation OTS prediction distance of zero, final model prediction distance of zero, and all forewarning times <1 hour. In order to achieve this goal, additional features will need to be explored that exploit topology and the distance metric that Takens' theorem guarantees. Some modifications for Stage-2 improve the results (e.g., the choice of p cutsets prior to the event as a variable, and use of a RBF kernel instead of a linear one). More search parameters in Stage-1 (e.g., those in Table 3) should lead to better results with enough CPU time. Additional graph dissimilarity measures may be helpful. We have discovered more features for Stage-1 that may be of use. More data is needed for a robust statistical validation of the model.

The choice of optimal features is very difficult. Theorems guide the choice of parameters, features, and the algorithm. Some combination of theorem-based feature selection and occasional intuition derived from experimentation is the only way to keep the cost of the research initiative practical. Feature selection is one of many hard problems involved in epilepsy prediction. When one adds features, one often needs more data to make meaningful statistical assertions. Other important choices involve the type of kernel and the thresholding strategy. Linear kernels and thresholding strategies may perform well while Radial Basis Function (RBF) kernels perform poorly and vice versa. Our previous work [4] used a voting method that performed well. There is no guarantee that a set of features will behave similarly with different kernels and strategies to determine the threshold. Other measurement functions are possible under Takens' theorem to create the phase-space states. Use of a single-class or multi-class SVM could also prove fruitful.

The results in Tables 6-7 and Figure 11 are encouraging, despite several limitations, which are discussed next. (1) We analyzed 60 datasets, 40 with epileptic events and 20 without events. Much more data (hundreds of datasets) are needed for strong statistical validation. (2) These data are from controlled clinical settings, rather than an uncontrolled (real-world) environ‐ ment. (3) The results depend on careful adjustment of training parameters. (4) Only physicianselected portions of the EEG are available, rather than the full monitoring period. (5) The present approach uses retrospective analysis of archival data on a desktop computer. Realworld forewarning requires analyst-independent, prospective analysis of real-time data on a portable device. (6) The results give forewarning times of 4 hours or less. A time-to-event estimate is needed. (7) All EEG involved temporal lobe epilepsy; other kinds of epilepsy need to be included. (8) A prospective analysis of long-term continuous data is the acid test for any predictive approach. Prospective data were unavailable for the present analysis. Clearly, much work remains to address these issues.

#### **8. Conclusions**

In Figure 11, the solid black line is the occurrence frequency (arbitrary units) in half-hour bins. The blue line is the cumulative distribution of forewarning versus time. The red H-bar with the star in the middle indicates the mean value of the forewarning times (approximately 1 hour) and the sample standard deviation. The result in Example 7 of Table 6 is better than random guessing or biased heuristics with D(final)=0, despite poorer cross validation accuracy than other examples. This example shows most of the forewarning times of ≤1.5 hours with a statistically significant accuracy. One can visually make a prediction by scanning Table 7 from left to right and looking for 2 contiguous plus values. When 2 values are found, the seizure is highly likely to occur. One may be tempted to reduce the forewarn time by increasing the value of positive values that trigger a forewarning, but that would likely result in bad OTS error,

Ideally, we would like to achieve an average cross validation OTS prediction distance of zero, final model prediction distance of zero, and all forewarning times <1 hour. In order to achieve this goal, additional features will need to be explored that exploit topology and the distance metric that Takens' theorem guarantees. Some modifications for Stage-2 improve the results (e.g., the choice of p cutsets prior to the event as a variable, and use of a RBF kernel instead of a linear one). More search parameters in Stage-1 (e.g., those in Table 3) should lead to better results with enough CPU time. Additional graph dissimilarity measures may be helpful. We have discovered more features for Stage-1 that may be of use. More data is needed for a robust

The choice of optimal features is very difficult. Theorems guide the choice of parameters, features, and the algorithm. Some combination of theorem-based feature selection and occasional intuition derived from experimentation is the only way to keep the cost of the research initiative practical. Feature selection is one of many hard problems involved in epilepsy prediction. When one adds features, one often needs more data to make meaningful statistical assertions. Other important choices involve the type of kernel and the thresholding strategy. Linear kernels and thresholding strategies may perform well while Radial Basis Function (RBF) kernels perform poorly and vice versa. Our previous work [4] used a voting method that performed well. There is no guarantee that a set of features will behave similarly with different kernels and strategies to determine the threshold. Other measurement functions are possible under Takens' theorem to create the phase-space states. Use of a single-class or

The results in Tables 6-7 and Figure 11 are encouraging, despite several limitations, which are discussed next. (1) We analyzed 60 datasets, 40 with epileptic events and 20 without events. Much more data (hundreds of datasets) are needed for strong statistical validation. (2) These data are from controlled clinical settings, rather than an uncontrolled (real-world) environ‐ ment. (3) The results depend on careful adjustment of training parameters. (4) Only physicianselected portions of the EEG are available, rather than the full monitoring period. (5) The

which is why it is not the value found by the second and third stage SVMs.

**7. Discussion**

48 Epilepsy Topics

statistical validation of the model.

multi-class SVM could also prove fruitful.

The present work uses Support Vector Machine analysis to extend earlier work by Hively *et al.* [4] for forewarning of epileptic seizures. The previous work obtained a prediction distance of 0.0559 and a maximum forewarning time of 5.1 hours. The present analysis divides the continuous data stream from one bipolar channel of scalp EEG into contiguous, non-overlap‐ ping windows; removes the muscular artifacts with a novel zero-phase quadratic filter; converts the artifact-filtered data into discrete symbols; applies the time-delay-embedding (Takens') theorem to create unique phase-space states that capture the brain dynamics; forms a graph from the nodes (phase-space states) and links (dynamical state-to-state transitions); extracts dissimilarity measures by pair-wise comparison of graphs (e.g., nodes in graph A that are not in B); uses these dissimilarity measures as features for a novel Support Vector Machine to classify the data as forewarning of a seizure event or not (i.e., not characteristic of the baseline, but characteristic of data near the event). The present work obtains a prediction distance as small as zero (sensitivity =1, specificity =1) with most of the forewarning times ≤1.5 hours. The best off-training-set error rate was.287 using 10-fold cross validation.

Our non-invasive (scalp) EEG analysis resulted in cross validation error rates comparable to other invasive EEG approaches. Additional accuracy could be obtained by applying this methodology to specific patients on a per patient basis for custom EEG models if the data were available. Modifications could conceivably be made to the algorithm to improve the compu‐ tational feasibility of per patient machine learning models. A research team could allow patients to start with group-based models that are less accurate while the patients collect and upload ambulatory data from their devices as they use them in real-world settings. Further‐ more, businesses could be compensated for creating patient specific models from patients' ambulatory data. Additionally, our algorithms have other applications as well, such as failure forewarning in machines [21] and bridges [22].

#### **Acknowledgements**

This manuscript has been authored by UT-Battelle, LLC, under contract DE-AC05-00OR22725 with the U.S. Department of Energy. The United States Government retains and the publisher, by accepting the article for publication, acknowledges that the United States Government retains a non-exclusive, paid-up, irrevocable, world-wide license to publish or reproduce the published form of this manuscript, or allow others to do so, for United States Government purposes.

[9] Jenkins JS. "The Mozart effect," *J. Royal Soc. Med.* (2001); 94:170-172

from scalp EEG," *IEEE Trans Biomed Eng*. (2003);50(5):584–93.

*TM-12961*, Oak Ridge National Laboratory, Oak Ridge, TN (1995).

National Laboratory, Oak Ridge, TN (2009).

EEG," *J Clin Neurophysiol*. (2001);18(3):223–45.

[16] Bondy JA, Murty USR, *Graph Theory*, Springer (2008).

relation," *Nuclear Physics B* (2007);782(3):219 – 240.

ure prediction," *Epilepsy Behav*. (2010);19(3):474–7.

*ORNL/TM-2011/260* (Oak Ridge National Laboratory).

*Intell. Syst. Technol.*( 2011);2(3):27:1–27.

delberg; 1981. p. 366–81.

10 Oct. 2013.

649-688.

2013.

[10] "10-20 System (EEG)."*Wikipedia*. Wikimedia Foundation, 22 July 2013. Web. 16 Aug.

Nonlinear Epilepsy Forewarning by Support Vector Machines

http://dx.doi.org/10.5772/57438

51

[11] Hively LM, Protopopescu VA, "Channel-consistent forewarning of epileptic events

[12] Hively LM, "Prognostication of Helicopter Failure," *ORNL/TM-2009-244*, Oak Ridge

[13] Protopopescu VA, Hively LM, Gailey PC, "Epileptic event forewarning from scalp

[14] Hively LM, *et al.,* "Nonlinear Analysis of EEG for Epileptic Seizures," *ORNL/*

[15] Takens F, "Detecting strange attractors in turbulence," In: Rand D, Young L-S, edi‐ tors. *Dynamical Systems and Turbulence*, Warwick 1980 [Internet]. Springer Berlin Hei‐

[17] Franzosi R, Pettini M, "Topology and phase transitions II. Theorem on a necessary

[18] Ng A. "Machine Learning: SVM Kernels." Coursera. Stanford University, n.d. Web.

[19] Chang C-C, Lin C-J, "LIBSVM: A library for support vector machines" *ACM Trans.*

[20] Arthurs S, Zaveri HP, Frei MG, Osorio I, "Patient and caregiver perspectives on seiz‐

[21] Protopopescu V, Hively LM, "Phase-space dissimilarity measures of nonlinear dy‐ namics: Industrial and biomedical applications," *Recent Res. Dev. Physics* (2005);6,

[22] Bubacz JA, Chmielewski HT, Pape AE, Depersio AJ, Hively LM, Abercrombie RK, "Phase Space Dissimilarity Measures for Structural Health Monitoring," (2011)

#### **Author details**

W.S. Ashbee1 , L.M. Hively2 and J.T. McDonald3

1 School of Computing at University of South Alabama, USA

2 Computational Sciences and Engineering Division at Oak Ridge National Lab, USA

3 School of Computing at University of South Alabama, USA

#### **References**


[9] Jenkins JS. "The Mozart effect," *J. Royal Soc. Med.* (2001); 94:170-172

retains a non-exclusive, paid-up, irrevocable, world-wide license to publish or reproduce the published form of this manuscript, or allow others to do so, for United States Government

and J.T. McDonald3

2 Computational Sciences and Engineering Division at Oak Ridge National Lab, USA

[1] Netoff T, Park Y, Parhi K. Seizure prediction using cost-sensitive support vector ma‐ chine. Engineering in Medicine and Biology Society (2009) EMBC 2009 Annual Inter‐

[2] Stacey W, Le Van Quyen M, Mormann F, Schulze-Bonhage A, "What is the present-

[3] Mirowski PW, LeCun Y, Madhavan D, Kuzniecky R. Comparing SVM and convolu‐ tional networks for epileptic seizure prediction from intracranial EEG. Machine

[4] Hively LM, McDonald JT, Munro NB, Cornelius E, "Forewarning of Epileptic Events from Scalp EEG," peer-reviewed proceedings paper for *Biomedical Science and Engi‐*

[5] Hively LM, Protopopescu VA, Munro NB, "Enhancements in epilepsy forewarning

[6] Percha B, Dzakpasu R, Żochowski M, and Parent J, "Transition from local to global phase synchrony in small world neural network and its possible implications for epi‐

[7] Van den Broeck C, Parrondo JMR, and Toral R, "Noise-induced nonequilibrium

[8] Pittau F, Tinuper P, Bisulli F, Naldi I, Cortelli P, Bisulli A, et al., "Videopolygraphic and functional {MRI} study of musicogenic epilepsy. A case report and literature re‐

day EEG evidence for a preictal state?" *Epilepsy Res.* (2011);97(3):243–51.

Learning for Signal Processing (2008) IEEE Workshop on MLSP, pp. 244–9.

via phase-space dissimilarity," *J Clin Neurophysiol*. (2005);22(6):402–9.

1 School of Computing at University of South Alabama, USA

3 School of Computing at University of South Alabama, USA

national Conference of the IEEE, pp. 3322–5.

*neering Conference at ORNL* (May 2013).

lepsy," *Phys. Rev. E* (2005);72, paper #031909

phase transition," *Phys. Rev. Lett.* (1994);73, 3395-3398

view," *Epilepsy & Behavior* (2008);13(4):685 – 692.

purposes.

50 Epilepsy Topics

**Author details**

, L.M. Hively2

W.S. Ashbee1

**References**


**Chapter 4**

**Juvenile Myoclonic Epilepsy: An Update**

We review the most important electroclinical aspects and possible subsyndromes of Juvenile myoclonic epilepsy (JME), as well as its genetic background, its pathophysiological and neuroimaging correlates, and treatment. JME is among the most common types of genetic epilepsies. The prevalence of JME in large cohorts has been estimated to be 5% to 10% of all epilepsies and around 18% of idiopathic generalized epilepsies but may be lower in some settings. There is a marked female predominance. Today JME is a widely recognized electro‐ clinical idiopathic generalized epilepsy syndrome. Onset is around the time of puberty. The most typical ictal phenomenon is bilateral myoclonia without loss of consciousness. Most patients also present with generalized tonic-clonic seizures (GTCS), and some with absence seizures. The typical circumstance at diagnosis is a first GTCS episode, after the patient has had myoclonia in the morning. Typically seizure episodes occur after awakening from a sleep period or in the evening relaxation period and are facilitated by sleep deprivation and sudden arousal. Diagnosis of JME can be made with the history of myoclonus plus a single GTCS plus generalized polyspike-waves or fast spike-waves on the EEG. The prevalence rate of photo‐ sensitivity (photoparoxysmal EEG response) in patients with JME ranges from 8 to 90%. Hyperventilation can induce absence seizures in patients with JME, while cognitive tasks are efficient in precipitating myoclonic seizures. Most patients have a good prognosis when treated with appropriate drugs, but with a well-known tendency to relapse after withdrawal. However, around 17% are able to discontinue medication and remain seizure-free thereafter. There is a small but still considerable subgroup of JME patients whose seizures are difficult to treat. Recent findings suggest that patients with JME have worse social adjustment in relevant aspects of their lives, works and familiar relationship. Differential diagnoses include the adolescent-onset progressive myoclonus epilepsies, or other forms of idiopathic generalized

> © 2014 The Author(s). Licensee InTech. This chapter is distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/3.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Boulenouar Mesraoua, Dirk Deleu, Hassan Al Hail,

Gayane Melikyan and Heinz Gregor Wieser

Additional information is available at the end of the chapter

http://dx.doi.org/10.5772/57431

**1. Introduction**

epilepsies of adolescence.

## **Juvenile Myoclonic Epilepsy: An Update**

Boulenouar Mesraoua, Dirk Deleu, Hassan Al Hail, Gayane Melikyan and Heinz Gregor Wieser

Additional information is available at the end of the chapter

http://dx.doi.org/10.5772/57431

#### **1. Introduction**

We review the most important electroclinical aspects and possible subsyndromes of Juvenile myoclonic epilepsy (JME), as well as its genetic background, its pathophysiological and neuroimaging correlates, and treatment. JME is among the most common types of genetic epilepsies. The prevalence of JME in large cohorts has been estimated to be 5% to 10% of all epilepsies and around 18% of idiopathic generalized epilepsies but may be lower in some settings. There is a marked female predominance. Today JME is a widely recognized electro‐ clinical idiopathic generalized epilepsy syndrome. Onset is around the time of puberty. The most typical ictal phenomenon is bilateral myoclonia without loss of consciousness. Most patients also present with generalized tonic-clonic seizures (GTCS), and some with absence seizures. The typical circumstance at diagnosis is a first GTCS episode, after the patient has had myoclonia in the morning. Typically seizure episodes occur after awakening from a sleep period or in the evening relaxation period and are facilitated by sleep deprivation and sudden arousal. Diagnosis of JME can be made with the history of myoclonus plus a single GTCS plus generalized polyspike-waves or fast spike-waves on the EEG. The prevalence rate of photo‐ sensitivity (photoparoxysmal EEG response) in patients with JME ranges from 8 to 90%. Hyperventilation can induce absence seizures in patients with JME, while cognitive tasks are efficient in precipitating myoclonic seizures. Most patients have a good prognosis when treated with appropriate drugs, but with a well-known tendency to relapse after withdrawal. However, around 17% are able to discontinue medication and remain seizure-free thereafter. There is a small but still considerable subgroup of JME patients whose seizures are difficult to treat. Recent findings suggest that patients with JME have worse social adjustment in relevant aspects of their lives, works and familiar relationship. Differential diagnoses include the adolescent-onset progressive myoclonus epilepsies, or other forms of idiopathic generalized epilepsies of adolescence.

© 2014 The Author(s). Licensee InTech. This chapter is distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/3.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

### **2. History of juvenile myoclonic epilepsy**

Juvenile Myoclonic Epilepsy (JME) has been recognized by early distinguished physicians as Theodore Herpin in 1867 [1] and Robot in 1899 [2]. However, it was not until 1957 that Janz and Christian gave the first and precise description of JME in 47 German patients [3]. Later on, Castells and Mendilaharsu described JME in 70 Uruguayans patients [4]. Following that, Delgado-Escueta and Enrile-Bacsal reported 43 cases of uncontrolled convulsive seizures because the syndrome of JME was not recognized in those cases. [5]. Since then JME was reported in different ethnic groups around the world (Asia, Europe, North and Latin America, Oceania and Africa).

of seizures occur between 12 and 18 years [17, 18]. Absence seizures (AS) are reported in 33.3% to 66.7%, myoclonic jerks (MJ) in 97%, and GTCS in 78.8% of the patients [14, 19]. In one study, AS antedated other types of seizures in all patients [20]. An earlier or later age of onset has also been reported [20-23]. Female preponderance for photosensitivity may explain an early

Juvenile Myoclonic Epilepsy: An Update http://dx.doi.org/10.5772/57431 55

**Figure 1.** Age of onset of 139 patients with myoclonic seizures (i.e., beginning cases of juvenile myoclonic epilepsy).

JME is probably the most common and the most characteristic form of the IGE of adolescent-

MJ occur prominently and spontaneously in the morning, after awakening; they are sudden, short-lasting, irregular, frequently symmetric; they may be self- limited (isolated) or may occur in clusters; if prolonged, these clusters may lead to a convulsive tonic-clonic seizure. MJ involve prominently the upper limbs; however the lower limbs and trunk are often not spared. Both distal and proximal jerky movements do occur; flexion of both forearms, flexion of both arms, flexion and abduction of the thighs and extension of the back, are typical. Patients may throw things out of their hands when the MJ involve or are restricted to the fingers; occasionally, the jerks are so intense that the patient falls to the ground (myoclonic-astatic seizures). Several authors reported some degree of asymmetry in MJ as well as focal features in the EEG, leading to the false diagnosis of focal epilepsy [26-28]. This issue will be discussed in the paragraph

onset of JME in female patients with photosensitivity [24, 25].

From Janz D et al. In: Epilepsy, J. Engel Jr and T A Pedley (ed), 1998. p.2391

**6. Clinical presentation**

**6.1. Myoclonic Jerks (MJ)**

below (*Electroencephalography*)

onset group. It is characterized by:

#### **3. Classification and definition of JME**

The ILAE Commission proposed and defined JME as a distinct syndrome of IGE: "Juvenile myoclonic epilepsy (impulsive petit mal) appears around puberty and is characterized by seizures with bilateral, single or repetitive, arrhythmic, irregular myoclonic jerks predomi‐ nantly in the arms. Jerks may cause some patients to fall suddenly. No disturbance of con‐ sciousness is noticeable. The disorder may be inherited, and sex distribution is equal (but see Introduction!). Often, there are GTCS and less often infrequent absences. The seizures usually occur shortly after awakening and are often precipitated by sleep deprivation. Interictal and ictal EEG have rapid, generalized, often irregular spike-waves and polyspike-waves; there is no close phase correlation between EEG spikes and jerks. Frequently, the patients are photo‐ sensitive. Response to appropriate drugs is good" [6].

#### **4. Incidence, prevalence, and sex ratio of JME**

JME is the most common form of genetic/idiopathic generalized epilepsy (IGE): The IGEs comprises 40% of epilepsies in the US, 20% in Mexico, 8% in Central America [7] and is much higher (between 68–82.6%) in the Arab countries [8]. JME is responsible for 6 to 12 % up to 30% of all epilepsies in hospital and clinics records [3, 9-11] and for 3% according to a door-to-door population survey [12]. The incidence of JME varies from 0.5 to 6.3%/100.000. Based on 1% of population risk for epilepsy by age 20 [13], the risk of JME in the general population would be 1 per 1000 to 2000. The prevalence of JME has been estimated to be 5% to 10% of all epilepsies and 18% of IGEs [14, 15]. Gender differences are evident for JME with a marked female predominance [15, 16].

#### **5. Age of onset**

As shown by Janz [17], one of its first investigators, JME has an age-related onset [Figure 1]. The seizures appear between 8 and 26 years with a mean age of onset at 14.2 years; the majority of seizures occur between 12 and 18 years [17, 18]. Absence seizures (AS) are reported in 33.3% to 66.7%, myoclonic jerks (MJ) in 97%, and GTCS in 78.8% of the patients [14, 19]. In one study, AS antedated other types of seizures in all patients [20]. An earlier or later age of onset has also been reported [20-23]. Female preponderance for photosensitivity may explain an early onset of JME in female patients with photosensitivity [24, 25].

**Figure 1.** Age of onset of 139 patients with myoclonic seizures (i.e., beginning cases of juvenile myoclonic epilepsy). From Janz D et al. In: Epilepsy, J. Engel Jr and T A Pedley (ed), 1998. p.2391

#### **6. Clinical presentation**

**2. History of juvenile myoclonic epilepsy**

**3. Classification and definition of JME**

sensitive. Response to appropriate drugs is good" [6].

**4. Incidence, prevalence, and sex ratio of JME**

Oceania and Africa).

54 Epilepsy Topics

predominance [15, 16].

**5. Age of onset**

Juvenile Myoclonic Epilepsy (JME) has been recognized by early distinguished physicians as Theodore Herpin in 1867 [1] and Robot in 1899 [2]. However, it was not until 1957 that Janz and Christian gave the first and precise description of JME in 47 German patients [3]. Later on, Castells and Mendilaharsu described JME in 70 Uruguayans patients [4]. Following that, Delgado-Escueta and Enrile-Bacsal reported 43 cases of uncontrolled convulsive seizures because the syndrome of JME was not recognized in those cases. [5]. Since then JME was reported in different ethnic groups around the world (Asia, Europe, North and Latin America,

The ILAE Commission proposed and defined JME as a distinct syndrome of IGE: "Juvenile myoclonic epilepsy (impulsive petit mal) appears around puberty and is characterized by seizures with bilateral, single or repetitive, arrhythmic, irregular myoclonic jerks predomi‐ nantly in the arms. Jerks may cause some patients to fall suddenly. No disturbance of con‐ sciousness is noticeable. The disorder may be inherited, and sex distribution is equal (but see Introduction!). Often, there are GTCS and less often infrequent absences. The seizures usually occur shortly after awakening and are often precipitated by sleep deprivation. Interictal and ictal EEG have rapid, generalized, often irregular spike-waves and polyspike-waves; there is no close phase correlation between EEG spikes and jerks. Frequently, the patients are photo‐

JME is the most common form of genetic/idiopathic generalized epilepsy (IGE): The IGEs comprises 40% of epilepsies in the US, 20% in Mexico, 8% in Central America [7] and is much higher (between 68–82.6%) in the Arab countries [8]. JME is responsible for 6 to 12 % up to 30% of all epilepsies in hospital and clinics records [3, 9-11] and for 3% according to a door-to-door population survey [12]. The incidence of JME varies from 0.5 to 6.3%/100.000. Based on 1% of population risk for epilepsy by age 20 [13], the risk of JME in the general population would be 1 per 1000 to 2000. The prevalence of JME has been estimated to be 5% to 10% of all epilepsies and 18% of IGEs [14, 15]. Gender differences are evident for JME with a marked female

As shown by Janz [17], one of its first investigators, JME has an age-related onset [Figure 1]. The seizures appear between 8 and 26 years with a mean age of onset at 14.2 years; the majority JME is probably the most common and the most characteristic form of the IGE of adolescentonset group. It is characterized by:

#### **6.1. Myoclonic Jerks (MJ)**

MJ occur prominently and spontaneously in the morning, after awakening; they are sudden, short-lasting, irregular, frequently symmetric; they may be self- limited (isolated) or may occur in clusters; if prolonged, these clusters may lead to a convulsive tonic-clonic seizure. MJ involve prominently the upper limbs; however the lower limbs and trunk are often not spared. Both distal and proximal jerky movements do occur; flexion of both forearms, flexion of both arms, flexion and abduction of the thighs and extension of the back, are typical. Patients may throw things out of their hands when the MJ involve or are restricted to the fingers; occasionally, the jerks are so intense that the patient falls to the ground (myoclonic-astatic seizures). Several authors reported some degree of asymmetry in MJ as well as focal features in the EEG, leading to the false diagnosis of focal epilepsy [26-28]. This issue will be discussed in the paragraph below (*Electroencephalography*)

Myoclonic status epilepticus (MSE) is not so rare. In MSE, consciousness may be intact. Drug withdrawal, sleep deprivation and alcohol intake are the main causes [29, 30].

The background activity is usually normal; some authors report theta slowing during poor seizure control [37], others found an increase in *Absolute Power* of delta, alpha and beta bands,

Juvenile Myoclonic Epilepsy: An Update http://dx.doi.org/10.5772/57431 57

Interictal EEG shows diffuse or generalized spike-wave (SW) and polyspike-wave (PSW)

**Figure 2.** Thirteen year-old girl with clinical JME since the age of 9 years showing interictal polyspike wave discharges

Localization related EEG abnormalities are found in 16.9-57.1 % of patients [14, 27, 28]. These focal abnormalities include unilateral discharges, paroxysms with unilateral onset, and frequently discharges with above 50% voltage asymmetries. In general these EEG changes are

Photosensitivity (photoparoxysmal EEG response), with or without MJ, in patients with JME

As mentioned above, in patients with JME, absence seizures can be induced by hyperventila‐

varies from 8 to 90% and is more frequent in females and adolescents [35] [Figure3]

predominantly seen at sleep onset and after provoked awakening [26].

while cognitive tasks can precipitate myoclonic seizures.

more evident in frontoparietal regions in patients with JME [38].

**7.1. Interictal EEG**

discharges at 3-6 Hz [Figure 2].

not associated with clinical manifestations.

tion [Figure 4],

#### **6.2. Generalized Tonic Clonic Seizures (GTCS)**

Frequently, in the outpatient clinic or emergency unit, a young patient is examined because he/she was a victim of a generalized seizure upon awakening. Often, when asked, the family reports that the GTCS followed repeated, severe MJ (generalized clonic-tonic-clonic seizure type) [5].

#### **6.3. Absence Seizures (AS)**

As indicated above, AS are generally reported in one-third of patients with JME [31, 32]. However, their frequency might be much higher (66.7%) [19]. There is general agreement that AS associated with JME are mild and short, when compared to the childhood absences and absences of *Juvenile Absence Epilepsy.* They are less severe with age and are often unnoticed by the patient [33].

In a recent prospective study, with long-term follow up of 257 patients with JME, Martinez-Juarez IE and coworkers encountered four JME groups: (1) Classic JME (72%), (2) Childhood Absence Epilepsy (CAE) evolving to JME (18%), (3) JME with Adolescent Absence (7%), and (4) JME with Astatic Seizures (3%). There was a female preponderance in the second group (CAE evolving to JME); the authors concluded that all 4 subtypes are chronic and probably lifelong [34].

#### **6.4. Precipitation of seizures**

As reported above, occurrence of MJ in the early morning is one of the hallmarks of JME. MJ and GTCS are induced by sleep deprivation, fatigue and excessive alcohol intake [35]. Sleep deprivation is understood as falling asleep late at night and getting up or awaken early in the morning (short sleep).

Seizure-provoking factors in JME are numerous; among them are: stress, fatigue, fever, sleep, flashing sunlight, music, reading, thinking, and excessive alcohol intake. In JME, photosensi‐ tivity (photoparoxysmal EEG response), is age-related and it varies considerably [35]. How‐ ever, only a small number of patients experiences seizures by photic stimulation in daily life. In patients with JME, absence seizures are induced by hyperventilation, while myoclonic seizures are provoked by cognitive tasks [35, 36].

#### **7. Electroencephalography**

JME is widely underdiagnosed despite a characteristic clinical picture and a distinct EEG profile. If a patient, who is suspected clinically to suffer from JME, has a normal EEG, a sleep EEG and an EEG on awakening should follow.

The background activity is usually normal; some authors report theta slowing during poor seizure control [37], others found an increase in *Absolute Power* of delta, alpha and beta bands, more evident in frontoparietal regions in patients with JME [38].

#### **7.1. Interictal EEG**

Myoclonic status epilepticus (MSE) is not so rare. In MSE, consciousness may be intact. Drug

Frequently, in the outpatient clinic or emergency unit, a young patient is examined because he/she was a victim of a generalized seizure upon awakening. Often, when asked, the family reports that the GTCS followed repeated, severe MJ (generalized clonic-tonic-clonic seizure

As indicated above, AS are generally reported in one-third of patients with JME [31, 32]. However, their frequency might be much higher (66.7%) [19]. There is general agreement that AS associated with JME are mild and short, when compared to the childhood absences and absences of *Juvenile Absence Epilepsy.* They are less severe with age and are often unnoticed by

In a recent prospective study, with long-term follow up of 257 patients with JME, Martinez-Juarez IE and coworkers encountered four JME groups: (1) Classic JME (72%), (2) Childhood Absence Epilepsy (CAE) evolving to JME (18%), (3) JME with Adolescent Absence (7%), and (4) JME with Astatic Seizures (3%). There was a female preponderance in the second group (CAE evolving to JME); the authors concluded that all 4 subtypes are chronic and probably

As reported above, occurrence of MJ in the early morning is one of the hallmarks of JME. MJ and GTCS are induced by sleep deprivation, fatigue and excessive alcohol intake [35]. Sleep deprivation is understood as falling asleep late at night and getting up or awaken early in the

Seizure-provoking factors in JME are numerous; among them are: stress, fatigue, fever, sleep, flashing sunlight, music, reading, thinking, and excessive alcohol intake. In JME, photosensi‐ tivity (photoparoxysmal EEG response), is age-related and it varies considerably [35]. How‐ ever, only a small number of patients experiences seizures by photic stimulation in daily life. In patients with JME, absence seizures are induced by hyperventilation, while myoclonic

JME is widely underdiagnosed despite a characteristic clinical picture and a distinct EEG profile. If a patient, who is suspected clinically to suffer from JME, has a normal EEG, a sleep

withdrawal, sleep deprivation and alcohol intake are the main causes [29, 30].

**6.2. Generalized Tonic Clonic Seizures (GTCS)**

type) [5].

56 Epilepsy Topics

the patient [33].

lifelong [34].

**6.4. Precipitation of seizures**

seizures are provoked by cognitive tasks [35, 36].

EEG and an EEG on awakening should follow.

**7. Electroencephalography**

morning (short sleep).

**6.3. Absence Seizures (AS)**

Interictal EEG shows diffuse or generalized spike-wave (SW) and polyspike-wave (PSW) discharges at 3-6 Hz [Figure 2].

**Figure 2.** Thirteen year-old girl with clinical JME since the age of 9 years showing interictal polyspike wave discharges not associated with clinical manifestations.

Localization related EEG abnormalities are found in 16.9-57.1 % of patients [14, 27, 28]. These focal abnormalities include unilateral discharges, paroxysms with unilateral onset, and frequently discharges with above 50% voltage asymmetries. In general these EEG changes are predominantly seen at sleep onset and after provoked awakening [26].

Photosensitivity (photoparoxysmal EEG response), with or without MJ, in patients with JME varies from 8 to 90% and is more frequent in females and adolescents [35] [Figure3]

As mentioned above, in patients with JME, absence seizures can be induced by hyperventila‐ tion [Figure 4],

while cognitive tasks can precipitate myoclonic seizures.

**7.2. Ictal EEG**

of brief polyspike wave sequences.

**8. Genetics of JME**

1-4 s.

poorly structured spike-wave sequence [Figure 5].

The characteristic ictal EEG manifestations of a MJ are a generalized burst of multiple spikes of short duration (0.5-2s). Frequently, however, the spikes are followed by slow waves with

Juvenile Myoclonic Epilepsy: An Update http://dx.doi.org/10.5772/57431 59

**Figure 5.** Seventeen year-old male with JME suffering from mild MJ of the upper extremities accompanied by bursts

Ictal EEG discharges of absences consist of multiple spikes usually preceding or superim‐ posed on a slow wave. These discharges often show a characteristic fragmentation and last from

JME is the most common cause of hereditary grand mal seizures in people with epilepsy in the population at large [5, 17]. It has both Mendelian inheritance and complex genetic inheri‐ tance [12, 39]. 49% of JME families have clinical and EEG traits suggesting an autosomal dominant inherited disease. Variants of JME genes, with small to modest effects, contribute to risk/susceptibility in the remaining 51% [5, 12, 39]. Linkage disequilibrium is understood as the occurrence, in a specific population, of both DNA markers (DNA microsatellites or SNPs, single nucleotide polymorphisms) and a JME mutation at a higher frequency than would be predicted by random chance. With the passage of time, linkage disequilibrium decays through recombinations and transmissions into thousands of generations resulting in the fact that the

**Figure 3.** Photoparoxysmal response in a 20 year old female patient with JME since the age of 11 years accompanied by jerky movements of both upper extremities.

**Figure 4.** Same patient as in Figure 3 showing generalized polyspike-wave complexes accompanied by mild impair‐ ment of cognition during hyperventilation.

#### **7.2. Ictal EEG**

The characteristic ictal EEG manifestations of a MJ are a generalized burst of multiple spikes of short duration (0.5-2s). Frequently, however, the spikes are followed by slow waves with poorly structured spike-wave sequence [Figure 5].

**Figure 5.** Seventeen year-old male with JME suffering from mild MJ of the upper extremities accompanied by bursts of brief polyspike wave sequences.

Ictal EEG discharges of absences consist of multiple spikes usually preceding or superim‐ posed on a slow wave. These discharges often show a characteristic fragmentation and last from 1-4 s.

#### **8. Genetics of JME**

**Figure 4.** Same patient as in Figure 3 showing generalized polyspike-wave complexes accompanied by mild impair‐

**Figure 3.** Photoparoxysmal response in a 20 year old female patient with JME since the age of 11 years accompanied

ment of cognition during hyperventilation.

by jerky movements of both upper extremities.

58 Epilepsy Topics

JME is the most common cause of hereditary grand mal seizures in people with epilepsy in the population at large [5, 17]. It has both Mendelian inheritance and complex genetic inheri‐ tance [12, 39]. 49% of JME families have clinical and EEG traits suggesting an autosomal dominant inherited disease. Variants of JME genes, with small to modest effects, contribute to risk/susceptibility in the remaining 51% [5, 12, 39]. Linkage disequilibrium is understood as the occurrence, in a specific population, of both DNA markers (DNA microsatellites or SNPs, single nucleotide polymorphisms) and a JME mutation at a higher frequency than would be predicted by random chance. With the passage of time, linkage disequilibrium decays through recombinations and transmissions into thousands of generations resulting in the fact that the epilepsy allele will have smaller and smaller genetic effects and will require other epilepsy alleles or environment to produce the epilepsy phenotype [40]. However, linkage disequili‐ brium is strongest and covers the widest region of a chromosome when the epilepsy allele is of recent origin, and has large genetic effects, e.g., Mendelian dominant or recessive effects.

5 Mendelian JME genes have been reported (http://omim.org and http:// www.ncbi.nlm.nih.gov/omim/). These are as follows: CACNB4 (calcium channel beta4 subunit) [41], CASR (calcium channel sensory receptor) [42], GABRA1 (GABA receptor alpha one subunit) [43], GABRD (GABA receptor delta subunit) [44], and Myoclonin1/EFHC1

Juvenile Myoclonic Epilepsy: An Update http://dx.doi.org/10.5772/57431 61

Also, three SNP susceptibility alleles of putative JME genes that contribute to the complex genetics of JME have been reported [12, 39, 46]: bromodomain-containing 2 (BRD2) [47], connexin 36 (Cx-36) [48], and malic enzyme2 (ME2) [49]. Additionally, more than 22 chromo‐

Familial segregation or association with disease identifies putative JME disease genes. In autosomal dominant JME, a candidate epilepsy gene should show at least one variant per affected individual; each candidate epilepsy gene should show homozygous mutations or compound heterozygous mutations in autosomal recessive JME [50-52]. Contributions of de novo mutations in the epilepsies have been demonstrated through studies of copy number variations (CNVs) which in fact contribute to genetic generalized epilepsies with complex inheritance, including JME. Consequently, pathogenic de novo mutations could be identified

It has been shown that mutations in Cchb4, the mouse homologue of human CACNB4 or mutations in GABRA1 are sufficient to produce the absence phenotype while mutations in Myoclonin1/EFHC1 or BRD2 are sufficient to produce a convulsive phenotype (myoclonic or

The identification of epilepsy alleles that cause JME could lead to new AED discoveries, to

There is an increasing interest in the behavioral and neuropsychological aspects of JME patients. Several studies have suggested specific cognitive deficits that explain some of the clinical and special behavioral findings in patients with JME. They also reported an increased incidence in psychiatric comorbidity in patients with JME. Pung and coworkers reported a circadian dysrhythmia in patients with JME which might explain the poor social outcome observed in those patients [56]. Several recent neuropsychological studies suggest that JME has a specific cognitive profile, with some deficiencies in areas related to the frontal lobes. These studies report impairment in word fluency and interference as well as dysfunctional planning abilities [57-59]. JME patients often show intolerance towards multiple tasks under time pressure. This might explain the occurrence of seizures in some of these patients during cognitive tasks. Also, patients with JME often fail to adhere to treatment plans. This might be linked to impairment in prospective memory. Interestingly Wandschneider and coworkers found this impairment also

(myoclonin1/one EF-hand containing gene) [45] [Table 1].

some loci linked to JME have been described [Table 1] [46].

in JME patients [51, 52].

clonic or tonic-clonic seizures).[41, 53-55].

early diagnosis and curative treatment of JME.

**9. Neuropsychological and behavioral studies in JME**

in their siblings indicating that it might be genetically determined [60].


a Human Genome Nomenclature Committee gene symbol in bold letters.

bMutation segregate with epilepsy affected members across 2 to 4 generation families or in singletons.

c SNP-associated variants of BRD2, Cx36 and ME2; AD, autosomal dominant; AR, autosomal recessive; JME, juvenile my‐ oclonic epilepsy; and pCAE, pyknoleptic childhood absence epilepsy.

**Table 1.** Juvenile myoclonic epilepsy genes and chromosome loci. (Modified from Delgado-Escueta AV,2004,2007)

5 Mendelian JME genes have been reported (http://omim.org and http:// www.ncbi.nlm.nih.gov/omim/). These are as follows: CACNB4 (calcium channel beta4 subunit) [41], CASR (calcium channel sensory receptor) [42], GABRA1 (GABA receptor alpha one subunit) [43], GABRD (GABA receptor delta subunit) [44], and Myoclonin1/EFHC1 (myoclonin1/one EF-hand containing gene) [45] [Table 1].

epilepsy allele will have smaller and smaller genetic effects and will require other epilepsy alleles or environment to produce the epilepsy phenotype [40]. However, linkage disequili‐ brium is strongest and covers the widest region of a chromosome when the epilepsy allele is of recent origin, and has large genetic effects, e.g., Mendelian dominant or recessive effects.

a

60 Epilepsy Topics

c

Human Genome Nomenclature Committee gene symbol in bold letters.

oclonic epilepsy; and pCAE, pyknoleptic childhood absence epilepsy.

bMutation segregate with epilepsy affected members across 2 to 4 generation families or in singletons.

SNP-associated variants of BRD2, Cx36 and ME2; AD, autosomal dominant; AR, autosomal recessive; JME, juvenile my‐

**Table 1.** Juvenile myoclonic epilepsy genes and chromosome loci. (Modified from Delgado-Escueta AV,2004,2007)

Also, three SNP susceptibility alleles of putative JME genes that contribute to the complex genetics of JME have been reported [12, 39, 46]: bromodomain-containing 2 (BRD2) [47], connexin 36 (Cx-36) [48], and malic enzyme2 (ME2) [49]. Additionally, more than 22 chromo‐ some loci linked to JME have been described [Table 1] [46].

Familial segregation or association with disease identifies putative JME disease genes. In autosomal dominant JME, a candidate epilepsy gene should show at least one variant per affected individual; each candidate epilepsy gene should show homozygous mutations or compound heterozygous mutations in autosomal recessive JME [50-52]. Contributions of de novo mutations in the epilepsies have been demonstrated through studies of copy number variations (CNVs) which in fact contribute to genetic generalized epilepsies with complex inheritance, including JME. Consequently, pathogenic de novo mutations could be identified in JME patients [51, 52].

It has been shown that mutations in Cchb4, the mouse homologue of human CACNB4 or mutations in GABRA1 are sufficient to produce the absence phenotype while mutations in Myoclonin1/EFHC1 or BRD2 are sufficient to produce a convulsive phenotype (myoclonic or clonic or tonic-clonic seizures).[41, 53-55].

The identification of epilepsy alleles that cause JME could lead to new AED discoveries, to early diagnosis and curative treatment of JME.

#### **9. Neuropsychological and behavioral studies in JME**

There is an increasing interest in the behavioral and neuropsychological aspects of JME patients. Several studies have suggested specific cognitive deficits that explain some of the clinical and special behavioral findings in patients with JME. They also reported an increased incidence in psychiatric comorbidity in patients with JME. Pung and coworkers reported a circadian dysrhythmia in patients with JME which might explain the poor social outcome observed in those patients [56]. Several recent neuropsychological studies suggest that JME has a specific cognitive profile, with some deficiencies in areas related to the frontal lobes. These studies report impairment in word fluency and interference as well as dysfunctional planning abilities [57-59]. JME patients often show intolerance towards multiple tasks under time pressure. This might explain the occurrence of seizures in some of these patients during cognitive tasks. Also, patients with JME often fail to adhere to treatment plans. This might be linked to impairment in prospective memory. Interestingly Wandschneider and coworkers found this impairment also in their siblings indicating that it might be genetically determined [60].

#### **10. Neuroimaging findings in JME**

A particular personality profile is associated with JME. Behavioral studies suggest a possible frontal lobe dysfunction [57]. Modern neuroimaging techniques have proven to be very useful in understanding the underlying mechanisms of JME. A PET study using H215O to measure cerebral blood flow in patients with IGE and a history of absence seizures showed that there was a significant focal increase in thalamic blood flow during absence seizures. This result suggests that the thalamus plays a key role in the pathogenesis of typical absence seizures [61]. In another study using 18F-FDG PET and a visual working memory paradigm in nine JME patients and 14 controls, in which pairs of abstract images were presented and subjects had to indicate (by pressing a button) whether the images were matching or not, Swartz and collea‐ gues showed that JME patients' performance was impaired during the working memory condition. The authors concluded that dysfunction in thalamo-fronto-cortical networks might account for poor working memory performance in JME patients. The decreased uptake of 18F-FDG in the ventral premotor cortex, the caudate, the dorsolateral prefrontal cortex bilaterally, and the left premotor area, was in favor of a widespread frontal impairment [62]. Using PET and the radioligand 11C-WAY-100635, Meschaks and colleagues observed reduced WAY-100635 binding potential in the dorsolateral prefrontal cortex, the raphe nuclei, and the hippocampus, but not in motor cortex. The observed reductions in serotonin 1A receptor binding suggest that the serotonin system is affected in JME, and also that serotonergic processes are involved in the pathophysiology of myoclonus in JME [63]. In another PET study, Ciumas and coworkers compared JME patients with patients suffering from generalized tonic-clonic seizures (GTCS): alterations in the dopamine system were found in both GTCS and JME [64].

**11. Animal model of JME**

for IGEs [77, 78].

**12. Management of JME**

to JME at the time of aggravation.

**12.1. AEDs treatment**

Genetic Absence Epilepsy Rats from Strasbourg (GAERS) is a well-established genetic model of absence epilepsy [73]; however, the baboon represents a more advanced non-human primate model of epilepsy, specifically of IGE [74]. It offers a natural model of photosensitive epilepsy with myoclonic and generalized tonic clonic seizures occurring spontaneously or provoked by intermittent photic stimulation [74]. In the baboon, seizures occur spontaneously or are triggered by ketamine or under other circumstances, such for example,fighting among

Juvenile Myoclonic Epilepsy: An Update http://dx.doi.org/10.5772/57431 63

In a recent study by Szabó and coworkers [76] involving a pedigree baboon colony, seizures were defined as generalized myoclonic or tonic-clonic; characteristically two thirds of the seizures occurred in the morning. Also, seizure onset occurred in adolescence (age, 5 y), the prevalence of recurrent seizures in this pedigree was 15%. Contrary to human recent findings, seizures in the baboon were more prevalent in male baboons, with a tendency of an early onset and more frequent seizures compared with female baboons. Electroencephalographically, on the baboon scalp, interictal epileptic discharges present as generalized spike-and-wave discharges of 4-6 Hz frequency. All the above clinical and EEG features in the baboon suggest similarities to juvenile myoclonic epilepsy in humans. The baboon also represents an excellent model for testing the efficacy and electrophysiological mechanisms of action of future AEDs

According to the international League Against Epilepsy, "In JME, response to appropriate drugs is good" [6]. However, 15 % of patients with JME might be drug resistant [79]. Today, with the available new AEDs the rate of drug resistance might be lower. The choice of AEDs is based on clinical experience and the available studies and trials. Several AEDs can be used with success in patients with JME. However, it is important to know, that some AEDs can aggravate myoclonic jerks. Valproate is still considered the first-line treatment in JME in male and females without childbearing potential. The dosage in adults varies from 1000mg to

2000mg/day. The control rate varies from 84.5% to 90% in different studies [5,,80-82].

Several studies have shown the efficacy of Lamotrigine (LTG) in the treatment of JME [83,84]. LTG is useful in younger women because of the potential teratogenicity of VPA [85], in patients with migraine (with aura) [86] and in patients with bipolar depression [87]. However, LTG appears to be less effective than VPA [82] and has the potential to exacerbate seizures in IGE and can aggravate MJ or GTCS [88]. The same authors [88] reported de novo appearance of MJ in IGE in five women among 93 patients treated with LTG (5.4%) with a phenotype close

baboons; also, baboons with seizures have normal brain anatomy [75].

One study using Magnetic Resonance Spectroscopy (proton MRS) found that N-acetyl aspartate (NAA) levels are reduced in the thalami of JME patients. This finding supports the idea that thalamic dysfunction is part of the underlying mechanism of epileptogenesis in JME [65]. Moreover, other interesting studies using 1H-MRS demonstrated a significantly reduced prefrontal concentration of NAA in JME patients compared with controls [66, 67]. This finding seems to be specific to JME, compared with other forms of IGE [67].

Using Functional MRI (fMRI), Vollmar et al. [68] investigated 30 JME patients with a chal‐ lenging working memory fMRI paradigm. The authors found an increased functional connec‐ tivity within the frontal and parietal lobes, between the motor system and areas of higher cognitive functions. They correlated those findings with the well-known fact that cognitive tasks can precipitate MJ in some JME patients [69].

Quantitative MRI has been used to demonstrate subtle but widespread cerebral structural changes in patients with IGE, particularly in patients with JME [70]. In a related study, Woermann and coworkers have shown that patients with JME have an increase in cortical gray matter in the mesial frontal lobes compared with healthy subjects [71].

Using T1weighted MRI and diffusion tensor imaging (DTI) O'Muircheartaigh et al. found a decreased mesial frontal gray matter volume and a reduced fractional anisotropy (FA) in the underlying white matter tracts. These findings may represent the anatomical basis for the reported neuropsychological and psychiatric changes seen in patients with JME [72].

#### **11. Animal model of JME**

**10. Neuroimaging findings in JME**

62 Epilepsy Topics

A particular personality profile is associated with JME. Behavioral studies suggest a possible frontal lobe dysfunction [57]. Modern neuroimaging techniques have proven to be very useful in understanding the underlying mechanisms of JME. A PET study using H215O to measure cerebral blood flow in patients with IGE and a history of absence seizures showed that there was a significant focal increase in thalamic blood flow during absence seizures. This result suggests that the thalamus plays a key role in the pathogenesis of typical absence seizures [61]. In another study using 18F-FDG PET and a visual working memory paradigm in nine JME patients and 14 controls, in which pairs of abstract images were presented and subjects had to indicate (by pressing a button) whether the images were matching or not, Swartz and collea‐ gues showed that JME patients' performance was impaired during the working memory condition. The authors concluded that dysfunction in thalamo-fronto-cortical networks might account for poor working memory performance in JME patients. The decreased uptake of 18F-FDG in the ventral premotor cortex, the caudate, the dorsolateral prefrontal cortex bilaterally, and the left premotor area, was in favor of a widespread frontal impairment [62]. Using PET and the radioligand 11C-WAY-100635, Meschaks and colleagues observed reduced WAY-100635 binding potential in the dorsolateral prefrontal cortex, the raphe nuclei, and the hippocampus, but not in motor cortex. The observed reductions in serotonin 1A receptor binding suggest that the serotonin system is affected in JME, and also that serotonergic processes are involved in the pathophysiology of myoclonus in JME [63]. In another PET study, Ciumas and coworkers compared JME patients with patients suffering from generalized tonic-clonic seizures (GTCS): alterations in the dopamine system were found in both GTCS and JME [64].

One study using Magnetic Resonance Spectroscopy (proton MRS) found that N-acetyl aspartate (NAA) levels are reduced in the thalami of JME patients. This finding supports the idea that thalamic dysfunction is part of the underlying mechanism of epileptogenesis in JME [65]. Moreover, other interesting studies using 1H-MRS demonstrated a significantly reduced prefrontal concentration of NAA in JME patients compared with controls [66, 67]. This finding

Using Functional MRI (fMRI), Vollmar et al. [68] investigated 30 JME patients with a chal‐ lenging working memory fMRI paradigm. The authors found an increased functional connec‐ tivity within the frontal and parietal lobes, between the motor system and areas of higher cognitive functions. They correlated those findings with the well-known fact that cognitive

Quantitative MRI has been used to demonstrate subtle but widespread cerebral structural changes in patients with IGE, particularly in patients with JME [70]. In a related study, Woermann and coworkers have shown that patients with JME have an increase in cortical gray

Using T1weighted MRI and diffusion tensor imaging (DTI) O'Muircheartaigh et al. found a decreased mesial frontal gray matter volume and a reduced fractional anisotropy (FA) in the underlying white matter tracts. These findings may represent the anatomical basis for the

reported neuropsychological and psychiatric changes seen in patients with JME [72].

seems to be specific to JME, compared with other forms of IGE [67].

matter in the mesial frontal lobes compared with healthy subjects [71].

tasks can precipitate MJ in some JME patients [69].

Genetic Absence Epilepsy Rats from Strasbourg (GAERS) is a well-established genetic model of absence epilepsy [73]; however, the baboon represents a more advanced non-human primate model of epilepsy, specifically of IGE [74]. It offers a natural model of photosensitive epilepsy with myoclonic and generalized tonic clonic seizures occurring spontaneously or provoked by intermittent photic stimulation [74]. In the baboon, seizures occur spontaneously or are triggered by ketamine or under other circumstances, such for example,fighting among baboons; also, baboons with seizures have normal brain anatomy [75].

In a recent study by Szabó and coworkers [76] involving a pedigree baboon colony, seizures were defined as generalized myoclonic or tonic-clonic; characteristically two thirds of the seizures occurred in the morning. Also, seizure onset occurred in adolescence (age, 5 y), the prevalence of recurrent seizures in this pedigree was 15%. Contrary to human recent findings, seizures in the baboon were more prevalent in male baboons, with a tendency of an early onset and more frequent seizures compared with female baboons. Electroencephalographically, on the baboon scalp, interictal epileptic discharges present as generalized spike-and-wave discharges of 4-6 Hz frequency. All the above clinical and EEG features in the baboon suggest similarities to juvenile myoclonic epilepsy in humans. The baboon also represents an excellent model for testing the efficacy and electrophysiological mechanisms of action of future AEDs for IGEs [77, 78].

#### **12. Management of JME**

#### **12.1. AEDs treatment**

According to the international League Against Epilepsy, "In JME, response to appropriate drugs is good" [6]. However, 15 % of patients with JME might be drug resistant [79]. Today, with the available new AEDs the rate of drug resistance might be lower. The choice of AEDs is based on clinical experience and the available studies and trials. Several AEDs can be used with success in patients with JME. However, it is important to know, that some AEDs can aggravate myoclonic jerks. Valproate is still considered the first-line treatment in JME in male and females without childbearing potential. The dosage in adults varies from 1000mg to 2000mg/day. The control rate varies from 84.5% to 90% in different studies [5,,80-82].

Several studies have shown the efficacy of Lamotrigine (LTG) in the treatment of JME [83,84]. LTG is useful in younger women because of the potential teratogenicity of VPA [85], in patients with migraine (with aura) [86] and in patients with bipolar depression [87]. However, LTG appears to be less effective than VPA [82] and has the potential to exacerbate seizures in IGE and can aggravate MJ or GTCS [88]. The same authors [88] reported de novo appearance of MJ in IGE in five women among 93 patients treated with LTG (5.4%) with a phenotype close to JME at the time of aggravation.

Levetiracetam (LEV) is highly effective in controlling seizures in JME as shown by the studies by Berkovic and co-workers [89], Noachtar and colleagues [90] and Rosenfeld and colleagues [91]. These randomized, double-blind, placebo-controlled studies showed a responder rate of 61% in patients with JME, with 20.8% of them becoming seizure-free. LEV should be one of the options in the treatment of JME [92] as a first line or add on, particularly in women of childbearing potential. However, LEV also may exaggerate myoclonus [93].

**13. Other treatments and approaches**

JME patients may benefit from this therapy.

**13.2. Vagal Nerve Stimulation (VNS)**

**13.3. Lifestyle, psychiatric treatment**

drug-resistant IGE patients.

**14. Conclusion**

Ketogenic diet has been utilized for a number of conditions. Recently, Kossoff and colleagues [108] have looked at the effectiveness of the use of diet for treatment for AED-resistant JME. The investigators used a modified Atkins diet as an adjunct therapy to treat 8 adolescents and adults patients with JME. Six (75%) of these patients had more than 50% seizure reduction after one month, five patients (63%), had a greater than 50% seizure improvement after three months; three patients reported increasing seizures frequency during periods of noncompli‐ ance. The authors concluded that the modified Atkins diet can be a useful therapy for young patients with AED-resistant JME. However, more patients need to be studied to assess which

Juvenile Myoclonic Epilepsy: An Update http://dx.doi.org/10.5772/57431 65

Only one study [109] reported on the role of VNS in drug resistant JME. In this study, 12 patients with drug resistant IGE were offered VNS. Among these patients 7 patients had JME: 5 of them responded to VNS and had reduced AED-treatment at follow up. Kostov and colleagues [109] concluded that adjunctive VNS therapy is a favourable treatment option for

In managing JME, life style is considered an important part of the treatment and this aspect should be discussed with the patient in order to obtain good seizure control. Patients should avoid all precipitating factors such as fatigue, sleep deprivation, alcohol and unnecessary drug intake. In particular, patients should avoid any potentially dangerous activities in the awak‐

There is also a high prevalence of psychiatric disorders in patients with JME, such as mood, anxiety, and personality disorders. Early recognition and treatment of these disturbances and

JME is a common form of IGE with a characteristic clinical and electroencephalographic profile. Usually, a sleep EEG or an EEG on awakening confirm the clinical suspicion. Despite this distinct clinical and EEG trait, JME is often not recognized as such; this might result in serious consequences for the sufferers: in particular, if potentially aggravating AEDs are used, especially Carbamazepine, Oxcarbazepine, Phenytoin; but also, in some patients, Lamotrigine, which might exacerbate absences and myoclonus. These AEDs are therefore contraindicated, although they can improve control of tonic-clonic seizures when these are refractory to other

ening period, such as taking a bath without observation, for example.

psychosocial difficulties play an important role in the prognosis of JME [110].

**13.1. Dietary therapy for JME**

If tolerated, Topiramate (TPM) can be useful in the treatment of JME particularly in overweight patients and in patients with associated migraine. Several authors have shown its efficacy as an add-on therapy in JME [94-96]. TPM was even reported as slightly more efficacious then VPA in a study by Levisohn and Holland in 2007 [97]. TPM may produce neuropsychiatric side effects particularly alteration of attention, and verbal fluency [98] and therefore may lead to treatment failure [82].

Few studies showed good efficacy of Zonisamide (ZNS) in patients with JME. One particular study by Kothare and colleagues looked at 15 patients with JME: 13 patients received ZNS as first monotherapy and 2 as add-on therapy. There were 80% of responders in the monotherapy group. 69% of patients were GTCS-free, 62% were seizure-free for MJ, and 38% were seizurefree for absences. The daily dose ranged between 200 and 500 mg [99].

Another study showed that ZNS treatment led to more than 50% reduction of seizure fre‐ quency in 83.3% of treated patients for GTCS and in 100% for MJ and absences [100].

Obeid and colleagues reported that Clonazepam was effective in controlling myoclonic jerks, but not the GTCS in JME patients [101]. In a later study, Panayiotopoulos [102] found both Clonazepam and Acetazolamide useful adjunctive drugs in JME, particularly if absences and myoclonus are associated. Mantoan and colleagues [92] found that Clonazepam can be combined with LTG in JME in order to avoid the myoclonic effects of LTG. The same authors, among others, have shown that few AEDs like Carbamazepine, Oxcarbazepine, and Phenytoin can exacerbate absences and myoclonus and even induce status epilepticus and should not be used in JME, although these drugs are able to control tonic-clonic seizures associated with JME when these are refractory to other medication [92,103-105].

Lacosamide may be effective in JME. However, larger, controlled studies confirming the usefulness of this AED are lacking [106].

#### **12.2. Discontinuing AEDs in patients with JME**

In a recent interesting study, Geithner and colleaguers [107] followed 31 patients with JME for as long as 25 years: Of these 31 patients, 67.7 % became seizure free. In 6 of these patients (28.6%), AEDs were discontinued with no more seizures. The most important factor that increased the chance for complete seizure freedom after stopping the AED was complete remission of the GTCS under a single anti-epileptic drug. However, the occurrence of a photoparoxysmal response increased the risk of seizure recurrence after stopping AEDs. The authors concluded that in order to maintain seizure freedom, lifelong antiepileptic drug treatment is not necessarily required by all patients with JME.

#### **13. Other treatments and approaches**

#### **13.1. Dietary therapy for JME**

Levetiracetam (LEV) is highly effective in controlling seizures in JME as shown by the studies by Berkovic and co-workers [89], Noachtar and colleagues [90] and Rosenfeld and colleagues [91]. These randomized, double-blind, placebo-controlled studies showed a responder rate of 61% in patients with JME, with 20.8% of them becoming seizure-free. LEV should be one of the options in the treatment of JME [92] as a first line or add on, particularly in women of

If tolerated, Topiramate (TPM) can be useful in the treatment of JME particularly in overweight patients and in patients with associated migraine. Several authors have shown its efficacy as an add-on therapy in JME [94-96]. TPM was even reported as slightly more efficacious then VPA in a study by Levisohn and Holland in 2007 [97]. TPM may produce neuropsychiatric side effects particularly alteration of attention, and verbal fluency [98] and therefore may lead

Few studies showed good efficacy of Zonisamide (ZNS) in patients with JME. One particular study by Kothare and colleagues looked at 15 patients with JME: 13 patients received ZNS as first monotherapy and 2 as add-on therapy. There were 80% of responders in the monotherapy group. 69% of patients were GTCS-free, 62% were seizure-free for MJ, and 38% were seizure-

Another study showed that ZNS treatment led to more than 50% reduction of seizure fre‐

Obeid and colleagues reported that Clonazepam was effective in controlling myoclonic jerks, but not the GTCS in JME patients [101]. In a later study, Panayiotopoulos [102] found both Clonazepam and Acetazolamide useful adjunctive drugs in JME, particularly if absences and myoclonus are associated. Mantoan and colleagues [92] found that Clonazepam can be combined with LTG in JME in order to avoid the myoclonic effects of LTG. The same authors, among others, have shown that few AEDs like Carbamazepine, Oxcarbazepine, and Phenytoin can exacerbate absences and myoclonus and even induce status epilepticus and should not be used in JME, although these drugs are able to control tonic-clonic seizures associated with JME

Lacosamide may be effective in JME. However, larger, controlled studies confirming the

In a recent interesting study, Geithner and colleaguers [107] followed 31 patients with JME for as long as 25 years: Of these 31 patients, 67.7 % became seizure free. In 6 of these patients (28.6%), AEDs were discontinued with no more seizures. The most important factor that increased the chance for complete seizure freedom after stopping the AED was complete remission of the GTCS under a single anti-epileptic drug. However, the occurrence of a photoparoxysmal response increased the risk of seizure recurrence after stopping AEDs. The authors concluded that in order to maintain seizure freedom, lifelong antiepileptic drug treatment is

quency in 83.3% of treated patients for GTCS and in 100% for MJ and absences [100].

childbearing potential. However, LEV also may exaggerate myoclonus [93].

free for absences. The daily dose ranged between 200 and 500 mg [99].

when these are refractory to other medication [92,103-105].

usefulness of this AED are lacking [106].

**12.2. Discontinuing AEDs in patients with JME**

not necessarily required by all patients with JME.

to treatment failure [82].

64 Epilepsy Topics

Ketogenic diet has been utilized for a number of conditions. Recently, Kossoff and colleagues [108] have looked at the effectiveness of the use of diet for treatment for AED-resistant JME. The investigators used a modified Atkins diet as an adjunct therapy to treat 8 adolescents and adults patients with JME. Six (75%) of these patients had more than 50% seizure reduction after one month, five patients (63%), had a greater than 50% seizure improvement after three months; three patients reported increasing seizures frequency during periods of noncompli‐ ance. The authors concluded that the modified Atkins diet can be a useful therapy for young patients with AED-resistant JME. However, more patients need to be studied to assess which JME patients may benefit from this therapy.

#### **13.2. Vagal Nerve Stimulation (VNS)**

Only one study [109] reported on the role of VNS in drug resistant JME. In this study, 12 patients with drug resistant IGE were offered VNS. Among these patients 7 patients had JME: 5 of them responded to VNS and had reduced AED-treatment at follow up. Kostov and colleagues [109] concluded that adjunctive VNS therapy is a favourable treatment option for drug-resistant IGE patients.

#### **13.3. Lifestyle, psychiatric treatment**

In managing JME, life style is considered an important part of the treatment and this aspect should be discussed with the patient in order to obtain good seizure control. Patients should avoid all precipitating factors such as fatigue, sleep deprivation, alcohol and unnecessary drug intake. In particular, patients should avoid any potentially dangerous activities in the awak‐ ening period, such as taking a bath without observation, for example.

There is also a high prevalence of psychiatric disorders in patients with JME, such as mood, anxiety, and personality disorders. Early recognition and treatment of these disturbances and psychosocial difficulties play an important role in the prognosis of JME [110].

#### **14. Conclusion**

JME is a common form of IGE with a characteristic clinical and electroencephalographic profile. Usually, a sleep EEG or an EEG on awakening confirm the clinical suspicion. Despite this distinct clinical and EEG trait, JME is often not recognized as such; this might result in serious consequences for the sufferers: in particular, if potentially aggravating AEDs are used, especially Carbamazepine, Oxcarbazepine, Phenytoin; but also, in some patients, Lamotrigine, which might exacerbate absences and myoclonus. These AEDs are therefore contraindicated, although they can improve control of tonic-clonic seizures when these are refractory to other medications. The following AEDs should not be used in JME: Gabapentin, Pregabalin, Tiagabine, and Vigabatrin; they can worsen seizures (Tiagabine and Vigabatrin might induce absence status epilepticus).

**References**

teurs.1899.

rol Latinoamer. 1958;4:23–48.

gy. 1984;34(3):285-294.

Epilepsia 1989;30:389-399.

*type*. Stuttgart: Thieme; 1977.

48–54.

386.

[1] Herpin T. *Des acces incomplets d'epilepsie*. Paris: Bailliere 1867.

[2] Rabot L. *De la myoclonie épileptique. Medical thesis, Paris.* Georges Carre et C.Naud,edi‐

Juvenile Myoclonic Epilepsy: An Update http://dx.doi.org/10.5772/57431 67

[3] Janz D, Christian W. Impulsive-petit mal. J Neurol (Z Nervenheilkd) 1957; 176: 344–

[4] Castells C, Mendilaharsu C. La epilepsia mioclónica bilateral y consciente. Acta Neu‐

[5] Delgado-Escueta AV, Enrile-Bacsal F. Juvenile myoclonic epilepsy of Janz. Neurolo‐

[6] Commission on Classification and Terminology of the International League Against Epilepsy.Proposal for revised classification of epilepsies and epileptic syndromes.

[7] Durón RM, Medina MT, Martínez-Juárez IE, Bailey JN, Perez-Gosiengfiao KT, Ra‐ mos-Ramírez R, López-Ruiz M, Alonso ME, Ortega RH, Pascual-Castroviejo I, Ma‐

[8] Benamer HT, Grosset DG. A systematic review of the epidemiology of epilepsy in

[9] Tsuboi T. *Primary generalized epilepsy with sporadic myoclonias of myoclonic petit mal*

[10] Murthy JM, Yangala R, Srinivas M. The syndromic classification of the International League Against Epilepsy: a hospital-based study from South India. *Epilepsia* 1998; 39:

[11] Jain S, Tripathi M, Srivastava AK, Narula A. Phenotypic analysis of juvenile myo‐

[12] Nicoletti A, Reggio A, Bartoloni A, Failla G, Sofia V, Bartalesi F, et al. Prevalence of

[13] Hauser WA, Annegers JF, Kurland LT. Incidence of epilepsy and unprovoked seiz‐

[14] Panayiotopoulos CP, Obeid T, Tahan AR. Juvenile myoclonic epilepsy: a 5-year pro‐

[15] Camfield CS, Striano P, Camfield PR. Epidemiology of juvenile myoclonic epilepsy.

clonic epilepsy in Indian families. *Acta Neurol Scand* 2003; 107: 356–362.

ures in Rochester, Minnesota:1935-1984.Epilepsia 1993;34:453-468.

epilepsy in rural Bolivia. *Neurology* 1999; 53: 2064–2069.

spective study. Epilepsia. 1994 Mar-Apr;35(2):285-296.

Epilepsy Behav. 2013;28 Suppl 1:S15-17.

chado-Salas J, Mija L, Delgado-Escueta AV. *Epilepsia*. 2005;46 Suppl 9:34-47.

Arab countries. Epilepsia. 2009;50(10):2301-2304.

Beside the pharmacological treatment, management of JME should also include the patient's lifestyle, with avoidance of sleep deprivation, alcohol excess and the treatment of the cognitive and psychiatric problems that is often associated with JME.

With correct diagnosis and appropriate AED treatment (such as Valproate, Levetiracetam), a small but important group of patients will be able to come off medication not requiring therefore lifelong AED medications.

In refractory cases of JME, modified Atkins diet might be useful. Vagus Nerve Stimulation, Callosotomy and Deep brain stimulation are rarely contemplated.

Neuroimaging, using advanced imaging techniques, suggests subtle structural and functional changes, mainly within the frontal lobes, in patients with JME.

These changes correlate with the observed neuropsychological deficits (frontal lobe dysfunc‐ tion) in patients with JME.

Genetically, JME is the most common cause of hereditary grand mal seizures and has both Mendelien (dominant or recessive trait) and complex genetic inheritance. During the last two decades a lot of discoveries have been made in this field. Finding more chromosome loci and more epilepsy-causing mutations for JME will continue to provide definitive evidence of the complex nature of this disease and of the existence of specific diseases within JME. Future

AEDs should be designed to counter major genes that cause JME.

#### **Author details**

Boulenouar Mesraoua1\*, Dirk Deleu1,2, Hassan Al Hail1,2, Gayane Melikyan1,2 and Heinz Gregor Wieser3

\*Address all correspondence to: boulenouar.mesraoua@wanadoo.fr

1 Department of Medicine /Neurology-Neurophysiology, Hamad Medical, Corporation, Doha, Qatar

2 Department of Clinical Neurology and Neurosciences, Weill Cornell Medical, College WCMC, Doha, Qatar

3 Department of Epileptology and Electroencephalography, Neurology Clinic, University of Zurich, Switzerland

#### **References**

medications. The following AEDs should not be used in JME: Gabapentin, Pregabalin, Tiagabine, and Vigabatrin; they can worsen seizures (Tiagabine and Vigabatrin might induce

Beside the pharmacological treatment, management of JME should also include the patient's lifestyle, with avoidance of sleep deprivation, alcohol excess and the treatment of the cognitive

With correct diagnosis and appropriate AED treatment (such as Valproate, Levetiracetam), a small but important group of patients will be able to come off medication not requiring

In refractory cases of JME, modified Atkins diet might be useful. Vagus Nerve Stimulation,

Neuroimaging, using advanced imaging techniques, suggests subtle structural and functional

These changes correlate with the observed neuropsychological deficits (frontal lobe dysfunc‐

Genetically, JME is the most common cause of hereditary grand mal seizures and has both Mendelien (dominant or recessive trait) and complex genetic inheritance. During the last two decades a lot of discoveries have been made in this field. Finding more chromosome loci and more epilepsy-causing mutations for JME will continue to provide definitive evidence of the complex nature of this disease and of the existence of specific diseases within JME. Future

and psychiatric problems that is often associated with JME.

Callosotomy and Deep brain stimulation are rarely contemplated.

changes, mainly within the frontal lobes, in patients with JME.

AEDs should be designed to counter major genes that cause JME.

\*Address all correspondence to: boulenouar.mesraoua@wanadoo.fr

Boulenouar Mesraoua1\*, Dirk Deleu1,2, Hassan Al Hail1,2, Gayane Melikyan1,2 and

1 Department of Medicine /Neurology-Neurophysiology, Hamad Medical, Corporation,

2 Department of Clinical Neurology and Neurosciences, Weill Cornell Medical, College

3 Department of Epileptology and Electroencephalography, Neurology Clinic, University of

absence status epilepticus).

66 Epilepsy Topics

therefore lifelong AED medications.

tion) in patients with JME.

**Author details**

Heinz Gregor Wieser3

WCMC, Doha, Qatar

Zurich, Switzerland

Doha, Qatar


[16] Mullins GM, O'Sullivan SS, Neligan A, McCarthy A, McNamara B, Galvin RJ, Swee‐ ney BJ. A study of idiopathic generalised epilepsy in an Irish population. Seizure. 2007;16(3):204-210.

[32] Syvertsen MR, Markhus R, Selmer KK, Nakken KO. Juvenile myoclonic epilepsy.

Juvenile Myoclonic Epilepsy: An Update http://dx.doi.org/10.5772/57431 69

[33] Panayiotopoulos CP, Obeid T, Waheed G. Differentiation of typical absence seizures in epileptic syndromes. A video EEG study of 224 seizures in 20 patients. Brain.

[34] Martínez-Juárez IE, Alonso ME, Medina MT, Durón RM, Bailey JN, López-Ruiz M, Ramos-Ramírez R, León L, Pineda G, Castroviejo IP, Silva R, Mija L, Perez-Gosieng‐ fiao K, Machado-Salas J, Delgado-Escueta AV. Juvenile myoclonic epilepsy subsyn‐ dromes: family studies and long-term follow-up. Brain. 2006;129(Pt 5):1269-1280. [35] Kasteleijn-Nolst Trenité DG, Schmitz B, Janz D, Delgado-Escueta AV, Thomas P, Hirsch E, Lerche H, Camfield C, Baykan B, Feucht M, Martínez-Juárez IE, Duron RM, Medina MT, Rubboli G, Jerney J, Hermann B, Yacubian E, Koutroumanidis M, Ste‐ phani U, Salas-Puig J, Reed RC, Woermann F, Wandschneider B, Bureau M, Gambar‐ della A, Koepp MJ, Gelisse P, Gurses C, Crespel A, Nguyen-Michel VH, Ferlazzo E, Grisar T, Helbig I, Koeleman BP, Striano P, Trimble M, Buono R, Cossette P, Represa A, Dravet C, Serafini A, Berglund IS, Sisodiya SM, Yamakawa K, Genton P. Consen‐ sus on diagnosis and management of JME: From founder's observations to current

[36] Beniczky S, Guaranha MS, Conradsen I, Singh MB, Rutar V, Lorber B, Braga P, Fres‐ sola AB, Inoue Y, Yacubian EM, Wolf P. Modulation of epileptiform EEG discharges in juvenile myoclonic epilepsy: an investigation of reflex epileptic traits. Epilepsia.

[37] Genton P, Salas Puig X, Tunon A, Lahoz C, Del Socorro Gonzales Sanchez M. Juve‐ nile myoclonic epilepsy and related syndromes: clinical and neurophysiological as‐ pects. In: Malafosse A, Genton P, Hirsch E, Marescaux C, Broglin, Bernasconi R, eds.

[38] Santiago-Rodríguez E, Harmony T, Cárdenas-Morales L, Hernández A, Fernández-Bouzas A. Analysis of background EEG activity in patients with juvenile myoclonic

[39] Delgado-Escueta AV. Advances in genetics of juvenile myoclonic epilepsies. Epilep‐

[40] Terwilliger JD, Göring HH. Gene mapping in the 20th and 21st centuries: statistical methods, data analysis, and experimental design. Hum biology, 72(1). Detroit, Michi‐

[41] Escayg A, De Waard M, Lee DD, et al. Coding and noncoding variation of the human calcium-channel beta4-subunit gene CACNB4 in patients with idiopathic generalized

epilepsy and episodic ataxia. Am J Hum Genet 2000;66(5):1531–1539

*Idiopathic generalized Epilepsies*. London:John Libbey,1994, p.253-265.

Tidsskr Nor Laegeforen. 2012;132(14):1610-1613.

trends. Epilepsy Behav. 2013;28 Suppl 1:S87-90.

1989;112 (Pt 4):1039-1056.

2012;53(5):832-839.

sy Curr 2007;7:61–67.

epilepsy. Seizure. 2008;17(5):437-45.

gan: Wayne State University Press; 2000. p. 63–132.


[32] Syvertsen MR, Markhus R, Selmer KK, Nakken KO. Juvenile myoclonic epilepsy. Tidsskr Nor Laegeforen. 2012;132(14):1610-1613.

[16] Mullins GM, O'Sullivan SS, Neligan A, McCarthy A, McNamara B, Galvin RJ, Swee‐ ney BJ. A study of idiopathic generalised epilepsy in an Irish population. Seizure.

[18] Obeid T, Panayiotopoulos CP. Juvenile myoclonic epilepsy: a study in Saudi Arabia.

[19] Mukhin KIu, Levin PG, Nikanorova MIu. Juvenile myoclonic epilepsy. Zh Nevrol

[20] Murthy JM, Rao CM, Meena AK.Clinical observations of juvenile myoclonic epilepsy

[21] Gram L, Alving J, Sagild JC, Dam M. Juvenile myoclonic epilepsy in unexpected age

[22] Gilliam F, Steinhoff BJ, Bittermann HJ, Kuzniecky R, Faught E, Abou-Khalil B. Adult myoclonic epilepsy: a distinct syndrome of idiopathic generalized epilepsy. Neurolo‐

[23] Jacob S, Martin D, Rajabally YA. Juvenile myoclonic epilepsy in an elderly patient.

[24] Wolf P, Goosses R. Relation of photosensitivity to epileptic syndromes.J Neurol Neu‐

[25] Lu Y, Waltz S, Stenzel K, Muhle H, Stephani U. Photosensitivity in epileptic syn‐ dromes of childhood and adolescence. Epileptic Disord. 2008;10(2):136-143.

[26] Serafini A, Rubboli G, Gigli GL, Koutroumanidis M, Gelisse P. Neurophysiology of

[27] Létourneau K, Cieuta-Walti C, Deacon C. Epileptiform asymmetries and treatment response in juvenile myoclonic epilepsy. Can J Neurol Sci. 2010;37(6):826-830.

[28] Jayalakshmi SS, Srinivasa Rao B, Sailaja S. Focal clinical and electroencephalographic features in patients with juvenile myoclonic epilepsy. Acta Neurol Scand.

[29] Salas Puig J, Tuñón A, Vidal JA, Mateos V, Guisasola LM, Lahoz CH. [Janz's juvenile myoclonic epilepsy: a little-known frequent syndrome. A study of 85 patients]. Med

[30] Larch J, Unterberger I, Bauer G, Reichsoellner J, Kuchukhidze G, Trinka E. Myoclonic status epilepticus in juvenile myoclonic epilepsy. Epileptic Disord. 2009;11(4):

[31] Janz D. Juvenile Myoclonic Epilepsy with typical absences. In:Dunca JS, Panayioto‐ poulos CP, ed. *Typical Absences and Related Syndromes*. London: Churchill,1994.

juvenile myoclonic epilepsy. Epilepsy Behav. 2013;28 Suppl 1:S30-39.

in 131 patients: a study in South India. Seizure. 1998;7(1):43-47.

2007;16(3):204-210.

68 Epilepsy Topics

*Epilepsia*. 1988;29(3):280-282.

gy. 2000;55(7):1030-1033.

2010;122(2):115-123.

309-314.

Clin (Barc). 1994;103(18):684-689.

Age Ageing. 2006;35(2):194-196.

rosurg Psychiatr 1986;49:1368-1391.

[17] Janz D. *Die Epilepsien*. Stuttgart:Thieme,1969.

Psikhiatr Im S S Korsakova. 1995;95(3):17-21.

groups. Epilepsy Res. 1988;2(2):137-140.


[42] Kapoor A, Satishchandra P, Ratnapriya R, et al. An idiopathic epilepsy syndrome linked to 3q13.3-q21 and missense mutations in the extracellular calcium sensing re‐ ceptor gene. Ann Neurol 2008;64(2):158–167.

[55] Mihalek RM, Banerjee PK, Korpi ER, et al. Attenuated sensitivity to neuroactive ste‐ roids in gamma-aminobutyrate type A receptor delta subunit knockout mice. Proc

Juvenile Myoclonic Epilepsy: An Update http://dx.doi.org/10.5772/57431 71

[56] Pung T, Schmitz B. Circadian rhythm and personality profile in juvenile myoclonic

[57] Devinsky O, Gershengorn J, Brown E, Perrine K, Vazquez B, Luciano D. Frontal func‐ tions in juvenile myoclonic epilepsy. Neuropsychiatry Neuropsychol Behav Neurol.

[58] Pascalicchio TF, de Araujo Filho GM, da Silva Noffs MA et al. Neuropsychological profile of patients with juvenile myoclonic epilepsy: a controlled study of 50 pa‐

[59] Piazzini A, Turner K, Vignoli A, Canger R, Canevini MP. Frontal cognitive dysfunc‐

[60] Wandschneider B, Kopp UA, Kliegel M et al. Prospective memory in patients with juvenile myoclonic epilepsy and their healthy siblings. Neurology, 2010; 75 (24):

[61] Prevett MC, Duncan JS, Jones T, Fish DR, Brooks DJ. Demonstration of thalamic acti‐ vation during typical absence seizures using H2(15)O and PET. Neurology, 1995; 45:

[62] Swartz BE, Simpkins F, Halgren E et al. Visual working memory in primary general‐

[63] Meschaks A, Lindstrom P, Halldin C, Farde L, Savic I. Regional reductions in seroto‐ nin 1A receptor binding in juvenile myoclonic epilepsy. Arch Neurol, 2005; 62 (6):

[64] Ciumas C, Wahlin TB, Espino C, Savic I. The dopamine system in idiopathic -615.615. [65] Mory SB, Li LM, Guerreiro CA, Cendes F. Thalamic dysfunction in juvenile myoclon‐

[66] Savic I, Osterman Y, Helms G. MRS shows syndrome differentiated metabolite changes in human-generalized epilepsies. NeuroImage, 2004; 21:163–172.

[67] Simister RJ, McLean MA, Barker GJ, Duncan JS. Proton MRS reveals frontal lobe me‐ tabolite abnormalities in idiopathic generalized epilepsy. Neurology, 2003; 61:897–

[68] Vollmar C, O'Muircheartaigh J, Barker GJ et al. Motor system hyperconnectivity in juvenile myoclonic epilepsy: a cognitive functional magnetic resonance imaging

ized epilepsy: an 18FDG-PET study. Neurology, 1996; 47:12031212.

ic epilepsy: a proton MRS study. Epilepsia, 2003; 44:1402–1405.

tion in juvenile myoclonic epilepsy. Epilepsia, 2008; 49:657-662.

Natl Acad Sci U S A 1999;96(22):12905–12910.

epilepsy. Epilepsia, 2006; 47 (Suppl. 2): 111-114.

tients.Epilepsy Behav, 2007;10: 263-267.

1997, 10:243-246.

2161-2167.

1396–1402.

946-950

902.

study. Brain, 2011; 134:1710–1719.


[55] Mihalek RM, Banerjee PK, Korpi ER, et al. Attenuated sensitivity to neuroactive ste‐ roids in gamma-aminobutyrate type A receptor delta subunit knockout mice. Proc Natl Acad Sci U S A 1999;96(22):12905–12910.

[42] Kapoor A, Satishchandra P, Ratnapriya R, et al. An idiopathic epilepsy syndrome linked to 3q13.3-q21 and missense mutations in the extracellular calcium sensing re‐

[43] Cossette P, Liu L, Brisebois K, et al. Mutation of GABRA1 in an autosomal dominant

[44] Dibbens LM, Feng HJ, Richards MC, et al. GABRD encoding a protein for extra- or peri-synaptic GABAA receptors is a susceptibility locus for generalized epilepsies.

[45] Suzuki T, Delgado-Escueta AV, Aguan K, et al. Mutations in EFHC1 cause juvenile

[46] Noebels JL, Avoli M, Rogawski MA, Olsen RW, Delgado-Escueta AV. The next dec‐ ade of research in the basic mechanisms of the epilepsies. In: Noebels JL, Avoli M, Rogawski MA, Olsen RW, Delgado-Escueta AV, editors. *Jasper's basic mechanisms of*

[47] Pal DK, Evgrafov OV, Tabares P, Zhang F, Durner M, Greenberg DA. BRD2 (RING3) is a probable major susceptibility gene for common juvenile myoclonic epilepsy.Am J

[48] Mas C, Taske N, Deutsch S, et al. Association of the connexin36 gene with juvenile

[49] Greenberg DA, Cayanis E, Strug L, et al. Malic enzyme 2 may underlie susceptibility to adolescent-onset idiopathic generalized epilepsy. Am J Hum Genet 2005;76(1):139–

[50] Roach JC, Glusman G, Smit AF, et al. Analysis of genetic inheritance in a family quar‐

[51] Robinson PN. Whole-exome sequencing for finding de novo mutations in sporadic

[52] Robinson PN, Krawitz P, Mundlos S. Strategies for exome and genome sequence da‐ ta analysis in disease-gene discovery projects. Clin Genet 2011;80(2):127–132.

[53] Feng HJ, Kang JQ, Song L, Dibbens L, Mulley J, Macdonald RL. Delta subunit sus‐ ceptibility variants E177A and R220H associated with complex epilepsy alter channel gating and surface expression of alpha4beta2delta GABAA receptors. J Neurosci

[54] Ding L, Feng HJ, Macdonald RL, Botzolakis EJ, Hu N, Gallagher MJ. GABA(A) recep‐ tor alpha1 subunit mutation A322D associated with autosomal dominant juvenile myoclonic epilepsy reduces the expression and alters the composition of wild type

tet by whole-genome sequencing. Science 2010;328(5978):636–639.

GABA(A) receptors. J Biol Chem 2010;285(34):26390–26405.

form of juvenile myoclonic epilepsy. Nat Genet 2002;31(2):184–189.

ceptor gene. Ann Neurol 2008;64(2):158–167.

Hum Mol Genet 2004;13(13):1315–1319.

Hum Genet 2003;73(2):261–270.

146.

70 Epilepsy Topics

2006;26(5):1499–1506.

myoclonic epilepsy. Nat Genet 2004;36(8):842–849.

*the epilepsies*. Oxford University Press; 2012. p. 3–11.

myoclonic epilepsy. J Med Genet 2004;41(7):e93.

mental retardation. Genome Biol 2010;11(12):144.


[69] Inoue, Y, Kubota H. Juvenile myoclonic epilepsy with praxis-induced seizures. *Juve‐ nile myoclonic epilepsy, The Janz Syndrome*. Wrightson Biomedical Publishing, Peters‐ field, UK (2000), pp. 73–81.

[82] Marson AG, Al-Kharusi AM, Alwaidh M, et al. The SANAD study of effectiveness of valproate, lamotrigine, or topiramate for generalised and unclassifiable epilepsy: an

Juvenile Myoclonic Epilepsy: An Update http://dx.doi.org/10.5772/57431 73

[83] Buchanan N. The use of lamotrigine in juvenile myoclonic epilepsy. Seizure, 1996;

[84] Morris GL, Hammer AE, Kustra RP, MessenheimerJA. Lamotrigine for patients with juvenile myoclonic epilepsy following prior treatment with valproate: results of an

[85] Wlodarczyk BJ, Palacios AM, George TM, Finnell RH. Antiepileptic drugs and preg‐

[86] Lampl C, Katsarava Z, Diener HC, Limmroth V. Lamotrigine reduces migraine aura and migraine attacks in patients with migraine with aura. J Neurol Neurosurg Psy‐

[87] Geddes JR, Calabrese JR, Goodwin GM. Lamotrigine for treatment of bipolar depres‐ sion: independent meta-analysis and meta-regression of individual patient data from

[88] Crespel A, Genton P, Berramdane M, et al. Lamotrigine associated with exacerbation or de novo myoclonus in idiopathic generalized epilepsies. Neurology 2005;65:762–

[89] Berkovic SF, Knowlton RC, Leroy RF, Schiemann J, Falter U. Levetiracetam N01057 Study Group. Placebo-controlled study of levetiracetam in idiopathic generalized ep‐

[90] Noachtar S, Andermann E, Meyvisch P, et al. Levetiracetam for the treatment of idio‐ pathic generalized epilepsy with myoclonic seizures. Neurology 2008;70:607–616. [91] Rosenfeld WE, Benbadis S, Edrich P, Tassinari CA, Hirsch E. Levetiracetam as addon therapy for idiopathic generalized epilepsy syndromes with onset during adoles‐ cence: analysis of two randomized, double-blind, placebo-controlled studies.Epilepsy

[92] Mantoan L, Walker M. Treatment options in juvenile myoclonic epilepsy. Curr Ttreat

[93] Babtain FA. Levetiracetam may worsen myoclonus in patients with juvenile myo‐

[94] Kellett MW, Smith DF, Stockton PA, Chadwick DW. Topiramate in clinical prac‐ tice:first year's post-licensing experience in a specialist epilepsy clinic. J Neurol Neu‐

[95] Prasad A, Kuzniecky RI, Knowlton RC, et al. Evolving antiepileptic drug treatment

clonic epilepsy: case reports. Clin Neuropharmacol. 2012;35(4):201-202.

in juvenile myoclonic epilepsy. Arch Neurol 2003;60:1100–1105.

unblinded randomised controlled trial. Lancet, 2007; 369:1016–1026.

open-label study. Epilepsy Behav, 2004; 5:509–512.

five randomised trials. Br J Psychiatry 2009;194:4–9.

nancy outcomes. Am J Med Genet A 2012;158A(8):2071–2090.

5:149–151.

764.

chiatry 2005;76:1730–1732.

ilepsy. Neurology 2007;69:1751–60

Options Neurol 2011;13(4):355–370.

rosurg Psychiatry 1999;66:759–763.

Res 2009;85:72–80.


[82] Marson AG, Al-Kharusi AM, Alwaidh M, et al. The SANAD study of effectiveness of valproate, lamotrigine, or topiramate for generalised and unclassifiable epilepsy: an unblinded randomised controlled trial. Lancet, 2007; 369:1016–1026.

[69] Inoue, Y, Kubota H. Juvenile myoclonic epilepsy with praxis-induced seizures. *Juve‐ nile myoclonic epilepsy, The Janz Syndrome*. Wrightson Biomedical Publishing, Peters‐

[70] Woermann FG, Sisodiya SM, Free SL, Duncan JS. Quantitative MRI in patients with idiopathic generalized epilepsy. Evidence of widespread cerebral structural changes.

[71] Woermann FG, Free SL, Koepp MJ, Sisodiya SM, Duncan JS. Abnormal cerebral structure in juvenile myoclonic epilepsy demonstrated with voxel-based analysis of

[72] O'Muircheartaigh J, Vollmar C, Barker GJ et al. Focal structural changes and cogni‐ tive dysfunction in juvenile myoclonic epilepsy.Neurology, 2011;76:34–40.

[73] Chipaux M, Vercueil L, Kaminska A, Mahon S, Charp S. Persistence of cortical censo‐ ry processing during absence seizures in human and an animal model: Evidence from EEG and intracellular recordings. PLoS One. 2013; 8(3): e58180.Published online 2013 March 4.doi: 10.1371/journal.pone. 0058180 10.1371/journal.pone.0058180.

[74] Killam EK. Photomyoclonic seizures in the baboon, *Papio papio*. Fed Proc 1979;

[75] Naquet R, Catier J, Menini C. Neurophysiology of photically induced epilepsy in *Pa‐ pio papio*. In: Meldrum BS, Marsden CD. Advances in neurology, vol 10:107-118. New

[76] Szabó CÁ, Knape KD, Leland MM, Cwikla DJ, Williams-Blangero S, Williams JT. Ep‐ idemiology and characterization of seizures in a pedigreed baboon colony. Comp

[77] Killam EK, Matsuzaki M, Killam KF. Studies of anticonvulsant compounds in the *Pa‐ pio papio* model of epilepsy. In: Sabelli H, editor. Chemical Modulation of Brain Func‐

[78] Meldrum BS, Anlezark G, Balzamo E, Horton RW, Trimble M. Photically induced ep‐ ilepsy in *Papio papio* as a model for drug studies. In: Meldrum BS, Marsden CD, edi‐

tors. Advances in Neurology. Vol 10. New York: Raven Press; 1975. p.119-128.

[79] Gelisse P, Genton P, Thomas P, Rey M, Samuelian JC, Dravet C. Clinical factors of drug resistance in juvenile myoclonic epilepsy. J Neurol Neurosurg Psychiatry,

[80] Penry JK, Dean JC, Riela AR. Juvenile myoclonic epilepsy: long-term response to

[81] Genton P, Gelisse P, Thomas P. Juvenile myoclonic epilepsy today: current definition and limits. B. Schmitz, T. Sander (eds.) *Juvenile myoclonic epilepsy: the Janz syndrome*,

Wrightson Medical Publishing, Petersfield and Philadelphia (2000), pp. 11–32.

field, UK (2000), pp. 73–81.

72 Epilepsy Topics

Brain 1999; 121:1661–1667.

38:2429–2433.

Med. 2012;62(6):535-538.

2001;70:240–243.

MRI. Brain,1999; 122:2101–2108.

York (NY): Raven Press; 1975. p.107-118.

tion. New York: Raven Press; 1973. p. 161–171

therapy. Epilepsia, 1989; 30 (Suppl. 4):19–23.


[96] Biton V, Bourgeois BFD. Topiramate in patients with juvenile myoclonic epilep‐ sy.Arch Neurol 2006;62:1705–1708.

[110] de Araujo Filho GM, Yacubian EM. Juvenile myoclonic epilepsy: psychiatric comor‐

Juvenile Myoclonic Epilepsy: An Update http://dx.doi.org/10.5772/57431 75

bidity and impact on outcome. Epilepsy Behav. 2013;28 Suppl 1:S74-80.


[110] de Araujo Filho GM, Yacubian EM. Juvenile myoclonic epilepsy: psychiatric comor‐ bidity and impact on outcome. Epilepsy Behav. 2013;28 Suppl 1:S74-80.

[96] Biton V, Bourgeois BFD. Topiramate in patients with juvenile myoclonic epilep‐

[97] Levisohn PM, Holland KD. Topiramate or valproate in patients with juvenile myo‐ clonic epilepsy: a randomized open-label comparison. Epilepsy Behav 2007;10:547–

[98] Sousa Pda S, Araújo Filho GM, Garzon E, Sakamoto AC, Yacubian EM. Topiramate for the treatment of juvenile myoclonic epilepsy. Arq Neuropsiquiatr 2005;63:733–

[99] Kothare SV, Valencia I, Khurana DS, Hardison H, Melvin JJ, Legido A. Efficacy and tolerability of zonisamide in juvenile myoclonic epilepsy. Epileptic Disord

[100] O'Rourke D, Flynn C, White M, Doherty C, Delanty N. Potential efficacy of zonisa‐ mide in refractory juvenile myoclonic epilepsy: retrospective evidence from an Irish

[101] Obeid T, Panayiotopoulos CP. Clonazepam in juvenile myoclonic epilepsy..Epilepsia.

[102] Panayiotopoulos CP. Treatment of typical absence seizures and related epileptic syn‐

[103] Thomas P, Valton L, Genton P. Absence and myoclonic status epilepticus precipitat‐ ed by antiepileptic drugs in idiopathic generalized epilepsy. Brain. 2006;129(Pt 5):

[104] Osorio I, Reed RC, Peltzer JN. Refractory idiopathic absence status epilepticus: a probable paradoxical effect of phenytoin and carbamazepine. Epilepsia, 2000; 7:887–

[105] Larch J, Unterberger I, Bauer G, Reichsoellner J, Kuchukhidze G, Trinka E. Myoclonic status epilepticus in juvenile myoclonic epilepsy. Epileptic Disord. 2009;11(4):

[106] Afra P, Adamolekun B. Lacosamide treatment of juvenile myoclonic epilepsy. Seiz‐

[107] Geithner J, Schneider F, Wang Z, Berneiser J, Herzer R, Kessler C, Runge U. Predic‐ tors for long-term seizure outcome in juvenile myoclonic epilepsy: 25-63 years of fol‐

[108] Kossoff EH, Henry BJ, Cervenka MC. Efficacy of dietary therapy for juvenile myo‐

[109] Kostov H, Larsson PG, Røste GK. Is vagus nerve stimulation a treatment option for patients with drug-resistant idiopathic generalized epilepsy? Acta Neurol Scand

compassionate-use case series. Ir Med J 2007;100:431–433.

dromes.Paediatr Drugs. 2001;3(5):379-403.

sy.Arch Neurol 2006;62:1705–1708.

552.

74 Epilepsy Topics

737.

2004;6:267–270.

1989;30(5):603-606.

1281-1292.

894

309-314.

ure. 2012;21(3):202-204.

Suppl. 2007;187:55-58.

low-up. Epilepsia. 2012;53(8):1379-1386.

clonic epilepsy. Epilepsy Behav. 2013;26(2):162-164.

**Chapter 5**

**Juvenile Myoclonic Epilepsy — A Maturation Syndrome**

Juvenile Myoclonic Epilepsy (JME) was first described by Herpin [1] in the middle of the 19th century, reporting seizure symptoms he observed in his adolescent son. While Rabot also described myoclonia in 1899 [2], it was not until 1957 when Janz and Christian [3] provided a detailed explanation of the complete syndrome, which was subsequently called by Antonio Delgado-Escueta [4] as "JME of Janz". Since then, JME has become a welldefined epilepsy syndrome that is recognized as one of the most common forms of genetic generalized epilepsy (GGE). JME is primarily characterized by the hallmark manifestation of myoclonic seizures (mainly upon awakening), although patients also often present with a combination of absence and generalized tonic-clonic seizures (GTCS) as well. Other prevalent features of JME include various types of reflex seizures, particularly with photosensitivity. Electrophysiologically, there are prominent generalized ictal and interic‐ tal discharges on scalp electroencephalography (EEG). In contrast to other GGE syn‐ dromes, such as Childhood Absence Epilepsy (CAE), and contrary to earlier assumptions, JME has been shown to be associated with cognitive and behavioral problems, a lifetime

In this chapter, we will review the electroclinical definition of JME along with treatments for its associated seizure types, cognitive and behavioral complications, and underlying pathophysiology. In contrast to long-standing assumptions that JME is due to frontal lobe hyperexcitability that primarily involves corticothalamic pathways, recent literature suggests that JME likely reflects an underlying developmental disorder affecting multiple

> © 2014 The Author(s). Licensee InTech. This chapter is distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/3.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

**Coming of Age**

http://dx.doi.org/10.5772/58372

C. Akos Szabo

**1. Introduction**

brain regions.

Russell D. Pella, Lakshmi Mukundan and

Additional information is available at the end of the chapter

risk of continued seizures, and medication resistance.

## **Juvenile Myoclonic Epilepsy — A Maturation Syndrome Coming of Age**

Russell D. Pella, Lakshmi Mukundan and C. Akos Szabo

Additional information is available at the end of the chapter

http://dx.doi.org/10.5772/58372

#### **1. Introduction**

Juvenile Myoclonic Epilepsy (JME) was first described by Herpin [1] in the middle of the 19th century, reporting seizure symptoms he observed in his adolescent son. While Rabot also described myoclonia in 1899 [2], it was not until 1957 when Janz and Christian [3] provided a detailed explanation of the complete syndrome, which was subsequently called by Antonio Delgado-Escueta [4] as "JME of Janz". Since then, JME has become a welldefined epilepsy syndrome that is recognized as one of the most common forms of genetic generalized epilepsy (GGE). JME is primarily characterized by the hallmark manifestation of myoclonic seizures (mainly upon awakening), although patients also often present with a combination of absence and generalized tonic-clonic seizures (GTCS) as well. Other prevalent features of JME include various types of reflex seizures, particularly with photosensitivity. Electrophysiologically, there are prominent generalized ictal and interic‐ tal discharges on scalp electroencephalography (EEG). In contrast to other GGE syn‐ dromes, such as Childhood Absence Epilepsy (CAE), and contrary to earlier assumptions, JME has been shown to be associated with cognitive and behavioral problems, a lifetime risk of continued seizures, and medication resistance.

In this chapter, we will review the electroclinical definition of JME along with treatments for its associated seizure types, cognitive and behavioral complications, and underlying pathophysiology. In contrast to long-standing assumptions that JME is due to frontal lobe hyperexcitability that primarily involves corticothalamic pathways, recent literature suggests that JME likely reflects an underlying developmental disorder affecting multiple brain regions.

© 2014 The Author(s). Licensee InTech. This chapter is distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/3.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

## **2. Clinical definition**

In an historical context, the syndrome has been previously referred to as "impulsive petit-mal" epilepsy or Janz syndrome [5]. The International League Against Epilepsy (ILAE) later designated the term "Juvenile Myoclonic Epilepsy" in 1975. Under the revised Classification of Epilepsies and Epileptic Syndromes [6], the ILAE defined JME as follows:

because primary centers achieve successful treatment via antiepileptic medications. Moreover, JME is often under-diagnosed and misdiagnosed [22, 23], which can present a challenge to estimating its prevalence. As an illustration, a French study [24] showed that the prevalence of JME in one geographic area was 0% between 1986 and 1994, which rose to nearly 50% between 1996 and 2000, likely due to increased recognition of the syndrome. Moreover, variations in clinical onset can present a challenge to estimating the prevalence of JME. Although JME the peak age of onset is in adolescence (12-18 years), similar to most GGE syndromes [10, 25], the age of onset can vary from 8 to 36 years. There are also *de novo* diagnoses in early adulthood, and JME can, though rarely, begin or be reactivated in advanced ages. For instance, a case report presented two patients from Turkey in whom JME began after the age

Juvenile Myoclonic Epilepsy — A Maturation Syndrome Coming of Age

http://dx.doi.org/10.5772/58372

79

**Myoclonic Seizures**: As noted above, myoclonic movements are one of the main symptoms of JME and consist of generalized seizures which are brief, irregular jerks of the head, trunk, and limbs that can be either symmetric or asymmetric, and may involve isolated regions of the body or the whole body. They usually predominate in the upper limbs (mostly distal muscles), although they occasionally involve muscles in the abdomen, paraspinal distributions, and lower extremity [27]. It is not uncommon for myoclonic seizures to be so subtle or brief that they are perceived as benign "inner shocks". As subtle as the may be, patients tend to more readily notice asymmetric jerks involving the dominant upper extremity because such movements noticeably impair daily functioning. The jerks, if violent, may cause the patient to drop or throw objects, or fall to the floor-which may be mistaken for nonpathological clums‐ iness. Myoclonic seizures can manifest as discrete, single events or they may occur in clusters. They are usually not associated with loss of consciousness but it is not uncommon for patients to lose awareness during the jerks. Moreover, clusters of myoclonic events can evolve into a GTCS and cause post-ictal confusion. They can occur during transition to sleep or during awakening from sleep, usually in the early morning hours [28]. This early morning pattern of seizures is associated with an increase in cortical excitability during that time of day, which has been noted in other patients with GGE, but to a greater extent in JME patients [29]. Developmentally, these jerks often subside in the fourth decade of life, although GTCS or

**Generalized Tonic-Clonic Seizures**: About 80% of the patients with JME have GTCS. These seizures are also more likely to occur if precipitated by sleep deprivation or alcohol intake. Usually, these events are immediately preceded by a series of myoclonic jerks and associated tongue biting prior to generalization. Because, as noted above, myoclonic events are often insidious, patients first seek treatment following an initial GTCS and are then subsequently diagnosed with JME. Typically, upon further evaluation, such patients acknowledge also that they have experienced myoclonic jerks, a description which adds additional clinical support for a JME diagnosis. Due to this association, some authors have proposed that clinicians should

of 70 [26].

**4. Clinical diagnosis**

absence seizures tend to persist [30].

Impulsive petit mal appears around puberty and is characterized by seizures with bilateral, single or repetitive, arrhythmic, irregular myoclonic jerks, predominantly in the arms. Jerks may cause some patients to fall suddenly. No disturbance of consciousness is noticeable. Often, there are GTCS and, less often infrequent absences. The seizures usually occur shortly after awakening and are often precipitated by sleep deprivation.

While the duration of myoclonic seizures are routinely short to the effect that it is often not possible to determine if individuals lose awareness during episodes, repetitive or clustered myoclonic seizures may be associated with altered levels of consciousness. Although GTCS and/or absence seizures occur less frequently in JME, repetitive and clustered seizures can lead to secondarily generalized tonic-clonic episodes. Whereas it is nearly pathognomonic for seizures to occur upon awakening and are precipitated by sleep deprivation, an additional diurnal pattern may also be apparent as myoclonic seizures can occur late in the afternoon or in the evening as well. Electroclinically, interictal EEG patterns consist of generalized 4-6 Hz spike-or polyspike-and-wave discharges, and a relatively high prevalence of photosensitivity.

### **3. Epidemiology**

JME accounts for approximately 4-11% of all epilepsies [7-12] with an incidence of 0.1 to 0.2 occurrences in 100,000 per year. In terms of gender effects, early studies reported that the incidence of JME was greater in male patients than females [13]; however, gender differences have not been consistent across studies, as some studies [9] have shown an equal distribution, and other researchers [10, 14-16] have reported a higher proportion (60%) of females. It is also a typically occurring GGE [12], with prevalence in GGE groups ranging from 5.5% [17] to representing the majority (45.5%) of patients [18]. Prevalence of JME as a proportion of GGE may also differ across the lifespan in some individuals as approximately 15% patients with childhood absence and juvenile absence syndromes develop JME as they age, especially those patients with photosensitive spike-and-wave findings on EEG [4, 19]. As there is a strong genetic component to JME, its prevalence is high among family members of JME patients as well, which has been shown to be the case for certain ethnic groups. Studies from Saudi Arabia, Turkey, and Iran have shown JME patients have a family association with rates that vary from 42.3 to 48.9% [18, 20, 21].

Not only are prevalence rates sensitive to genetic influences, the proportion of patients in particular clinic settings may vary as well. For example, the prevalence in primary epilepsy clinics is high, while it tends to be much lower in tertiary referral centers, due in large part because primary centers achieve successful treatment via antiepileptic medications. Moreover, JME is often under-diagnosed and misdiagnosed [22, 23], which can present a challenge to estimating its prevalence. As an illustration, a French study [24] showed that the prevalence of JME in one geographic area was 0% between 1986 and 1994, which rose to nearly 50% between 1996 and 2000, likely due to increased recognition of the syndrome. Moreover, variations in clinical onset can present a challenge to estimating the prevalence of JME. Although JME the peak age of onset is in adolescence (12-18 years), similar to most GGE syndromes [10, 25], the age of onset can vary from 8 to 36 years. There are also *de novo* diagnoses in early adulthood, and JME can, though rarely, begin or be reactivated in advanced ages. For instance, a case report presented two patients from Turkey in whom JME began after the age of 70 [26].

#### **4. Clinical diagnosis**

**2. Clinical definition**

78 Epilepsy Topics

**3. Epidemiology**

42.3 to 48.9% [18, 20, 21].

In an historical context, the syndrome has been previously referred to as "impulsive petit-mal" epilepsy or Janz syndrome [5]. The International League Against Epilepsy (ILAE) later designated the term "Juvenile Myoclonic Epilepsy" in 1975. Under the revised Classification

Impulsive petit mal appears around puberty and is characterized by seizures with bilateral, single or repetitive, arrhythmic, irregular myoclonic jerks, predominantly in the arms. Jerks may cause some patients to fall suddenly. No disturbance of consciousness is noticeable. Often, there are GTCS and, less often infrequent absences. The seizures usually occur shortly after

While the duration of myoclonic seizures are routinely short to the effect that it is often not possible to determine if individuals lose awareness during episodes, repetitive or clustered myoclonic seizures may be associated with altered levels of consciousness. Although GTCS and/or absence seizures occur less frequently in JME, repetitive and clustered seizures can lead to secondarily generalized tonic-clonic episodes. Whereas it is nearly pathognomonic for seizures to occur upon awakening and are precipitated by sleep deprivation, an additional diurnal pattern may also be apparent as myoclonic seizures can occur late in the afternoon or in the evening as well. Electroclinically, interictal EEG patterns consist of generalized 4-6 Hz spike-or polyspike-and-wave discharges, and a relatively high prevalence of photosensitivity.

JME accounts for approximately 4-11% of all epilepsies [7-12] with an incidence of 0.1 to 0.2 occurrences in 100,000 per year. In terms of gender effects, early studies reported that the incidence of JME was greater in male patients than females [13]; however, gender differences have not been consistent across studies, as some studies [9] have shown an equal distribution, and other researchers [10, 14-16] have reported a higher proportion (60%) of females. It is also a typically occurring GGE [12], with prevalence in GGE groups ranging from 5.5% [17] to representing the majority (45.5%) of patients [18]. Prevalence of JME as a proportion of GGE may also differ across the lifespan in some individuals as approximately 15% patients with childhood absence and juvenile absence syndromes develop JME as they age, especially those patients with photosensitive spike-and-wave findings on EEG [4, 19]. As there is a strong genetic component to JME, its prevalence is high among family members of JME patients as well, which has been shown to be the case for certain ethnic groups. Studies from Saudi Arabia, Turkey, and Iran have shown JME patients have a family association with rates that vary from

Not only are prevalence rates sensitive to genetic influences, the proportion of patients in particular clinic settings may vary as well. For example, the prevalence in primary epilepsy clinics is high, while it tends to be much lower in tertiary referral centers, due in large part

of Epilepsies and Epileptic Syndromes [6], the ILAE defined JME as follows:

awakening and are often precipitated by sleep deprivation.

**Myoclonic Seizures**: As noted above, myoclonic movements are one of the main symptoms of JME and consist of generalized seizures which are brief, irregular jerks of the head, trunk, and limbs that can be either symmetric or asymmetric, and may involve isolated regions of the body or the whole body. They usually predominate in the upper limbs (mostly distal muscles), although they occasionally involve muscles in the abdomen, paraspinal distributions, and lower extremity [27]. It is not uncommon for myoclonic seizures to be so subtle or brief that they are perceived as benign "inner shocks". As subtle as the may be, patients tend to more readily notice asymmetric jerks involving the dominant upper extremity because such movements noticeably impair daily functioning. The jerks, if violent, may cause the patient to drop or throw objects, or fall to the floor-which may be mistaken for nonpathological clums‐ iness. Myoclonic seizures can manifest as discrete, single events or they may occur in clusters. They are usually not associated with loss of consciousness but it is not uncommon for patients to lose awareness during the jerks. Moreover, clusters of myoclonic events can evolve into a GTCS and cause post-ictal confusion. They can occur during transition to sleep or during awakening from sleep, usually in the early morning hours [28]. This early morning pattern of seizures is associated with an increase in cortical excitability during that time of day, which has been noted in other patients with GGE, but to a greater extent in JME patients [29]. Developmentally, these jerks often subside in the fourth decade of life, although GTCS or absence seizures tend to persist [30].

**Generalized Tonic-Clonic Seizures**: About 80% of the patients with JME have GTCS. These seizures are also more likely to occur if precipitated by sleep deprivation or alcohol intake. Usually, these events are immediately preceded by a series of myoclonic jerks and associated tongue biting prior to generalization. Because, as noted above, myoclonic events are often insidious, patients first seek treatment following an initial GTCS and are then subsequently diagnosed with JME. Typically, upon further evaluation, such patients acknowledge also that they have experienced myoclonic jerks, a description which adds additional clinical support for a JME diagnosis. Due to this association, some authors have proposed that clinicians should strongly consider a JME diagnosis, until proven otherwise, when teens present with an initial unprovoked GTCS [31].

develop nonpharmaceutical therapeutic interventions, which is important for the treatment of

Juvenile Myoclonic Epilepsy — A Maturation Syndrome Coming of Age

http://dx.doi.org/10.5772/58372

81

Using a questionnaire, da Silva Sousa [35] found 23% of JME patients surveyed reported having reflex seizures with thinking and concentration, 20% with praxis, 11% with speaking in public, 15% with visual stimuli, 7% with reading, 5% with calculating and writing, 5% with music, and 3% with drawing. In a subsequent study by the same group, patients were continuously monitored for 4 to 6 hours by video-EEG while neuropsychological and physiological triggers were presented [37]. These triggers had a provocative effect in 38%, with praxis being most common trigger. There were also inhibitory effects of tasks in over 90%. 40% of the patients had no effects on ictal or interictal epileptic discharges. A more recent multicenter, video-EEG study, controlling for spontaneous fluctuation of ictal and interictal epileptic discharges, found the provocative effect of neuropsychological and physiological triggers was decreased from 22 to 18%, while the rate of inhibition was decreased from over 90% to 29% [38]. The inhibition was thought to be a non-specific effect of arousal and mental activation, while the provocative

In clinical practice, if reflex epilepsy is suspected from convergence of other clinical data as noted above, clinical procedures can often helpful in identifying suspected triggers. Detailed neuropsychological testing can be conducted with patients who have seizures induced by thinking and completing complex mental activities [39]. Also, functional MRI can be used to study signal changes in the networks involved in generating reflex seizures during tasks.

Because the prevalence of photosensitivity in JME ranges from 25 to 40% [40], clinicians often opt to elicit reflex responses via intermittent photic stimulation (ILS) with concomitant routine EEG with video. Photic stimulation typically consists of flashing light at various frequencies in front of the patient, with seizures being most likely to be elicited by frequencies of 12-18Hz in individuals who are photosensitive. Various visual patterns may be used to provoke seizures as well, and include alternating, oscillating, black or white, or linear responses. Beyond formal testing procedures via ILS, photoparoxysmal and convulsive responses along with ictal and interictal epileptic discharges can be triggered by reading and praxis in patients with JME [13, 40, 41]. Patients who are photosensitive may have the following responses: photic driving, non-convulsive photoparoxysmal episodes, or photoconvulsive responses. Notably, photosensitivity in JME patients is increased in the early morning hours soon after awakening, which is consistent with the diurnal pattern of seizures in these patients [42]. There are no differences in age or gender of patients with reflex seizures in general; however, in patients with photosensitivity there is a clear predominance in adolescence and in the female sex [42]. In terms of the biological mechanisms of reflex seizures in JME patients, particular cerebral regions have been implicated. Myoclonic seizures in JME are expressed in the primary motor and supplementary motor cortices [43], which are extensively connected to the primary sensory and association cortices. It is not clear whether visual stimuli generates a response from the sensory cortices that propagates to functionally connected cortical and subcortical structures, or if a there is parallel synchronization of larger networks induced by the stimuli [44, 45]. In the case of photosensitivity, visual stimulation likely synchronizes frontoparietal cortices prior to the onset of the epileptiform discharge [44-46], a phenomenon suspected on

reflex seizures, in addition to an AED regimen [36].

triggers were task-specific.

**Absence Seizures**: Absence seizures are characterized by a brief loss of awareness (i.e., a few seconds) without any motor manifestations. Such seizures are relatively uncommon in patients with JME. For example, Janz reported that 28% of his patients with JME had absence seizures [3, 13]. Commonly, children who initially experience absence seizures may develop myoclonus or GTCS within 1 to 9 years of their seizure onset, and then may subsequently be diagnosed with JME. The absence seizures of JME differ from those of other GGE, such as CAE or Juvenile Absence Epilepsy (JAE) as they are shorter in duration and associated with a lesser degree of altered consciousness [32].

**Myoclonic Status Epilepticus**: Myoclonic status epilepticus is rare in JME and can present in a variety of ways. Namely, patients can display prolonged myoclonic events, a combination of myoclonic seizures and GTCS, or GTCS that that can follow prolonged absence seizures. Risk factors for this clinical phenomenon include AED withdrawal, sleep deprivation, alcohol intake, and suboptimal therapy [33, 34].

**Precipitating Factors**: As sleep deprivation is the usual precipitant of seizures in JME [35], adolescents or a young adults often experience myoclonic or GTC seizures precipitated by late night studying or socialization. Because of its strong association with precipitating seizures in JME patients, sleep deprivation is typically employed as an activation procedure to provoke characteristic EEG changes that are diagnostically relevant (4-6Hz generalized polyspike-andwave discharges). Also associated with sleep patterns, sudden and provoked awakenings, pose additional increased risk for seizures in JME. Two-thirds of patients with JME have at least one provoking factor [28].

#### **5. Reflex seizures associated with JME**

Reflex seizures are temporally preceded by some type of external stimuli and may occur exclusively, or in conjunction with, spontaneous seizures. Common external triggers include alcohol use, flashing lights, heat, bathing, and eating. They can also be less frequently elicited by internal stimuli such as stress, fever, hyperventilation, thinking, fatigue, menstrual cycle, and sleep. The proclivity of certain stimuli for seizure provocation is often age-dependent. For instance, fever is a more common provoking stimulus in children than in adults.

Regardless of the suspected triggers, it is import to obtain a detailed history from patients and family members to determine if there is a reflex component to seizures. This should include querying about specific triggers, seizure semiology (partial or generalized), family history of reflex seizures, and whether unprovoked seizures occur as well. As patients rarely lose awareness with myoclonic seizures, they may have adequate awareness of their subjective triggers. Knowing the patterns of responses in patients along with prevalence of triggers is, therefore, key to initially investigating a reflex component prior to enlisting formal testing. Moreover, identification of triggering factors leads to finding ways to avoid precipitants and develop nonpharmaceutical therapeutic interventions, which is important for the treatment of reflex seizures, in addition to an AED regimen [36].

strongly consider a JME diagnosis, until proven otherwise, when teens present with an initial

**Absence Seizures**: Absence seizures are characterized by a brief loss of awareness (i.e., a few seconds) without any motor manifestations. Such seizures are relatively uncommon in patients with JME. For example, Janz reported that 28% of his patients with JME had absence seizures [3, 13]. Commonly, children who initially experience absence seizures may develop myoclonus or GTCS within 1 to 9 years of their seizure onset, and then may subsequently be diagnosed with JME. The absence seizures of JME differ from those of other GGE, such as CAE or Juvenile Absence Epilepsy (JAE) as they are shorter in duration and associated with a lesser degree of

**Myoclonic Status Epilepticus**: Myoclonic status epilepticus is rare in JME and can present in a variety of ways. Namely, patients can display prolonged myoclonic events, a combination of myoclonic seizures and GTCS, or GTCS that that can follow prolonged absence seizures. Risk factors for this clinical phenomenon include AED withdrawal, sleep deprivation, alcohol

**Precipitating Factors**: As sleep deprivation is the usual precipitant of seizures in JME [35], adolescents or a young adults often experience myoclonic or GTC seizures precipitated by late night studying or socialization. Because of its strong association with precipitating seizures in JME patients, sleep deprivation is typically employed as an activation procedure to provoke characteristic EEG changes that are diagnostically relevant (4-6Hz generalized polyspike-andwave discharges). Also associated with sleep patterns, sudden and provoked awakenings, pose additional increased risk for seizures in JME. Two-thirds of patients with JME have at

Reflex seizures are temporally preceded by some type of external stimuli and may occur exclusively, or in conjunction with, spontaneous seizures. Common external triggers include alcohol use, flashing lights, heat, bathing, and eating. They can also be less frequently elicited by internal stimuli such as stress, fever, hyperventilation, thinking, fatigue, menstrual cycle, and sleep. The proclivity of certain stimuli for seizure provocation is often age-dependent. For

Regardless of the suspected triggers, it is import to obtain a detailed history from patients and family members to determine if there is a reflex component to seizures. This should include querying about specific triggers, seizure semiology (partial or generalized), family history of reflex seizures, and whether unprovoked seizures occur as well. As patients rarely lose awareness with myoclonic seizures, they may have adequate awareness of their subjective triggers. Knowing the patterns of responses in patients along with prevalence of triggers is, therefore, key to initially investigating a reflex component prior to enlisting formal testing. Moreover, identification of triggering factors leads to finding ways to avoid precipitants and

instance, fever is a more common provoking stimulus in children than in adults.

unprovoked GTCS [31].

80 Epilepsy Topics

altered consciousness [32].

intake, and suboptimal therapy [33, 34].

least one provoking factor [28].

**5. Reflex seizures associated with JME**

Using a questionnaire, da Silva Sousa [35] found 23% of JME patients surveyed reported having reflex seizures with thinking and concentration, 20% with praxis, 11% with speaking in public, 15% with visual stimuli, 7% with reading, 5% with calculating and writing, 5% with music, and 3% with drawing. In a subsequent study by the same group, patients were continuously monitored for 4 to 6 hours by video-EEG while neuropsychological and physiological triggers were presented [37]. These triggers had a provocative effect in 38%, with praxis being most common trigger. There were also inhibitory effects of tasks in over 90%. 40% of the patients had no effects on ictal or interictal epileptic discharges. A more recent multicenter, video-EEG study, controlling for spontaneous fluctuation of ictal and interictal epileptic discharges, found the provocative effect of neuropsychological and physiological triggers was decreased from 22 to 18%, while the rate of inhibition was decreased from over 90% to 29% [38]. The inhibition was thought to be a non-specific effect of arousal and mental activation, while the provocative triggers were task-specific.

In clinical practice, if reflex epilepsy is suspected from convergence of other clinical data as noted above, clinical procedures can often helpful in identifying suspected triggers. Detailed neuropsychological testing can be conducted with patients who have seizures induced by thinking and completing complex mental activities [39]. Also, functional MRI can be used to study signal changes in the networks involved in generating reflex seizures during tasks.

Because the prevalence of photosensitivity in JME ranges from 25 to 40% [40], clinicians often opt to elicit reflex responses via intermittent photic stimulation (ILS) with concomitant routine EEG with video. Photic stimulation typically consists of flashing light at various frequencies in front of the patient, with seizures being most likely to be elicited by frequencies of 12-18Hz in individuals who are photosensitive. Various visual patterns may be used to provoke seizures as well, and include alternating, oscillating, black or white, or linear responses. Beyond formal testing procedures via ILS, photoparoxysmal and convulsive responses along with ictal and interictal epileptic discharges can be triggered by reading and praxis in patients with JME [13, 40, 41]. Patients who are photosensitive may have the following responses: photic driving, non-convulsive photoparoxysmal episodes, or photoconvulsive responses. Notably, photosensitivity in JME patients is increased in the early morning hours soon after awakening, which is consistent with the diurnal pattern of seizures in these patients [42]. There are no differences in age or gender of patients with reflex seizures in general; however, in patients with photosensitivity there is a clear predominance in adolescence and in the female sex [42].

In terms of the biological mechanisms of reflex seizures in JME patients, particular cerebral regions have been implicated. Myoclonic seizures in JME are expressed in the primary motor and supplementary motor cortices [43], which are extensively connected to the primary sensory and association cortices. It is not clear whether visual stimuli generates a response from the sensory cortices that propagates to functionally connected cortical and subcortical structures, or if a there is parallel synchronization of larger networks induced by the stimuli [44, 45]. In the case of photosensitivity, visual stimulation likely synchronizes frontoparietal cortices prior to the onset of the epileptiform discharge [44-46], a phenomenon suspected on the basis of large cortical blood flow increases even before the appearance of hyperventilationinduced generalized spike-and-wave discharges in absence epilepsy [47, 48]. Given the variety of aforementioned triggers, it is expected that reflex seizures are provoked by stimulation of different sensory cortices, such as the primary somatosensory cortex, primary auditory cortex, or related association areas. For instance, somatosensory or cognitively evoked seizures are evoked in physiologically activated cortical areas that overlap with hyperexcitable cortices giving rise to ictal or interictal epileptic discharges [49].

#### **6. Electrographic findings**

Given the rate of reflex seizures in JME, it is clinically indicated to conduct a sleep-deprived EEG with activation procedures, such as photic stimulation and hyperventilation. Although a positive EEG highly supports a clinical diagnosis of JME, a negative EEG does not definitively rule out the diagnosis. When examining EEG results, there are typical EEG patterns that occur in JME that should be considered.

Legend: The discharge is originally 5 Hz, slowing down to 3 Hz. No clinical signs were noted.

Juvenile Myoclonic Epilepsy — A Maturation Syndrome Coming of Age

http://dx.doi.org/10.5772/58372

83

**Figure 1.** Generalized Interictal Epileptic Discharge in Drowsiness

**Figure 2.** Generalized Polyspikes Activated in Stage II Sleep

**Interictal Pattern**: The interictal background activity is usually considered normal in patients with JME [4, 7, 50, 51]. However, Genton et al. [52] found that routine EEGs were normal in 27% of cases and misleading or nonspecific in 20% of cases; although, 54% showed generalized interictal epileptic discharges (IEDs). Other studies have reported percentages of abnormali‐ ties, consisting of generalized polyspike-and-wave or spike-and-wave complexes in 75 to 85% of probands [9, 53, 54]. The typical interictal EEG finding in JME is a 3.5-6 Hz spike-and polyspike-and-wave pattern with a frontocentral predominance that lasts up to 15 to 20 seconds [50, 55]. While a pattern of more prolonged IEDs with a 2-3 Hz frequency is not specific to patients with absence seizures, there may be an increased risk of clinical absence seizures if a 3Hz spike-and-wave interictal pattern is noted. IEDs are primarily detected during drowsi‐ ness and or sleep onset, or upon awakening [56, 57] (see Figures 1-3), during ILS (photopar‐ oxysmal response) (see Figure 4), and hyperventilation. They may be associated with myoclonic seizures or GTCS (photoconvulsive response). Family members of JME are often affected or carry the abnormal EEG traits, as 80% percent of symptomatic siblings and 6% of asymptomatic siblings have diffuse 4-6 Hz spike-and-wave complexes [58]. Photoparoxysmal EEG response is another heritable trait of JME, and can be found in 20 to 60% of near relatives of probands [59].

**Ictal Pattern**: The ictal EEG during myoclonic jerks (see Figure 5) typically reveals a sudden onset of brief (less than 0.5 sec) bursts of 10-16 Hz polyspike-and-wave discharges often followed by 1-3 Hz slow waves [60]. The number of spike-and-wave complexes ranges from 5 to 20 per discharge and correlates with the intensity, rather than the duration of each seizure [3]. Absence seizures in patients with JME are usually associated with 3 Hz spikeand-wave activity, which is sometimes preceded by 4-6 Hz polyspike-and-wave discharg‐ es that decrease in frequency to 3 Hz as the patient loses consciousness. These spike-andwave discharges are usually shorter in duration than those observed in childhood and juvenile absence epilepsies [61].

#### Juvenile Myoclonic Epilepsy — A Maturation Syndrome Coming of Age http://dx.doi.org/10.5772/58372 83


Legend: The discharge is originally 5 Hz, slowing down to 3 Hz. No clinical signs were noted.

**Figure 1.** Generalized Interictal Epileptic Discharge in Drowsiness

the basis of large cortical blood flow increases even before the appearance of hyperventilationinduced generalized spike-and-wave discharges in absence epilepsy [47, 48]. Given the variety of aforementioned triggers, it is expected that reflex seizures are provoked by stimulation of different sensory cortices, such as the primary somatosensory cortex, primary auditory cortex, or related association areas. For instance, somatosensory or cognitively evoked seizures are evoked in physiologically activated cortical areas that overlap with hyperexcitable cortices

Given the rate of reflex seizures in JME, it is clinically indicated to conduct a sleep-deprived EEG with activation procedures, such as photic stimulation and hyperventilation. Although a positive EEG highly supports a clinical diagnosis of JME, a negative EEG does not definitively rule out the diagnosis. When examining EEG results, there are typical EEG patterns that occur

**Interictal Pattern**: The interictal background activity is usually considered normal in patients with JME [4, 7, 50, 51]. However, Genton et al. [52] found that routine EEGs were normal in 27% of cases and misleading or nonspecific in 20% of cases; although, 54% showed generalized interictal epileptic discharges (IEDs). Other studies have reported percentages of abnormali‐ ties, consisting of generalized polyspike-and-wave or spike-and-wave complexes in 75 to 85% of probands [9, 53, 54]. The typical interictal EEG finding in JME is a 3.5-6 Hz spike-and polyspike-and-wave pattern with a frontocentral predominance that lasts up to 15 to 20 seconds [50, 55]. While a pattern of more prolonged IEDs with a 2-3 Hz frequency is not specific to patients with absence seizures, there may be an increased risk of clinical absence seizures if a 3Hz spike-and-wave interictal pattern is noted. IEDs are primarily detected during drowsi‐ ness and or sleep onset, or upon awakening [56, 57] (see Figures 1-3), during ILS (photopar‐ oxysmal response) (see Figure 4), and hyperventilation. They may be associated with myoclonic seizures or GTCS (photoconvulsive response). Family members of JME are often affected or carry the abnormal EEG traits, as 80% percent of symptomatic siblings and 6% of asymptomatic siblings have diffuse 4-6 Hz spike-and-wave complexes [58]. Photoparoxysmal EEG response is another heritable trait of JME, and can be found in 20 to 60% of near relatives

**Ictal Pattern**: The ictal EEG during myoclonic jerks (see Figure 5) typically reveals a sudden onset of brief (less than 0.5 sec) bursts of 10-16 Hz polyspike-and-wave discharges often followed by 1-3 Hz slow waves [60]. The number of spike-and-wave complexes ranges from 5 to 20 per discharge and correlates with the intensity, rather than the duration of each seizure [3]. Absence seizures in patients with JME are usually associated with 3 Hz spikeand-wave activity, which is sometimes preceded by 4-6 Hz polyspike-and-wave discharg‐ es that decrease in frequency to 3 Hz as the patient loses consciousness. These spike-andwave discharges are usually shorter in duration than those observed in childhood and

giving rise to ictal or interictal epileptic discharges [49].

**6. Electrographic findings**

82 Epilepsy Topics

in JME that should be considered.

of probands [59].

juvenile absence epilepsies [61].

**Figure 2.** Generalized Polyspikes Activated in Stage II Sleep


Legend: Note movement artifact on ECG channel.

Accurate identification of EEG patterns is crucial to avoid misdiagnosis of JME (e.g., misiden‐ tified as a partial seizure). As an illustration, some researchers have reported that clinical and EEG asymmetries can lead to a delay in the diagnosis of JME by an average of 2 years [62]. Thus, it is important to understand that although JME is classified as a GGE, myoclonic seizures and GTCS may be associated with a combination of focal clinical and/or EEG findings. However, focal abnormalities are rarely found on routine neurological exams or MRI studies. Further, the challenges of making a diagnoses by electroclinical data is best exemplified by evidence of versive seizures or circling seizures that are associated with generalized discharges

Juvenile Myoclonic Epilepsy — A Maturation Syndrome Coming of Age

http://dx.doi.org/10.5772/58372

85

Data from JME patients in the UK found that 36.7% of EEGs had focal slow waves, spikes, and sharp waves as well as focal onset of a generalized discharge. In more than half of the JME patients, at least one EEG showed focal abnormalities [65]. 18 patients with JME were studied in Poland which demonstrated that 8 patients had focal abnormalities on EEG that previously led to misdiagnosis [66]. Further, a large study of JME patients in India [67] found asymmetrical clinical presentations in 17% of patients and focal EEG abnormalities were seen over 45% of patients. Of these patients, focal findings seen on EEG included amplitude asymmetry or a lateralized onset of generalized discharges in 45% and independent focal EEG abnormalities

**Figure 5.** Myoclonic Seizure

on EEG [63, 64].

**7. Focal features of EEG**

**Figure 3.** Fragmented Spikes in Slow-Wave Sleep


**Figure 4.** Photoparoxysmal Response

Juvenile Myoclonic Epilepsy — A Maturation Syndrome Coming of Age http://dx.doi.org/10.5772/58372 85

Legend: Note movement artifact on ECG channel.

**Figure 5.** Myoclonic Seizure

**Figure 3.** Fragmented Spikes in Slow-Wave Sleep

84 Epilepsy Topics

**Figure 4.** Photoparoxysmal Response

#### **7. Focal features of EEG**

Accurate identification of EEG patterns is crucial to avoid misdiagnosis of JME (e.g., misiden‐ tified as a partial seizure). As an illustration, some researchers have reported that clinical and EEG asymmetries can lead to a delay in the diagnosis of JME by an average of 2 years [62]. Thus, it is important to understand that although JME is classified as a GGE, myoclonic seizures and GTCS may be associated with a combination of focal clinical and/or EEG findings. However, focal abnormalities are rarely found on routine neurological exams or MRI studies. Further, the challenges of making a diagnoses by electroclinical data is best exemplified by evidence of versive seizures or circling seizures that are associated with generalized discharges on EEG [63, 64].

Data from JME patients in the UK found that 36.7% of EEGs had focal slow waves, spikes, and sharp waves as well as focal onset of a generalized discharge. In more than half of the JME patients, at least one EEG showed focal abnormalities [65]. 18 patients with JME were studied in Poland which demonstrated that 8 patients had focal abnormalities on EEG that previously led to misdiagnosis [66]. Further, a large study of JME patients in India [67] found asymmetrical clinical presentations in 17% of patients and focal EEG abnormalities were seen over 45% of patients. Of these patients, focal findings seen on EEG included amplitude asymmetry or a lateralized onset of generalized discharges in 45% and independent focal EEG abnormalities in 33% of patients. A video-EEG study from the Cleveland Clinic Foundation [68] reported that of the patients with only JME, 54% exhibited either focal semiologic or electroencephalo‐ graphic features or a combination thereof. Focal myoclonic seizures (i.e., unilateral myoclonic jerks, version, and asymmetric limb posturing) were recorded in six patients whose ictal EEG showed a generalized seizure pattern. Two patients had lateralized upper extremity extensor posturing ("sign of four") with evolution of a GTCS. They also reported that one patient had primary GTCS presenting after successful resection of a parietal tumor, and two patients had temporal lobe epilepsy in addition to JME. Taken together, the findings of focal or multifocal clinical or EEG abnormalities suggest the possibility of a symptomatic etiology, and may reflect an underlying genetically mediated neurodevelopmental disorder.

possibility of remission upon discontinuation of antiepileptic medications. In patients with CAE, the development of GTCS or myoclonic seizures may indicate a poorer prognosis, and

Juvenile Myoclonic Epilepsy — A Maturation Syndrome Coming of Age

http://dx.doi.org/10.5772/58372

87

**Juvenile Absence Epilepsy (JAE)**: Similar to JME, JAE is a GGE that mainly presents during adolescence, usually at the age of 10 to 17 years. Although the central semiology is signified by absence seizures, there is often less impairment of consciousness than in CAE, and the seizures do not tend to occur in clusters. On the other hand, GTCS seizures are more common (reported in 80% of patients) with JAE, similar to JME. While some authors have contended that myoclonic jerks do not occur in JAE [32], the ILAE has included them in the definition of JAE [6]. As such, 15 to 25% of JAE patients have myoclonic jerks-they are infrequent, mild, and of random temporal distribution [74]. In comparison to JME, myoclonic movements usually

**Myoclonic Absence**: This rare epilepsy syndrome begins in preadolescence and is character‐ ized by prolonged absence seizures lasting 10-60 seconds in duration with rhythmical bilateral myoclonias affecting the extremities, often synchronously with 3 Hz polyspike-and-wave discharges [75]. This syndrome is associated with cognitive dysfunction and is medically

**Generalized Tonic-Clonic Seizures Upon Awakening**: The onset of this epilepsy syndrome is also in adolescence [6]. As in JME, the GTCS tend to occur upon awakening. There is still debate on whether this syndrome should be classified with JME, due to electroclinical similarities and possible genetic associations [76]. However, patients with GTCS Upon

**Progressive Myoclonic Epilepsy (PME)**: Progressive myoclonic epilepsies are a group of genetic disorders which are characterized by myoclonus, seizures, and advancing cognitive and neurological decline [77]. Examples of PMEs include Lafora disease, neuronal ceroid lipofuscinosis, Unverricht Lundborg disease, myoclonic epilepsy with ragged red fibers (MERRF), Gaucher disease, and dentatorubral pallidoluysian atrophy. They can start in childhood, adolescence, or young adulthood and tend to be associated with multiple neuro‐ logical and cognitive deficits. The myoclonus associated with these disorders can be induced by action or stimulation and tends to be multifocal. ILS tends to activate myoclonus at frequencies below 6 Hz. These diseases are associated with progressive neurological deterio‐ ration along with worsening seizure control. The diagnosis is more difficult to determine,

primarily during the early stages of the disease when seizure severity is limited [78].

Response to treatment is drug-specific in the case of JME. Valproic acid (VPA) is considered the first line of therapy in patients with GGE, especially JME; however, it is not FDA approved for this condition [79, 80]. Rather than randomized control trials, evidence supporting the choice of VPA has primarily come from observational studies. While VPA is a broad spectrum

increases the probability that such patients will develop JME.

occur in the afternoon rather than in the morning after awakening.

Awakening do not exhibit myoclonic or absence seizures.

refractory in a majority of the patients.

**9. Treatment of JME**

It has also been suggested that asymmetric EEG findings in such patients increase the proba‐ bility of decreased response to standard antiepileptic therapies. A group of Canadian re‐ searchers [69] presented data showing 90% of their JME patients had generalized epileptiform discharges on EEG and 57% of patients had asymmetric EEG abnormalities, including focal epileptic discharges or slowing. Of their patients, over 39% were medically refractory to at least one AED, and a poor treatment response was observed to a greater degree in patients with EEG asymmetries. Nevertheless, the clinical significance of focal clinical or EEG abnor‐ malities has not received much attention in the literature.

#### **8. Differential diagnosis**

**Doose Syndrome**: Myoclonic-astatic epilepsy was first described by Dr. Hermann Doose in 1970 [70]. The ILAE classified it formally as symptomatic generalized epilepsy, and it was renamed as epilepsy with myoclonic-atonic seizures. It is much less prevalent than JME (1 to 2% of all epilepsies) and has a younger onset age (7 months to 6 years, peak of 2 to 4 years) with a greater prevalence among males. Unlike JME, there is an association with cognitive impairment, which is severe in some cases of Doose Syndrome. This syndrome is characterized by symmetric myoclonic jerks followed by sudden atonia. During a myoclonic atonic seizure, the EEG consists of irregular generalized 2-3Hz or more spike-and polyspike-and-wave discharges. The atonic seizures coincide with subsequent slow waves. The interictal EEG background similarly consists of frequent 2-3 Hz generalized spike-and polyspike-and-wave discharges, contrasted 4-6 Hz discharges in JME. Up to 50% of patients may achieve seizure freedom and continue to have normal development, but the remainder may have severe impairments in cognitive functioning, behavioral problems, and ataxia [71].

**Childhood Absence Epilepsy (CAE)**: CAE is a type of GGE that has an age of onset of 4 to 10 years with a peak between 5 to 7 years [72]. The characteristic seizure type is an absence seizure during which there is brief loss of consciousness for a few seconds with associated behavioral arrest, eyelid myoclonia, and rare hand and facial automatisms [73]. They may occur multiple times a day. While absence seizures may also occur in JME, their frequency and duration is far less than in CAE. The interictal EEG of CAE is characterized generalized spike-and polyspikeand-wave discharges of 3Hz frequency [6]. Prognosis for CAE is considered excellent, with possibility of remission upon discontinuation of antiepileptic medications. In patients with CAE, the development of GTCS or myoclonic seizures may indicate a poorer prognosis, and increases the probability that such patients will develop JME.

**Juvenile Absence Epilepsy (JAE)**: Similar to JME, JAE is a GGE that mainly presents during adolescence, usually at the age of 10 to 17 years. Although the central semiology is signified by absence seizures, there is often less impairment of consciousness than in CAE, and the seizures do not tend to occur in clusters. On the other hand, GTCS seizures are more common (reported in 80% of patients) with JAE, similar to JME. While some authors have contended that myoclonic jerks do not occur in JAE [32], the ILAE has included them in the definition of JAE [6]. As such, 15 to 25% of JAE patients have myoclonic jerks-they are infrequent, mild, and of random temporal distribution [74]. In comparison to JME, myoclonic movements usually occur in the afternoon rather than in the morning after awakening.

**Myoclonic Absence**: This rare epilepsy syndrome begins in preadolescence and is character‐ ized by prolonged absence seizures lasting 10-60 seconds in duration with rhythmical bilateral myoclonias affecting the extremities, often synchronously with 3 Hz polyspike-and-wave discharges [75]. This syndrome is associated with cognitive dysfunction and is medically refractory in a majority of the patients.

**Generalized Tonic-Clonic Seizures Upon Awakening**: The onset of this epilepsy syndrome is also in adolescence [6]. As in JME, the GTCS tend to occur upon awakening. There is still debate on whether this syndrome should be classified with JME, due to electroclinical similarities and possible genetic associations [76]. However, patients with GTCS Upon Awakening do not exhibit myoclonic or absence seizures.

**Progressive Myoclonic Epilepsy (PME)**: Progressive myoclonic epilepsies are a group of genetic disorders which are characterized by myoclonus, seizures, and advancing cognitive and neurological decline [77]. Examples of PMEs include Lafora disease, neuronal ceroid lipofuscinosis, Unverricht Lundborg disease, myoclonic epilepsy with ragged red fibers (MERRF), Gaucher disease, and dentatorubral pallidoluysian atrophy. They can start in childhood, adolescence, or young adulthood and tend to be associated with multiple neuro‐ logical and cognitive deficits. The myoclonus associated with these disorders can be induced by action or stimulation and tends to be multifocal. ILS tends to activate myoclonus at frequencies below 6 Hz. These diseases are associated with progressive neurological deterio‐ ration along with worsening seizure control. The diagnosis is more difficult to determine, primarily during the early stages of the disease when seizure severity is limited [78].

#### **9. Treatment of JME**

in 33% of patients. A video-EEG study from the Cleveland Clinic Foundation [68] reported that of the patients with only JME, 54% exhibited either focal semiologic or electroencephalo‐ graphic features or a combination thereof. Focal myoclonic seizures (i.e., unilateral myoclonic jerks, version, and asymmetric limb posturing) were recorded in six patients whose ictal EEG showed a generalized seizure pattern. Two patients had lateralized upper extremity extensor posturing ("sign of four") with evolution of a GTCS. They also reported that one patient had primary GTCS presenting after successful resection of a parietal tumor, and two patients had temporal lobe epilepsy in addition to JME. Taken together, the findings of focal or multifocal clinical or EEG abnormalities suggest the possibility of a symptomatic etiology, and may reflect

It has also been suggested that asymmetric EEG findings in such patients increase the proba‐ bility of decreased response to standard antiepileptic therapies. A group of Canadian re‐ searchers [69] presented data showing 90% of their JME patients had generalized epileptiform discharges on EEG and 57% of patients had asymmetric EEG abnormalities, including focal epileptic discharges or slowing. Of their patients, over 39% were medically refractory to at least one AED, and a poor treatment response was observed to a greater degree in patients with EEG asymmetries. Nevertheless, the clinical significance of focal clinical or EEG abnor‐

**Doose Syndrome**: Myoclonic-astatic epilepsy was first described by Dr. Hermann Doose in 1970 [70]. The ILAE classified it formally as symptomatic generalized epilepsy, and it was renamed as epilepsy with myoclonic-atonic seizures. It is much less prevalent than JME (1 to 2% of all epilepsies) and has a younger onset age (7 months to 6 years, peak of 2 to 4 years) with a greater prevalence among males. Unlike JME, there is an association with cognitive impairment, which is severe in some cases of Doose Syndrome. This syndrome is characterized by symmetric myoclonic jerks followed by sudden atonia. During a myoclonic atonic seizure, the EEG consists of irregular generalized 2-3Hz or more spike-and polyspike-and-wave discharges. The atonic seizures coincide with subsequent slow waves. The interictal EEG background similarly consists of frequent 2-3 Hz generalized spike-and polyspike-and-wave discharges, contrasted 4-6 Hz discharges in JME. Up to 50% of patients may achieve seizure freedom and continue to have normal development, but the remainder may have severe

**Childhood Absence Epilepsy (CAE)**: CAE is a type of GGE that has an age of onset of 4 to 10 years with a peak between 5 to 7 years [72]. The characteristic seizure type is an absence seizure during which there is brief loss of consciousness for a few seconds with associated behavioral arrest, eyelid myoclonia, and rare hand and facial automatisms [73]. They may occur multiple times a day. While absence seizures may also occur in JME, their frequency and duration is far less than in CAE. The interictal EEG of CAE is characterized generalized spike-and polyspikeand-wave discharges of 3Hz frequency [6]. Prognosis for CAE is considered excellent, with

impairments in cognitive functioning, behavioral problems, and ataxia [71].

an underlying genetically mediated neurodevelopmental disorder.

malities has not received much attention in the literature.

**8. Differential diagnosis**

86 Epilepsy Topics

Response to treatment is drug-specific in the case of JME. Valproic acid (VPA) is considered the first line of therapy in patients with GGE, especially JME; however, it is not FDA approved for this condition [79, 80]. Rather than randomized control trials, evidence supporting the choice of VPA has primarily come from observational studies. While VPA is a broad spectrum antiepileptic medication, it is very effective in 80% of patients [81]. Nonetheless, even in seizure-free patients, discontinuation of VPA leads to a very high rate of relapses [9]. A chart review from Duke University [82] of 33 JME patients identified resistance to VPA in 30%. The VPA refractory patients had a higher prevalence of EEG asymmetries (40% vs. 10%), atypical seizure characteristics including auras and postictal confusion (30% vs. 4%), and intellectual deficiency (20% vs. 0%). Clinical characteristics combined with EEG data may help in predict‐ ing which JME patients will respond favorably to VPA.

Clonazepam (CLN) was one of the first medications to be approved for myoclonic seizures and GTCS. While VPA is as an effective treatment for most patients with JME, one study showed [9] additional efficacy with CLN adjunctive therapy to control myoclonic seizures. However, CLN monotherapy did not consistently prevent GTCS. The authors concluded that adding CLN may allow the dose of VPA to be reduced in patients demonstrating dosedependent VPA side effects. The addition of CLN can also be utilized in combination with

Juvenile Myoclonic Epilepsy — A Maturation Syndrome Coming of Age

http://dx.doi.org/10.5772/58372

89

Levetiracetam (LVT) was more recently approved for the treatment of myoclonic seizures associated with JME. Similar to CLN, it can reduce the myoclonic seizures and GTCS [93], possibly with even better long-term efficacy and fewer side effects. Recently, several studies found LVT to be effective in all types of seizures in JME as concluded in a randomized double-blinded placebo controlled trial [94, 95]. Hence, LVT is another worthy alternative to VPA, especially in women of childbearing age. Other medications that may be effec‐ tive in JME include zonisamide (ZNS), felbamate (FBM), lacosamide (LCM), and cloba‐ zam (CLB); however, systematic trials of those medications are lacking. Nonetheless, a retrospective chart review demonstrated that ZNS monotherapy was effective and well-

In contrast to effective AED treatments, there are a number of antiepileptic medications that may aggravate certain seizure types associated with JME, particularly myoclonic and absence seizures. Among commonly prescribed anticonvulsants, carbamazepine appears to have the strongest aggravating potential for myoclonic seizures, whereas the aggravating effect of phenytoin (PHT) is less prominent [97]. A retrospective study of patients with GGE, showed that oxcarbazepine (OXC) also aggravated myoclonic and absence seizures, and increased interictal epileptic discharges on EEG [98]. A retrospective study of adult GGE patients who developed video-EEG documented status epilepticus were all found to be taking either CBZ, PHT, vigabatrin (VGB), or gabapentin (GBP). Potential precipitating factors included dose increase of CBZ or PHT, or the initiation of CBZ, VGB, or GBP, as well as a decrease of phenobarbital (PB) dosage. Withdrawal of the aggravating agents and adjustment to medica‐ tion resulted in full seizure control in that study. Therefore, it is important to make the appropriate diagnosis and medication selection to avoid potentially worsening certain seizure types. It is also important to consider the age, gender, comorbid conditions, and drug inter‐

As with a number of other chronic medical conditions (i.e., diabetes, hypertension, obesity), lifestyle factors have an important role to play in managing epilepsy disease burden. To illustrate the problems of how psychosocial factors influence the treatment of JME, Baykan et al. [30] followed 48 patients with JME for a mean of 20 years and found that 16.7% of patients had "pseudo-resistance" due to medication non-adherence and lifestyle related issues. Sharpe and Buchanan [99] studied 36 patients with JME at a tertiary referral hospital

actions when initiating antiepileptic therapy in JME patients.

other antiepileptic medications as well.

tolerated in JME patients [96].

**10. Lifestyle modifications**

More recent support of VPA's relative efficacy over other antiepileptic medications was demonstrated The Standard and New Anti-Epileptic Drugs trials (SANAD) [83]. SANAD was a concurrent pragmatic parallel group, open-label, randomized trial which examined seizure control, tolerability, quality of life, and economic outcomes of standard antiepileptic medica‐ tions used for GGE. VPA was compared with lamotrigine (LTG) and topiramate (TPM), and was significantly better for time to treatment failure (i.e., seizure control, side effects, addition of another AED). For time to 12-month remission, VPA was significantly better than LTG for GGE. Overall, VPA was better tolerated than TPM and more efficacious than LTG. In children with GGE, a retrospective observational study [84] examined the long-term effectiveness of LTG compared to VPA monotherapy in newly diagnosed patients. After 12 months of treatment, 69% patients continued LTG compared to 89% that adhered to their VPA regimen; after 24 months rates were 57% and 83%, respectively. Valproate showed equal efficacy in all GGE syndromes, whereas LTG showed better efficacy in CAE and JAE syndromes, than in JME.

Although VPA can be used safely by a number of patients, including those with comorbid psychiatric disease or underlying psychiatric vulnerability [85], there are some limitations. In particular, there is a high risk of teratogenicity as well as delayed cognitive development in children (pregnancy class D) [86, 87]. Exposure to VPA in utero has been associated with maladaptive behavior and decreased socialization skills in childhood [88, 89] as well. In cases where VPA is used during pregnancy, either because of unplanned pregnancy or because alternative treatment options of equivalent efficacy are unavailable, appropriate counseling, precautionary measures, and monitoring should be provided. Although LTG may have a lower efficacy, and has even been reported to increase myoclonic seizures in some patients, it is a popular alternative to VPA in women in childbearing age for the treatment of JME.

LTG and TPM have FDA approval for primary GTCS in the United States. Hence, more studies, albeit with small numbers of participants, have been conducted with those agents. An openlabel study [90], designed to evaluate LTG monotherapy as a possible alternative in patients with JME who previously failed VPA, found LTG to be as effective and better tolerated compared with VPA. A small (*N*=28) randomized, controlled trial [91] compared TPM and VPA in adolescents and adults during a 12-week maintenance period, which resulted in 67% seizure freedom in the TPM group and 57% in the VPA group. The researchers concluded that TPM may be an effective, well-tolerated alternative to VPA as the groups had similar rates of adverse effects. Similarly, a prospective, open-label, randomized observational study in Korea [92] compared the efficacy and tolerability between TPM and VPA in patients with JME and did not find differences in efficacy, although the side effect profile of TPM was more favorable. Clonazepam (CLN) was one of the first medications to be approved for myoclonic seizures and GTCS. While VPA is as an effective treatment for most patients with JME, one study showed [9] additional efficacy with CLN adjunctive therapy to control myoclonic seizures. However, CLN monotherapy did not consistently prevent GTCS. The authors concluded that adding CLN may allow the dose of VPA to be reduced in patients demonstrating dosedependent VPA side effects. The addition of CLN can also be utilized in combination with other antiepileptic medications as well.

Levetiracetam (LVT) was more recently approved for the treatment of myoclonic seizures associated with JME. Similar to CLN, it can reduce the myoclonic seizures and GTCS [93], possibly with even better long-term efficacy and fewer side effects. Recently, several studies found LVT to be effective in all types of seizures in JME as concluded in a randomized double-blinded placebo controlled trial [94, 95]. Hence, LVT is another worthy alternative to VPA, especially in women of childbearing age. Other medications that may be effec‐ tive in JME include zonisamide (ZNS), felbamate (FBM), lacosamide (LCM), and cloba‐ zam (CLB); however, systematic trials of those medications are lacking. Nonetheless, a retrospective chart review demonstrated that ZNS monotherapy was effective and welltolerated in JME patients [96].

In contrast to effective AED treatments, there are a number of antiepileptic medications that may aggravate certain seizure types associated with JME, particularly myoclonic and absence seizures. Among commonly prescribed anticonvulsants, carbamazepine appears to have the strongest aggravating potential for myoclonic seizures, whereas the aggravating effect of phenytoin (PHT) is less prominent [97]. A retrospective study of patients with GGE, showed that oxcarbazepine (OXC) also aggravated myoclonic and absence seizures, and increased interictal epileptic discharges on EEG [98]. A retrospective study of adult GGE patients who developed video-EEG documented status epilepticus were all found to be taking either CBZ, PHT, vigabatrin (VGB), or gabapentin (GBP). Potential precipitating factors included dose increase of CBZ or PHT, or the initiation of CBZ, VGB, or GBP, as well as a decrease of phenobarbital (PB) dosage. Withdrawal of the aggravating agents and adjustment to medica‐ tion resulted in full seizure control in that study. Therefore, it is important to make the appropriate diagnosis and medication selection to avoid potentially worsening certain seizure types. It is also important to consider the age, gender, comorbid conditions, and drug inter‐ actions when initiating antiepileptic therapy in JME patients.

#### **10. Lifestyle modifications**

antiepileptic medication, it is very effective in 80% of patients [81]. Nonetheless, even in seizure-free patients, discontinuation of VPA leads to a very high rate of relapses [9]. A chart review from Duke University [82] of 33 JME patients identified resistance to VPA in 30%. The VPA refractory patients had a higher prevalence of EEG asymmetries (40% vs. 10%), atypical seizure characteristics including auras and postictal confusion (30% vs. 4%), and intellectual deficiency (20% vs. 0%). Clinical characteristics combined with EEG data may help in predict‐

More recent support of VPA's relative efficacy over other antiepileptic medications was demonstrated The Standard and New Anti-Epileptic Drugs trials (SANAD) [83]. SANAD was a concurrent pragmatic parallel group, open-label, randomized trial which examined seizure control, tolerability, quality of life, and economic outcomes of standard antiepileptic medica‐ tions used for GGE. VPA was compared with lamotrigine (LTG) and topiramate (TPM), and was significantly better for time to treatment failure (i.e., seizure control, side effects, addition of another AED). For time to 12-month remission, VPA was significantly better than LTG for GGE. Overall, VPA was better tolerated than TPM and more efficacious than LTG. In children with GGE, a retrospective observational study [84] examined the long-term effectiveness of LTG compared to VPA monotherapy in newly diagnosed patients. After 12 months of treatment, 69% patients continued LTG compared to 89% that adhered to their VPA regimen; after 24 months rates were 57% and 83%, respectively. Valproate showed equal efficacy in all GGE syndromes, whereas LTG showed better efficacy in CAE and JAE syndromes, than in

Although VPA can be used safely by a number of patients, including those with comorbid psychiatric disease or underlying psychiatric vulnerability [85], there are some limitations. In particular, there is a high risk of teratogenicity as well as delayed cognitive development in children (pregnancy class D) [86, 87]. Exposure to VPA in utero has been associated with maladaptive behavior and decreased socialization skills in childhood [88, 89] as well. In cases where VPA is used during pregnancy, either because of unplanned pregnancy or because alternative treatment options of equivalent efficacy are unavailable, appropriate counseling, precautionary measures, and monitoring should be provided. Although LTG may have a lower efficacy, and has even been reported to increase myoclonic seizures in some patients, it is a

popular alternative to VPA in women in childbearing age for the treatment of JME.

LTG and TPM have FDA approval for primary GTCS in the United States. Hence, more studies, albeit with small numbers of participants, have been conducted with those agents. An openlabel study [90], designed to evaluate LTG monotherapy as a possible alternative in patients with JME who previously failed VPA, found LTG to be as effective and better tolerated compared with VPA. A small (*N*=28) randomized, controlled trial [91] compared TPM and VPA in adolescents and adults during a 12-week maintenance period, which resulted in 67% seizure freedom in the TPM group and 57% in the VPA group. The researchers concluded that TPM may be an effective, well-tolerated alternative to VPA as the groups had similar rates of adverse effects. Similarly, a prospective, open-label, randomized observational study in Korea [92] compared the efficacy and tolerability between TPM and VPA in patients with JME and did not find differences in efficacy, although the side effect profile of TPM was more favorable.

ing which JME patients will respond favorably to VPA.

JME.

88 Epilepsy Topics

As with a number of other chronic medical conditions (i.e., diabetes, hypertension, obesity), lifestyle factors have an important role to play in managing epilepsy disease burden. To illustrate the problems of how psychosocial factors influence the treatment of JME, Baykan et al. [30] followed 48 patients with JME for a mean of 20 years and found that 16.7% of patients had "pseudo-resistance" due to medication non-adherence and lifestyle related issues. Sharpe and Buchanan [99] studied 36 patients with JME at a tertiary referral hospital and found that most patients with poor seizure control had provoked seizures only, which prompted the authors to emphasize the importance of lifestyle changes in managing seizures in JME.

**12. Non-medical treatments**

options.

myoclonic seizures.

Alternative surgical therapies do not represent standard of care and are not approved for JME. However, in medically-refractory patients, clinicians often employ additional treatment

Juvenile Myoclonic Epilepsy — A Maturation Syndrome Coming of Age

http://dx.doi.org/10.5772/58372

91

**Vagal Nerve Stimulator**: While vagal nerve stimulators (VNS) are approved for the treatment of medically refractory partial seizures, its use in JME has not been well documented. None‐ theless, the most recent AAN guidelines on VNS for epilepsy [112] asserted that VNS was effective in children with medically-refractory GGE. A greater than 50% reduction in seizure frequency was reported in 55% of the 470 children with partial or generalized epilepsy in the reviewed literature (13 class III studies). Based on those findings, the authors concluded that VNS could be considered adjunctive treatment. No such recommendations have been made

One of the few JME specific studies of VNS [113] indicated that adjunctive therapy with VNS might be considered a favorable option for treatment of refractory cases of GGE, and may be the only alternative to refractory JME. A total of 12 cases with GGE and VNS implantation were followed for a mean of 23 months. It was found that the rate of seizure total reduction was 61%: a 62% reduction of GTCS, 58% reduction in absences, and 40% reduction of myoclonic seizures. Five of the seven patients in that study who had a JME diagnoses were responders, two of which became seizure free. One patient in that study was diagnosed with JME and failed conventional treatment that included 8 AEDs and the ketogenic diet. Subsequent to VNS implantation, she had >75% reduction in GTCS along with >50% reduction in absence and

**Corpus Callosotomy**: There are even fewer studies that have examined the effectiveness of corpus callosotomy for medically refractory cases of JME. A case series (*N*=11) from Brazil [114] included GGE patients who underwent extensive one-stage callosal transections. At least 75% reduction in frequency of GTCS was noted in all patients; three patients had complete remission of absences and the other patients had a >90% reduction in the frequency of absence seizures. Only one patient who had myoclonic seizures prior to surgery remitted completely after surgery. Postoperative EEG recordings showed disruption of bilateral synchrony of interictal epileptic discharges in all patients, with minimal neurological deficits. While the authors reported that the patients only showed a decline of 4 points in mean FSIQ standard score, a more detailed examination of the patients' psychosocial status and cognitive func‐ tioning was not reported. A Japanese case report [115] of medically refractory JME showed that an anterior corpus callosotomy resulted in a desynchronization of generalized spike-andwave discharges and a complete resolution of myoclonic seizures. An anterior corpus callos‐ otomy may be effective for seizure reduction in some cases of refractory GGE due to the

regarding adult treatment as more trials are needed in that population.

disruption of interhemispheric synchronization [116].

Thus, perhaps the most obvious lifestyle factors worth noting in terms of JME pertain to patients' exposure to stimuli s associated with increased seizure frequency (i.e., fatigue, stress, sleep deprivation, alcohol intake, drug use, photic stimulation, stress, AED withdrawal/noncompliance) [9, 100-103]. From a disease management perspective, behavioral medicine strategies that are promoted by integrated, multidisciplinary teams may prove useful [104]. Elements of such programs often include steps to improve medication adherence, treat psychological problems and manage stress, implement sleep hygiene interventions, and develop methods to cope with exposure to possible seizure triggers [104-106]. Moreover, administration of psychoeducation should be considered a minimum standard of care by providing patients with information regarding identifiable triggers, risks of sleep deprivation, possible problems with exposure to stroboscopic flashes, relationship of alcohol intake and seizure control, as well as ways to increase AED adherence to prevent further seizures [107]. Although, some authors [103] have suggested that it is more efficient to use higher doses of AEDs in the initial treatment of JME, rather than implement behavior change strategies, systematic study of the effects of multidisciplinary treatment on lifestyle factors in epilepsy has remained an understudied area. However, initial studies in JME have suggested that there may be a desirable effect on seizure control, and likely improvements in quality of life. If some of these measures are put into place, it is likely that the number of patients with refractory seizures will be reduced [108, 109].

#### **11. Cognitive and behavioral tasks**

There have been studies suggesting that cognitive tasks can provoke or inhibit of epileptiform discharges in patients with JME [110]. These uncommon precipitating factors, such as mental and motor hand tasks, may be under-recognized in JME [35]. A Japanese study involving 480 patients with epilepsy tested the effect of cognitive tasks on EEG (i.e., neuropsychological EEG activation) consisting of reading, speaking, writing, written calculation, mental calculation, and spatial construction. They found that these cognitive tasks have an inhibitory effect on EEG discharges in the majority of epilepsy patients (64%), although they have a provocative effect in other patients (8%). The seizures caused by the provocative effect of these tasks were found to be precipitated by action programming or thinking activity (linguistic and praxic) [111]. Following this, Beniczky et al. [38] studied 60 patients with JME and found that the provocative effect of the cognitive tasks is task-specific, whereas the inhibitory effect seems to be related to cognitive activation in general. Another study involving 76 JME patients had similar results, suggesting that the inhibitory effect of these tasks support non-pharmacologic therapeutic interventions in JME [37].

#### **12. Non-medical treatments**

and found that most patients with poor seizure control had provoked seizures only, which prompted the authors to emphasize the importance of lifestyle changes in managing seizures

Thus, perhaps the most obvious lifestyle factors worth noting in terms of JME pertain to patients' exposure to stimuli s associated with increased seizure frequency (i.e., fatigue, stress, sleep deprivation, alcohol intake, drug use, photic stimulation, stress, AED withdrawal/noncompliance) [9, 100-103]. From a disease management perspective, behavioral medicine strategies that are promoted by integrated, multidisciplinary teams may prove useful [104]. Elements of such programs often include steps to improve medication adherence, treat psychological problems and manage stress, implement sleep hygiene interventions, and develop methods to cope with exposure to possible seizure triggers [104-106]. Moreover, administration of psychoeducation should be considered a minimum standard of care by providing patients with information regarding identifiable triggers, risks of sleep deprivation, possible problems with exposure to stroboscopic flashes, relationship of alcohol intake and seizure control, as well as ways to increase AED adherence to prevent further seizures [107]. Although, some authors [103] have suggested that it is more efficient to use higher doses of AEDs in the initial treatment of JME, rather than implement behavior change strategies, systematic study of the effects of multidisciplinary treatment on lifestyle factors in epilepsy has remained an understudied area. However, initial studies in JME have suggested that there may be a desirable effect on seizure control, and likely improvements in quality of life. If some of these measures are put into place, it is likely that the number of patients with refractory

There have been studies suggesting that cognitive tasks can provoke or inhibit of epileptiform discharges in patients with JME [110]. These uncommon precipitating factors, such as mental and motor hand tasks, may be under-recognized in JME [35]. A Japanese study involving 480 patients with epilepsy tested the effect of cognitive tasks on EEG (i.e., neuropsychological EEG activation) consisting of reading, speaking, writing, written calculation, mental calculation, and spatial construction. They found that these cognitive tasks have an inhibitory effect on EEG discharges in the majority of epilepsy patients (64%), although they have a provocative effect in other patients (8%). The seizures caused by the provocative effect of these tasks were found to be precipitated by action programming or thinking activity (linguistic and praxic) [111]. Following this, Beniczky et al. [38] studied 60 patients with JME and found that the provocative effect of the cognitive tasks is task-specific, whereas the inhibitory effect seems to be related to cognitive activation in general. Another study involving 76 JME patients had similar results, suggesting that the inhibitory effect of these tasks support non-pharmacologic

in JME.

90 Epilepsy Topics

seizures will be reduced [108, 109].

**11. Cognitive and behavioral tasks**

therapeutic interventions in JME [37].

Alternative surgical therapies do not represent standard of care and are not approved for JME. However, in medically-refractory patients, clinicians often employ additional treatment options.

**Vagal Nerve Stimulator**: While vagal nerve stimulators (VNS) are approved for the treatment of medically refractory partial seizures, its use in JME has not been well documented. None‐ theless, the most recent AAN guidelines on VNS for epilepsy [112] asserted that VNS was effective in children with medically-refractory GGE. A greater than 50% reduction in seizure frequency was reported in 55% of the 470 children with partial or generalized epilepsy in the reviewed literature (13 class III studies). Based on those findings, the authors concluded that VNS could be considered adjunctive treatment. No such recommendations have been made regarding adult treatment as more trials are needed in that population.

One of the few JME specific studies of VNS [113] indicated that adjunctive therapy with VNS might be considered a favorable option for treatment of refractory cases of GGE, and may be the only alternative to refractory JME. A total of 12 cases with GGE and VNS implantation were followed for a mean of 23 months. It was found that the rate of seizure total reduction was 61%: a 62% reduction of GTCS, 58% reduction in absences, and 40% reduction of myoclonic seizures. Five of the seven patients in that study who had a JME diagnoses were responders, two of which became seizure free. One patient in that study was diagnosed with JME and failed conventional treatment that included 8 AEDs and the ketogenic diet. Subsequent to VNS implantation, she had >75% reduction in GTCS along with >50% reduction in absence and myoclonic seizures.

**Corpus Callosotomy**: There are even fewer studies that have examined the effectiveness of corpus callosotomy for medically refractory cases of JME. A case series (*N*=11) from Brazil [114] included GGE patients who underwent extensive one-stage callosal transections. At least 75% reduction in frequency of GTCS was noted in all patients; three patients had complete remission of absences and the other patients had a >90% reduction in the frequency of absence seizures. Only one patient who had myoclonic seizures prior to surgery remitted completely after surgery. Postoperative EEG recordings showed disruption of bilateral synchrony of interictal epileptic discharges in all patients, with minimal neurological deficits. While the authors reported that the patients only showed a decline of 4 points in mean FSIQ standard score, a more detailed examination of the patients' psychosocial status and cognitive func‐ tioning was not reported. A Japanese case report [115] of medically refractory JME showed that an anterior corpus callosotomy resulted in a desynchronization of generalized spike-andwave discharges and a complete resolution of myoclonic seizures. An anterior corpus callos‐ otomy may be effective for seizure reduction in some cases of refractory GGE due to the disruption of interhemispheric synchronization [116].

#### **13. Prognosis**

As with most GGE syndromes, JME responds well to an appropriate AED regimen, demon‐ strating a 70 to 80% response rate [117]. However, the long-term prognosis remains AEDdependent as there is an 80% recurrence risk after AED withdrawal. It is thought that CAE typically has a better prognosis and spontaneous remission that JME even when untreated, although this can be complicated by the development of myoclonus. In fact, a Canadian study [19] noted that 65% of children with CAE experienced remission after 10 to 18 years after their diagnosis, although 44% of patients who did not experience remission developed JME.

**14. Cognitive dysfunction**

linkage, family studies of cognition have surfaced as well.

Consideration of neuropsychological findings in JME has a brief history beginning with largely anecdotal reports suggesting patients displayed a combination of behaviors consistent with a "frontal syndrome" [3] in the context of normal intellectual functioning. As the following literature illustrates, the initial assertions that JME patients display a unitary "frontal syn‐ drome" are challenged by data from a number of novel empirical designs. For instance, consideration of non-frontal factors is considered germane in light of the involvement of multiple brain regions that extend beyond structural frontal anatomy in JME [123]. Several themes of investigation have surfaced regarding cognition including delineation of neuropsy‐ chological profiles in JME, group performance differences, disease specific and non-disease factors that may affect cognitive functioning, along with neuroanatomical and functional correlates of task performance. Because of the idiopathic nature of JME and its strong genetic

Juvenile Myoclonic Epilepsy — A Maturation Syndrome Coming of Age

http://dx.doi.org/10.5772/58372

93

Given that JME has been conceptualized as a disorder affecting frontal brain regions, some of the first empirically-driven studies focused on the relationship of JME and performance on tasks that theoretically require significant frontal lobe recruitment. In this vein, JME groups have been assessed by a number of clinical and basic research cognitive measures that task frontal systems. For instance, a small study showed that a group of adult JME patients (*n*=9) demonstrated an abnormal number of errors and omissions on a task that requires individuals to identify previously learned complex visual information [124]. Such match-to-sample measures often have immediate and delayed portions that task visual working memory systems, which an aspect of executive functioning. The magnitude of abnormality was generally not as large as performance by a group of patients with frontal lobe epilepsy (FLE; *n*=15), although it was typically lower than control participants' (*n*=14) proficiency. In a followup study with a similar task, researchers from that group [125] reported that adult JME patients (*n*=9) had fewer correct responses than controls (*n*=14) on the same delayed match to sample paradigm. The JME patients performed normally on the immediate match to sample tasks, which generally relates to attentional functioning, but slower psychomotor speed performance than controls. However, it is not clear if the same patient and control groups were used across both of the studies, which decreases the generalizability of the findings. Nevertheless, these two projects served as stringboards for subsequent applied and experimental designs.

While it is important that the latter studies showed commonalities between theoretically similar patient groups (i.e., groups with suspected frontal lobe abnormalities) in order to provide additional validity evidence for a "frontal" construct, it is also relevant to include data from groups that are theoretically divergent (i.e., controls, various non-frontal clinical samples). As such, neurocognitive performance of a mixed sample [126] of patients diagnosed with temporal lobe epilepsy (TLE; *n*=15) and JME (*n*=15) was contrasted on a number of tasks related to frontal lobe functions. JME patients had impairments on a wide range of cognitive measures, as defined as >1.5 standard deviations below the mean for the normative comparison group. Of the 15 JME patients used in their analyses, one patient showed no impairment on testing, while half of the remaining sample had impairments on <3 tests and the other half was

A long-term population-based study [14] that included 24 patients (majority women) who developed JME by 16 years of age were followed for up to 25 years after seizure onset. Eight patients (36%) developed convulsive status epilepticus and three patients (12.5%) had intractable seizures. In this cohort, 17% of patients had full remission of all seizure types and only myoclonic seizures persisted in 13%. Thus, about one-third of patients had remission of disabling seizures without the need for continuation of AEDs. Moreover, although JME rarely undergoes spontaneous remission, there may be an age effect that increases the probability of spontaneous remission as individuals reach the fourth decade of life. In another long-term study [30, 118] that followed 48 patients with JME (mean age of 40 years) for approximately 20 years, the authors indicated that seizure severity and myoclonic frequency seemed to change across the lifespan. For instance, myoclonia went into remission in the fifth decade of life in 54% of patients. Five patients discontinued AED treatment and six patients had lower AED dosage; 10 out of 11 of these patients did not relapse during the mean follow up of 8 years.

One of the longest outcome studies of JME [119] followed 31 patients for an average of 39 years. Over two-thirds became seizure free, a third of whom could be taken off AEDs altogether. Predictors of poor outcome were identified as presence of GTCS preceded by myoclonic seizures, longer duration of epilepsy, and the need for AED polytherapy. Predictors of seizure freedom included complete control of GTCS. Other factors may also influence remission as well. In a study of 32 JME patients in Japan, those with focal interictal epileptic discharges on EEG and or discharges activated by cognitive stimuli, had the least favorable outcome [120]. However, value of EEG in predicting outcome of JME remains controversial [121]. A study from Brazil [122] showed that patients with persistent seizures had an earlier age of onset, higher prevalence of personality disorders, and higher incidence of sensitivity to praxis-and verbally-induced ictal or interictal epileptic discharges.

Hence, it appears that a subgroup of patients can be identified as being at a higher risk of refractory seizures. JME patients with an earlier onset of epilepsy, GTCS evolving after a buildup of myoclonic seizures, focal EEG abnormalities or sensitivity to cognitive activation, and cognitive or behavioral problems may be more likely to remain refractory to antiepileptic medications. Ideally, the significance of these factors needs to be validated prospectively by a multicenter, multinational study.

#### **14. Cognitive dysfunction**

**13. Prognosis**

92 Epilepsy Topics

As with most GGE syndromes, JME responds well to an appropriate AED regimen, demon‐ strating a 70 to 80% response rate [117]. However, the long-term prognosis remains AEDdependent as there is an 80% recurrence risk after AED withdrawal. It is thought that CAE typically has a better prognosis and spontaneous remission that JME even when untreated, although this can be complicated by the development of myoclonus. In fact, a Canadian study [19] noted that 65% of children with CAE experienced remission after 10 to 18 years after their diagnosis, although 44% of patients who did not experience remission developed JME.

A long-term population-based study [14] that included 24 patients (majority women) who developed JME by 16 years of age were followed for up to 25 years after seizure onset. Eight patients (36%) developed convulsive status epilepticus and three patients (12.5%) had intractable seizures. In this cohort, 17% of patients had full remission of all seizure types and only myoclonic seizures persisted in 13%. Thus, about one-third of patients had remission of disabling seizures without the need for continuation of AEDs. Moreover, although JME rarely undergoes spontaneous remission, there may be an age effect that increases the probability of spontaneous remission as individuals reach the fourth decade of life. In another long-term study [30, 118] that followed 48 patients with JME (mean age of 40 years) for approximately 20 years, the authors indicated that seizure severity and myoclonic frequency seemed to change across the lifespan. For instance, myoclonia went into remission in the fifth decade of life in 54% of patients. Five patients discontinued AED treatment and six patients had lower AED dosage; 10 out of 11 of these patients did not relapse during the mean follow up of 8 years.

One of the longest outcome studies of JME [119] followed 31 patients for an average of 39 years. Over two-thirds became seizure free, a third of whom could be taken off AEDs altogether. Predictors of poor outcome were identified as presence of GTCS preceded by myoclonic seizures, longer duration of epilepsy, and the need for AED polytherapy. Predictors of seizure freedom included complete control of GTCS. Other factors may also influence remission as well. In a study of 32 JME patients in Japan, those with focal interictal epileptic discharges on EEG and or discharges activated by cognitive stimuli, had the least favorable outcome [120]. However, value of EEG in predicting outcome of JME remains controversial [121]. A study from Brazil [122] showed that patients with persistent seizures had an earlier age of onset, higher prevalence of personality disorders, and higher incidence of sensitivity to praxis-and

Hence, it appears that a subgroup of patients can be identified as being at a higher risk of refractory seizures. JME patients with an earlier onset of epilepsy, GTCS evolving after a buildup of myoclonic seizures, focal EEG abnormalities or sensitivity to cognitive activation, and cognitive or behavioral problems may be more likely to remain refractory to antiepileptic medications. Ideally, the significance of these factors needs to be validated prospectively by a

verbally-induced ictal or interictal epileptic discharges.

multicenter, multinational study.

Consideration of neuropsychological findings in JME has a brief history beginning with largely anecdotal reports suggesting patients displayed a combination of behaviors consistent with a "frontal syndrome" [3] in the context of normal intellectual functioning. As the following literature illustrates, the initial assertions that JME patients display a unitary "frontal syn‐ drome" are challenged by data from a number of novel empirical designs. For instance, consideration of non-frontal factors is considered germane in light of the involvement of multiple brain regions that extend beyond structural frontal anatomy in JME [123]. Several themes of investigation have surfaced regarding cognition including delineation of neuropsy‐ chological profiles in JME, group performance differences, disease specific and non-disease factors that may affect cognitive functioning, along with neuroanatomical and functional correlates of task performance. Because of the idiopathic nature of JME and its strong genetic linkage, family studies of cognition have surfaced as well.

Given that JME has been conceptualized as a disorder affecting frontal brain regions, some of the first empirically-driven studies focused on the relationship of JME and performance on tasks that theoretically require significant frontal lobe recruitment. In this vein, JME groups have been assessed by a number of clinical and basic research cognitive measures that task frontal systems. For instance, a small study showed that a group of adult JME patients (*n*=9) demonstrated an abnormal number of errors and omissions on a task that requires individuals to identify previously learned complex visual information [124]. Such match-to-sample measures often have immediate and delayed portions that task visual working memory systems, which an aspect of executive functioning. The magnitude of abnormality was generally not as large as performance by a group of patients with frontal lobe epilepsy (FLE; *n*=15), although it was typically lower than control participants' (*n*=14) proficiency. In a followup study with a similar task, researchers from that group [125] reported that adult JME patients (*n*=9) had fewer correct responses than controls (*n*=14) on the same delayed match to sample paradigm. The JME patients performed normally on the immediate match to sample tasks, which generally relates to attentional functioning, but slower psychomotor speed performance than controls. However, it is not clear if the same patient and control groups were used across both of the studies, which decreases the generalizability of the findings. Nevertheless, these two projects served as stringboards for subsequent applied and experimental designs.

While it is important that the latter studies showed commonalities between theoretically similar patient groups (i.e., groups with suspected frontal lobe abnormalities) in order to provide additional validity evidence for a "frontal" construct, it is also relevant to include data from groups that are theoretically divergent (i.e., controls, various non-frontal clinical samples). As such, neurocognitive performance of a mixed sample [126] of patients diagnosed with temporal lobe epilepsy (TLE; *n*=15) and JME (*n*=15) was contrasted on a number of tasks related to frontal lobe functions. JME patients had impairments on a wide range of cognitive measures, as defined as >1.5 standard deviations below the mean for the normative comparison group. Of the 15 JME patients used in their analyses, one patient showed no impairment on testing, while half of the remaining sample had impairments on <3 tests and the other half was impaired on >4 tests. Performance was most frequently impaired on a task that requires one to identify relationships between objects on a conceptual level, which was significantly discrepant from the TLE group. Additionally, patients were administered a task that requires individuals to quickly and sequentially alternate responses according to specific rules (i.e., Trailmaking Test Part B). On this task, the JME participants had lower scores than the TLE patients.

concept formation, psychomotor speed, and fine motor dexterity and speed. While visual inspection of the data suggested that JME patients performed lower than the group with absence seizures on a measure of language fluency, the group with absence seizures appeared to perform lower in all other domains. However, statistical group comparisons were not

Juvenile Myoclonic Epilepsy — A Maturation Syndrome Coming of Age

http://dx.doi.org/10.5772/58372

95

Just as there has been variability in finding group differences between patient samples, findings from studies contrasting JME patients from healthy controls have been mixed and contradictory. In addition to the few clinical comparison group designs noted above, a number of other studies have shown that JME patients have statistically significantly lower proficiency than controls across a host of measures including intelligence [134, 135], verbal IQ [135], working memory [135], digit span [136-139], sustained attention [134], processing speed [134, 135], Trailmaking Parts A and B [137-139], Trailmaking Part B [140], mental flexibility [134, 137, 141, 142], response inhibition [134, 143], Stroop Test [137, 139, 140], Stroop Interference [136], speeded color reading [143], verbal abstraction [141], concept formation [136, 137, 142], perseverations[140], clock and cube drawing [143], sematic fluency [136, 138, 140, 143-145], phonemic fluency [137, 138, 140, 141, 144, 145], naming [135, 141], verbal learning and memory [138, 140, 143], visual memory [140, 141], and prospective memory [136]. On the other hand, a number of studies have also failed to find group differences on tasks of IQ [138, 143, 144], verbal intelligence [141, 145], auditory working memory and attention span [141, 143-146], spatial span [144, 145], spatial working memory [147], psychomotor speed [146], motor speed [134], mental flexibility [146], Trailmaking Parts A and B [145], response inhibition[141, 144, 145], figural fluency [134, 145], semantic fluency [141, 146], phonemic fluency [146], reading [134], naming [143], language comprehension[144], line orientation [143], facial recognition [143], memory [134], verbal learning [141, 144, 145], visual memory [144, 145], and design learning [141]. Another more recent study [148] reported no difference between JME patients and controls on a range of neuropsychological measures. Beyond functioning on objective cognitive tasks, another avenue for research is to examine patients' perceptions of cognitive dysfunction. Such a design has shown the JME patients rate a higher level of self-reported

Given that the variability in findings across studies likely relates to a number of primary and secondary factors, the influence of the effects of such factors is also likely varied. For instance, one small study [128] showed that it may be important to consider the influence of seizure frequency on JME patients' abilities. In general, the JME patients in their sample did not perform differently from healthy controls on tasks of verbal attention, verbal working memory, or phonemic verbal fluency. On the other hand, controls consistently performed higher than JME patients on tasks related to psychomotor speed, cognitive flexibility, categorical verbal fluency, planning, along with abstraction and categorization. The patients who had been seizure free (*n*=11) for a year were also compared with patients who continued to have seizures (*n*=11). The groups showed no differences on any of the cognitive testing, although JME patients scored significantly lower on three of four indices of a decision-making task than controls; seizure status in the JME group was related to performance. However, when compared with controls, the patients who were not seizure-free also showed additional

conducted to differentiate any of the clinical groups from one another.

executive dysfunction than controls [144].

Additional studies have contrasted JME groups on test batteries that primarily include executive function tasks that have a relationship with frontal lobe involvement. Piazzini et al. [127] reported data from an adult sample of patients with JME (*n*=50), TLE (*n*=40), FLE (*n*=40), and controls (*n*=40). In that study, JME patients performed statistically significantly lower than TLE patients and controls, yet scored similarly to FLE on the Wisconsin Card Sorting Test and a semantic verbal fluency test. A group in Austria [128] published data from two studies [129, 130] that included a decision-making task and a number of neuropsychological variables in patients. Their results showed that JME patients performed similarly to mesial TLE patients on nearly all variables except they had lower semantic verbal fluency scores and slower psychomotor speed. The groups were otherwise similar on measures of verbal attention, verbal working memory, cognitive flexibility, planning, along with abstraction and categorization.

Other results [131] indicated that children with CAE (*n*=28), another GGE, largely performed worse than individuals with JME (*n*=11) on tasks of visual sustained attention, the Stroop Test, and Trailmaking Test. No differences were noted for verbal memory. Another group [132] sampled children and adolescents who had a similar level of normal intellectual functioning and were diagnosed with either recent-onset JME (*n*=20) or recent-onset benign childhood epilepsy with centrotemporal spikes (BCECTS; *n*=12). A sample of first-degree cousins (*n*=51) were also utilized as control participants. The researchers focused their cognitive assessment on objective measures of executive functioning that included subtests from the Delis-Kaplan Executive Function System (D-KEFS). They also examined behaviors that were subjectively rated by parents on the Behavior Rating Inventory of Executive Function (BRIEF). In their samples, the JME group performed significantly poorer than the control group on the D-KEFS Inhibition subtest. Parent report on the BRIEF indicated the JME group had more pathological ratings on the Behavioral Regulation and Metacognition scales than controls as well. However, there were no significant differences between the BCECTS and JME group on any measures. The latter study highlights that executive differences exist early in the disease course of JME, although the magnitude of executive dysfunction may not be larger than that in other genetic epilepsies.

There have also been reports [133] of cognitive functioning in mixed samples of children and adolescents with normal intellectual functioning diagnosed with GGE (i.e., JME & Absence) and genetic localization-related epilepsies (i.e., BECTS & non-BECTS). Although that study included 26 JME patients and a number of other patients with IPE, the JME group was compared only with the 72 healthy children, which indicated that the JME group performed below the mean of the control group in all cognitive domains (e.g., intelligence, academic, language, memory, executive functioning, fine motor dexterity and speed, cognitive process‐ ing speed). In particular, JME patients' lowest performance was in arithmetic, inhibition, concept formation, psychomotor speed, and fine motor dexterity and speed. While visual inspection of the data suggested that JME patients performed lower than the group with absence seizures on a measure of language fluency, the group with absence seizures appeared to perform lower in all other domains. However, statistical group comparisons were not conducted to differentiate any of the clinical groups from one another.

impaired on >4 tests. Performance was most frequently impaired on a task that requires one to identify relationships between objects on a conceptual level, which was significantly discrepant from the TLE group. Additionally, patients were administered a task that requires individuals to quickly and sequentially alternate responses according to specific rules (i.e., Trailmaking Test Part B). On this task, the JME participants had lower scores than the TLE

Additional studies have contrasted JME groups on test batteries that primarily include executive function tasks that have a relationship with frontal lobe involvement. Piazzini et al. [127] reported data from an adult sample of patients with JME (*n*=50), TLE (*n*=40), FLE (*n*=40), and controls (*n*=40). In that study, JME patients performed statistically significantly lower than TLE patients and controls, yet scored similarly to FLE on the Wisconsin Card Sorting Test and a semantic verbal fluency test. A group in Austria [128] published data from two studies [129, 130] that included a decision-making task and a number of neuropsychological variables in patients. Their results showed that JME patients performed similarly to mesial TLE patients on nearly all variables except they had lower semantic verbal fluency scores and slower psychomotor speed. The groups were otherwise similar on measures of verbal attention, verbal working memory, cognitive flexibility, planning, along with abstraction and categorization.

Other results [131] indicated that children with CAE (*n*=28), another GGE, largely performed worse than individuals with JME (*n*=11) on tasks of visual sustained attention, the Stroop Test, and Trailmaking Test. No differences were noted for verbal memory. Another group [132] sampled children and adolescents who had a similar level of normal intellectual functioning and were diagnosed with either recent-onset JME (*n*=20) or recent-onset benign childhood epilepsy with centrotemporal spikes (BCECTS; *n*=12). A sample of first-degree cousins (*n*=51) were also utilized as control participants. The researchers focused their cognitive assessment on objective measures of executive functioning that included subtests from the Delis-Kaplan Executive Function System (D-KEFS). They also examined behaviors that were subjectively rated by parents on the Behavior Rating Inventory of Executive Function (BRIEF). In their samples, the JME group performed significantly poorer than the control group on the D-KEFS Inhibition subtest. Parent report on the BRIEF indicated the JME group had more pathological ratings on the Behavioral Regulation and Metacognition scales than controls as well. However, there were no significant differences between the BCECTS and JME group on any measures. The latter study highlights that executive differences exist early in the disease course of JME, although the magnitude of executive dysfunction may not be larger than that in other genetic

There have also been reports [133] of cognitive functioning in mixed samples of children and adolescents with normal intellectual functioning diagnosed with GGE (i.e., JME & Absence) and genetic localization-related epilepsies (i.e., BECTS & non-BECTS). Although that study included 26 JME patients and a number of other patients with IPE, the JME group was compared only with the 72 healthy children, which indicated that the JME group performed below the mean of the control group in all cognitive domains (e.g., intelligence, academic, language, memory, executive functioning, fine motor dexterity and speed, cognitive process‐ ing speed). In particular, JME patients' lowest performance was in arithmetic, inhibition,

patients.

94 Epilepsy Topics

epilepsies.

Just as there has been variability in finding group differences between patient samples, findings from studies contrasting JME patients from healthy controls have been mixed and contradictory. In addition to the few clinical comparison group designs noted above, a number of other studies have shown that JME patients have statistically significantly lower proficiency than controls across a host of measures including intelligence [134, 135], verbal IQ [135], working memory [135], digit span [136-139], sustained attention [134], processing speed [134, 135], Trailmaking Parts A and B [137-139], Trailmaking Part B [140], mental flexibility [134, 137, 141, 142], response inhibition [134, 143], Stroop Test [137, 139, 140], Stroop Interference [136], speeded color reading [143], verbal abstraction [141], concept formation [136, 137, 142], perseverations[140], clock and cube drawing [143], sematic fluency [136, 138, 140, 143-145], phonemic fluency [137, 138, 140, 141, 144, 145], naming [135, 141], verbal learning and memory [138, 140, 143], visual memory [140, 141], and prospective memory [136]. On the other hand, a number of studies have also failed to find group differences on tasks of IQ [138, 143, 144], verbal intelligence [141, 145], auditory working memory and attention span [141, 143-146], spatial span [144, 145], spatial working memory [147], psychomotor speed [146], motor speed [134], mental flexibility [146], Trailmaking Parts A and B [145], response inhibition[141, 144, 145], figural fluency [134, 145], semantic fluency [141, 146], phonemic fluency [146], reading [134], naming [143], language comprehension[144], line orientation [143], facial recognition [143], memory [134], verbal learning [141, 144, 145], visual memory [144, 145], and design learning [141]. Another more recent study [148] reported no difference between JME patients and controls on a range of neuropsychological measures. Beyond functioning on objective cognitive tasks, another avenue for research is to examine patients' perceptions of cognitive dysfunction. Such a design has shown the JME patients rate a higher level of self-reported executive dysfunction than controls [144].

Given that the variability in findings across studies likely relates to a number of primary and secondary factors, the influence of the effects of such factors is also likely varied. For instance, one small study [128] showed that it may be important to consider the influence of seizure frequency on JME patients' abilities. In general, the JME patients in their sample did not perform differently from healthy controls on tasks of verbal attention, verbal working memory, or phonemic verbal fluency. On the other hand, controls consistently performed higher than JME patients on tasks related to psychomotor speed, cognitive flexibility, categorical verbal fluency, planning, along with abstraction and categorization. The patients who had been seizure free (*n*=11) for a year were also compared with patients who continued to have seizures (*n*=11). The groups showed no differences on any of the cognitive testing, although JME patients scored significantly lower on three of four indices of a decision-making task than controls; seizure status in the JME group was related to performance. However, when compared with controls, the patients who were not seizure-free also showed additional significant differences on measures of cognitive flexibility, planning, and facets of a decisionmaking task than the seizure-free patients. Another study [146] examining performance on the same decision-making task reported no difference between controls and JME patients as a group, although fewer patients who were not seizure-free showed improvement and learning on the task over time. These findings suggest that seizure frequency may be an important modulator of cognitive abilities in JME patients.

relationship between cognition and abnormal EEG findings has also been demonstrated elsewhere [137, 138]. In contrast, JME patients with at-rest epileptiform discharges (*n*=11) [142] have been shown to demonstrate worse abstract reasoning concept formation, and mental

Juvenile Myoclonic Epilepsy — A Maturation Syndrome Coming of Age

http://dx.doi.org/10.5772/58372

97

In addition to demonstrating divergence and convergence amongst clinic groups and anatomy considerations, the genetic involvement of JME may also elucidate predictable patterns of cognitive performance. In a unique approach to demonstrating potential genetic vulnerabili‐ ties that underlie cognitive performance, there is evidence [131] that first degree relatives of JME patients have lower performance on tasks of sustained attention than relatives of indi‐ viduals diagnosed with CAE or TLE. Another small sample of young adults with JME performed similarly to their siblings on a number of tasks related to executive, language, verbal memory, phonemic and semantic fluency, and general intellectual functioning [144]. However, the JME group scored in a range suggesting more self-reported symptoms related to behav‐ ioral, motivational, cognitive, and emotional factors than their sibling [144]. Somnez et al. [143] also reported a number of similarities in cognitive performance for patients with JME who had a relative with epilepsy and those who did not have a relative with epilepsy. Nevertheless, the authors indicated that patients with a family history of seizures were found to be "less successful in general cognitive evaluation," in particular on a spatial perceptual task, forward auditory digit span repetition, and speeded reading measure. A more recent sibling study of cognitive differences in JME patients [136] included data indicating that a group of siblings showed no significant differences on any cognitive measure when compared with the JME group. Moreover, the siblings' performance was not discrepant from a group of healthy controls except for an aspect that indicates JME participants generated more responses that were counter to the test rules on a measure of prospective memory (i.e., remembering to remember). Taken together, findings in these studies may reflect a familial genetic vulnera‐ bility for subtle cognitive abnormalities in JME family members who do not have a history of

In general, the aggregate of cognitive studies in JME patients has indicated that JME patients do not typically perform at the same level as individuals from control or normative groups. Moreover, review of the literature suggests that more studies show JME patients have poorer performance on frontally mediated tasks (i.e., processing speed, response inhibition, and verbal fluency) Than those that do not show such difference. So, too, virtually no study has shown that JME groups perform better than controls on any number of tasks. However, the totality of results has not been consistent and a number of studies have shown that patients with JME sometimes display cognitive abnormalities in other neuropsychological domains as well. Additionally, the literature is weighted toward including mainly tasks related to frontal lobe functioning at the expense of investigating other cognitive domains, which results in literature bias. There are also conflicting data across the studies comparing controls with JME patients to the degree that it is difficult to definitively conclude there is a specific JME cognitive endophenotype, such as a "frontal syndrome." It is likely that, based on the number of inconsistent results for both fontal and non-frontal task performances, any given JME patient may display a range of cognitive abnormalities-the expression of which is likely dependent

flexibility.

seizures.

Overall ability level, such as intelligence, has been shown to relate to performance across cognitive domains [149, 150]. This is likely to be the case with JME patients as well and may influence the significance level of research findings, particularly in small samples. As an example, heterogeneous cognitive performance in JME patients was noted in a published abstract [151], indicating that drug-resistant patients who performed in the impaired range on tasks of executive functioning had a high rate of impairments on tests in other domains. However, there were only limited published data available to review from that report. Other variables have been associated with cognitive performance in JME patients as well such as age of epilepsy onset [138], duration of epilepsy [135, 138], and educational level [135]. Addition‐ ally, JME patients on a regimen of multiple AEDs may perform worse than patients on monotherapy on tests of psychomotor speed, cognitive flexibility, and phonemic fluency [128]. On the other hand, some studies have shown no relationship with cognitive performance in JME groups related to duration of epilepsy [127], education level [138], the frequency of seizures [127, 138], treatment status [127], the type of seizures [127], age [138], sex [138], family history [138], or previous intake of an AED [138].

In addition to demography, there may be biological determinants of particular cognitive performance profiles in JME. In a magnetic resonance spectroscopy (MRS) study of the brain [152], researchers indicated JME patients with reduced frontal N-Acetylaspartic acid had lower mental flexibility compared with JME patients with normal levels. Other biological influences may relate to cognitive functioning as well. In JME patients, frontal and thalamic volumes have been associated with executive task performance [132]. Those researchers did not show similar anatomical relationships with cognitive performance in patients diagnosed with other epilepsies [132]. Similar work [141] has revealed an association of fractional anisotropy (FA), a measure of white matter integrity on diffusion tensor imaging (DTI), in anterior supplemen‐ tary motor area (SMA) regions with scores on a picture naming task. FA values in the posterior cingulate region and corresponding gray matter volume (GMV) negatively predicted scores on the Trailmaking task. However, no other FA values were correlated with any other clinical variables or neuropsychological testing scores [141]. In terms of functional paradigms, fMRI patterns have differed for JME patients versus controls during tasks that require a high level of attention, concentration, and working memory, suggesting motor cortex involvement even though there were no group differences in the actual outcome of the task [147]. This suggests that there may be alterations in or pathological changes to cerebral regional recruitment during task performances as a function of disease state.

Regarding EEG studies, mixed findings have been demonstrated for the relationship of EEG patterns on cognitive performance. One group [143] examined the influence of paroxysmal EEG findings on task performance, but did not indicate a relationship. A non-significant relationship between cognition and abnormal EEG findings has also been demonstrated elsewhere [137, 138]. In contrast, JME patients with at-rest epileptiform discharges (*n*=11) [142] have been shown to demonstrate worse abstract reasoning concept formation, and mental flexibility.

significant differences on measures of cognitive flexibility, planning, and facets of a decisionmaking task than the seizure-free patients. Another study [146] examining performance on the same decision-making task reported no difference between controls and JME patients as a group, although fewer patients who were not seizure-free showed improvement and learning on the task over time. These findings suggest that seizure frequency may be an important

Overall ability level, such as intelligence, has been shown to relate to performance across cognitive domains [149, 150]. This is likely to be the case with JME patients as well and may influence the significance level of research findings, particularly in small samples. As an example, heterogeneous cognitive performance in JME patients was noted in a published abstract [151], indicating that drug-resistant patients who performed in the impaired range on tasks of executive functioning had a high rate of impairments on tests in other domains. However, there were only limited published data available to review from that report. Other variables have been associated with cognitive performance in JME patients as well such as age of epilepsy onset [138], duration of epilepsy [135, 138], and educational level [135]. Addition‐ ally, JME patients on a regimen of multiple AEDs may perform worse than patients on monotherapy on tests of psychomotor speed, cognitive flexibility, and phonemic fluency [128]. On the other hand, some studies have shown no relationship with cognitive performance in JME groups related to duration of epilepsy [127], education level [138], the frequency of seizures [127, 138], treatment status [127], the type of seizures [127], age [138], sex [138], family

In addition to demography, there may be biological determinants of particular cognitive performance profiles in JME. In a magnetic resonance spectroscopy (MRS) study of the brain [152], researchers indicated JME patients with reduced frontal N-Acetylaspartic acid had lower mental flexibility compared with JME patients with normal levels. Other biological influences may relate to cognitive functioning as well. In JME patients, frontal and thalamic volumes have been associated with executive task performance [132]. Those researchers did not show similar anatomical relationships with cognitive performance in patients diagnosed with other epilepsies [132]. Similar work [141] has revealed an association of fractional anisotropy (FA), a measure of white matter integrity on diffusion tensor imaging (DTI), in anterior supplemen‐ tary motor area (SMA) regions with scores on a picture naming task. FA values in the posterior cingulate region and corresponding gray matter volume (GMV) negatively predicted scores on the Trailmaking task. However, no other FA values were correlated with any other clinical variables or neuropsychological testing scores [141]. In terms of functional paradigms, fMRI patterns have differed for JME patients versus controls during tasks that require a high level of attention, concentration, and working memory, suggesting motor cortex involvement even though there were no group differences in the actual outcome of the task [147]. This suggests that there may be alterations in or pathological changes to cerebral regional recruitment during

Regarding EEG studies, mixed findings have been demonstrated for the relationship of EEG patterns on cognitive performance. One group [143] examined the influence of paroxysmal EEG findings on task performance, but did not indicate a relationship. A non-significant

modulator of cognitive abilities in JME patients.

96 Epilepsy Topics

history [138], or previous intake of an AED [138].

task performances as a function of disease state.

In addition to demonstrating divergence and convergence amongst clinic groups and anatomy considerations, the genetic involvement of JME may also elucidate predictable patterns of cognitive performance. In a unique approach to demonstrating potential genetic vulnerabili‐ ties that underlie cognitive performance, there is evidence [131] that first degree relatives of JME patients have lower performance on tasks of sustained attention than relatives of indi‐ viduals diagnosed with CAE or TLE. Another small sample of young adults with JME performed similarly to their siblings on a number of tasks related to executive, language, verbal memory, phonemic and semantic fluency, and general intellectual functioning [144]. However, the JME group scored in a range suggesting more self-reported symptoms related to behav‐ ioral, motivational, cognitive, and emotional factors than their sibling [144]. Somnez et al. [143] also reported a number of similarities in cognitive performance for patients with JME who had a relative with epilepsy and those who did not have a relative with epilepsy. Nevertheless, the authors indicated that patients with a family history of seizures were found to be "less successful in general cognitive evaluation," in particular on a spatial perceptual task, forward auditory digit span repetition, and speeded reading measure. A more recent sibling study of cognitive differences in JME patients [136] included data indicating that a group of siblings showed no significant differences on any cognitive measure when compared with the JME group. Moreover, the siblings' performance was not discrepant from a group of healthy controls except for an aspect that indicates JME participants generated more responses that were counter to the test rules on a measure of prospective memory (i.e., remembering to remember). Taken together, findings in these studies may reflect a familial genetic vulnera‐ bility for subtle cognitive abnormalities in JME family members who do not have a history of seizures.

In general, the aggregate of cognitive studies in JME patients has indicated that JME patients do not typically perform at the same level as individuals from control or normative groups. Moreover, review of the literature suggests that more studies show JME patients have poorer performance on frontally mediated tasks (i.e., processing speed, response inhibition, and verbal fluency) Than those that do not show such difference. So, too, virtually no study has shown that JME groups perform better than controls on any number of tasks. However, the totality of results has not been consistent and a number of studies have shown that patients with JME sometimes display cognitive abnormalities in other neuropsychological domains as well. Additionally, the literature is weighted toward including mainly tasks related to frontal lobe functioning at the expense of investigating other cognitive domains, which results in literature bias. There are also conflicting data across the studies comparing controls with JME patients to the degree that it is difficult to definitively conclude there is a specific JME cognitive endophenotype, such as a "frontal syndrome." It is likely that, based on the number of inconsistent results for both fontal and non-frontal task performances, any given JME patient may display a range of cognitive abnormalities-the expression of which is likely dependent upon various factors that have not been adequately described in the literature. As such, it will be important for researchers to continue to investigate etiological contributors to cognitive functioning in JME patients that account for the influence of psychosocial variables, neuro‐ biological functions, and various other metrics of individual differences and disease charac‐ teristics.

prevalence of Axis I (30%) and Axis II (26%) diagnoses has also been reported to be high, and a number (47%) have current or lifetime prevalence of some type of disorder. Concerning specific disorders in studies where standardized diagnostic interviews were used, results are as follows: anxiety disorders (21-23.8%) [165, 167], generalized anxiety disorder (19-23%) [165, 166, 168], depression (17-20.9%) [165-168], and somatoform disorders (5.6-7%) [165, 166, 168]. Those studies also indicated that less than 5% of patients have a substance abuse disorder, psychotic disorder, obsessive compulsive disorder, dysthymia, specific phobias, or attention deficit disorder [165-168]. Beyond Axis I disorders, it has been noted that 9 to 20% of samples have met criteria for a personality disorder according to structured diagnostic interview [166, 167]. Histrionic, paranoid personality, and borderline personality disorders were most prevalent. There has been a high rate of Axis I (11.3-19%) [164, 165] and Axis II (23%) [164]

Juvenile Myoclonic Epilepsy — A Maturation Syndrome Coming of Age

http://dx.doi.org/10.5772/58372

99

Across other various diagnostic schemes, such as retrospective chart review and clinical judgment, prevalence of any mental disorder varies from 26.5 to 34% [109, 169, 170]. Different researchers have also reported various combinations of mental disorders in samples such that 25.3% in one study were determined to have either an anxiety, phobia, or somatization disorder and 18.1% had a mood disorder [140]. Frequency of specific diagnoses for depression (8.6-14.5%) [109, 170] and anxiety disorders 15.5% [109] have also been reported. Similar to studies using structured diagnostic interviews, other designs have indicated a low prevalence (<5%) for psychotic disorders [169] and obsessive compulsive disorder [109, 169]. However, some data points have been outliers in that less than 5% of samples have been diagnosed with depression and anxiety disorders as well as depression [169]. In addition to diagnostic rates, patients with JME have scored significantly higher than controls on measures of symptoms related to depression, anxiety [171], internalization (24%), and externalization (16%) [172]. Prevalence of personality disorders in studies with less controlled diagnostic procedures have varied widely and have comprised 3 to 14% of samples [109, 169, 170]. Data from the afore‐ mentioned studies have indicated that groups of patients with JME have a higher incidence of mental disorders than members of control groups. Such studies have tended not to report

In addition to comparing control groups, research has been conducted examining differential prevalence of mental health problems in JME groups with that found in other groups of epilepsy patients. In a study of 157 patients with GGE [173], there were no significant differ‐ ences in rates of psychiatric diagnoses between JME patients (23%) and diagnoses occurring in other GGE syndromes. In comparison to patients with partial epilepsies, varying degrees of concordance have been noted. In larger samples [168], similar numbers of psychiatric disorders have been found in JME patients (49%) and patients with refractory MTS (50%). In contrast, an early study using a structured diagnostic interview [174] found only 22% of the JME patients met criteria for a mental disorder, while 55% of patients with TLE met criteria for a mood disorder. Comparison of prevalence has suggested that JME patients have a lower rate of psychotic disorders than patients with MTS, although anxiety disorder may be more prevalent in JME groups (23%). That research group [165] also compared a larger cohort and did not find differing levels of psychiatric diagnoses (i.e., mood, anxiety or somatoform)

psychiatric comorbidity in those patients who have one diagnosis as well.

comorbidity rates.

#### **15. Psychiatric complications**

Psychiatric disorders are common in the general population [153] and the presence of a neurological disorder has been associated with increased prevalence [154]. Patients with epilepsy, as a group, have a high rate of psychopathology [155] and present with unique psychiatric problems that may complicate proper diagnosis and treatment. For instance, patients with epilepsy may experience psychiatric symptoms that are caused by, maintained, or exacerbated by discrete epileptiform activity. The most striking incidence of this relates to peri-ictal states that cause intense psychiatric reactions that include any number of symptoms consistent with anxiety, fear and panic, negative emotionality, and even psychotic phenomena [156]. Moreover, psychiatric factors in epilepsy have been related to challenges that stem from the functional impact of the disease that serve to restrict, modify, or impair individuals' functional status [157]. In that regard, limitations, such as restricted job duties and driving cessation, have been related to a higher incidence of mental health and psychological adjust‐ ment challenges along with lower rated quality of life [158].

While much is known about the psychosocial aspects of some common epilepsy syndromes (i.e., TLE), there has been less prominent study in individuals who have been diagnosed with other less frequently occurring forms of epilepsy. Within the JME literature, there are oft mentioned assertions that individuals with JME possess characteristics including irresponsi‐ bility, labile behavior, poor discipline, quixotic temperaments, emotional regulation difficul‐ ties, and egocentrism, although few empirical data have directly addressed these psychological traits in a systematic fashion [159]. Early retrospective study identified that there was a high portion (36.4%) of "character neurosis disorder" in JME patients [160]. Other indications [161] were that approximately 29% had some type of psychiatric disorder. Applying personality typologies that are specific to epilepsy patients has proven challenging [162], although research-driven examinations have begun to more clearly identify the types of interictal psychiatric complications that frequently occur in patients with JME. Within that empirical approach, researchers have reported that a combination of personality features, psychiatric symptoms, and contemporary psychiatric disorders occur in JME patients. To a lesser extent, relationships with extra-disease factors, such as social functioning have also been investigated in JME patients.

In terms of research approaches, differential prevalence designs have shown that JME patients have a higher rate of psychiatric diagnoses than controls. Standardized diagnostic interviews, such as the Structured Clinical Interview for DSM-IV [163], have demonstrated rates of psychopathology in JME patient clinical samples ranging from 35 to 62% [164-167]. Lifetime prevalence of Axis I (30%) and Axis II (26%) diagnoses has also been reported to be high, and a number (47%) have current or lifetime prevalence of some type of disorder. Concerning specific disorders in studies where standardized diagnostic interviews were used, results are as follows: anxiety disorders (21-23.8%) [165, 167], generalized anxiety disorder (19-23%) [165, 166, 168], depression (17-20.9%) [165-168], and somatoform disorders (5.6-7%) [165, 166, 168]. Those studies also indicated that less than 5% of patients have a substance abuse disorder, psychotic disorder, obsessive compulsive disorder, dysthymia, specific phobias, or attention deficit disorder [165-168]. Beyond Axis I disorders, it has been noted that 9 to 20% of samples have met criteria for a personality disorder according to structured diagnostic interview [166, 167]. Histrionic, paranoid personality, and borderline personality disorders were most prevalent. There has been a high rate of Axis I (11.3-19%) [164, 165] and Axis II (23%) [164] psychiatric comorbidity in those patients who have one diagnosis as well.

upon various factors that have not been adequately described in the literature. As such, it will be important for researchers to continue to investigate etiological contributors to cognitive functioning in JME patients that account for the influence of psychosocial variables, neuro‐ biological functions, and various other metrics of individual differences and disease charac‐

Psychiatric disorders are common in the general population [153] and the presence of a neurological disorder has been associated with increased prevalence [154]. Patients with epilepsy, as a group, have a high rate of psychopathology [155] and present with unique psychiatric problems that may complicate proper diagnosis and treatment. For instance, patients with epilepsy may experience psychiatric symptoms that are caused by, maintained, or exacerbated by discrete epileptiform activity. The most striking incidence of this relates to peri-ictal states that cause intense psychiatric reactions that include any number of symptoms consistent with anxiety, fear and panic, negative emotionality, and even psychotic phenomena [156]. Moreover, psychiatric factors in epilepsy have been related to challenges that stem from the functional impact of the disease that serve to restrict, modify, or impair individuals' functional status [157]. In that regard, limitations, such as restricted job duties and driving cessation, have been related to a higher incidence of mental health and psychological adjust‐

While much is known about the psychosocial aspects of some common epilepsy syndromes (i.e., TLE), there has been less prominent study in individuals who have been diagnosed with other less frequently occurring forms of epilepsy. Within the JME literature, there are oft mentioned assertions that individuals with JME possess characteristics including irresponsi‐ bility, labile behavior, poor discipline, quixotic temperaments, emotional regulation difficul‐ ties, and egocentrism, although few empirical data have directly addressed these psychological traits in a systematic fashion [159]. Early retrospective study identified that there was a high portion (36.4%) of "character neurosis disorder" in JME patients [160]. Other indications [161] were that approximately 29% had some type of psychiatric disorder. Applying personality typologies that are specific to epilepsy patients has proven challenging [162], although research-driven examinations have begun to more clearly identify the types of interictal psychiatric complications that frequently occur in patients with JME. Within that empirical approach, researchers have reported that a combination of personality features, psychiatric symptoms, and contemporary psychiatric disorders occur in JME patients. To a lesser extent, relationships with extra-disease factors, such as social functioning have also been investigated

In terms of research approaches, differential prevalence designs have shown that JME patients have a higher rate of psychiatric diagnoses than controls. Standardized diagnostic interviews, such as the Structured Clinical Interview for DSM-IV [163], have demonstrated rates of psychopathology in JME patient clinical samples ranging from 35 to 62% [164-167]. Lifetime

teristics.

98 Epilepsy Topics

in JME patients.

**15. Psychiatric complications**

ment challenges along with lower rated quality of life [158].

Across other various diagnostic schemes, such as retrospective chart review and clinical judgment, prevalence of any mental disorder varies from 26.5 to 34% [109, 169, 170]. Different researchers have also reported various combinations of mental disorders in samples such that 25.3% in one study were determined to have either an anxiety, phobia, or somatization disorder and 18.1% had a mood disorder [140]. Frequency of specific diagnoses for depression (8.6-14.5%) [109, 170] and anxiety disorders 15.5% [109] have also been reported. Similar to studies using structured diagnostic interviews, other designs have indicated a low prevalence (<5%) for psychotic disorders [169] and obsessive compulsive disorder [109, 169]. However, some data points have been outliers in that less than 5% of samples have been diagnosed with depression and anxiety disorders as well as depression [169]. In addition to diagnostic rates, patients with JME have scored significantly higher than controls on measures of symptoms related to depression, anxiety [171], internalization (24%), and externalization (16%) [172]. Prevalence of personality disorders in studies with less controlled diagnostic procedures have varied widely and have comprised 3 to 14% of samples [109, 169, 170]. Data from the afore‐ mentioned studies have indicated that groups of patients with JME have a higher incidence of mental disorders than members of control groups. Such studies have tended not to report comorbidity rates.

In addition to comparing control groups, research has been conducted examining differential prevalence of mental health problems in JME groups with that found in other groups of epilepsy patients. In a study of 157 patients with GGE [173], there were no significant differ‐ ences in rates of psychiatric diagnoses between JME patients (23%) and diagnoses occurring in other GGE syndromes. In comparison to patients with partial epilepsies, varying degrees of concordance have been noted. In larger samples [168], similar numbers of psychiatric disorders have been found in JME patients (49%) and patients with refractory MTS (50%). In contrast, an early study using a structured diagnostic interview [174] found only 22% of the JME patients met criteria for a mental disorder, while 55% of patients with TLE met criteria for a mood disorder. Comparison of prevalence has suggested that JME patients have a lower rate of psychotic disorders than patients with MTS, although anxiety disorder may be more prevalent in JME groups (23%). That research group [165] also compared a larger cohort and did not find differing levels of psychiatric diagnoses (i.e., mood, anxiety or somatoform) between JME patients and those with MTS, although the presence of psychotic disorders was associated with MTS group membership. Indeed, that group also published data [175] from the same sample that indicated significantly more MTS patients (11.6%) had at least two core symptoms of psychosis compared with 4.8% of those with JME. However, the proportion of individuals with post-ictal psychosis or interictal psychosis was similar between the groups. With regard to symptom severity, a sample of 20 JME and 20 TLE patients had scores on measures of stress and depression [176] that were similar.

is likely that lower psychological stress promotes adherence, and the effect of being adherent leads to better seizure control, which also likely results in better psychosocial functioning and fewer psychiatric symptoms. As psychological stress has been inconsistently shown to be related with seizure control, researchers have implemented psychological interventions expecting this will affect seizure outcome. In one such study [109], 58 JME patients with uncontrolled seizures receiving a "rational AED regimen" participated in psychological intervention with the goal to eliminate seizure precipitants. Treatment modalities in that study included an anti-stress program or Cognitive Behavioral Therapy intervention. The results from that study indicated that patients showed a reduction of seizure activity across three time epochs as treatment progressed and this also coincided with ratings of psychiatric functioning. This is particularly relevant as there are indications that psychological factors might relate to seizure control and perceptions of seizure control. For instance, in a survey of JME patients, 62 of 75 respondents (83%) reported that they viewed stress as the most frequent seizure precipitant [35]. These findings provide preliminary evidence that medication adherence and psychological treatment may have important roles in influencing emotional well-being in these

Juvenile Myoclonic Epilepsy — A Maturation Syndrome Coming of Age

http://dx.doi.org/10.5772/58372

101

Along with rates of psychiatric problems, other functional outcomes have been examined. For example, long-term follow-up of JME patients [14] has indicated that 65% to 77% of patients reported being "very satisfied" with their health, work, friendships, and social life. However, 61% were prescribed a psychiatric medication, nearly one third were unemployed, and 74% reported at least one "unfavorable" social outcome (i.e., school suspensions, truancy, fighting, criminal offense, social isolation, social impulsiveness, unplanned pregnancy). The authors noted that such outcomes occur at similar rates as other childhood onset epilepsies. Likewise, other data has indicated that only 22.9% of patients may find employment [140]. JME patients [182] also have adjustment challenges in work and family life. De Araujo Filho et al. [166] noted that JME patients, compared with controls, also had a lower DSM-IV Global Assessment of

Functioning score and they reported a higher number of psychosocial problems.

and limbic structures compared with JME patients with no psychiatric disorder.

As neurobiological theories of psychiatric disorders continue to be advanced [183, 184], there have also been more studies of cerebral functioning in patients with epilepsy and comorbid psychiatric conditions [185]. In that vein, a group from Brazil has been at the forefront of examining characteristics of brain structure and function in JME patients with and without psychiatric diagnoses. In particular, they [180] have shown small to medium correlations for NAA/Cr in the left medial primary motor region and right thalamus, and small to medium correlations for GLX/Cr in the right medial primary motor and left lateral primary motor areas for psychiatric JME patients. That group [179] also showed that 16 JME patients with a cluster B personality disorder had reduction of GMV in right thalamus compared with JME patients with no psychiatric problems. The JME group with personality disorders also had bilateral increases in GMV in the middle frontal gyrus and right orbitofrontal cortex, and decreased white matter in the posterior corpus callosum. Additional study [181] of cluster B personality disorders in patients with JME showed bilateral morphological changes in thalamic, frontal,

patients.

Further research on various broad spectrum measures of personality traits [172] has revealed a number of disparate findings. Study on JME patients' responses on metrics of personality characteristics has indicated JME patients [172] experience significantly higher levels of a "repressive defensiveness" trait than the test normative sample, although a number of other variables were not significantly elevated. Other research [177] has compared responses of JME patients and controls on the Minnesota Multiphasic Personality Inventory [178], and another group [176] did not find group differences on a Five Factor personality inventory in JME and TLE patients. Another study [171] of personality features in JME explored possible endophe‐ notypic expressions of personality features that are not part of contemporary psychiatric diagnostic categorization. As such, compared with controls, JME patients were shown to have a high rate of novelty seeking, low rate of harm avoidance, and low rate of self-directedness (e.g., lack of goal direction and incongruent habits). Taken together, these disparate and nonsignificant findings have led to conclusions that there may not be a specific personality profile associated with JME [172, 176].

Finding factors that influence or modify the expression of mental health problems in JME has also proven elusive. In a retrospective chart review [169] of 155 JME patients, it was shown that psychiatric disorders were more prevalent in patients with medically resistant seizures (58.3%) as opposed to non-resistant (19%). Similarly, anxiety disorders in JME patients have been associated with lack of seizure control and a history of having several lifetime GTCSs [167]. Moreover, other psychiatric and personality disorders have been associated with seizure frequency [166]. In contrast, others have not found associations of frequency of psychiatric diagnoses with duration of epilepsy [164, 168], type of seizures [164, 168], seizure frequency [164, 168], or number and type of AEDs [168]. Researchers [164] have also not found factors associated with psychiatric comorbidity including age and medication adherence.

Regarding personality features [177], patient endorsements on measures of personality have not been related to age of epilepsy, diagnosis onset, or seizure frequency. In JME patients with a personality disorder compared with JME patients without a personality disorder, there have been no group associations with disease duration [166], age of onset, or "adequate treatment" [179-181]. However, there have been significant associations of personality facets, such as novelty seeking, with early age of epilepsy onset and higher frequency of myoclonic seizures [171]. Disease chronicity has also been shown to relate to personality features, such as restraint [172].

In contrast to describing risk factors, researchers have identified potentially protective factors against anxiety and personality disorders that include being treated with an AED for more than 2 years [166, 168]. Although the direction of causality for this association is not known, it is likely that lower psychological stress promotes adherence, and the effect of being adherent leads to better seizure control, which also likely results in better psychosocial functioning and fewer psychiatric symptoms. As psychological stress has been inconsistently shown to be related with seizure control, researchers have implemented psychological interventions expecting this will affect seizure outcome. In one such study [109], 58 JME patients with uncontrolled seizures receiving a "rational AED regimen" participated in psychological intervention with the goal to eliminate seizure precipitants. Treatment modalities in that study included an anti-stress program or Cognitive Behavioral Therapy intervention. The results from that study indicated that patients showed a reduction of seizure activity across three time epochs as treatment progressed and this also coincided with ratings of psychiatric functioning. This is particularly relevant as there are indications that psychological factors might relate to seizure control and perceptions of seizure control. For instance, in a survey of JME patients, 62 of 75 respondents (83%) reported that they viewed stress as the most frequent seizure precipitant [35]. These findings provide preliminary evidence that medication adherence and psychological treatment may have important roles in influencing emotional well-being in these patients.

between JME patients and those with MTS, although the presence of psychotic disorders was associated with MTS group membership. Indeed, that group also published data [175] from the same sample that indicated significantly more MTS patients (11.6%) had at least two core symptoms of psychosis compared with 4.8% of those with JME. However, the proportion of individuals with post-ictal psychosis or interictal psychosis was similar between the groups. With regard to symptom severity, a sample of 20 JME and 20 TLE patients had scores on

Further research on various broad spectrum measures of personality traits [172] has revealed a number of disparate findings. Study on JME patients' responses on metrics of personality characteristics has indicated JME patients [172] experience significantly higher levels of a "repressive defensiveness" trait than the test normative sample, although a number of other variables were not significantly elevated. Other research [177] has compared responses of JME patients and controls on the Minnesota Multiphasic Personality Inventory [178], and another group [176] did not find group differences on a Five Factor personality inventory in JME and TLE patients. Another study [171] of personality features in JME explored possible endophe‐ notypic expressions of personality features that are not part of contemporary psychiatric diagnostic categorization. As such, compared with controls, JME patients were shown to have a high rate of novelty seeking, low rate of harm avoidance, and low rate of self-directedness (e.g., lack of goal direction and incongruent habits). Taken together, these disparate and nonsignificant findings have led to conclusions that there may not be a specific personality profile

Finding factors that influence or modify the expression of mental health problems in JME has also proven elusive. In a retrospective chart review [169] of 155 JME patients, it was shown that psychiatric disorders were more prevalent in patients with medically resistant seizures (58.3%) as opposed to non-resistant (19%). Similarly, anxiety disorders in JME patients have been associated with lack of seizure control and a history of having several lifetime GTCSs [167]. Moreover, other psychiatric and personality disorders have been associated with seizure frequency [166]. In contrast, others have not found associations of frequency of psychiatric diagnoses with duration of epilepsy [164, 168], type of seizures [164, 168], seizure frequency [164, 168], or number and type of AEDs [168]. Researchers [164] have also not found factors

associated with psychiatric comorbidity including age and medication adherence.

Regarding personality features [177], patient endorsements on measures of personality have not been related to age of epilepsy, diagnosis onset, or seizure frequency. In JME patients with a personality disorder compared with JME patients without a personality disorder, there have been no group associations with disease duration [166], age of onset, or "adequate treatment" [179-181]. However, there have been significant associations of personality facets, such as novelty seeking, with early age of epilepsy onset and higher frequency of myoclonic seizures [171]. Disease chronicity has also been shown to relate to

In contrast to describing risk factors, researchers have identified potentially protective factors against anxiety and personality disorders that include being treated with an AED for more than 2 years [166, 168]. Although the direction of causality for this association is not known, it

measures of stress and depression [176] that were similar.

associated with JME [172, 176].

100 Epilepsy Topics

personality features, such as restraint [172].

Along with rates of psychiatric problems, other functional outcomes have been examined. For example, long-term follow-up of JME patients [14] has indicated that 65% to 77% of patients reported being "very satisfied" with their health, work, friendships, and social life. However, 61% were prescribed a psychiatric medication, nearly one third were unemployed, and 74% reported at least one "unfavorable" social outcome (i.e., school suspensions, truancy, fighting, criminal offense, social isolation, social impulsiveness, unplanned pregnancy). The authors noted that such outcomes occur at similar rates as other childhood onset epilepsies. Likewise, other data has indicated that only 22.9% of patients may find employment [140]. JME patients [182] also have adjustment challenges in work and family life. De Araujo Filho et al. [166] noted that JME patients, compared with controls, also had a lower DSM-IV Global Assessment of Functioning score and they reported a higher number of psychosocial problems.

As neurobiological theories of psychiatric disorders continue to be advanced [183, 184], there have also been more studies of cerebral functioning in patients with epilepsy and comorbid psychiatric conditions [185]. In that vein, a group from Brazil has been at the forefront of examining characteristics of brain structure and function in JME patients with and without psychiatric diagnoses. In particular, they [180] have shown small to medium correlations for NAA/Cr in the left medial primary motor region and right thalamus, and small to medium correlations for GLX/Cr in the right medial primary motor and left lateral primary motor areas for psychiatric JME patients. That group [179] also showed that 16 JME patients with a cluster B personality disorder had reduction of GMV in right thalamus compared with JME patients with no psychiatric problems. The JME group with personality disorders also had bilateral increases in GMV in the middle frontal gyrus and right orbitofrontal cortex, and decreased white matter in the posterior corpus callosum. Additional study [181] of cluster B personality disorders in patients with JME showed bilateral morphological changes in thalamic, frontal, and limbic structures compared with JME patients with no psychiatric disorder.

Overall, the current data do not consistently support a specific combination of behavioral or psychiatric symptoms among JME patients suggestive of a personality syndrome. In general, rates of psychiatric diagnoses in JME patients have been shown to be higher than the general population. Further, preliminary data suggest that JME patients have a higher rate of anxiety symptoms and less prevalence of psychotic symptoms compared with TLE patients. However, while there are a number of studies of psychiatric functioning in other epilepsies, such as TLE, psychiatric data in JME patients remain inadequate. Moreover, a number of studies typically lack satisfactory sample size, control and differing patients groups, psychosocial contributors, neuroimaging (i.e., structural and functional) studies, or neurophysiological metrics. Addi‐ tionally, there is an absence in the literature regarding genetic, sibling, or family studies that address psychiatric variables. Such data will further help advance causal models of psycho‐ pathology and influence treatment implications. Although many details regarding psychiatric problems in JME are not well defined, the combined data has immediate implications for clinical care.

190]. EFHC1 was originally shown to be a microtubule–associated protein (MAP), localized to the centrosome and the mitotic spindle. Loss of function of EFHC1 disrupts mitotic spindle organization and impairs radial migration of neurons to the developing cortex. Moreover, recent evidence points to impaired radial glia scaffold formation and altered morphology of both radially (mainly excitatory) and tangentially (mainly inhibitory) migrating neurons. This disruption of migration results in fewer neurons and impaired cortical migration of both excitatory and inhibitory neurons. Impaired morphology, i.e., the inability to transform from

Juvenile Myoclonic Epilepsy — A Maturation Syndrome Coming of Age

http://dx.doi.org/10.5772/58372

103

Susceptibility genes may also play an important role in the expression of epilepsy phenotypes. Although channelopathies only account for 3% of JME [187], their genetic associations with epilepsy are likely important markers for disease characteristics. GABRA2 mutations, in addition to SCN1A and SCN1B mutations, have been liked to pedigrees with generalized epilepsy with febrile seizures, whose members may present with many types of epilepsy

There are also several susceptibility single-nucleotide polymorphism (SNP) loci that contribute to the risk of developing GGE and JME, which has been suggested by genome-wide association analyses. For GGE, in general, at least three candidate genes have been noted: 1) VRK (2p16.1): a loci that may affect cortical development, 2) SCNA1 (2q24.3): a loci for coding the α-subunit of the neuronal voltage-gated sodium channel, mutations of which are the most common cause of channelopathy-related epilepsy, and 3) PNPO (17q21.32): a loci that is involved in pyridox‐ ine oxidation to an active-cofactor involved in neurotransmitter metabolism. Susceptibility candidates for specific GGE syndromes, such as absence epilepsy or JME, were also identified. The linkage associated with JME is at 1q43, flanked by two regions with high recombination rates and covering the CHRM3 gene, which codes for the M3 muscarinic acetylcholine receptor. While nicotinic acetylcholine receptor mutations have been linked to autosomal dominant frontal lobe epilepsy, it is not clear how muscarinic receptors contribute to epileptogenesis. In summary, the genetic effects of JME are manifold and various genotypes can result in the same epilepsy syndrome. This may be due to a combination of factors, including the lack of accurate epilepsy syndrome differences in subphenotypes or failure to identify epigenetic

**Pathology**: Brain pathology studies in cases of JME are rare (i.e., approximately 20 cases) as patients usually have an average life expectancy. Meencke and Janz [191] published brain pathology in eight patients with GGE, most of whom had myoclonic seizures. They found microdysgenesis in seven patients, defined as persistent uni-or bipolar cells in the subpial region, increased cell density in the stratum moleculare, portrusions of neurons against the pial membrane, containing well-differentiated neurons, columnar neuronal architecture, and increased neurons in the subcortical white matter. Atypical neurons were also found in the hippocampus and cerebellum [192]. There was no evidence of gliosis or chronic neuronal loss. Those findings were criticized for a lack of control group pathology samples [193], and the authors [191] published an additional number of brains from a mixed group of GGE. Using that latter group of specimens, the authors reported that there was a frontal-to-parietal decrease of "microdysplasia", which was absent in the occipital lobes. In 2000, another group

multipolar to bipolar cells, leads to the neurons remaining in the intermediate zone.

syndromes, including JME [186].

mechanisms.

#### **16. Pathophysiology**

Based upon evidence of broad cognitive dysfunction and psychiatric problems along with the findings that there are various seizures triggers and EEG findings, mounting evidence has suggested that JME affects the brain in a generalized manner. Nonetheless, the contention that JME is a disease of the frontal lobe and or thalamus has been asserted consistently by the literature, even in the face of contradictory neuroimaging and neuropsychological findings. Recent developments in genetics and neuroimaging have opened new avenues to understand the mechanisms underlying GGEs and expand the conceptualization of JME.

**Genetics**: The genetic inheritance of JME is complex as about 20 chromosomal loci have been linked to the disease. While several channelopathies have been associated with multiple GGE phenotypes, particular mutations have also been specifically implicated in JME. For instance, GABRA1 mutation reducing GABA-receptor function and expression [186], and EFHC1 mutations affecting mitotic spindle organization [187, 188], are both candidate mechanisms.

Autosomal dominant inheritance of a GABRA-1 mutation was reported in a large French-Canadian family with seven members exhibiting JME [186]. GABRA1 on chromosome 5q34 encodes for the α1-subunit of the γ-aminobutyric acid receptor subtype A (GABAA-receptor), linked to a chloride-ionophore. Activation of this receptor allows chloride to enter the neuron, leading to hyperpolarization. The GABRA1 mutation leads to a loss of function of the GABAAreceptor in vitro, either due to altered gating properties or decreased expression of receptors at the cell surface, resulting in decreased inhibition.

EFHC1 is located on chromosome 6p11 and was the first gene to be identified with specific linkage to JME [58]. EHFC1 mutations are more commonly associated with JME than channe‐ lopathies, accounting for 9% of JME cases [189]. They are inherited in an autosomal dominant fashion, but also account for singletons and sporadic cases. The putative role of EFHC1 is evolving, but appears to affect neuronal division and migration during corticogenesis [187, 190]. EFHC1 was originally shown to be a microtubule–associated protein (MAP), localized to the centrosome and the mitotic spindle. Loss of function of EFHC1 disrupts mitotic spindle organization and impairs radial migration of neurons to the developing cortex. Moreover, recent evidence points to impaired radial glia scaffold formation and altered morphology of both radially (mainly excitatory) and tangentially (mainly inhibitory) migrating neurons. This disruption of migration results in fewer neurons and impaired cortical migration of both excitatory and inhibitory neurons. Impaired morphology, i.e., the inability to transform from multipolar to bipolar cells, leads to the neurons remaining in the intermediate zone.

Overall, the current data do not consistently support a specific combination of behavioral or psychiatric symptoms among JME patients suggestive of a personality syndrome. In general, rates of psychiatric diagnoses in JME patients have been shown to be higher than the general population. Further, preliminary data suggest that JME patients have a higher rate of anxiety symptoms and less prevalence of psychotic symptoms compared with TLE patients. However, while there are a number of studies of psychiatric functioning in other epilepsies, such as TLE, psychiatric data in JME patients remain inadequate. Moreover, a number of studies typically lack satisfactory sample size, control and differing patients groups, psychosocial contributors, neuroimaging (i.e., structural and functional) studies, or neurophysiological metrics. Addi‐ tionally, there is an absence in the literature regarding genetic, sibling, or family studies that address psychiatric variables. Such data will further help advance causal models of psycho‐ pathology and influence treatment implications. Although many details regarding psychiatric problems in JME are not well defined, the combined data has immediate implications for

Based upon evidence of broad cognitive dysfunction and psychiatric problems along with the findings that there are various seizures triggers and EEG findings, mounting evidence has suggested that JME affects the brain in a generalized manner. Nonetheless, the contention that JME is a disease of the frontal lobe and or thalamus has been asserted consistently by the literature, even in the face of contradictory neuroimaging and neuropsychological findings. Recent developments in genetics and neuroimaging have opened new avenues to understand

**Genetics**: The genetic inheritance of JME is complex as about 20 chromosomal loci have been linked to the disease. While several channelopathies have been associated with multiple GGE phenotypes, particular mutations have also been specifically implicated in JME. For instance, GABRA1 mutation reducing GABA-receptor function and expression [186], and EFHC1 mutations affecting mitotic spindle organization [187, 188], are both candidate mechanisms. Autosomal dominant inheritance of a GABRA-1 mutation was reported in a large French-Canadian family with seven members exhibiting JME [186]. GABRA1 on chromosome 5q34 encodes for the α1-subunit of the γ-aminobutyric acid receptor subtype A (GABAA-receptor), linked to a chloride-ionophore. Activation of this receptor allows chloride to enter the neuron, leading to hyperpolarization. The GABRA1 mutation leads to a loss of function of the GABAAreceptor in vitro, either due to altered gating properties or decreased expression of receptors

EFHC1 is located on chromosome 6p11 and was the first gene to be identified with specific linkage to JME [58]. EHFC1 mutations are more commonly associated with JME than channe‐ lopathies, accounting for 9% of JME cases [189]. They are inherited in an autosomal dominant fashion, but also account for singletons and sporadic cases. The putative role of EFHC1 is evolving, but appears to affect neuronal division and migration during corticogenesis [187,

the mechanisms underlying GGEs and expand the conceptualization of JME.

at the cell surface, resulting in decreased inhibition.

clinical care.

102 Epilepsy Topics

**16. Pathophysiology**

Susceptibility genes may also play an important role in the expression of epilepsy phenotypes. Although channelopathies only account for 3% of JME [187], their genetic associations with epilepsy are likely important markers for disease characteristics. GABRA2 mutations, in addition to SCN1A and SCN1B mutations, have been liked to pedigrees with generalized epilepsy with febrile seizures, whose members may present with many types of epilepsy syndromes, including JME [186].

There are also several susceptibility single-nucleotide polymorphism (SNP) loci that contribute to the risk of developing GGE and JME, which has been suggested by genome-wide association analyses. For GGE, in general, at least three candidate genes have been noted: 1) VRK (2p16.1): a loci that may affect cortical development, 2) SCNA1 (2q24.3): a loci for coding the α-subunit of the neuronal voltage-gated sodium channel, mutations of which are the most common cause of channelopathy-related epilepsy, and 3) PNPO (17q21.32): a loci that is involved in pyridox‐ ine oxidation to an active-cofactor involved in neurotransmitter metabolism. Susceptibility candidates for specific GGE syndromes, such as absence epilepsy or JME, were also identified. The linkage associated with JME is at 1q43, flanked by two regions with high recombination rates and covering the CHRM3 gene, which codes for the M3 muscarinic acetylcholine receptor. While nicotinic acetylcholine receptor mutations have been linked to autosomal dominant frontal lobe epilepsy, it is not clear how muscarinic receptors contribute to epileptogenesis.

In summary, the genetic effects of JME are manifold and various genotypes can result in the same epilepsy syndrome. This may be due to a combination of factors, including the lack of accurate epilepsy syndrome differences in subphenotypes or failure to identify epigenetic mechanisms.

**Pathology**: Brain pathology studies in cases of JME are rare (i.e., approximately 20 cases) as patients usually have an average life expectancy. Meencke and Janz [191] published brain pathology in eight patients with GGE, most of whom had myoclonic seizures. They found microdysgenesis in seven patients, defined as persistent uni-or bipolar cells in the subpial region, increased cell density in the stratum moleculare, portrusions of neurons against the pial membrane, containing well-differentiated neurons, columnar neuronal architecture, and increased neurons in the subcortical white matter. Atypical neurons were also found in the hippocampus and cerebellum [192]. There was no evidence of gliosis or chronic neuronal loss. Those findings were criticized for a lack of control group pathology samples [193], and the authors [191] published an additional number of brains from a mixed group of GGE. Using that latter group of specimens, the authors reported that there was a frontal-to-parietal decrease of "microdysplasia", which was absent in the occipital lobes. In 2000, another group compared five GGE patients, who died of SUDEP, to a nonepileptic control group [194]. This study did not find an increase of the "microdysplasia". Based upon the small numbers of patients with heterogeneous electroclinical features and syndromal diagnoses, it is difficult to draw any conclusions regarding pathology findings that are specific to JME.

It has been shown that JME patients have symmetric FA reductions in the bilateral superior and anterior corona radiata, genu and body of the corpus callosum, along and with middle and superior frontal tracts [139]. White matter tract MD increases coincided with the FA reduction, although the left superior parietal lobe was also affected. There were no increases in FA or decreases in MD found in JME patients. Frequency of GTCS was correlated with the

Juvenile Myoclonic Epilepsy — A Maturation Syndrome Coming of Age

http://dx.doi.org/10.5772/58372

105

Several studies have shown similar reductions of FA [141] and increases in MD [198], mainly concentrated in thalamocortical pathways and the corpus callosum; pathways that are responsible for propagation and synchronization of spike-and-wave discharges. These DTI changes have been correlated with GMV reduction in the supplementary motor area (SMA) and posterior cingulate cortices. The SMA FA was correlated with deficits for word naming tasks and expression, while the posterior cingulate changes predicted cognitive inhibition scores [141] and deficits in frontal lobe executive motor functioning. A subsequent study by the same group demonstrated a correlation between DTI and fMRI-based measures of structural and functional connectivity between prefrontal cognitive cortex and motor cortices [198]. Connectivity was decreased between prefrontal and frontopolar regions and increased between occipital cortex and SMA. The authors suggested that the frontal connectivity changes may be related to cognitive deficits seen in JME, and occipitomotor connectivity may be related to photosensitivity, although association with these clinical phenotypes were not investigated.

The neuroanatomic basis of GMV and DTI changes is not clear. It is suspected that the decreased connectivity may be due to altered cortical organization, but as many of these changes are correlated with disease activity, degenerative changes cannot be ruled out. So far, the studies have shown disparate findings, thus, leading to an inconclusive picture of cerebral differences in JME. However, the aggregate of study results suggests diffuse abnormalities with morphological changes that include regions beyond the frontal lobes. Moreover, patho‐ logical studies are needed to confirm changes in cortical organization and connectivity. To improve on this knowledge base, an animal model of JME could be utilized to investigate a number of clinicopathological data points including morphometric changes and histopatho‐

*Functional – Interictal PET, MRS, EEG-fMRI*: Interictal PET studies using flourodeoxyglucose (FDG) evaluate metabolism over a 40-minute uptake time. While FDG-PET is not ideal for evaluating transient ictal state or interictal discharges it does evaluate a more chronic func‐ tional status. In an early study, Swartz et al. [125] indicated that JME patients showed decreased metabolism in the prefrontal and premotor cortices, which was associated with behavioral and cognitive dysfunction. Increased metabolism in the bilateral thalami was correlated with increased ictal or interictal epileptic discharges in another study of JME patients [199]. These studies demonstrated predominantly frontal lobe dysfunction, and an obvious activation of the thalami by ictal and interictal epileptic discharges. The absence of cortical activations suggests more fluctuation in metabolism due to epileptic discharges, decreased neuronal activity in the interictal state, or a general decrease of metabolism due to the disruptive effect

severity of the microstructural changes.

logic findings.

of epileptic discharges on cognitive activity.

**Neuroimaging**: In contrast to pathology studies, a number of morphometric imaging ap‐ proaches since the 1990s have contributed to the neurobiological understanding of GGE and JME. Such studies have demonstrated differences between JME patients and normal controls, and suggest that there are subtle pathological changes difficult to assess by routine histology. Further, while JME manifests as generalized electroclinical activity, structural imaging has indicated circumscribed, multiregional abnormalities. Similarly, functional imaging studies have also demonstrated a predominance of symmetrical multiregional involvement (i.e., cerebral blood flow [CBF] or blood oxygen level dependent [BOLD] changes) prior to and during absence seizures or prolonged spike-and-wave discharges.

*Structural Neuroimaging:* Routine clinical structural imaging scans, whether computerized tomography (CT) or magnetic resonance imaging (MRI), are considered normal in individuals with JME. In contrast, more sensitive techniques, such as voxel-based morphometry (VBM), can show subtle variations in gray matter concentration or volume that differ in JME patients compared with healthy controls. An early quantitative MRI volumetric study in GGE partici‐ pants showed increases in normalized gray matter volume (GMV) compared to healthy controls, but also possible differences between other epilepsy syndromes as well [195]. In a subsequent VBM study, statistical parametric mapping demonstrated increased mesial frontal GMV in JME patients. Individual patients also demonstrated additional changes in parieto‐ temporal regions along with decreases frontally in individual patients analyzed by a volume of interest method [195].

More recent VBM studies in JME patients have demonstrated multiregional volumetric and morphometric brain changes [196, 197]. Lin et al. [196] showed increased right frontal GMV and but reduced bilateral thalami, insula, and cerebellar hemisphere volume. Differences in GMV as a function of photosensitivity were also detected, as there were decreases in occipital lobe, left inferior frontal lobe, and hippocampal GMV in photosensitive patients compared with non-photosensitive JME patients. Altered cortical morphology differences in terms of surface area metrics, and not cortical thickness, was noted in the left hemisphere: insula, cingulate, occipital pole, middle temporal, and fusiform gyri [197]. Similar differences were also noted in the right hemisphere: insula, inferior temporal gyrus, and precuneus. Mean cortical curvature measurements in JME patients were also different from controls in the bilateral insula, left cingulate, and right inferior temporal gyrus.

Diffusion tensor imaging (DTI) provides a complementary approach to traditional morpho‐ metric and volumetric analyses, as it accounts for white matter connectivity in cortical and subcortical regions. The most important parameters assessed by DTI are fractional anisotropy (FA) and mean diffusivity (MD), which provide information about the microstructural integrity of the white matter pathways. In particular, FA is influenced by myelin integrity and fiber density, whereas MD is correlated with microscopic membrane disruption and extracel‐ lular fluid accumulation.

It has been shown that JME patients have symmetric FA reductions in the bilateral superior and anterior corona radiata, genu and body of the corpus callosum, along and with middle and superior frontal tracts [139]. White matter tract MD increases coincided with the FA reduction, although the left superior parietal lobe was also affected. There were no increases in FA or decreases in MD found in JME patients. Frequency of GTCS was correlated with the severity of the microstructural changes.

compared five GGE patients, who died of SUDEP, to a nonepileptic control group [194]. This study did not find an increase of the "microdysplasia". Based upon the small numbers of patients with heterogeneous electroclinical features and syndromal diagnoses, it is difficult to

**Neuroimaging**: In contrast to pathology studies, a number of morphometric imaging ap‐ proaches since the 1990s have contributed to the neurobiological understanding of GGE and JME. Such studies have demonstrated differences between JME patients and normal controls, and suggest that there are subtle pathological changes difficult to assess by routine histology. Further, while JME manifests as generalized electroclinical activity, structural imaging has indicated circumscribed, multiregional abnormalities. Similarly, functional imaging studies have also demonstrated a predominance of symmetrical multiregional involvement (i.e., cerebral blood flow [CBF] or blood oxygen level dependent [BOLD] changes) prior to and

*Structural Neuroimaging:* Routine clinical structural imaging scans, whether computerized tomography (CT) or magnetic resonance imaging (MRI), are considered normal in individuals with JME. In contrast, more sensitive techniques, such as voxel-based morphometry (VBM), can show subtle variations in gray matter concentration or volume that differ in JME patients compared with healthy controls. An early quantitative MRI volumetric study in GGE partici‐ pants showed increases in normalized gray matter volume (GMV) compared to healthy controls, but also possible differences between other epilepsy syndromes as well [195]. In a subsequent VBM study, statistical parametric mapping demonstrated increased mesial frontal GMV in JME patients. Individual patients also demonstrated additional changes in parieto‐ temporal regions along with decreases frontally in individual patients analyzed by a volume

More recent VBM studies in JME patients have demonstrated multiregional volumetric and morphometric brain changes [196, 197]. Lin et al. [196] showed increased right frontal GMV and but reduced bilateral thalami, insula, and cerebellar hemisphere volume. Differences in GMV as a function of photosensitivity were also detected, as there were decreases in occipital lobe, left inferior frontal lobe, and hippocampal GMV in photosensitive patients compared with non-photosensitive JME patients. Altered cortical morphology differences in terms of surface area metrics, and not cortical thickness, was noted in the left hemisphere: insula, cingulate, occipital pole, middle temporal, and fusiform gyri [197]. Similar differences were also noted in the right hemisphere: insula, inferior temporal gyrus, and precuneus. Mean cortical curvature measurements in JME patients were also different from controls in the

Diffusion tensor imaging (DTI) provides a complementary approach to traditional morpho‐ metric and volumetric analyses, as it accounts for white matter connectivity in cortical and subcortical regions. The most important parameters assessed by DTI are fractional anisotropy (FA) and mean diffusivity (MD), which provide information about the microstructural integrity of the white matter pathways. In particular, FA is influenced by myelin integrity and fiber density, whereas MD is correlated with microscopic membrane disruption and extracel‐

draw any conclusions regarding pathology findings that are specific to JME.

during absence seizures or prolonged spike-and-wave discharges.

bilateral insula, left cingulate, and right inferior temporal gyrus.

of interest method [195].

104 Epilepsy Topics

lular fluid accumulation.

Several studies have shown similar reductions of FA [141] and increases in MD [198], mainly concentrated in thalamocortical pathways and the corpus callosum; pathways that are responsible for propagation and synchronization of spike-and-wave discharges. These DTI changes have been correlated with GMV reduction in the supplementary motor area (SMA) and posterior cingulate cortices. The SMA FA was correlated with deficits for word naming tasks and expression, while the posterior cingulate changes predicted cognitive inhibition scores [141] and deficits in frontal lobe executive motor functioning. A subsequent study by the same group demonstrated a correlation between DTI and fMRI-based measures of structural and functional connectivity between prefrontal cognitive cortex and motor cortices [198]. Connectivity was decreased between prefrontal and frontopolar regions and increased between occipital cortex and SMA. The authors suggested that the frontal connectivity changes may be related to cognitive deficits seen in JME, and occipitomotor connectivity may be related to photosensitivity, although association with these clinical phenotypes were not investigated.

The neuroanatomic basis of GMV and DTI changes is not clear. It is suspected that the decreased connectivity may be due to altered cortical organization, but as many of these changes are correlated with disease activity, degenerative changes cannot be ruled out. So far, the studies have shown disparate findings, thus, leading to an inconclusive picture of cerebral differences in JME. However, the aggregate of study results suggests diffuse abnormalities with morphological changes that include regions beyond the frontal lobes. Moreover, patho‐ logical studies are needed to confirm changes in cortical organization and connectivity. To improve on this knowledge base, an animal model of JME could be utilized to investigate a number of clinicopathological data points including morphometric changes and histopatho‐ logic findings.

*Functional – Interictal PET, MRS, EEG-fMRI*: Interictal PET studies using flourodeoxyglucose (FDG) evaluate metabolism over a 40-minute uptake time. While FDG-PET is not ideal for evaluating transient ictal state or interictal discharges it does evaluate a more chronic func‐ tional status. In an early study, Swartz et al. [125] indicated that JME patients showed decreased metabolism in the prefrontal and premotor cortices, which was associated with behavioral and cognitive dysfunction. Increased metabolism in the bilateral thalami was correlated with increased ictal or interictal epileptic discharges in another study of JME patients [199]. These studies demonstrated predominantly frontal lobe dysfunction, and an obvious activation of the thalami by ictal and interictal epileptic discharges. The absence of cortical activations suggests more fluctuation in metabolism due to epileptic discharges, decreased neuronal activity in the interictal state, or a general decrease of metabolism due to the disruptive effect of epileptic discharges on cognitive activity.

Magnetic resonance spectroscopy (MRS) can be used to measure neuronal function and concentrations of neurotransmitters in the cerebral cortex and subcortical structures. One study showed reduced N-acetyl aspartate (NAA) in the medial frontal lobes (and not in the occipital lobes) of individuals with JME [200]. The same group [152] showed NAA reduction was specific to JME as compared to individuals diagnosed with GTCS upon Awakening, but was similar to patients with CAE and JAE, [201]. However, investigation of cortical regions was limited and did not involve the lateral frontal, sensorimotor, or parietal areas.

frequently as the stimulus direction and intensity can be optimized to specifically activate hand muscles. TMS can be applied using single or paired pulses. Single pulse TMS is used to measure the resting motor threshold and the cortical silent period (CSP). Furthermore, the paired pulse stimulation paradigm is used measure short-latency intracortical inhibition and facilitation.

Juvenile Myoclonic Epilepsy — A Maturation Syndrome Coming of Age

http://dx.doi.org/10.5772/58372

107

Early studies demonstrated an increase in CSP in individuals with GGE, suggesting increased intracortical inhibition [204]. Sleep deprivation in individuals with GGE resulted in a greater change in cortical excitability compared to patients with partial epilepsies or controls, based upon paired-pulse studies [205]. ILS appears to decrease CSP in healthy controls, but not in photosensitive or nonphotosensitive individuals with GGE [206]. These findings are somewhat counterintuitive, as it would be expected that hyperexcitability would be observed in the motor

Antiepileptic medications can reduce cortical excitability, but there also seems to be variable findings across studies of various GGE syndromes and stimulation protocols. A meta-analysis, showed normal rMT threshold in all GGE patients, except for individuals with JME, who had decreased motor thresholds compared to healthy controls [207], suggesting a degree of cortical excitability in the patients. In a subsequent study, the same group showed that rMT was decreased for photosensitive individuals with GGE compared to those without, although this may be due to syndrome-specific bias, as photosensitivity is most commonly associated with

**Animal Models**: Due to its high rate of photosensitivity, the bamboo species, *Papio hamadryas papio*, has long been investigated as a model for epileptic photosensitivity [209]. Researchers established the electrophysiological mechanisms of the model by intracranial depth electrode recordings during photoepileptic responses that occurred with or without lesioned pathways. The ictal and interictal epileptic discharges have been shown to conform to a generalized cerebral distribution, and photic stimulation in the baboons elicits both myoclonic seizures and GTCS. Unlike humans, who respond to ILS at different frequencies (i.e., 12-18Hz), this species is maximally photosensitive at frequencies of 20-25 Hz. However, like humans, they are the most sensitivity to seizure activity in mornings, shortly after awakening. In order to study the seizures in a controlled way, researchers typically elicit activation by ILS, often after pretreatment with allylglycine, a GABA synthesis inhibitor, or by administering proconvul‐

Observations of the baboon's photoparoxysmal responses have indicated that there are generalized spike-and-wave complexes that occur maximally in the frontocentral regions, particularly in the mesial frontal surfaces [43]. During ILS, the occipital driving responses remain localized to occipito-parietal areas with occipital IEDs being rare and temporally unsustained. Fischer-Williams's et al. [43] early study showed that subcortical structures, thalamus, basal ganglia, and brainstem were only secondarily affected by frontocentral IEDs, usually in association with high-amplitude or repetitive discharges. The amygdala, hippo‐

In addition, paroxysmal visually evoked potential (PVEP) studies, using single flash stimuli administered after a 10 second train of 25 Hz ILS, have enabled researchers to track discrete

campus, and uncus were not involved in the photoepileptic responses.

cortices of GGE patients.

sants (e.g., pentylenetetrazol) [210].

JME [208].

Functional connectivity is a novel method evaluating regional covariance of BOLD signal over time. Most of the functional connectivity data is derived from generalized spike-and-wave discharges in individuals with absence epilepsy. Such studies have examined BOLD changes before, during, and after spike-and-wave discharges. While the thalamocortical networks projecting mainly to the medial frontal lobe are briefly activated during these discharges, more sustained activations occur in the parietal lobes prior to the discharges. As the discharges evolve, increases in BOLD dissipate, resulting in decreased frontoparietal BOLD signal. The brief discharges associated with JME are less likely to cause similar degrees of BOLD changes. Therefore, evaluating more stable measures such as functional connectivity, can provide more information regarding the networks.

One group has contributed greatly to evaluating connectivity both structurally and function‐ ally in individuals with JME [141, 198]. As mentioned above, they have demonstrated that functional connectivity is closely correlated with structural connectivity in the medial frontal cortices in individuals with JME [198]. They showed that during verbal fluency tasks, there is diffuse symmetrical activation of the SMA, bilateral inferior frontal gyri, left premotor area, left thalamus and bilateral putamen, as well as bilateral ventral visual areas. The thalamic region, showing altered connectivity on DTI, was connected to cortices largely overlapping the areas of functional activation. Overall, the individuals with JME showed increased taskdependent connectivity with respect to the frontal cortices compared to controls. Based upon these findings, the authors concluded that the thalamus serves an important function in increased frontal connectivity and coherence.

However, another study of functional connectivity in GGE, which included a large number of patients with JME, showed that there was no alteration of thalamic or mesial frontal connec‐ tivity when no discharges occurred [202]. Although the investigative techniques have differed between studies and selection biases may be present, the latter study suggests that generalized spike-and-wave discharges may be generated by healthy networks in response to abnormal connectivity and cortical synchronization in disparate brain regions. Indeed, another study showed alterations in default mode network (DMN) connectivity (in the absence of ictal or interictal epileptic discharges) between posterior and anterior nodes in patients with GGE [203]. Given the current limitations, additional methods, such as intracranial electrophysio‐ logical recordings and subsequent pathological examination would help elucidate the mechanisms underlying the CBF changes.

**Transcranial Magnetic Stimulation**: Transcranial magnetic stimulation (TMS) leads to a brief depolarization of cortical neurons, and can target brain regions such as the motor and language cortices. Motor cortex stimulation, such as the primary hand motor area, is conducted frequently as the stimulus direction and intensity can be optimized to specifically activate hand muscles. TMS can be applied using single or paired pulses. Single pulse TMS is used to measure the resting motor threshold and the cortical silent period (CSP). Furthermore, the paired pulse stimulation paradigm is used measure short-latency intracortical inhibition and facilitation.

Magnetic resonance spectroscopy (MRS) can be used to measure neuronal function and concentrations of neurotransmitters in the cerebral cortex and subcortical structures. One study showed reduced N-acetyl aspartate (NAA) in the medial frontal lobes (and not in the occipital lobes) of individuals with JME [200]. The same group [152] showed NAA reduction was specific to JME as compared to individuals diagnosed with GTCS upon Awakening, but was similar to patients with CAE and JAE, [201]. However, investigation of cortical regions

Functional connectivity is a novel method evaluating regional covariance of BOLD signal over time. Most of the functional connectivity data is derived from generalized spike-and-wave discharges in individuals with absence epilepsy. Such studies have examined BOLD changes before, during, and after spike-and-wave discharges. While the thalamocortical networks projecting mainly to the medial frontal lobe are briefly activated during these discharges, more sustained activations occur in the parietal lobes prior to the discharges. As the discharges evolve, increases in BOLD dissipate, resulting in decreased frontoparietal BOLD signal. The brief discharges associated with JME are less likely to cause similar degrees of BOLD changes. Therefore, evaluating more stable measures such as functional connectivity, can provide more

One group has contributed greatly to evaluating connectivity both structurally and function‐ ally in individuals with JME [141, 198]. As mentioned above, they have demonstrated that functional connectivity is closely correlated with structural connectivity in the medial frontal cortices in individuals with JME [198]. They showed that during verbal fluency tasks, there is diffuse symmetrical activation of the SMA, bilateral inferior frontal gyri, left premotor area, left thalamus and bilateral putamen, as well as bilateral ventral visual areas. The thalamic region, showing altered connectivity on DTI, was connected to cortices largely overlapping the areas of functional activation. Overall, the individuals with JME showed increased taskdependent connectivity with respect to the frontal cortices compared to controls. Based upon these findings, the authors concluded that the thalamus serves an important function in

However, another study of functional connectivity in GGE, which included a large number of patients with JME, showed that there was no alteration of thalamic or mesial frontal connec‐ tivity when no discharges occurred [202]. Although the investigative techniques have differed between studies and selection biases may be present, the latter study suggests that generalized spike-and-wave discharges may be generated by healthy networks in response to abnormal connectivity and cortical synchronization in disparate brain regions. Indeed, another study showed alterations in default mode network (DMN) connectivity (in the absence of ictal or interictal epileptic discharges) between posterior and anterior nodes in patients with GGE [203]. Given the current limitations, additional methods, such as intracranial electrophysio‐ logical recordings and subsequent pathological examination would help elucidate the

**Transcranial Magnetic Stimulation**: Transcranial magnetic stimulation (TMS) leads to a brief depolarization of cortical neurons, and can target brain regions such as the motor and language cortices. Motor cortex stimulation, such as the primary hand motor area, is conducted

was limited and did not involve the lateral frontal, sensorimotor, or parietal areas.

information regarding the networks.

106 Epilepsy Topics

increased frontal connectivity and coherence.

mechanisms underlying the CBF changes.

Early studies demonstrated an increase in CSP in individuals with GGE, suggesting increased intracortical inhibition [204]. Sleep deprivation in individuals with GGE resulted in a greater change in cortical excitability compared to patients with partial epilepsies or controls, based upon paired-pulse studies [205]. ILS appears to decrease CSP in healthy controls, but not in photosensitive or nonphotosensitive individuals with GGE [206]. These findings are somewhat counterintuitive, as it would be expected that hyperexcitability would be observed in the motor cortices of GGE patients.

Antiepileptic medications can reduce cortical excitability, but there also seems to be variable findings across studies of various GGE syndromes and stimulation protocols. A meta-analysis, showed normal rMT threshold in all GGE patients, except for individuals with JME, who had decreased motor thresholds compared to healthy controls [207], suggesting a degree of cortical excitability in the patients. In a subsequent study, the same group showed that rMT was decreased for photosensitive individuals with GGE compared to those without, although this may be due to syndrome-specific bias, as photosensitivity is most commonly associated with JME [208].

**Animal Models**: Due to its high rate of photosensitivity, the bamboo species, *Papio hamadryas papio*, has long been investigated as a model for epileptic photosensitivity [209]. Researchers established the electrophysiological mechanisms of the model by intracranial depth electrode recordings during photoepileptic responses that occurred with or without lesioned pathways. The ictal and interictal epileptic discharges have been shown to conform to a generalized cerebral distribution, and photic stimulation in the baboons elicits both myoclonic seizures and GTCS. Unlike humans, who respond to ILS at different frequencies (i.e., 12-18Hz), this species is maximally photosensitive at frequencies of 20-25 Hz. However, like humans, they are the most sensitivity to seizure activity in mornings, shortly after awakening. In order to study the seizures in a controlled way, researchers typically elicit activation by ILS, often after pretreatment with allylglycine, a GABA synthesis inhibitor, or by administering proconvul‐ sants (e.g., pentylenetetrazol) [210].

Observations of the baboon's photoparoxysmal responses have indicated that there are generalized spike-and-wave complexes that occur maximally in the frontocentral regions, particularly in the mesial frontal surfaces [43]. During ILS, the occipital driving responses remain localized to occipito-parietal areas with occipital IEDs being rare and temporally unsustained. Fischer-Williams's et al. [43] early study showed that subcortical structures, thalamus, basal ganglia, and brainstem were only secondarily affected by frontocentral IEDs, usually in association with high-amplitude or repetitive discharges. The amygdala, hippo‐ campus, and uncus were not involved in the photoepileptic responses.

In addition, paroxysmal visually evoked potential (PVEP) studies, using single flash stimuli administered after a 10 second train of 25 Hz ILS, have enabled researchers to track discrete electrophysiological events through cortical-subcortical networks. In one study, earliest activation of frontocentral IEDs occurred 20 to 30 milliseconds after the flash, with a subse‐ quent eyelid or facial myoclonus occurring 10 to 12 milliseconds later [211]. Motor symptoms tend to occur only when the amplitude of the cortical discharges exceeded 200 microvolts [212]. Following a PVEP, activation of thalamic nuclei (e.g., ventralis lateralis, centrum medianum, & lateralis posterior) occurs at cortical discharge amplitudes exceeding 400 microvolts. Silva-Barrat et al. [213] also demonstrated that photosensitive and asymptomatic control baboons show difference PVEP responses in peristriate and parietal regions, but not in the striate or cortical responses. It was also shown that frontocentral ictal or interictal epileptic discharges were not activated by ILS following bilateral occipital lobe ablation, while destruction of the superior colliculus or pulvinar unilaterally caused only transient suppression of photoepileptic responses [214]. In terms of other lesion study, corpus callosum sectioning in combination with unilateral stimulation of the occipital lobe resulted in frontocentral discharges and seizures that remained ipsilateral to the activated occipital lobe [215].

included the parieto-occipital region, parietal lobe, premotor area, and orbitofrontal cortex. This was unexpected as the scalp EEG demonstrated only generalized discharges, and the multifocal discharges appeared to occur both parietally and frontally, reflecting a more diffuse

Juvenile Myoclonic Epilepsy — A Maturation Syndrome Coming of Age

http://dx.doi.org/10.5772/58372

109

Previous examinations of pathology have not shown cortical abnormalities in these animals, suggesting developmental changes or seizure-induced injury [225]. Comparison of neuronal counts from the molecular layer, cortical layer 6 and subcortically, in three sulci, namely the cingulate, intraparietal, and lunate sulcus between asymptomatic photosensitive and healthy control baboons did not demonstrate cortical neuronal reductions, but did detect increased numbers of neurons in the subcortical white matter of the anterior cingulate sulcus [226]. Neuronal flow cytometry, on the other hand, demonstrated cortical neuronal reductions, particularly in the frontal regions related to motor functions and somatosensory cortex [227]. The motor cortices most involved were the face and hand regions, areas most commonly involved in the manifestation of myoclonic seizures. The decreased cell counts are likely to result in decreased functional connectivity, especially in the motor cortices. The electrophy‐ siological effects of the neuron reductions are unclear, but may be related to the extent that inhibitory and excitatory neurons are affected. More detailed pathology for specific types of neurons needs to be pursued, and a qualitative evaluation of cortical organization also needs

JME is a commonly occurring and electroclinically well-defined neurological disease. As with most GGEs, it is relatively straightforward to diagnose and treat; however, there is still uncertainty regarding features which may influence a poor response to antiepileptic medica‐ tions and complicate overall prognosis. Compared to other types of GGE, JME appears to have a structural etiology. Thus, although it shares commonalities with other GGEs, JME breaks that mold, as the condition has been associated with multiregional and/or asymmetric electroclinical findings. While the nature and development of that structural and functional etiology does not appear to be frontal lobe specific, research has yet to define a replicable model for the disease. Moreover, as JME sub-syndromes [228] may have overlapping features with other epilepsies, establishment of cognitive signs and symptoms that are specific to JME has

As a result of the apparent time-course of syndrome onset along with differing prevalence and prognosis across the lifespan, in conjunction with the subtle cerebral abnormalities as noted above, some cases of JME likely reflect a set of underlying neurodevelopmental processes. Nevertheless, there remains a lack of knowledge regarding histopathology, electroclinicial characteristics and propagation patterns, genetic contributors to phenotypic expression, and disease biomarkers. Given its similarities to human GGE syndromes (i.e., similar seizure types and electrophysiological characteristics) and methodical constraints in human research, recent contributions from the animal literature suggest that continued investigation of these factors

pathology.

to be considered.

**17. Conclusions**

continued to elude researchers.

Although the epilepsy in these baboons appears similar to JME in many ways, early researchers failed to establish a model for JME from characteristics seen with the baboons. However, the phenotypic expression of several hundred baboons has recently been described clinically and with scalp EEG [216]. The onset of the syndrome typically occurs in adolescence, there is a preponderance of myoclonic and GTCS, and there is a similar diurnal pattern (i.e., predomi‐ nately in the morning) to JME [217]. The myoclonic seizures tend to affect the face and arms. The GTCS tend to occur in a sporadic fashion, and can be triggered by handling or ketamine. The interictal epileptic discharges are generalized in distribution with 4-6 Hz in frequency, although 2-3 Hz discharges have been noted in younger baboons [218]. Younger baboons can also present with absence seizures, and even spike-and-wave stupor [216]. Although the baboon model of epilepsy and JME share many clinical characteristics, including a genetic etiology, the mode of inheritance and underlying genotypes are not known.

Szabó et al. [219, 220] have also evaluated structural and functional neuroimaging in a large baboon colony. Their findings indicated that while routine structural MRI studies are normal, morphometric analyses revealed decreased central and intraparietal sulci along with cingulate sulci in baboons that showed IEDs [221]. The smaller sulcal areas may reflect an underlying developmental anomaly, resulting in decreased U-fibers, rather than a process due to aging or degenerative disease.

Functional studies have aimed to describe epileptic neuronal networks. IED rate and associ‐ ated cerebral blood flow changes on H2 15O-PET have demonstrated co-activations of the premotor, perirolandic, insular-parietal, and occipital cortices, areas that are also observed in human GGE networks [222]. Resting state fMRI has shown altered connectivity of the motor, but not the visual cortices, and DMN connectivity was altered in the epileptic baboons as well [223]. This group has also used intracranial video-EEG in combination with depth, grid, and strip electrodes in order to improve spatial resolution [224]. Results from this approach indicated that frequent multiregional IEDs occur, which appears to trigger generalized discharges. The researchers also recorded myoclonic and GTCS, most of which were triggered multifocally. The focal regions that were most active, according to invasive monitoring, included the parieto-occipital region, parietal lobe, premotor area, and orbitofrontal cortex. This was unexpected as the scalp EEG demonstrated only generalized discharges, and the multifocal discharges appeared to occur both parietally and frontally, reflecting a more diffuse pathology.

Previous examinations of pathology have not shown cortical abnormalities in these animals, suggesting developmental changes or seizure-induced injury [225]. Comparison of neuronal counts from the molecular layer, cortical layer 6 and subcortically, in three sulci, namely the cingulate, intraparietal, and lunate sulcus between asymptomatic photosensitive and healthy control baboons did not demonstrate cortical neuronal reductions, but did detect increased numbers of neurons in the subcortical white matter of the anterior cingulate sulcus [226]. Neuronal flow cytometry, on the other hand, demonstrated cortical neuronal reductions, particularly in the frontal regions related to motor functions and somatosensory cortex [227]. The motor cortices most involved were the face and hand regions, areas most commonly involved in the manifestation of myoclonic seizures. The decreased cell counts are likely to result in decreased functional connectivity, especially in the motor cortices. The electrophy‐ siological effects of the neuron reductions are unclear, but may be related to the extent that inhibitory and excitatory neurons are affected. More detailed pathology for specific types of neurons needs to be pursued, and a qualitative evaluation of cortical organization also needs to be considered.

#### **17. Conclusions**

electrophysiological events through cortical-subcortical networks. In one study, earliest activation of frontocentral IEDs occurred 20 to 30 milliseconds after the flash, with a subse‐ quent eyelid or facial myoclonus occurring 10 to 12 milliseconds later [211]. Motor symptoms tend to occur only when the amplitude of the cortical discharges exceeded 200 microvolts [212]. Following a PVEP, activation of thalamic nuclei (e.g., ventralis lateralis, centrum medianum, & lateralis posterior) occurs at cortical discharge amplitudes exceeding 400 microvolts. Silva-Barrat et al. [213] also demonstrated that photosensitive and asymptomatic control baboons show difference PVEP responses in peristriate and parietal regions, but not in the striate or cortical responses. It was also shown that frontocentral ictal or interictal epileptic discharges were not activated by ILS following bilateral occipital lobe ablation, while destruction of the superior colliculus or pulvinar unilaterally caused only transient suppression of photoepileptic responses [214]. In terms of other lesion study, corpus callosum sectioning in combination with unilateral stimulation of the occipital lobe resulted in frontocentral discharges and seizures

Although the epilepsy in these baboons appears similar to JME in many ways, early researchers failed to establish a model for JME from characteristics seen with the baboons. However, the phenotypic expression of several hundred baboons has recently been described clinically and with scalp EEG [216]. The onset of the syndrome typically occurs in adolescence, there is a preponderance of myoclonic and GTCS, and there is a similar diurnal pattern (i.e., predomi‐ nately in the morning) to JME [217]. The myoclonic seizures tend to affect the face and arms. The GTCS tend to occur in a sporadic fashion, and can be triggered by handling or ketamine. The interictal epileptic discharges are generalized in distribution with 4-6 Hz in frequency, although 2-3 Hz discharges have been noted in younger baboons [218]. Younger baboons can also present with absence seizures, and even spike-and-wave stupor [216]. Although the baboon model of epilepsy and JME share many clinical characteristics, including a genetic

Szabó et al. [219, 220] have also evaluated structural and functional neuroimaging in a large baboon colony. Their findings indicated that while routine structural MRI studies are normal, morphometric analyses revealed decreased central and intraparietal sulci along with cingulate sulci in baboons that showed IEDs [221]. The smaller sulcal areas may reflect an underlying developmental anomaly, resulting in decreased U-fibers, rather than a process due to aging or

Functional studies have aimed to describe epileptic neuronal networks. IED rate and associ‐

premotor, perirolandic, insular-parietal, and occipital cortices, areas that are also observed in human GGE networks [222]. Resting state fMRI has shown altered connectivity of the motor, but not the visual cortices, and DMN connectivity was altered in the epileptic baboons as well [223]. This group has also used intracranial video-EEG in combination with depth, grid, and strip electrodes in order to improve spatial resolution [224]. Results from this approach indicated that frequent multiregional IEDs occur, which appears to trigger generalized discharges. The researchers also recorded myoclonic and GTCS, most of which were triggered multifocally. The focal regions that were most active, according to invasive monitoring,

15O-PET have demonstrated co-activations of the

etiology, the mode of inheritance and underlying genotypes are not known.

degenerative disease.

108 Epilepsy Topics

ated cerebral blood flow changes on H2

that remained ipsilateral to the activated occipital lobe [215].

JME is a commonly occurring and electroclinically well-defined neurological disease. As with most GGEs, it is relatively straightforward to diagnose and treat; however, there is still uncertainty regarding features which may influence a poor response to antiepileptic medica‐ tions and complicate overall prognosis. Compared to other types of GGE, JME appears to have a structural etiology. Thus, although it shares commonalities with other GGEs, JME breaks that mold, as the condition has been associated with multiregional and/or asymmetric electroclinical findings. While the nature and development of that structural and functional etiology does not appear to be frontal lobe specific, research has yet to define a replicable model for the disease. Moreover, as JME sub-syndromes [228] may have overlapping features with other epilepsies, establishment of cognitive signs and symptoms that are specific to JME has continued to elude researchers.

As a result of the apparent time-course of syndrome onset along with differing prevalence and prognosis across the lifespan, in conjunction with the subtle cerebral abnormalities as noted above, some cases of JME likely reflect a set of underlying neurodevelopmental processes. Nevertheless, there remains a lack of knowledge regarding histopathology, electroclinicial characteristics and propagation patterns, genetic contributors to phenotypic expression, and disease biomarkers. Given its similarities to human GGE syndromes (i.e., similar seizure types and electrophysiological characteristics) and methodical constraints in human research, recent contributions from the animal literature suggest that continued investigation of these factors in a baboon population will offer a unique avenue to further refine human models of JME, particularly in the setting of photosensitivity. Furthermore, pharmacological development would likely be assisted by employing trials of agents in animals that show behaviors consis‐ tent with human phenotypes.

[8] Hauser WA, Annegers JF, Rocca WA. Descriptive epidemiology of epilepsy: contri‐ butions of population-based studies from Rochester, Minnesota. Mayo Clinic Pro‐

Juvenile Myoclonic Epilepsy — A Maturation Syndrome Coming of Age

http://dx.doi.org/10.5772/58372

111

[9] Panayiotopoulos CP, Obeid T, Tahan AR. Juvenile myoclonic epilepsy: a 5-year pro‐

[10] Figueredo R, Trevisol-Bittencourt PC, Ferro JB. [Clinical-epidemiological study of pa‐ tients with juvenile myoclonic epilepsy in Santa Catarina State, Brazil]. Arquivos De

[11] Kun LN, Ling LW, Wah YW, Lian TT. Epidemiologic study of epilepsy in young Sin‐

[12] Murthy JM, Yangala R, Srinivas M. The syndromic classification of the International League Against Epilepsy: a hospital-based study from South India. Epilepsia.

[13] Janz D, Durner M. Juvenile Myoclonic Epilepsy. In: Engel J, Pedley, T.A., editor. Epi‐ lepsy: a comprehensive textbook. Philadelphia, PA: Lippincott-Raven Press; 1998. p.

[14] Camfield CS, Camfield PR. Juvenile myoclonic epilepsy 25 years after seizure onset:

[15] Camfield CS, Striano P, Camfield PR. Epidemiology of juvenile myoclonic epilepsy.

[16] Mullins GM, O'Sullivan SS, Neligan A, McCarthy A, McNamara B, Galvin RJ, et al. A study of idiopathic generalised epilepsy in an Irish population. Seizure. 2007;16(3):

[17] Manonmani V, Tan CT. A study of newly diagnosed epilepsy in Malaysia. Singapore

[18] Asadi-Pooya AA, Emami M, Sperling MR. A clinical study of syndromes of idiopath‐ ic (genetic) generalized epilepsy. Journal of the Neurological Sciences. 2013;324(1-2):

[19] Wirrell EC, Camfield CS, Camfield PR, Gordon KE, Dooley JM. Long-term prognosis of typical childhood absence epilepsy: remission or progression to juvenile myoclon‐

[20] Obeid T, Panayiotopoulos CP. Juvenile myoclonic epilepsy: a study in Saudi Arabia.

[21] Sozmen V, Baybas S, Dirican A, Koksal A, Ozturk M. Frequency of epilepsies in fami‐ ly members of patients with different epileptic syndromes. European Neurology.

a population-based study. Neurology. 2009;73(13):1041-5.

Epilepsy & Behavior. 2013;28(Suppl 1):S15-S7.

Medical Journal. 1999;40(1):32-5.

ic epilepsy. Neurology. 1996;47(4):912-8.

Epilepsia. 1988;29(3):280-2.

2011;65(1):4-9.

ceedings Mayo Clinic. 1996;71(6):576-86.

Neuro-Psiquiatria. 1999;57(2B):401-4.

1998;39(1):48-54.

2389–400.

204-10.

113-7.

spective study. Epilepsia. 1994;35(2):285-96.

gaporean men. Epilepsia. 1999;40(10):1384-7.

In terms of treatment, various combinations of agents have been shown effective, although some cases of JME result in intractability. Overall, regardless of intractability, JME has the potential for negatively impacting quality of life. For instance, recent studies indicated discernible interictal effects on cognition and behavior, even in patients that are relatively wellcontrolled. Thus, contrary to previous assertions and clinical lore, JME does not appear to be a "benign" disease.

#### **Author details**

Russell D. Pella, Lakshmi Mukundan and C. Akos Szabo\*

\*Address all correspondence to: szabo@uthscsa.edu

University of Texas Health Science Center at San Antonio, USA

#### **References**


[8] Hauser WA, Annegers JF, Rocca WA. Descriptive epidemiology of epilepsy: contri‐ butions of population-based studies from Rochester, Minnesota. Mayo Clinic Pro‐ ceedings Mayo Clinic. 1996;71(6):576-86.

in a baboon population will offer a unique avenue to further refine human models of JME, particularly in the setting of photosensitivity. Furthermore, pharmacological development would likely be assisted by employing trials of agents in animals that show behaviors consis‐

In terms of treatment, various combinations of agents have been shown effective, although some cases of JME result in intractability. Overall, regardless of intractability, JME has the potential for negatively impacting quality of life. For instance, recent studies indicated discernible interictal effects on cognition and behavior, even in patients that are relatively wellcontrolled. Thus, contrary to previous assertions and clinical lore, JME does not appear to be

[1] Herpin TH. Des asces incomplets de l'epilepsie. Paris, France: J Balliere et Fils; 1867.

[3] Janz D, Christian W. Impulsiv-petit mal. Deutsche Zeitschrift fu°r Nervenheilkunde

[4] Delgado-Escueta AV, Enrile-Bacsal F. Juvenile myoclonic epilepsy of Janz. Neurolo‐

[5] Lund M, Reintoft H, Simonsen N. [A controlled social and psychological investiga‐ tion of patients with juvenile myoclonic epilepsy]. Ugeskrift For Laeger. 1975;137(41):

[6] Commission on Classification and Terminology of the International League Against Epilepsy. Proposal for revised classification of epilepsies and epileptic syndromes.

[7] Genton P, Gelisse P. Juvenile myoclonic epilepsy. Archives of Neurology. 2001;58(9):

[2] Rabot LDL. De la myoclonie épileptique. Paris, France: Thèse; 1899.

tent with human phenotypes.

a "benign" disease.

110 Epilepsy Topics

**Author details**

**References**

1957;176:346–86.

2415-8.

1487-90.

gy. 1984;34(3):285-94.

Epilepsia. 1989;30(4):389-99.

Russell D. Pella, Lakshmi Mukundan and C. Akos Szabo\*

University of Texas Health Science Center at San Antonio, USA

\*Address all correspondence to: szabo@uthscsa.edu


[22] Grünewald RA, Chroni E, Panayiotopoulos CP. Delayed diagnosis of juvenile myo‐ clonic epilepsy. Journal of Neurology, Neurosurgery & Psychiatry. 1992;55(6):497-9.

[36] Wolf P. From precipitation to inhibition of seizures: rationale of a therapeutic para‐

Juvenile Myoclonic Epilepsy — A Maturation Syndrome Coming of Age

http://dx.doi.org/10.5772/58372

113

[37] Guaranha MS, da Silva Sousa P, de Araújo-Filho GM, Lin K, Guilhoto LM, Caboclo LO, et al. Provocative and inhibitory effects of a video-EEG neuropsychologic proto‐

[38] Beniczky S, Guaranha MSB, Conradsen I, Singh MB, Rutar V, Lorber B, et al. Modu‐ lation of epileptiform EEG discharges in juvenile myoclonic epilepsy: an investiga‐

[39] Matsuoka H, Takahashi T, Sasaki M, Matsumoto K, Yoshida S, Numachi Y, et al. Neuropsychological EEG activation in patients with epilepsy. Brain: A Journal of

[40] Kasteleijn-Nolst Trenité DGA. Provoked and reflex seizures: surprising or common?

[41] Mayer T, Wolf P. Perioral reflex myoclonias in reading epilepsy and juvenile myo‐ clonic epilepsy. In: Wolf P IY, Zifkin B editor. Reflex Epilepsies: Progress in Under‐

[42] Kasteleijn-Nolst Trenité DG. Photosensitivity in epilepsy. Electrophysiological and clinical correlates. Acta Neurologica Scandinavica Supplementum. 1989;125:3-149.

[43] Fischer-Williams M, Poncet M, Riche D, Naquet R. Light-induced epilepsy in the ba‐ boon, Papio papio: Cortical and depth recordings. Electroencephalography & Clini‐

[44] Parra J, Kalitzin SN, Iriarte J, Blanes W, Velis DN, Lopes da Silva FH. Gamma-band phase clustering and photosensitivity: is there an underlying mechanism common to photosensitive epilepsy and visual perception? Brain: A Journal of Neurology.

[45] Varotto GELSGF. Enhanced frontocentral EEG connectivity in photosensitive gener‐ alized epilepsies: A partial directed coherence study. Epilepsia. 2012;53(2):359-67.

[46] Moeller F, Siebner HR, Ahlgrimm N, Wolff S, Muhle H, Granert O, et al. fMRI activa‐ tion during spike and wave discharges evoked by photic stimulation. NeuroImage.

[47] Bai X, Vestal M, Berman R, Negishi M, Spann M, Vega C, et al. Dynamic time course of typical childhood absence seizures: EEG, behavior, and functional magnetic reso‐ nance imaging. The Journal of Neuroscience: The Official Journal of The Society for

[48] Benuzzi F, Mirandola L, Pugnaghi M, Farinelli V, Tassinari CA, Capovilla G, et al. Increased cortical BOLD signal anticipates generalized spike and wave discharges in

col in juvenile myoclonic epilepsy. Epilepsia. 2009;50(11):2446-55.

tion of reflex epileptic traits. Epilepsia. 2012;53(5):832-9.

standing. New Barnet, Herts, UK: John Libbey; 2004. p. 62–3.

digm. Epilepsia. 2005;46 Suppl 1:15-6.

Neurology. 2000;123(2):318-30.

Epilepsia. 2012;53 Suppl 4:105-13.

cal Neurophysiology. 1968;25(6):557-69.

2003;126(Pt 5):1164-72.

2009;48(4):682-95.

Neuroscience. 2010;30(17):5884-93.


[36] Wolf P. From precipitation to inhibition of seizures: rationale of a therapeutic para‐ digm. Epilepsia. 2005;46 Suppl 1:15-6.

[22] Grünewald RA, Chroni E, Panayiotopoulos CP. Delayed diagnosis of juvenile myo‐ clonic epilepsy. Journal of Neurology, Neurosurgery & Psychiatry. 1992;55(6):497-9.

[23] Gunatilake SB, Seneviratne SL. Juvenile myoclonic epilepsy: a study in Sri Lanka.

[24] Genton P, Gelisse P, Thomas P. Juvenile myoclonic epilepsy today. In: Schmitz B, Sander, T, editor. Juvenile Myoclonic Epilepsy: The Janz Syndrome. Berlin, Germany:

[25] Serratosa JM, Delgado-Escueta AV. Juvenile Myoclonic Epilepsy. In: Wyllie E, editor. The treatment of epilepsy: principles and practices. Philadelphia, PA: Lea & Febiger;

[26] Tóth V, Rásonyi G, Fogarasi A, Kovács N, Auer T, Janszky J. Juvenile myoclonic epi‐ lepsy starting in the eighth decade. Epileptic Disorders: International Epilepsy Jour‐

[27] Hirano Y, Oguni H, Funatsuka M, Imai K, Osawa M. Differentiation of myoclonic seizures in epileptic syndromes: A video-polygraphic study of 26 patients. Epilepsia.

[28] Kasteleijn-Nolst Trenité DGA, de Weerd A, Beniczky S. Chronodependency and pro‐ vocative factors in juvenile myoclonic epilepsy. Epilepsy & Behavior. 2013;28(Suppl

[29] Badawy RA, Macdonell RA, Jackson GD, Berkovic SF. Why do seizures in general‐

[30] Baykan B, Altindag EA, Bebek N, Ozturk AY, Aslantas B, Gurses C, et al. Myoclonic seizures subside in the fourth decade in juvenile myoclonic epilepsy. Neurology.

[31] Genton P, Thomas P, Kasteleijn-Nolst Trenité DGA, Medina MT, Salas-Puig J. Clini‐ cal aspects of juvenile myoclonic epilepsy. Epilepsy & Behavior. 2013;28(Suppl 1):S8-

[32] Sadleir LG, Scheffer IE, Smith S, Carstensen B, Carlin J, Connolly MB, et al. Factors influencing clinical features of absence seizures. Epilepsia. 2008;49(12):2100-7.

[33] Larch J, Unterberger I, Bauer G, Reichsoellner J, Kuchukhidze G, Trinka E. Myoclonic status epilepticus in juvenile myoclonic epilepsy. Epileptic Disorders: International

[34] Crespel A, Velizarova R, Genton P, Gélisse P. [Juvenile myoclonic epilepsy with re‐ current myoclonic status: efficacy of valproate]. Thérapie. 2009;64(5):321-3.

[35] da Silva Sousa P, Lin K, Garzon E, Sakamoto AC, Yacubian EMT. Self-perception of factors that precipitate or inhibit seizures in juvenile myoclonic epilepsy. Seizure.

Epilepsy Journal With Videotape. 2009;11(4):309-14.

ized epilepsy often occur in the morning? Neurology. 2009;73(3):218-22.

Seizure. 2000;9(3):221-3.

1993. p. 552-70.

112 Epilepsy Topics

2009;50(6):1525-35.

2008;70(22 Part 2):2123-9.

1):S25-S9.

S14.

2005;14(5):340-6.

Hippokrates-Verlag; 2000. p. 11-32.

nal With Videotape. 2007;9(3):341-5.


adolescents and adults with idiopathic generalized epilepsies. Epilepsia. 2012;53(4): 622-30.

[62] Lancman ME, Asconapé JJ, Penry JK. Clinical and EEG asymmetries in juvenile myo‐

Juvenile Myoclonic Epilepsy — A Maturation Syndrome Coming of Age

http://dx.doi.org/10.5772/58372

115

[63] Lancman ME, Asconapé JJ, Golimstok A. Circling seizures in a case of juvenile myo‐

[64] Ferrie CD. Idiopathic Generalized Epilepsies Imitating Focal Epilepsies. Epilepsia.

[65] Aliberti V, Grünewald RA, Panayiotopoulos CP, Chroni E. Focal electroencephalo‐ graphic abnormalities in juvenile myoclonic epilepsy. Epilepsia. 1994;35(2):297-301.

[66] Niedzielska K, Kuran W, Romaniak A. [Clinical-electrophysiological pattern of juve‐ nile myoclonic epilepsy]. Neurologia I Neurochirurgia Polska. 1995;29(5):675-85. [67] Jayalakshmi SS, Srinivasa Rao B, Sailaja S. Focal clinical and electroencephalographic features in patients with juvenile myoclonic epilepsy. Acta Neurologica Scandinavi‐

[68] Usui N, Kotagal P, Matsumoto R, Kellinghaus C, Lüders HO. Focal Semiologic and Electroencephalographic Features in Patients with Juvenile Myoclonic Epilepsy. Epi‐

[69] Létourneau K, Cieuta-Walti C, Deacon C. Epileptiform asymetries and treatment re‐ sponse in juvenile myoclonic epilepsy. The Canadian Journal of Neurological Scien‐

[70] Kelley SA, Kossoff EH. Doose syndrome (myoclonic-astatic epilepsy): 40 years of progress. Developmental Medicine & Child Neurology. 2010;52(11):988-93.

[71] Caraballo RH, Chamorro N, Darra F, Fortini S, Arroyo H. Epilepsy with myoclonic atonic seizures: an electroclinical study of 69 patients. Pediatric Neurology.

[72] Fisher RS, Boas WvE, Blume W, Elger C, Genton P, Lee P, et al. Epileptic Seizures and Epilepsy: Definitions Proposed by the International League Against Epilepsy (ILAE) and the International Bureau for Epilepsy (IBE). Epilepsia. 2005;46(4):470-2.

[73] Adie WJ. Pyknolepsy: a form of epilepsy in children, with a good prognosis. Sect

[74] Panayiotopoulos CP. Typical absence seizures and related epileptic syndromes: As‐ sessment of current state and directions for future research. Epilepsia. 2008;49(12):

[75] Tassinari CA, Bureau M. Epilepsy with myoclonic absences. In: Roger J DC, Bureau M, Dreifuss FE, Wolf P, editor. Epileptic syndromes in infancy, childhood and ado‐

lescence. London, England: John Libbery Eurotext; 1985. p. 121-9.

ces Le Journal Canadien Des Sciences Neurologiques. 2010;37(6):826-30.

clonic epilepsy. Epilepsia. 1994;35(2):302-6.

clonic epilepsy. Epilepsia. 1994;35(2):317-8.

2005;46:91-5.

ca. 2010;122(2):115-23.

2013;48(5):355-62.

Neur. 1924;17:19-25.

2131-9.

lepsia. 2005;46(10):1668-76.


[62] Lancman ME, Asconapé JJ, Penry JK. Clinical and EEG asymmetries in juvenile myo‐ clonic epilepsy. Epilepsia. 1994;35(2):302-6.

adolescents and adults with idiopathic generalized epilepsies. Epilepsia. 2012;53(4):

[49] Ferlazzo E, Zifkin BG, Andermann E, Andermann F. Cortical triggers in generalized reflex seizures and epilepsies. Brain: A Journal of Neurology. 2005;128(Part 4):700-10.

[50] Asconapé J, Penry JK. Some clinical and EEG aspects of benign juvenile myoclonic

[51] So GM, Thiele EA, Sanger T, Schmid R, Riviello JJ, Jr. Electroencephalogram and clin‐ ical focalities in juvenile myoclonic epilepsy. Journal of Child Neurology. 1998;13(11):

[52] Genton P, Gonzalez Sanchez MS, Saltarelli A, Bureau M, Dravet C, Roger J. [Mislead‐ ing aspects of the standard electroencephalogram in juvenile myoclonic epilepsy: a retrospective study of 56 consecutive cases]. Neurophysiologie clinique=Clinical neu‐

[53] Baise-Zung C, Guilhoto LMFF, Grossmann RM. Juvenile myoclonic epilepsy: nonclassic electroencephalographical presentation in adult patients. European Journal of

[54] Syvertsen MR, Markhus R, Selmer KK, Nakken KO. [Juvenile myoclonic epilepsy].

[55] Genton P, Puig A, Tunon C, Lahoz CH, Sanchez M. Juvenile myoclonic epilepsy and related syndromes: clinical and neurophysiological aspects. In: Malafosse A, Genton, P, Hirsch, E, Marescaux, C, Broglin, D, Bernasconi, R, editor. Idiopathic generalized epilepsies: clinical, experimental and genetic aspects. London, UK: John Libbey and

[56] Fittipaldi F, Currà A, Fusco L, Ruggieri S, Manfredi M. EEG discharges on awaken‐ ing: a marker of idiopathic generalized epilepsy. Neurology. 2001;56(1):123-6.

[57] Janz D. Epilepsy with impulsive petit mal (juvenile myoclonic epilepsy). Acta Neuro‐

[58] Delgado-Escueta AV, Greenberg DA, Treiman L, Liu A, Sparkes RS, Barbetti A, et al. Mapping the gene for juvenile myoclonic epilepsy. Epilepsia. 1989;30 Suppl 4:S8.

[59] Koeleman BPC, de Kovel CGF, Kasteleijn-Nolst Trenité DGA. Photoparoxysmal EEG response and genetic dissection of juvenile myoclonic epilepsy. Epilepsy & Behavior.

[60] Serafini A, Rubboli G, Gigli GL, Koutroumanidis M, Gelisse P. Neurophysiology of juvenile myoclonic epilepsy. Epilepsy & Behavior. 2013;28(Suppl 1):S30-S9.

[61] Panayiotopoulos CP. A clinical guide to epileptic syndromes and their treatment 2nd

622-30.

114 Epilepsy Topics

541-5.

epilepsy. Epilepsia. 1984;25(1):108-14.

rophysiology. 1995;25(5):283-90. Epub 1995.

Tidsskr Nor Laegeforen. 2012;132(14):1610-3.

Neurology. 2006;13(2):171-5.

Company Ltd; 1994. p. 253–65.

2013;28(Suppl 1):S69-S71.

logica Scandinavica. 1985;72(5):449-59.

ed. United Kingdom: Springer; 2007.


[76] Durner M, Sander T, Greenberg DA, Johnson K, Beck-Mannagetta G, Janz D. Locali‐ zation of idiopathic generalized epilepsy on chromosome 6p in families of juvenile myoclonic epilepsy patients. Neurology. 1991;41(10):1651-5.

[89] Duncan S. Teratogenesis of sodium valproate. Current Opinion in Neurology.

Juvenile Myoclonic Epilepsy — A Maturation Syndrome Coming of Age

http://dx.doi.org/10.5772/58372

117

[90] Morris GL, Hammer AE, Kustra RP, Messenheimer JA. Lamotrigine for patients with juvenile myoclonic epilepsy following prior treatment with valproate: results of an

[91] Levisohn PM, Holland KD. Topiramate or valproate in patients with juvenile myo‐ clonic epilepsy: A randomized open-label comparison. Epilepsy & Behavior.

[92] Park KM, Kim SH, Nho SK, Shin KJ, Park J, Ha SY, et al. A randomized open-label observational study to compare the efficacy and tolerability between topiramate and valproate in juvenile myoclonic epilepsy. Journal of Clinical Neuroscience: Official

[93] Specchio LM, Gambardella A, Giallonardo AT, Michelucci R, Specchio N, Boero G, et al. Open label, long-term, pragmatic study on levetiracetam in the treatment of juve‐

[94] Noachtar S, Andermann E, Meyvisch P, Andermann F, Gough WB, Schiemann-Del‐ gado J. Levetiracetam for the treatment of idiopathic generalized epilepsy with myo‐

[95] Rosenfeld WE, Benbadis S, Edrich P, Tassinari CA, Hirsch E. Levetiracetam as addon therapy for idiopathic generalized epilepsy syndromes with onset during adoles‐ cence: analysis of two randomized, double-blind, placebo-controlled studies.

[96] Kothare SV, Valencia I, Khurana DS, Hardison H, Melvin JJ, Legido A. Efficacy and tolerability of zonisamide in juvenile myoclonic epilepsy. Epileptic Disorders: Inter‐

[97] Genton P, Gelisse P, Thomas P, Dravet C. Do carbamazepine and phenytoin aggra‐

[98] Gelisse P, Genton P, Kuate C, Pesenti A, Baldy-Moulinier M, Crespel A. Worsening of Seizures by Oxcarbazepine in Juvenile Idiopathic Generalized Epilepsies. Epilep‐

[99] Sharpe C, Buchanan N. Juvenile myoclonic epilepsy: diagnosis, management and

[100] Mantoan L, Walker M. Treatment options in juvenile myoclonic epilepsy. Current

[101] Penry JK, Dean JC, Riela AR. Juvenile myoclonic epilepsy: long-term response to

Journal of the Neurosurgical Society of Australasia. 2013;20(8):1079-82.

nile myoclonic epilepsy. Epilepsy Research. 2006;71(1):32-9.

national Epilepsy Journal With Videotape. 2004;6(4):267-70.

vate juvenile myoclonic epilepsy? Neurology. 2000;55(8):1106-9.

outcome. The Medical Journal of Australia. 1995;162(3):133-4.

Treatment Options In Neurology. 2011;13(4):355-70.

therapy. Epilepsia. 1989;30 Suppl 4:S19.

clonic seizures. Neurology. 2008;70(8):607-16.

Epilepsy Research. 2009;85(1):72-80.

sia. 2004;45(10):1282-6.

open-label study. Epilepsy & Behavior. 2004;5(4):509-12.

2007;20(2):175-80.

2007;10(4):547-52.


[89] Duncan S. Teratogenesis of sodium valproate. Current Opinion in Neurology. 2007;20(2):175-80.

[76] Durner M, Sander T, Greenberg DA, Johnson K, Beck-Mannagetta G, Janz D. Locali‐ zation of idiopathic generalized epilepsy on chromosome 6p in families of juvenile

[77] Berkovic SF, So NK, Andermann F. Progressive myoclonus epilepsies: clinical and neurophysiological diagnosis. Journal of Clinical Neurophysiology. 1991;8(3):261-71.

[78] Shahwan A, Farrell M, Delanty N. Progressive myoclonic epilepsies: a review of ge‐

[79] French JA, Kanner AM, Bautista J, Abou-Khalil B, Browne T, Harden CL, et al. Effica‐ cy and tolerability of the new antiepileptic drugs I: treatment of new onset epilepsy: report of the Therapeutics and Technology Assessment Subcommittee and Quality Standards Subcommittee of the American Academy of Neurology and the American

[80] Auvin S. Treatment of Juvenile Myoclonic Epilepsy. CNS Neuroscience & Therapeu‐

[81] Calleja S, Salas-Puig J, Ribacoba R, Lahoz CH. Evolution of juvenile myoclonic epi‐ lepsy treated from the outset with sodium valproate. Seizure. 2001;10(6):424-7.

[82] Fernando-Dongas MC, Radtke RA, VanLandingham KE, Husain AM. Characteristics of valproic acid resistant juvenile myoclonic epilepsy. Seizure. 2000;9(6):385-8.

[83] Marson AG, Al-Kharusi AM, Alwaidh M, Appleton R, Baker GA, Chadwick DW, et al. The SANAD study of effectiveness of valproate, lamotrigine, or topiramate for generalised and unclassifiable epilepsy: An unblinded randomised controlled trial.

[84] Mazurkiewicz-Bełdzińska M, Szmuda M, Matheisel A. Long-term efficacy of val‐ proate versus lamotrigine in treatment of idiopathic generalized epilepsies in chil‐

[85] Ben-Menachem E, Schmitz B, Tomson T, Vajda F. Role of valproate across the ages. Treatment of epilepsy in adults. Acta Neurologica Scandinavica. 2006;114:14-27.

[86] Adab N, Kini U, Vinten J, Ayres J, Baker G, Clayton-Smith J, et al. The longer term outcome of children born to mothers with epilepsy. Journal of Neurology, Neurosur‐

[87] Morrow J, Russell A, Guthrie E, Parsons L, Robertson I, Waddell R, et al. Malforma‐ tion risks of antiepileptic drugs in pregnancy: a prospective study from the UK Epi‐ lepsy and Pregnancy Register. Journal of Neurology, Neurosurgery & Psychiatry.

[88] Vinten J, Bromley RL, Taylor J, Adab N, Kini U, Baker GA. The behavioral conse‐ quences of exposure to antiepileptic drugs in utero. Epilepsy & Behavior. 2009;14(1):

myoclonic epilepsy patients. Neurology. 1991;41(10):1651-5.

netic and therapeutic aspects. Lancet Neurology. 2005;4(4):239-48.

Epilepsy Society. Neurology. 2004;62(8):1252-60.

tics. 2008;14(3):227-33.

116 Epilepsy Topics

The Lancet. 2007;369(9566):1016-26.

gery & Psychiatry. 2004;75(11):1575-83.

2006;77(2):193-8.

197-201.

dren and adolescents. Seizure. 2010;19(3):195-7.


[102] Shian WJ, Chi CS. Juvenile myoclonic epilepsy. Zhonghua Yi Xue Za Zhi=Chinese Medical Journal; Free China Ed. 1994;53(6):375-8.

[115] Iimura Y, Sugano H, Nakajima M, Arai H. Stepwise synchronization through the cor‐ pus callosum is one cause of myoclonic jerks. World Neurosurgery. 2012;77(2):

Juvenile Myoclonic Epilepsy — A Maturation Syndrome Coming of Age

http://dx.doi.org/10.5772/58372

119

[116] Jenssen S, Sperling MR, Tracy JI, Nei M, Joyce L, David G, et al. Corpus callosotomy

[117] Faught E. Treatment of refractory primary generalized epilepsy. Reviews In Neuro‐

[118] Baykan B, Martínez-Juárez IE, Altindag EA, Camfield CS, Camfield PR. Lifetime prognosis of juvenile myoclonic epilepsy. Epilepsy & Behavior. 2013;28(Suppl 1):S18-

[119] Geithner JFZJRCU. Predictors for long-term seizure outcome in juvenile myoclonic

[120] Matsuoka H. The seizure prognosis of juvenile myoclonic epilepsy. The Japanese

[121] Seneviratne U, Cook M, D'Souza W. The prognosis of idiopathic generalized epilep‐

[122] Guaranha MSB, Filho GMdA, Lin K, Guilhoto LMFF, Caboclo LOSF, Yacubian EMT. Prognosis of juvenile myoclonic epilepsy is related to endophenotypes. Seizure.

[123] Ristić AJ, Ostojić J, Kozić D, Vojvodić NM, Popović LM, Janković S, et al. Hippocam‐ pal metabolic dysfunction in juvenile myoclonic epilepsy: 3D multivoxel spectrosco‐

[124] Swartz BE, Halagren E, Simpkins F, Syndulko K. Primary memory in patients with frontal and primary generalized epilepsy. Journal of Epilepsy. 1994;7(3):232-41. [125] Swartz BE, Simpkins F, Halgren E, Mandelkern M, Brown C, Krisdakumtorn T, et al. Visual working memory in primary generalized epilepsy: an 18FDG-PET study.

[126] Devinsky O, Gershengorn J, Brown E, Perrine K, Vazquez B, Luciano D. Frontal func‐ tions in juvenile myoclonic epilepsy. Neuropsychiatry, Neuropsychology, & Behav‐

[127] Piazzini A, Turner K, Vignoli A, Canger R, Canevini MP. Frontal cognitive dysfunc‐

[128] Zamarian L, Höfler J, Kuchukhidze G, Delazer M, Bonatti E, Kemmler G, et al. Deci‐ sion making in juvenile myoclonic epilepsy. Journal of Neurology. 2013;260(3):

tion in juvenile myoclonic epilepsy. Epilepsia. 2008;49(4):657-62.

py study. Journal of the Neurological Sciences. 2011;305(1-2):139-42.

epilepsy: 25-63 years of follow-up. Epilepsia. 2012;53(8):1379-86.

Journal of Psychiatry and Neurology. 1992;46(2):293-6.

in refractory idiopathic generalized epilepsy. Seizure. 2006;15(8):621-9.

logical Diseases. 2004;1 Suppl 1:S34-S43.

sy. Epilepsia. 2012;53(12):2079-90.

Neurology. 1996;47(5):1203-12.

ioral Neurology. 1997;10(4):243-6.

2011;20(1):42-8.

839-46.

399.E5-8.

S24.


[115] Iimura Y, Sugano H, Nakajima M, Arai H. Stepwise synchronization through the cor‐ pus callosum is one cause of myoclonic jerks. World Neurosurgery. 2012;77(2): 399.E5-8.

[102] Shian WJ, Chi CS. Juvenile myoclonic epilepsy. Zhonghua Yi Xue Za Zhi=Chinese

[103] Sokic D, Ristic AJ, Vojvodic N, Jankovic S, Sindjelic AR. Frequency, causes and phe‐ nomenology of late seizure recurrence in patients with juvenile myoclonic epilepsy

[104] Dahl JC. A Behaviour Medicine Approach to Epilepsy-Time for a Paradigm Shift?

[105] Guilfoyle SM, Follansbee-Junger K, Modi AC. Development and preliminary imple‐ mentation of a psychosocial service into standard medical care for pediatric epilepsy.

[106] Lundgren T, Dahl J, Hayes S. Evaluation of mediators of change in the treatment of epilepsy with acceptance and commitment therapy. Journal of Behavioral Medicine.

[107] Kasteleijn-Nolst Trenité DGA, Schmitz B, Janz D, Delgado-Escueta AV, Thomas P, Hirsch E, et al. Consensus on diagnosis and management of JME: From founder's ob‐

[108] Crespel A, Gelisse P, Reed RC, Ferlazzo E, Jerney J, Schmitz B, et al. Management of juvenile myoclonic epilepsy. Epilepsy & Behavior. 2013;28(Suppl 1):S81-S6.

[109] Martinović Z. Adjunctive behavioural treatment in adolescents and young adults

[110] da Silva Sousa P, Lin K, Garzon E, Ceiki Sakamoto A, Yacubian EMT. Language-and praxis-induced jerks in patients with juvenile myoclonic epilepsy. Epileptic Disor‐

[111] Matsuoka H, Nakamura M, Ohno T, Shimabukuro J, Suzuki T, Numachi Y, et al. The Role of Cognitive–Motor Function in Precipitation and Inhibition of Epileptic Seiz‐

[112] Morris GL, 3rd, Gloss D, Buchhalter J, Mack KJ, Nickels K, Harden C. Evidencebased guideline update: Vagus nerve stimulation for the treatment of epilepsy: Re‐ port of the Guideline Development Subcommittee of the American Academy of

[113] Kostov H, Larsson PG, Røste GK. Is vagus nerve stimulation a treatment option for patients with drug-resistant idiopathic generalized epilepsy? Acta Neurologica Scan‐

[114] Cukiert A, Burattini JA, Mariani PP, Cukiert CM, Argentoni-Baldochi M, Baise-Zung C, et al. Outcome after extended callosal section in patients with primary idiopathic

ders: International Epilepsy Journal with Videotape. 2005;7(2):115-21.

servations to current trends. Epilepsy & Behavior. 2013;28(Suppl 1):S87-S90.

Medical Journal; Free China Ed. 1994;53(6):375-8.

after a long period of remission. Seizure. 2007;16(6):533-7.

Scandinavian Journal of Behaviour Therapy. 1999;28(3):97.

Clinical Practice in Pediatric Psychology. 2013;1(3):276-88.

with juvenile myoclonic epilepsy. Seizure. 2001;10(1):42-7.

2008;31(3):225-35.

118 Epilepsy Topics

ures. Epilepsia. 2005;46:17-20.

dinavica. 2007;115:55-8.

Neurology. Neurology. 2013;81(16):1453-9.

generalized epilepsy. Epilepsia. 2009;50(6):1377-80.


[129] Zamarian L, Trinka E, Bonatti E, Kuchukhidze G, Bodner T, Benke T, et al. Executive functions in chronic mesial temporal lobe epilepsy. Epilepsy Research and Treat‐ ment. 2011;Vol. 2011(ArtileID 596174):11 pages.

[141] O'Muircheartaigh J, Vollmar C, Barker GJ, Kumari V, Symms MR, Thompson P, et al. Focal structural changes and cognitive dysfunction in juvenile myoclonic epilepsy.

Juvenile Myoclonic Epilepsy — A Maturation Syndrome Coming of Age

http://dx.doi.org/10.5772/58372

121

[142] Lavandier N, de Toffol B, Gillet P, Hommet C, Corcia P, Autret A. Syndrome dysexe ´cutif, e´pilepie myoclonique juve´nile et anomalies paroxystiques intercritiques. Rev

[143] Sonmez F, Atakli D, Sari H, Atay T, Arpaci B. Cognitive function in juvenile myo‐

[144] Iqbal N, Caswell HL, Hare DJ, Pilkington O, Mercer S, Duncan S. Neuropsychologi‐ cal profiles of patients with juvenile myoclonic epilepsy and their siblings: a prelimi‐ nary controlled experimental video-EEG case series. Epilepsy & Behavior. 2009;14(3):

[145] Roebling R, Scheerer N, Uttner I, Gruber O, Kraft E, Lerche H. Evaluation of cogni‐ tion, structural, and functional MRI in juvenile myoclonic epilepsy. Epilepsia.

[146] Wandschneider B, Centeno M, Vollmar C, Stretton J, O'Muircheartaigh J, Thompson PJ, et al. Risk-taking behavior in juvenile myoclonic epilepsy. Epilepsia. 2013. [147] Vollmar C, O'Muircheartaigh J, Barker GJ, Symms MR, Thompson P, Kumari V, et al. Motor system hyperconnectivity in juvenile myoclonic epilepsy: a cognitive function‐ al magnetic resonance imaging study. Brain: A Journal of Neurology. 2011;134(Pt 6):

[148] Lin K, de Araujo Filho GM, Pascalicchio TF, Silva I, Tudesco ISS, Guaranha MSB, et al. Hippocampal atrophy and memory dysfunction in patients with juvenile myo‐

[149] Brooks BL. A Study of low scores in Canadian children and adolescents on the Wechsler Intelligence Scale for Children, Fourth Edition (WISC-IV). Child Neuropsy‐

[150] Steinberg BA, Bieliauskas LA, Smith GE, Langellotti C, Ivnik RJ. Mayo's Older Amer‐ icans Normative Studies: Age-and IQ-Adjusted Norms for the Boston Naming Test, the MAE Token Test, and the Judgment of Line Orientation Test. The Clinical Neuro‐

[151] Walsh J, Marson AG, Smith PEM, Rees MI, Baker GA, Thomas RH. Neuroticism and executive function in drug-refractory juvenile myoclonic epilepsy. Epilepsy & Behav‐

[152] Savic I, Österman Y, Helms G. MRS shows syndrome differentiated metabolite

changes in human-generalized epilepsies. NeuroImage. 2004;21(1):163.

clonic epilepsy. Epilepsy & Behavior. 2004;5(3):329-36.

clonic epilepsy. Epilepsy & Behavior. 2013;29(1):247-51.

Neurology. 2011;76(1):34-40.

Neurol. 2002;158(Suppl. 1):164.

516-21.

1710-9.

2009;50(11):2456-65.

chology. 2011;17(3):281-9.

ior. 2013;28(Suppl 1).

psychologist. 2005;19(3-4):280-328.


[141] O'Muircheartaigh J, Vollmar C, Barker GJ, Kumari V, Symms MR, Thompson P, et al. Focal structural changes and cognitive dysfunction in juvenile myoclonic epilepsy. Neurology. 2011;76(1):34-40.

[129] Zamarian L, Trinka E, Bonatti E, Kuchukhidze G, Bodner T, Benke T, et al. Executive functions in chronic mesial temporal lobe epilepsy. Epilepsy Research and Treat‐

[130] Delazer M, Zamarian L, Bonatti E, Kuchukhidze G, Koppelstaetter F, Bodner T, et al. Decision making under ambiguity and under risk in mesial temporal lobe epilepsy.

[131] Levav M, Mirsky AF, Herault J, Xiong L, Amir N, Andermann E. Familial association of neuropsychological traits in patients with generalized and partial seizure disor‐ ders. Journal of Clinical and Experimental Neuropsychology. 2002;24(3):311-26. [132] Pulsipher DT, Seidenberg M, Guidotti L, Tuchscherer VN, Morton J, Sheth RD, et al. Thalamofrontal circuitry and executive dysfunction in recent-onset juvenile myo‐

[133] Jackson DC, Dabbs K, Walker NM, Jones JE, Hsu DA, Stafstrom CE, et al. The neuro‐ psychological and academic substrate of new/recent-onset epilepsies. The Journal of

[134] Guidotti L, Seidenberg M, Pulsipher DT, Morton J, Dow C, Sheth RD, et al. Cognition and MRI volumetrics in children with recent onset juvenile myoclonic epilepsy. Epi‐

[135] Pascalicchio TF, de Araujo Filho GM, da Silva Noffs MH, Lin K, Caboclo LOSF, Vi‐ dal-Dourado M, et al. Neuropsychological profile of patients with juvenile myoclonic epilepsy: a controlled study of 50 patients. Epilepsy & Behavior. 2007;10(2):263-7. [136] Wandschneider B, Kopp UA, Kliegel M, Stephani U, Kurlemann G, Janz D, et al. Pro‐ spective memory in patients with juvenile myoclonic epilepsy and their healthy sib‐

[137] Moschetta SP, Valente KD. Juvenile myoclonic epilepsy: the impact of clinical varia‐ bles and psychiatric disorders on executive profile assessed with a comprehensive

[138] Kim S-Y, Hwang Y-H, Lee H-W, Suh C-K, Kwon S-H, Park S-P. Cognitive impair‐ ment in juvenile myoclonic epilepsy. Journal of Clinical Neurology (Seoul, Korea).

[139] Kim JH, Suh S-i, Park S-Y, Seo W-K, Koh I, Koh S-B, et al. Microstructural white mat‐ ter abnormality and frontal cognitive dysfunctions in juvenile myoclonic epilepsy.

[140] Jovic NJ, Vucinic DM, Ignjatovic PM, Kosac A. Cognitive and psychiatric issues in ju‐

neuropsychological battery. Epilepsy & Behavior. 2012;25(4):682-6.

venile myoclonic epilepsy. Epilepsy & Behavior. 2013;28(2):324.

ment. 2011;Vol. 2011(ArtileID 596174):11 pages.

Neuropsychologia. 2010;48(194-200).

clonic epilepsy. Epilepsia. 2009;50(5):1210-9.

Pediatrics. 2013;162(5):1047-53.e1.

lings. Neurology. 2010;75(24):2161-7.

lepsia. 2006;47:110.

120 Epilepsy Topics

2007;3(2):86-92.

Epilepsia. 2012;53(8):1371-8.


[153] Kessler RC, Berglund P, Demler O, Jin R, Merikangas KR, Walters EE. Lifetime prev‐ alence and age-of-anset distributions of DSM-IV disorders in the National Comorbid‐ ity Survey Replication. Archives of General Psychiatry. 2005;62(6):593-602.

[167] de Araujo Filho GM, Pascalicchio TF, Lin K, Sousa PS, Yacubian EMT. Neuropsychi‐ atric profiles of patients with juvenile myoclonic epilepsy treated with valproate or

Juvenile Myoclonic Epilepsy — A Maturation Syndrome Coming of Age

http://dx.doi.org/10.5772/58372

123

[168] de Araújo Filho GM, Rosa VP, Lin K, Caboclo LOSF, Sakamoto AC, Yacubian EMT. Psychiatric comorbidity in epilepsy: a study comparing patients with mesial tempo‐ ral sclerosis and juvenile myoclonic epilepsy. Epilepsy & Behavior. 2008;13(1):

[169] Gélisse P, Genton P, Thomas P, Rey M, Samuelian JC, Dravet C. Clinical factors of drug resistance in juvenile myoclonic epilepsy. Journal of Neurology, Neurosurgery

[170] Salas-Puig J, Camara da Silva A, Davet C, J R. L´épilepsie myoclonique juvenile dans

[171] Moschetta S, Fiore LA, Fuentes D, Gois J, Valente KD. Personality traits in patients with juvenile myoclonic epilepsy. Epilepsy & Behavior. 2011;21(4):473-7.

[172] Plattner B, Pahs G, Kindler J, Williams RP, Hall RE, Mayer H, et al. Juvenile myoclon‐ ic epilepsy: a benign disorder? Personality traits and psychiatric symptoms. Epilepsy

[173] Akanuma N, Hara E, Adachi N, Hara K, Koutroumanidis M. Psychiatric comorbidity in adult patients with idiopathic generalized epilepsy. Epilepsy & Behavior.

[174] Perini GI, Tosin C, Carraro C, Bernasconi G, Canevini MP, Canger R, et al. Interictal mood and personality disorders in temporal lobe epilepsy and juvenile myoclonic

epilepsy. Journal of Neurology, Neurosurgery & Psychiatry. 1996;61(6):601-5.

[175] de Araújo Filho GM, da Silva JM, Mazetto L, Marchetti RL, Yacubian EMT. Psycho‐ ses of epilepsy: A study comparing the clinical features of patients with focal versus

[176] Pung T, Schmitz B. Circadian rhythm and personality profile in juvenile myoclonic

[177] Karachristianou S, Katsarou Z, Bostantjopoulou S, Economou A, Garyfallos G, Delin‐ ikopoulou E. Personality profile of patients with juvenile myoclonic epilepsy. Epilep‐

[178] Butcher JN, Dahlstrom WG, Graham JR, Tellegen A, Kaemmer B. Manual for the re‐ standardized Minnesota Multiphasic Personality Inventory: MMPI—2. An adminis‐ trative and interpretive guide. Minneapolis: University of Minnesota Press; 1989.

[179] de Araújo Filho GM, Jackowski AP, Lin K, Guaranha MSB, Guilhoto LMFF, da Silva HH, et al. Personality traits related to juvenile myoclonic epilepsy: MRI reveals pre‐

generalized epilepsies. Epilepsy & Behavior. 2011;20(4):655-8.

la population du centre. Saint-Paul Epilepsies 1990;2:108–13.

topiramate. Epilepsy & Behavior. 2006;8(3):606-9.

196-201.

& Psychiatry. 2001;70(2):240-3.

& Behavior. 2007;10(4):560-4.

epilepsy. Epilepsia. 2006;47:111-4.

sy & Behavior. 2008;13(4):654-7.

2008;13(1):248-51.


[167] de Araujo Filho GM, Pascalicchio TF, Lin K, Sousa PS, Yacubian EMT. Neuropsychi‐ atric profiles of patients with juvenile myoclonic epilepsy treated with valproate or topiramate. Epilepsy & Behavior. 2006;8(3):606-9.

[153] Kessler RC, Berglund P, Demler O, Jin R, Merikangas KR, Walters EE. Lifetime prev‐ alence and age-of-anset distributions of DSM-IV disorders in the National Comorbid‐

[154] Marrie RA, Fisk JD, Bo Nancy Y, Stella L, Elliott L, Caetano P, et al. Mental comor‐ bidity and multiple sclerosis: validating administrative data to support population-

[155] Kessler RC, Lane MC, Shahly V, Stang PE. Accounting for comorbidity in assessing the burden of epilepsy among US adults: results from the National Comorbidity Sur‐

[156] Mula M. Epilepsy-induced behavioral changes during the ictal phase. Epilepsy & Be‐

[157] Gandy M, Sharpe L, Perry KN, Miller L, Thayer Z, Boserio J, et al. The psychosocial correlates of depressive disorders and suicide risk in people with epilepsy. Journal of

[158] Gilliam F, Hecimovic H, Sheline Y. Psychiatric comorbidity, health, and function in

[159] de Araújo Filho GM, Yacubian EMT. Juvenile myoclonic epilepsy: psychiatric comor‐ bidity and impact on outcome. Epilepsy & Behavior. 2013;28(Suppl 1):S74-S80. [160] Lund M, Reintoft H, Simonsen N. [A controlled sociological and psychological study on patients with juvenile myoclonus epilepsy]. Der Nervenarzt. 1976;47(12):708-12.

[161] Tsuboi T. Primary generalized epilepsy with sporadic myoclonias of myoclonic petit

[162] Devinsky O, Najjar S. Evidence against the existence of a temporal lobe epilepsy per‐

[163] First MB, Spitzer RL, Gibbon M, Williams JBW. Structured Clinical Interview for DSM-IV-TR Axis I Disorders, Research Version, Patient Edition. (SCID-I/P). New

[164] Trinka E, Kienpointner G, Unterberger I, Luef G, Bauer G, Doering LB, et al. Psychi‐ atric comorbidity in juvenile myoclonic epilepsy. Epilepsia. 2006;47(12):2086-91. [165] de Araujo Filho G, Mazetto L, da Silva JM, Caboclo LO, Yacubian EM. Psychiatric co‐ morbidity in patients with two prototypes of focal versus generalized epilepsy syn‐

[166] de Araújo Filho GM, Pascalicchio TF, Sousa PdS, Lin K, Ferreira Guilhoto LMF, Ya‐ cubian EMT. Psychiatric disorders in juvenile myoclonic epilepsy: a controlled study

York: Biometrics Research, New York State Psychiatric Institute; 2002.

vey Replication (NCS-R). Molecular Psychiatry. 2012;17(7):748-58.

ity Survey Replication. Archives of General Psychiatry. 2005;62(6):593-602.

based surveillance. BMC Neurology. 2013;13(1):1-8.

Psychosomatic Research. 2013;74(3):227-32.

epilepsy. Epilepsy & Behavior. 2003;4 Suppl 4:S26-S30.

mal type. Stuttgart: Thieme Medical Publishers; 1977.

dromes. Seizure. 2011;20(5):383-6.

sonality syndrome. Neurology. 1999;53(5, Suppl 2):S13-S25.

of 100 patients. Epilepsy & Behavior. 2007;10(3):437-41.

havior. 2013.

122 Epilepsy Topics


frontal abnormalities through a voxel-based morphometry study. Epilepsy & Behav‐ ior. 2009;15(2):202-7.

[191] Meencke HJ, Janz D. The significance of microdysgenesia in primary generalized epi‐ lepsy: an answer to the considerations of Lyon and Gastaut. Epilepsia. 1985;26(4):

Juvenile Myoclonic Epilepsy — A Maturation Syndrome Coming of Age

http://dx.doi.org/10.5772/58372

125

[192] Veith G, Wicke R. Cerebrale Differenzierungsstorungen bei Epilepsie. Westdtsch Ver.

[193] Lyon G, Gastaut H. Considerations on the significance attributed to unusual cerebral histological findings recently described in eight patients with primary generalized

[194] Opeskin K, Kalnins RM, Halliday G, Cartwright H, Berkovic SF. Idiopathic general‐ ized epilepsy: lack of significant microdysgenesis. Neurology. 2000;55(8):1101-6.

[195] Woermann FG, Free SL, Koepp MJ, Sisodiya SM, Duncan JS. Abnormal cerebral structure in juvenile myoclonic epilepsy demonstrated with voxel-based analysis of

[196] Lin K, Jackowski AP, Carrete H, Jr., de Araújo Filho GM, Silva HH, Guaranha MSB, et al. Voxel-based morphometry evaluation of patients with photosensitive juvenile

[197] Ronan L, Alhusaini S, Scanlon C, Doherty CP, Delanty N, Fitzsimons M. Widespread cortical morphologic changes in juvenile myoclonic epilepsy: evidence from structur‐

[198] Vollmar C, O'Muircheartaigh J, Symms MR, Barker GJ, Thompson P, Kumari V, et al. Altered microstructural connectivity in juvenile myoclonic epilepsy: The missing

[199] Kim JH, Im KC, Kim JS, Lee S-A, Kang JK. Correlation of interictal spike-wave with thalamic glucose metabolism in juvenile myoclonic epilepsy. Neuroreport.

[200] Savic I, Lekvall A, Greitz D, Helms G. MR spectroscopy shows reduced frontal lobe concentrations of N-acetyl aspartate in patients with juvenile myoclonic epilepsy. Ep‐

[201] Simister RJ, McLean MA, Barker GJ, Duncan JS. Proton MRS reveals frontal lobe me‐ tabolite abnormalities in idiopathic generalized epilepsy. Neurology. 2003;61(7):

[202] Moeller F, Maneshi M, Pittau F, Gholipour T, Bellec P, Dubeau F, et al. Functional connectivity in patients with idiopathic generalized epilepsy. Epilepsia. 2011;52(3):

[203] McGill ML, Devinsky O, Kelly C, Milham M, Castellanos FX, Quinn BT, et al. Default mode network abnormalities in idiopathic generalized epilepsy. Epilepsy & Behav‐

MRI. Brain: A Journal of Neurology. 1999;122 (Pt 11):2101-8.

myoclonic epilepsy. Epilepsy Research. 2009;86(2-3):138-45.

368-71.

1968:515-34.

epilepsy. Epilepsia. 1985;26:365-7.

al MRI. Epilepsia. 2012;53(4):651-8.

link. Neurology. 2012;78(20):1555-9.

2005;16(11):1151-5.

897-902.

515-22.

ilepsia. 2000;41(3):290-6.

ior. 2012;23(3):353-9.


[191] Meencke HJ, Janz D. The significance of microdysgenesia in primary generalized epi‐ lepsy: an answer to the considerations of Lyon and Gastaut. Epilepsia. 1985;26(4): 368-71.

frontal abnormalities through a voxel-based morphometry study. Epilepsy & Behav‐

[180] de Araújo Filho GM, Lin K, Lin J, Peruchi MM, Caboclo LOSF, Guaranha MSB, et al. Are personality traits of juvenile myoclonic epilepsy related to frontal lobe dysfunc‐

[181] de Araujo Filho GM, de Araujo TB, Sato JR, Silva Id, Lin K, Júnior HC, et al. Person‐ ality traits in juvenile myoclonic epilepsy: evidence of cortical abnormalities from a

[182] Moschetta S, Valente KD. Impulsivity and seizure frequency, but not cognitive defi‐ cits, impact social adjustment in patients with juvenile myoclonic epilepsy. Epilepsia.

[183] Staufenbiel SM, Penninx BWJH, Spijker AT, Elzinga BM, van Rossum EFC. Hair cor‐ tisol, stress exposure, and mental health in humans: A systematic review. Psycho‐

[184] Vargas C, López-Jaramillo C, Vieta E. A systematic literature review of resting state network--functional MRI in bipolar disorder. Journal of Affective Disorders.

[185] Busch RM, Frazier T, Chapin JS, Hamrahian AH, Diehl B, Alexopoulos A, et al. Role of cortisol in mood and memory in patients with intractable temporal lobe epilepsy.

[186] Cossette P, Liu L, Brisebois K, Dong H, Lortie A, Vanasse M, et al. Mutation of GA‐ BRA1 in an autosomal dominant form of juvenile myoclonic epilepsy. Nature Genet‐

[187] de Nijs L, Wolkoff N, Coumans B, Delgado-Escueta AV, Grisar T, Lakaye B. Muta‐ tions of EFHC1, linked to juvenile myoclonic epilepsy, disrupt radial and tangential migrations during brain development. Human Molecular Genetics. 2012;21(23):

[188] de Nijs L, Wolkoff N, Grisar T, Lakaye B. Juvenile myoclonic epilepsy as a possible neurodevelopmental disease: Role of EFHC1 or Myoclonin1. Epilepsy & Behavior.

[189] Delgado-Escueta AV. Advances in Genetics of Juvenile Myoclonic Epilepsies. Epilep‐

[190] de Nijs L, Léon C, Nguyen L, LoTurco JJ, Delgado-Escueta AV, Grisar T, et al. EFHC1 interacts with microtubules to regulate cell division and cortical development. Na‐

surface morphometry study. Epilepsy & Behavior. 2013;27(2):385-92.

tions? A proton MRS study. Epilepsia. 2009;50(5):1201-9.

ior. 2009;15(2):202-7.

124 Epilepsy Topics

2013;54(5):866-70.

2013;150(3):727-35.

ics. 2002;31(2):184.

2013;28(s1):S58-S60.

sy Currents. 2007;7(3):61-7.

ture Neuroscience. 2009;12(10):1266-74.

5106-17.

Neurology. 2012;78(14):1064-8.

neuroendocrinology. 2013;38(8):1220-35.


[204] Macdonell RA. Prolonged cortical silent period after transcranial magnetic stimula‐ tion in generalized epilepsy. Neurology. 2001;57:706-8.

[217] Szabó CÁ, Knape KD, Leland MM, Cwikla DJ, Williams-Blangero S, Williams JT. Ep‐ idemiology and characterization of seizures in a pedigreed baboon colony. Compara‐

Juvenile Myoclonic Epilepsy — A Maturation Syndrome Coming of Age

http://dx.doi.org/10.5772/58372

127

[218] Szabó CA, Leland MM, Knape K, Elliott JJ, Haines V, Williams JT. Clinical and EEG phenotypes of epilepsy in the baboon (Papio hamadryas spp.). Epilepsy Research.

[219] Szabó CÁ, Salinas FS, Narayana S. Functional PET Evaluation of the Photosensitive

[220] Szabó CÁ, Narayana S, Kochunov PV, Franklin C, Knape K, Davis MD, et al. PET imaging in the photosensitive baboon: Case-controlled study. Epilepsia. 2007;48(2):

[221] Szabó CA, Kochunov P, Knape KD, McCoy KJM, Leland MM, Lancaster JL, et al. Cortical sulcal areas in baboons (Papio hamadryas spp.) with generalized interictal

[222] Szabó CÁ, Narayana S, Franklin C, Knape KD, Davis MD, Fox PT, et al. "Resting" CBF in the epileptic baboon: Correlation with ketamine dose and interictal epileptic

[223] Salinas FS, Fox PT, Szabó CA. Using resting-state fMRI to demonstrate functional connectivity differences between epileptic and control non-human primates. Presen‐

[224] Szabó CÁ, Salinas FS, Leland MM, Caron J-L, Hanes MA, Knape KD, et al. Baboon model of generalized epilepsy: Continuous intracranial video-EEG monitoring with

[225] Ticku MK, Lee JC, Murk S, Mhatre MC, Story JL, Kagan-Hallet K, et al. Inhibitory and excitatory amino acid receptors, c-fos expression, and calcium-binding proteins in the brain of baboons (Papio hamadryas) that exhibit 'spontaneous' grand mal epi‐

[226] Phelix CF, Szabó CÁ. Quantification of heterotopic neurons in baboons with general‐ ized interictal epileptic discharges. Poster session presented at: American Epilepsy

[227] Young NA, Szabó CÁ, Phelix CF, Flaherty DK, Balaram P, Foust-Yeoman KB, et al. Epileptic baboons have lower numbers of neurons in specific areas of cortex. Pro‐ ceedings of the National Academy of Sciences of the United States of America. 2013.

[228] Martínez-Juárez IE, Alonso ME, Medina MT, Durón RM, Bailey JN, López-Ruiz M, et al. Juvenile myoclonic epilepsy subsyndromes: family studies and long-term follow-

epileptic discharges on scalp EEG. Epilepsy Research. 2011;93(2/3):91-5.

Baboon. The Open Neuroimaging Journal. 2011;5:206-15.

discharges. Epilepsy Research. 2008;82(1):59-65.

tation at: NC3R Primate Welfare Meeting; 2013; London, UK.

subdural electrodes. Epilepsy Research. 2012;101(1/2):46-55.

lepsy. Epilepsy Research Supplement. 1992;9:141-9.

Society 66th Annual Meeting; 2012; San Diego, CA.

up. Brain: A Journal of Neurology. 2006;129(Pt 5):1269-80.

tive Medicine. 2012;62(6):535-8.

2005;65(1-2):71-80.

245-53.


[217] Szabó CÁ, Knape KD, Leland MM, Cwikla DJ, Williams-Blangero S, Williams JT. Ep‐ idemiology and characterization of seizures in a pedigreed baboon colony. Compara‐ tive Medicine. 2012;62(6):535-8.

[204] Macdonell RA. Prolonged cortical silent period after transcranial magnetic stimula‐

[205] Badawy RA, Curatolo JM, Newton M, Berkovic SF, Macdonell RA. Changes in corti‐ cal excitability differentiate generalized and focal epilepsy. Annals of Neurology.

[206] Groppa S, Siebner HR, Kurth C, Stephani U, Siniatchkin M. Abnormal response of motor cortex to photic stimulation in idiopathic generalized epilepsy. Epilepsia.

[207] Brigo F, Storti M, Benedetti MD, Rossini F, Nardone R, Tezzon F, et al. Resting motor threshold in idiopathic generalized epilepsies: a systematic review with meta-analy‐

[208] Brigo F, Igwe SC. Corticospinal excitability in juvenile myoclonic epilepsy: The story

[209] Killam EK. Photomyoclonic seizures in the baboon, Papio papio. Federation Proceed‐

[210] Meldrum BS, Anlezark G, Balzamo E, Horton RW, Trimble M. Photically induced ep‐ ilepsy in Papio papio as a model for drug studies. Advances In Neurology.

[211] Menini C, Stutzmann JM, Laurent H, Naquet R. Paroxysmal visual evoked potentials (PVEPs) in Papio papio. I. Morphological and topographical characteristics. Compar‐ ison with paroxysmal discharges. Electroencephalography & Clinical Neurophysiol‐

[212] Silva-Barrat C, Ménini C, Bryère P, Naquet R. Multiunitary activity analysis of corti‐ cal and subcortical structures in paroxysmal discharges and grand mal seizures in photosensitive baboons. Electroencephalography & Clinical Neurophysiology.

[213] Silva-Barrat C, Ménini C. The influence of intermittent light stimulation on potentials evoked by single flashes in photosensitive and non-photosensitive Papio papio. Elec‐

[214] Menini C, Silva-Barrat C. The photosensitive epilepsy of the baboon. A model of gen‐

[215] Fukuda H, Valin A, Bryère P, Riche D, Wada JA, Naquet R. Role of the forebrain commissure and hemispheric independence in photosensitive response of epileptic baboon, Papio papio. Electroencephalography & Clinical Neurophysiology.

[216] Szabó CÁ, Knape KD, Leland MM, Williams JT. Electroclinical phenotypes in a ped‐

troencephalography & Clinical Neurophysiology. 1984;57(5):448-61.

eralized reflex epilepsy. Advances In Neurology. 1998;75:29-47.

igreed baboon colony. Epilepsy Research. 2013;105(1-2):77-85.

tion in generalized epilepsy. Neurology. 2001;57:706-8.

sis. Epilepsy Research. 2012;101(1-2):3-13.

2007;61(4):324-31.

126 Epilepsy Topics

2008;49(12):2022-9.

so far. Seizure. 2013.

1975;10:119-32.

ings. 1979;38(10):2429-33.

ogy. 1980;50(5-6):356-64.

1986;64(5):455-68.

1988;69(4):363-70.


**Chapter 6**

**Reflex Epilepsy**

Raidah Saleem AlBaradie

http://dx.doi.org/10.5772/58462

**1. Introduction**

are genetic in origin [2].

need a diagnosis of epilepsy.

**Pathophysiology**

particular stimulus.

morning [4].

Additional information is available at the end of the chapter

A reflex seizure is a condition in which seizures can be provoked habitually by an external stimulus or, less commonly, internal mental processes, or by activity of the patient. It is most commonly precipitated by visual stimuli. Other somatosensory occurrences, including thinking, reading, listening to music, and eating may also induce reflex seizures [1]. Reflex epilepsies are quite uncommon, occurring in only 5% of all epilepsies. Most of these epilepsies

The definition of reflex epilepsy was recognized initially by on International League Against Epilepsy classification in 1989 [3]. The classification in 2001 formed reflex seizure and epilepsy definitions. Three types of reflex seizure met clinically embrace pure reflex epilepsy, reflex seizures that happen in generalized or focal epilepsy syndromes that are also connected with spontaneous seizures, and isolated reflex seizures arising in conditions that do not essentially

The occurrence of seizures in people with epilepsy is rarely predictable. Elements that aggravate seizures may differ from person to person and may include sleep deprivation, systemic illness, or ingestion of particular food products [2]. These factors typically do not activate seizures in a consistent pattern and they may lower seizure threshold in patients with unprovoked seizures. In comparison, reflex seizures denote a time-dependent response to a

An example of a trigger which more reliably causes interictal epileptic discharges (IEDs) or clinical seizures in photosensitive individuals is intermittent light stimulation (ILS). 12-to 18- Hz frequencies in photic stimulation are more likely to produce seizures than others, and the degree of photosensitivity may depend on the time of day, where it is increased early in the

> © 2014 The Author(s). Licensee InTech. This chapter is distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/3.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

**Chapter 6**

## **Reflex Epilepsy**

Raidah Saleem AlBaradie

Additional information is available at the end of the chapter

http://dx.doi.org/10.5772/58462

#### **1. Introduction**

A reflex seizure is a condition in which seizures can be provoked habitually by an external stimulus or, less commonly, internal mental processes, or by activity of the patient. It is most commonly precipitated by visual stimuli. Other somatosensory occurrences, including thinking, reading, listening to music, and eating may also induce reflex seizures [1]. Reflex epilepsies are quite uncommon, occurring in only 5% of all epilepsies. Most of these epilepsies are genetic in origin [2].

The definition of reflex epilepsy was recognized initially by on International League Against Epilepsy classification in 1989 [3]. The classification in 2001 formed reflex seizure and epilepsy definitions. Three types of reflex seizure met clinically embrace pure reflex epilepsy, reflex seizures that happen in generalized or focal epilepsy syndromes that are also connected with spontaneous seizures, and isolated reflex seizures arising in conditions that do not essentially need a diagnosis of epilepsy.

#### **Pathophysiology**

The occurrence of seizures in people with epilepsy is rarely predictable. Elements that aggravate seizures may differ from person to person and may include sleep deprivation, systemic illness, or ingestion of particular food products [2]. These factors typically do not activate seizures in a consistent pattern and they may lower seizure threshold in patients with unprovoked seizures. In comparison, reflex seizures denote a time-dependent response to a particular stimulus.

An example of a trigger which more reliably causes interictal epileptic discharges (IEDs) or clinical seizures in photosensitive individuals is intermittent light stimulation (ILS). 12-to 18- Hz frequencies in photic stimulation are more likely to produce seizures than others, and the degree of photosensitivity may depend on the time of day, where it is increased early in the morning [4].

© 2014 The Author(s). Licensee InTech. This chapter is distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/3.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Reflex seizures are also provoked by the stimulation of other primary sensory cortices, such as the primary auditory, or somatosensory cortices, and by activation of premotor, pericingu‐ late (SMA), and parietal lobe association cortices. It is proposed that the stimulus may create an abnormal response directly in the sensory or association cortices, with a synchronized discharge spreading functionally connected cortical or subcortical structures or a physiological reaction is in charge for the initiation of synchronization of larger networks or functionally connected epileptogenic cortex. In addition, forming of real connectivity during the photo‐ paroxysmal response indicated the frontocentral cortices were already synchronized prior to the appearance of the ictal or interictal discharge [5].

phalomyopathies (mitochondrial disorders complicated by cognitive decline and progressive

Reflex Epilepsy

131

http://dx.doi.org/10.5772/58462

Females have more common photosensitivity but there is no sex predilection in reflex epilep‐

Seizures can be provoked by visual stimulus such as flickering light, removal of visual fixation or light intensity, complex visual patterns, viewing particular objects, or other visual stimuli [15]. The most common type of visually induced seizure is photosensitive seizures. Photosen‐ sitivity is an abnormal visual sensitivity of the brain in response to flickering light sources. It is expressed in the electroencephalography (EEG) as a generalized spikes/polyspikes and wave discharge (photoparoxysmal response) produced by intermittent photic stimulation, or clinical seizures in vulnerable individuals [16]. The prevalence of photosensitivity in patients with epilepsy ranges from 2% to 20%. There are three groups of Photosensitivite subjects divided based on their response to (ILS) and it is more commonly associated with idiopathic general‐

**1.** Individuals who develop seizures only when they are exposed to light stimulus.

Photosensitivity is nearly twice as common in females as in males. 25% of patients lose their photosensitivity in their 20s and 30s. Genetic tendency play an important role in photosensi‐ tivity. Regional occipital cortical hyperexcitability is noted in functional magnetic resonance imaging (MRI) and magnetoencephalography in photosensitive patients [17]. It is proposed that there is hyperexcitability of the visual cortex in photosensitive patients as noted from human and animal data. When a sufficient large area from the visual cortex is stimulated, it will lead to an epileptiform discharge and a seizure might be provoked by mechanism requires the physiologic activation of a critical area of cortical tissue especially the parvocellar more than mangocellar pathway.The most common light source that plays a role in photosensitive seizures is television more than computer monitors and video games. The reflection of sunlight directly or intermittently on a road lined with trees, lamplights, and colorful and bright

The treatment of photosensitive epilepsy can be achieved with or without combined antiepi‐ leptic drugs by avoiding the stimulus, stimulus modification such as avoidance of clear sources of blinking lights and video games, avoiding extended game play, increase distance from the television set, and using a remote control are all important and useful strategies. At times,

**2.** Individuals who experience seizures with or without light stimulus.

**3.** Asymptomatic individuals with a photosensitive reaction on EEG.

blinking lights are other stimuli in photosensitive subjects [17].

weakness) [14].

**Epidemiology**

**2. Precipitating factors**

ized epilepsy, which constitutes 20-40% of all epilepsy:

**2.1. Visual stimulus**

sies [1].

Some studies showed diminished inhibition of the motor cortex during ILS which was seen by transcranial magnetic stimulation in photosensitive individuals. This decrease seems to denote synchronization of the frontocentral cortices or the absence of an inhibitory modulation by an intervening cortical region. In addition, it was noted in patients who were studied for resective surgery that a temporal lobe epileptogenic zone was activated by ILS without clear generation of the seizure from the occipital lobe. This support that there are mechanisms other than spread being crucial to induce "reflex seizures [6]. This was also supported by animal models studies, where it demonstrated an abnormal cortical development in the baboon in the histology and the morphology with reflex seizures [7].

#### **Etiology**

The cause of reflex epilepsy can be acquired or inherited. Reflex seizures rarely caused by acquired cerebral lesions. The most common etiologies which affect the brain frontal, temporal or parietal association cortices are strokes, encephalitis, or cortical dysplasia. In comparison to trauma and meningitis, which incline to affect more the anterior frontal and temporal lobes. The most common signs of acquired reflex epilepsy are startle seizures, which is presented clinically by sudden myoclonic or tonic contraction of the truncal and extremity musculature. This type of seizures can be triggered by somatosensory, auditory, or proprioceptive (move‐ ment) stimuli. It is correlated by a brief electroencephalographic discharge, which might be followed by a partial seizure. The perirolandic cortices and mesial frontoparietal networks are commonly implicated in the generation of reflex seizures [8].

Whereas the inherited reflex epilepsy is supported by animal models as stated earlier such as audio genic seizures characterize genetic reflex epilepsies in predisposed strains of mice, rats, and birds [9]. Recent studies demonstrated a link of photosensitivity to bands 7q32 and 16p13 by one group [10] and was linked to 6p21 and 13q31 by another [11]. Also, children with chromosomal abnormalities have been shown to have augmented affinity to photosensitivity. Hot water epilepsy (HWE) was linked to band 4q24-q28 in one family and to band 10q21-q22 in 6 families [12]. Patients with autosomal dominant temporal lobe epilepsy have seizures triggered by speech and auditory stimuli which is associated with mutations in the LGI1 gene in chromosome 10q22-q24 [13]. Several genetic factors predispose to reflex epilepsies, some related to channelopathies, other affecting brain development such as neurodegenerative disorders. They are stated to as progressive myoclonic epilepsies which includ Unverricht-Lundborg disease, Lafora disease, neuronal ceroid lipofuscinosis, and mitochondrial ence‐ phalomyopathies (mitochondrial disorders complicated by cognitive decline and progressive weakness) [14].

#### **Epidemiology**

Reflex seizures are also provoked by the stimulation of other primary sensory cortices, such as the primary auditory, or somatosensory cortices, and by activation of premotor, pericingu‐ late (SMA), and parietal lobe association cortices. It is proposed that the stimulus may create an abnormal response directly in the sensory or association cortices, with a synchronized discharge spreading functionally connected cortical or subcortical structures or a physiological reaction is in charge for the initiation of synchronization of larger networks or functionally connected epileptogenic cortex. In addition, forming of real connectivity during the photo‐ paroxysmal response indicated the frontocentral cortices were already synchronized prior to

Some studies showed diminished inhibition of the motor cortex during ILS which was seen by transcranial magnetic stimulation in photosensitive individuals. This decrease seems to denote synchronization of the frontocentral cortices or the absence of an inhibitory modulation by an intervening cortical region. In addition, it was noted in patients who were studied for resective surgery that a temporal lobe epileptogenic zone was activated by ILS without clear generation of the seizure from the occipital lobe. This support that there are mechanisms other than spread being crucial to induce "reflex seizures [6]. This was also supported by animal models studies, where it demonstrated an abnormal cortical development in the baboon in the

The cause of reflex epilepsy can be acquired or inherited. Reflex seizures rarely caused by acquired cerebral lesions. The most common etiologies which affect the brain frontal, temporal or parietal association cortices are strokes, encephalitis, or cortical dysplasia. In comparison to trauma and meningitis, which incline to affect more the anterior frontal and temporal lobes. The most common signs of acquired reflex epilepsy are startle seizures, which is presented clinically by sudden myoclonic or tonic contraction of the truncal and extremity musculature. This type of seizures can be triggered by somatosensory, auditory, or proprioceptive (move‐ ment) stimuli. It is correlated by a brief electroencephalographic discharge, which might be followed by a partial seizure. The perirolandic cortices and mesial frontoparietal networks are

Whereas the inherited reflex epilepsy is supported by animal models as stated earlier such as audio genic seizures characterize genetic reflex epilepsies in predisposed strains of mice, rats, and birds [9]. Recent studies demonstrated a link of photosensitivity to bands 7q32 and 16p13 by one group [10] and was linked to 6p21 and 13q31 by another [11]. Also, children with chromosomal abnormalities have been shown to have augmented affinity to photosensitivity. Hot water epilepsy (HWE) was linked to band 4q24-q28 in one family and to band 10q21-q22 in 6 families [12]. Patients with autosomal dominant temporal lobe epilepsy have seizures triggered by speech and auditory stimuli which is associated with mutations in the LGI1 gene in chromosome 10q22-q24 [13]. Several genetic factors predispose to reflex epilepsies, some related to channelopathies, other affecting brain development such as neurodegenerative disorders. They are stated to as progressive myoclonic epilepsies which includ Unverricht-Lundborg disease, Lafora disease, neuronal ceroid lipofuscinosis, and mitochondrial ence‐

the appearance of the ictal or interictal discharge [5].

histology and the morphology with reflex seizures [7].

commonly implicated in the generation of reflex seizures [8].

**Etiology**

130 Epilepsy Topics

Females have more common photosensitivity but there is no sex predilection in reflex epilep‐ sies [1].

### **2. Precipitating factors**

#### **2.1. Visual stimulus**

Seizures can be provoked by visual stimulus such as flickering light, removal of visual fixation or light intensity, complex visual patterns, viewing particular objects, or other visual stimuli [15]. The most common type of visually induced seizure is photosensitive seizures. Photosen‐ sitivity is an abnormal visual sensitivity of the brain in response to flickering light sources. It is expressed in the electroencephalography (EEG) as a generalized spikes/polyspikes and wave discharge (photoparoxysmal response) produced by intermittent photic stimulation, or clinical seizures in vulnerable individuals [16]. The prevalence of photosensitivity in patients with epilepsy ranges from 2% to 20%. There are three groups of Photosensitivite subjects divided based on their response to (ILS) and it is more commonly associated with idiopathic general‐ ized epilepsy, which constitutes 20-40% of all epilepsy:


Photosensitivity is nearly twice as common in females as in males. 25% of patients lose their photosensitivity in their 20s and 30s. Genetic tendency play an important role in photosensi‐ tivity. Regional occipital cortical hyperexcitability is noted in functional magnetic resonance imaging (MRI) and magnetoencephalography in photosensitive patients [17]. It is proposed that there is hyperexcitability of the visual cortex in photosensitive patients as noted from human and animal data. When a sufficient large area from the visual cortex is stimulated, it will lead to an epileptiform discharge and a seizure might be provoked by mechanism requires the physiologic activation of a critical area of cortical tissue especially the parvocellar more than mangocellar pathway.The most common light source that plays a role in photosensitive seizures is television more than computer monitors and video games. The reflection of sunlight directly or intermittently on a road lined with trees, lamplights, and colorful and bright blinking lights are other stimuli in photosensitive subjects [17].

The treatment of photosensitive epilepsy can be achieved with or without combined antiepi‐ leptic drugs by avoiding the stimulus, stimulus modification such as avoidance of clear sources of blinking lights and video games, avoiding extended game play, increase distance from the television set, and using a remote control are all important and useful strategies. At times, covering one eye and rotating if the screen flashes or if myoclonic jerks occur is useful. Use of 100 Hz television sets found to diminish sensitivity in many patients: the screen is naturally less provocative than a 50 Hz screen however screen content may still be stimulating. When needed, the drug of choice is valproate in monotherapy. Antiepileptic drugs that can be used in this condition such as clobazam, lamotrigine, topiramate, and levetiracetam might be helpful [18].

seen in few seizures. A normal computed tomography and MRI of the brain are seen in patients with HWE but in few reported cases a focal malformation of the left parietal cortex was detected with brain MRI. The underlying mechanism of HWE remains uncertain. It was suggested that repeated pouring of water (a kindling effect) and the temperature of the water (a facilitative or triggering factor) during bathing in genetically or anatomically vulnerable persons play a role in the pathophysiology and this starting stimulus is considered to be complex. It may include a mixture of factors, such as contact of scalp with hot water, the

Reflex Epilepsy

133

http://dx.doi.org/10.5772/58462

The management of HWE is to decrease the temperature of the water and altering the method of bathing. These may be enough to control the seizures. Adding an antiepileptic drug can be considered if the mentioned precautions fail. The used antiepileptic drugs include carbama‐ zepine, phenytoin, phenobarbital, sodium valproate, oxcarbazepine, lamotrigine, clobazam, or levetiracetam. A self-abort the attack by distracting maneuvers, like listening to music, chanting the name of God, or remembering their dear ones might be used [24]. HWE usually

Audiogenic seizures have been noted in many animal types and happen usually in employed mouse and rodent models of genetically determined epilepsy. Auditory stimuli are less common precipitants of reflex seizures where sounds may produce seizures in cases of startle

Musicogenic epilepsy (ME) is rare and it is considered to be reflex-evoked or sensory-evoked epilepsy. ME is a seizures induced by hearing certain sounds such as a specific type or piece of music. The seizure can be induced also while the subject is exposed to the musical trigger during sleep or merely thinking about it. In other patients, an affective component of the stimulus is obvious, in addition a nonmusical sounds, such as whirring machinery, may be actual causes in others. The prevalence is one case per 10,000,000 populations [26]. It can be underdiagnosed because the latency between stimulus and seizure onset has been found up to several minutes. It occurs more in males than females with the mean age of onset of the seizures is 14 years. This type of seizures was reported in infancy. The most common type of this seizure is a complex partial with secondary generalization with ictal EEG onset in the mesial temporal region. Cerebral single-photon emission computed tomography (SPECT) in a patient with musicogenic epilepsy demonstrated a right temporal focus. The Treatment can be achieved by the use of anti-epileptic drugs or surgical intervention for medically refractory

Movement-induced reflex seizures such as nonketotic hyperglycemi are reflex seizures most likely to be seen by general neurologists, internists, or other medical specialists in the hospital setting which usually resolve with normalization of the metabolic disturbance. Postanoxic

ME with temporal lobe scars and glial temporal lesions [27].

temperature of the water, and the specific cortical area of stimulation [23].

carry good prognosis.

epilepsy [2].

**2.4. Movement**

**2.3. Auditory stimulation**

Photosensitive epilepsies usually carry very good prognosis, about 25% of patients with these conditions will lose their photosensitivity in their third decade. Most such patients will relapse if they discontinue the antiepileptic drugs early [19]. It worth to state that photosensitivity can be seen in idiopathic generalized epilepsies such as juvenile myoclonic epilepsy, and in crytogenic generalized epilepsies such as severe myoclonic epilepsy of infancy (Dravet syndrome), or with degenerative gray matter encephalopathies such as Lafora's disease, Unverricht-Lundborg disease, Kufs' disease, the neuronal ceroid lipofuscinoses, and in others progressive myoclonus epilepsies [19].

#### **2.2. Somatosensory stimulus**

It may include light touch, tapping, or immersion in hot water. The seizures can be provoked by touch may occur in infancy or childhood is called startle epilepsy or reflex myoclonic epilepsy [20]. The type of seizures inclines to be generalized and less commonly, partial-onset seizures which are activated by touch due to the activation of a sensorimotor cortex.

An important example of somatosensory stimulus induced seizure is hot water epilepsy (HWE) where seizures can be triggered by bathing with hot water pouring over the head, face, or neck. It was first described in 1945. It is the second most common type of reflex epilepsy after photosensitive epilepsy and it is considered to be rare. The seizure is induced when the individual is exposed to water warmer than 37 °C. HWE constitute 3.6-6.9% of all epilepsy cases. HWE occur mostly in infants and children and a male/female ratio of 2-3/1. But also it can be seen after 40 years of age. Familial HWE cases with more than one affected member have been reported in 7-15% of Indian patients, and 1-27% of these patients reported a history of febrile convulsions [21].

One of the important causes of HWE is the genetic etiology. Genome-wide linkage analysis of Indian families delivered proof of linkage for the disorder at 10q21.3-q22.3 and recognized a 15 Mb disease-associated haplotype in four out of six families analyzed [21]. It was noted in a study proven hyperthermic kindling in rats after several episodes of hot water stimulations, there was progressive epileptic activity displayed during lowering of rectal temperature thresholds from 41.5 to 40.0 ºC, drop in latency for developing seizures from 185 to 118 sec and increase in duration of hippocampal seizure discharge from 15 to 140 sec with gradual increase in difficulty of EEG after discharges and neuronal sprouting observed in supragranular molecular layer and in stratum lacunosum [22].

The EEG is usually normal interictally and in 15-20% might reveal diffuse abnormalities such as lateralized or localized spike discharges in the anterior temporal regions. There are technical limitations and difficulties in obtaining an ictal EEG records but a temporal lobe onset was seen in few seizures. A normal computed tomography and MRI of the brain are seen in patients with HWE but in few reported cases a focal malformation of the left parietal cortex was detected with brain MRI. The underlying mechanism of HWE remains uncertain. It was suggested that repeated pouring of water (a kindling effect) and the temperature of the water (a facilitative or triggering factor) during bathing in genetically or anatomically vulnerable persons play a role in the pathophysiology and this starting stimulus is considered to be complex. It may include a mixture of factors, such as contact of scalp with hot water, the temperature of the water, and the specific cortical area of stimulation [23].

The management of HWE is to decrease the temperature of the water and altering the method of bathing. These may be enough to control the seizures. Adding an antiepileptic drug can be considered if the mentioned precautions fail. The used antiepileptic drugs include carbama‐ zepine, phenytoin, phenobarbital, sodium valproate, oxcarbazepine, lamotrigine, clobazam, or levetiracetam. A self-abort the attack by distracting maneuvers, like listening to music, chanting the name of God, or remembering their dear ones might be used [24]. HWE usually carry good prognosis.

#### **2.3. Auditory stimulation**

covering one eye and rotating if the screen flashes or if myoclonic jerks occur is useful. Use of 100 Hz television sets found to diminish sensitivity in many patients: the screen is naturally less provocative than a 50 Hz screen however screen content may still be stimulating. When needed, the drug of choice is valproate in monotherapy. Antiepileptic drugs that can be used in this condition such as clobazam, lamotrigine, topiramate, and levetiracetam might be

Photosensitive epilepsies usually carry very good prognosis, about 25% of patients with these conditions will lose their photosensitivity in their third decade. Most such patients will relapse if they discontinue the antiepileptic drugs early [19]. It worth to state that photosensitivity can be seen in idiopathic generalized epilepsies such as juvenile myoclonic epilepsy, and in crytogenic generalized epilepsies such as severe myoclonic epilepsy of infancy (Dravet syndrome), or with degenerative gray matter encephalopathies such as Lafora's disease, Unverricht-Lundborg disease, Kufs' disease, the neuronal ceroid lipofuscinoses, and in others

It may include light touch, tapping, or immersion in hot water. The seizures can be provoked by touch may occur in infancy or childhood is called startle epilepsy or reflex myoclonic epilepsy [20]. The type of seizures inclines to be generalized and less commonly, partial-onset

An important example of somatosensory stimulus induced seizure is hot water epilepsy (HWE) where seizures can be triggered by bathing with hot water pouring over the head, face, or neck. It was first described in 1945. It is the second most common type of reflex epilepsy after photosensitive epilepsy and it is considered to be rare. The seizure is induced when the individual is exposed to water warmer than 37 °C. HWE constitute 3.6-6.9% of all epilepsy cases. HWE occur mostly in infants and children and a male/female ratio of 2-3/1. But also it can be seen after 40 years of age. Familial HWE cases with more than one affected member have been reported in 7-15% of Indian patients, and 1-27% of these patients reported a history

One of the important causes of HWE is the genetic etiology. Genome-wide linkage analysis of Indian families delivered proof of linkage for the disorder at 10q21.3-q22.3 and recognized a 15 Mb disease-associated haplotype in four out of six families analyzed [21]. It was noted in a study proven hyperthermic kindling in rats after several episodes of hot water stimulations, there was progressive epileptic activity displayed during lowering of rectal temperature thresholds from 41.5 to 40.0 ºC, drop in latency for developing seizures from 185 to 118 sec and increase in duration of hippocampal seizure discharge from 15 to 140 sec with gradual increase in difficulty of EEG after discharges and neuronal sprouting observed in supragranular

The EEG is usually normal interictally and in 15-20% might reveal diffuse abnormalities such as lateralized or localized spike discharges in the anterior temporal regions. There are technical limitations and difficulties in obtaining an ictal EEG records but a temporal lobe onset was

seizures which are activated by touch due to the activation of a sensorimotor cortex.

helpful [18].

132 Epilepsy Topics

progressive myoclonus epilepsies [19].

**2.2. Somatosensory stimulus**

of febrile convulsions [21].

molecular layer and in stratum lacunosum [22].

Audiogenic seizures have been noted in many animal types and happen usually in employed mouse and rodent models of genetically determined epilepsy. Auditory stimuli are less common precipitants of reflex seizures where sounds may produce seizures in cases of startle epilepsy [2].

Musicogenic epilepsy (ME) is rare and it is considered to be reflex-evoked or sensory-evoked epilepsy. ME is a seizures induced by hearing certain sounds such as a specific type or piece of music. The seizure can be induced also while the subject is exposed to the musical trigger during sleep or merely thinking about it. In other patients, an affective component of the stimulus is obvious, in addition a nonmusical sounds, such as whirring machinery, may be actual causes in others. The prevalence is one case per 10,000,000 populations [26]. It can be underdiagnosed because the latency between stimulus and seizure onset has been found up to several minutes. It occurs more in males than females with the mean age of onset of the seizures is 14 years. This type of seizures was reported in infancy. The most common type of this seizure is a complex partial with secondary generalization with ictal EEG onset in the mesial temporal region. Cerebral single-photon emission computed tomography (SPECT) in a patient with musicogenic epilepsy demonstrated a right temporal focus. The Treatment can be achieved by the use of anti-epileptic drugs or surgical intervention for medically refractory ME with temporal lobe scars and glial temporal lesions [27].

#### **2.4. Movement**

Movement-induced reflex seizures such as nonketotic hyperglycemi are reflex seizures most likely to be seen by general neurologists, internists, or other medical specialists in the hospital setting which usually resolve with normalization of the metabolic disturbance. Postanoxic myoclonus (Lance-Adams) may also represent a movement-induced seizure in the medical patient population [28].

is considered symptomatic epilepsy related to localization. It is localized to temporolimbic, extralimbic, perirolandic, or suprasylvian. In the temporolimbic type, complex partial seizure develops towards the end of the meal. The extralimbic, perirolandic, and suprasylvian types are similar to simple reflex epilepsy and it has a very short latency. Clinically, they may have simple partial seizure, hemiparesis, and mental retardation. The ictal and interictal EEG findings support clinic seizures. The underlying lesion can be due to static encephalopathy, or progressive lesions, like a deep localized glioma. The seizure can be controlled by antiepileptic drug therapy such as clobazam and epilepsy surgery might be considered for refractory

Reflex Epilepsy

135

http://dx.doi.org/10.5772/58462

Various stimuli are important in aggravating reflex seizures. The physiological mechanism in reflex epilepsies is still not well defined. However, the cortical hyperexcitability is the most frequently vital factor. The hyperexcitability of different cortical areas may be due to a genetic tendency or to an acquired lesion. The diagnosis of is reflex epilepsy essential for a well understanding how brain works and management for patients. The study of these seizures will benefit the possible insights of somatomotor processing, language mechanisms and the physiology of ideation. Getting proper history and clinical data from the patient will help in better understanding and managing patients with reflex epilepsy. During the electrophysio‐ logical recording of the patient who is believed to have reflex seizures, by giving supposed stimuli, seizure type may be noted and the treatment should be controlled properly. Genetic

Consultant Pediatric Neurologist/Epileptologist, King Fahd Specialist Hospital-Dammam,

[1] Striano S, Coppola A, Del Gaudio L, Striano P. Reflex seizures and reflex epilepsies: old models for understanding mechanisms of epileptogenesis. Epilepsy Res. Jun

[2] Kasteleijn-Nolst Trenité DG. Provoked and reflex seizures: surprising or common?

counseling should be given in patients with strong family history of epilepsy.

patients [32].

**3. Conclusion**

**Author details**

Saudi Arabia

**References**

Raidah Saleem AlBaradie

2012; 100(1-2):1-11.

Epilepsia. Sep 2012;53 Suppl 4:105-13.

#### **2.5. Complex mental processes**

Some of the most unusual and intriguing disorders in neurology are the reflex epilepsies in which seizures are provoked by complex actions or mental processes. Examples of these triggers include reading, eating, micturition, tooth brushing, walking, answering the tele‐ phone, and thinking.

Reading epilepsy is an interesting syndrome which was first described in 1956 by Bickford et al. It is characterized by a feeling of movements in the jaw or throat while reading or myoclonic movements of the jaw which may lead to a generalized tonic-clonic seizure when reading remains. Jaw jerks are the most important mark of reading epilepsy but it can also manifest by an abrupt loss of consciousness, blank staring spells, paroxysmal alexia or dyslexia, and prolonged stuttering. In addition a language-related tasks other than reading, such as awk‐ ward talking, writing, difficult calculations, playing chess or card games, singing, and recitation can also induce seizures [29]

The underlying cause such as neuroanatomical and biochemical basis of reading epilepsy is not clear. The epileptogenic component of the reading process is different between patients such as eye movements, comprehension, emotional content, speech production, and proprio‐ ceptive feedback. A release of endogenous opioids during reading-induced partial seizures in areas of brain involved in normal reading has been found. This directed to the theory that there are networks of cortical areas parallel subserving both cognitive functions and epileptic activity [30]. This is evident by the data from a combined EEG/electromyography-functional MRI study which showed a network of cortical and subcortical areas that are in close proximity with functional areas relevant for language and motor functions has been shown to have significant blood oxygen level dependent changes time-locked with seizure activity [30].

It is important to differentiate between 'primary' or 'specific' reading epilepsy, with seizures only in relation to reading, from a 'secondary', non-specific variety, with seizures when reading. It can be divided into the subgroups idiopathic (primary) and less-frequently seen cryptogenic/symptomatic epilepsy. In primary reading epilepsy, only seizures produced by reading develop without spontaneous seizures.

It occurs more in males with the age of onset is in adolescence and young adulthood. A strong family history of seizures has been stated in 40-50% of patients. EEG is normal in 80% of patients interictally. A spike and wave discharges are seen in 11%, and temporal paroxysmal discharges in 5%. Ictally, 77% have epileptiform discharges consisting of short bursts of sharp waves, spikes, or spike and wave complexes that are bilateral and symmetrical in 32%, bilateral but asymmetrical in 38%, and unilateral or focal in 30%. The seizures are well controlled with valproate, clonazepam, or by modifying the stimulus. The prognosis is usually good [31]

Eating epilepsy rare and it is may cause a seizures that can be triggered by parts of anticipating food, eating itself, or the post-prandial period. It usually occurs in the second decade of life, with a male preponderance. The patients also may have spontaneous seizures. Eating epilepsy is considered symptomatic epilepsy related to localization. It is localized to temporolimbic, extralimbic, perirolandic, or suprasylvian. In the temporolimbic type, complex partial seizure develops towards the end of the meal. The extralimbic, perirolandic, and suprasylvian types are similar to simple reflex epilepsy and it has a very short latency. Clinically, they may have simple partial seizure, hemiparesis, and mental retardation. The ictal and interictal EEG findings support clinic seizures. The underlying lesion can be due to static encephalopathy, or progressive lesions, like a deep localized glioma. The seizure can be controlled by antiepileptic drug therapy such as clobazam and epilepsy surgery might be considered for refractory patients [32].

#### **3. Conclusion**

myoclonus (Lance-Adams) may also represent a movement-induced seizure in the medical

Some of the most unusual and intriguing disorders in neurology are the reflex epilepsies in which seizures are provoked by complex actions or mental processes. Examples of these triggers include reading, eating, micturition, tooth brushing, walking, answering the tele‐

Reading epilepsy is an interesting syndrome which was first described in 1956 by Bickford et al. It is characterized by a feeling of movements in the jaw or throat while reading or myoclonic movements of the jaw which may lead to a generalized tonic-clonic seizure when reading remains. Jaw jerks are the most important mark of reading epilepsy but it can also manifest by an abrupt loss of consciousness, blank staring spells, paroxysmal alexia or dyslexia, and prolonged stuttering. In addition a language-related tasks other than reading, such as awk‐ ward talking, writing, difficult calculations, playing chess or card games, singing, and

The underlying cause such as neuroanatomical and biochemical basis of reading epilepsy is not clear. The epileptogenic component of the reading process is different between patients such as eye movements, comprehension, emotional content, speech production, and proprio‐ ceptive feedback. A release of endogenous opioids during reading-induced partial seizures in areas of brain involved in normal reading has been found. This directed to the theory that there are networks of cortical areas parallel subserving both cognitive functions and epileptic activity [30]. This is evident by the data from a combined EEG/electromyography-functional MRI study which showed a network of cortical and subcortical areas that are in close proximity with functional areas relevant for language and motor functions has been shown to have significant blood oxygen level dependent changes time-locked with seizure activity [30].

It is important to differentiate between 'primary' or 'specific' reading epilepsy, with seizures only in relation to reading, from a 'secondary', non-specific variety, with seizures when reading. It can be divided into the subgroups idiopathic (primary) and less-frequently seen cryptogenic/symptomatic epilepsy. In primary reading epilepsy, only seizures produced by

It occurs more in males with the age of onset is in adolescence and young adulthood. A strong family history of seizures has been stated in 40-50% of patients. EEG is normal in 80% of patients interictally. A spike and wave discharges are seen in 11%, and temporal paroxysmal discharges in 5%. Ictally, 77% have epileptiform discharges consisting of short bursts of sharp waves, spikes, or spike and wave complexes that are bilateral and symmetrical in 32%, bilateral but asymmetrical in 38%, and unilateral or focal in 30%. The seizures are well controlled with valproate, clonazepam, or by modifying the stimulus. The prognosis is usually good [31]

Eating epilepsy rare and it is may cause a seizures that can be triggered by parts of anticipating food, eating itself, or the post-prandial period. It usually occurs in the second decade of life, with a male preponderance. The patients also may have spontaneous seizures. Eating epilepsy

patient population [28].

134 Epilepsy Topics

phone, and thinking.

**2.5. Complex mental processes**

recitation can also induce seizures [29]

reading develop without spontaneous seizures.

Various stimuli are important in aggravating reflex seizures. The physiological mechanism in reflex epilepsies is still not well defined. However, the cortical hyperexcitability is the most frequently vital factor. The hyperexcitability of different cortical areas may be due to a genetic tendency or to an acquired lesion. The diagnosis of is reflex epilepsy essential for a well understanding how brain works and management for patients. The study of these seizures will benefit the possible insights of somatomotor processing, language mechanisms and the physiology of ideation. Getting proper history and clinical data from the patient will help in better understanding and managing patients with reflex epilepsy. During the electrophysio‐ logical recording of the patient who is believed to have reflex seizures, by giving supposed stimuli, seizure type may be noted and the treatment should be controlled properly. Genetic counseling should be given in patients with strong family history of epilepsy.

#### **Author details**

Raidah Saleem AlBaradie

Consultant Pediatric Neurologist/Epileptologist, King Fahd Specialist Hospital-Dammam, Saudi Arabia

#### **References**


[3] Commission on Classification and Terminology of the International League Against Epilepsy. Proposal for revised classification of epilepsies and epileptic syndromes. Epilepsia 1989;30(4):389-99.

[17] Wilkins AJ, Bonanni P, Porciatti V, Guerrini R. Physiology of human photosensitivi‐

Reflex Epilepsy

137

http://dx.doi.org/10.5772/58462

[18] Zifkin BG, Kasteleijn-Nolst Trenité D. Reflex epilepsy and reflex seizures of the visu‐

[19] Lu Y, Waltz S, Stenzel K, Muhle H, Stephani U. Photosensitivity in epileptic syn‐ dromes of childhood and adolescence. Epileptic Disord 2008;10(2):136-43.

[20] Ricci S, Cusmai R, Fusco L, Vigevano F. Reflex myoclonic epilepsy in infancy: a new age-dependent idiopathic epileptic syndrome related to startle reaction. Epilepsia.

[22] Grosso S, Farnetani MA, Francione S, Galluzzi P, Vatti G, Cordelli DM, et al.. Hot water epilepsy and focal malformation of the parietal cortex development. Brain Dev

[23] Ilbay G, Sahin D, Ates N. Changes in blood-brain barrier permeability during hot

[24] Savitha MR, Krishnamurthy B, Ashok DA, Ramachandra NB. Self abortion of attacks

[25] Mehta AD, Ettinger AB, Perrine K, Dhawan V, Patil A, Jain SK, et al. Seizure propa‐ gation in a patient with musicogenic epilepsy. Epilepsy Behav 2009;14(2):421-4.

[27] [47] Tayah TF, Abou-Khalil B, Gilliam FG, Knowlton RC, Wushensky CA, Gallagher MJ. Musicogenic seizures can arise from multiple temporal lobe foci: intracranial

[28] Lance JW, Adams RD. The syndrome of intention or action myoclonus as a sequel to

[29] Bickford RG, Whean JL, Klass DW, Corbin KB. Reading epilepsy: clinical and electro‐ encephalographic studies of a new syndrome. Trans Am Neurol Assoc. 1956;(81st

[30] Koepp MJ, Richardson MP, Brooks DJ, Duncan JS. Focal cortical release of endoge‐ nous opioids during reading-induced seizures. Lancet 1998;352(9132):952-5.

[31] Valenti MP, Tinuper P, Cerullo A, Carcangiu R, Marini C. Reading epilepsy in a pa‐ tient with previous idiopathic focal epilepsy with centrotemporal spikes. Epileptic

[32] Senanayake N. 'Eating epilepsy'-a reappraisal. Epilepsy Res 1990;5(1):74-9.

al system: a clinical review. Epileptic Disord 2000;2(3):129-36.

[21] Satishchandra P. Hot-water epilepsy. Epilepsia 2003;44 Suppl 1:29-32.

water-induced seizures in rats. Neurol Sci 2003;24(4):232-5.

EEG analyses of three patients. Epilepsia 2006;47(8):1402-6.

hypoxic encephalopathy. Brain. Mar 1963;86:111-36. [Medline].

[26] Critchley M. Musicogenic epilepsy. Brain 1937;60:13-27.

in patients with Hot Water Epilepsy. Indian Pediatr 2007;44(4):295-8.

ty. Epilepsia 2004;45 Suppl 1:7-13.

Apr 1995;36(4):342-8. [Medline].

2004;26(7): 490-3.

Meeting):100-2. [Medline].

Disord. Sep 1999;1(3):167-71. [Medline].


[3] Commission on Classification and Terminology of the International League Against Epilepsy. Proposal for revised classification of epilepsies and epileptic syndromes.

[4] Kasteleijn-Nolst Trenite, DGA. Ictal SPECT in a case of pure musicogenic epilepsy. Photosensitivity in epilepsy: Electrophysiological and clinical correlates. Acta Neurol

[5] Varotto G, Visani E, Canafoglia L, Franceschetti S, Avanzini G, Panzica F. Enhanced frontocentral EEG connectivity in photosensitive generalized epilepsies: a partial di‐

[6] Groppa S, Siebner HR, Kurth C, Stephani U, Siniatchkin M. Abnormal responses of motor cortex stimulation in idiopathic generalized epilepsy. Epilepsia.

[7] Young NA, Szabo CA, Phelix CF,Flaherty DK, Balaram P, Foust KB, et al. Regional

[8] Garcia-Morales I, Maestu F, Perez-Jimenez MA, Elices E, Ortiz T, Alvarez-Linera J. A clinical and magnetoencephalography study of MRI-negative startle epilepsy. Epi‐

[9] Batini C, Teillet MA, Naquet R. An avian model of genetic reflex epilepsy. Arch Ital

[10] Pinto D, Westland B, de Haan GJ, et al. Genome-wide linkage scan of epilepsy-relat‐ ed photoparoxysmal electroencephalographic response: evidence for linkage on chromosomes 7q32 and 16p13. Hum Mol Genet. Jan 1 2005;14(1):171-8. [Medline]. [11] Pinto D, Kasteleijn-Nolst Trenite DG, Cordell HJ, et al. Explorative two-locus linkage analysis suggests a multiplicative interaction between the 7q32 and 16p13 myoclonic seizures-related photosensitivity loci. Genet Epidemiol. Jan 2007;31(1):42-50. [Med‐

[12] Ratnapriya R, Satishchandra P, Dilip S, Gadre G, Anand A. Familial autosomal domi‐ nant reflex epilepsy triggered by hot water maps to 4q24-q28. Hum Genet. Nov

[13] Brodtkorb E, Michler RP, Gu W, Steinlein OK. Speech-induced aphasic seizures in epilepsy caused by LGI1 mutation. Epilepsia. Jun 2005;46(6):963-6. [Medline]. [14] de Siqueira LF. Progressive myoclonic epilepsies: review of clinical, molecular and

[15] Chuang YC, Chang WN, Lin TK, Lu CH, Chen SD, Huang CR. Game-related seizures presenting with two types of clinical features. Seizure. Mar 2006;15(2):98-105. [Med‐

[16] Covanis A. Photosensitivity in idiopathic generalized epilepsies. Epilepsia 2005;46

therapeutic aspects. J Neurol. Oct 2010;257(10):1612-9. [Medline].

rected coherence study. Epilepsia. Feb 2012;53(2):359-67. [Medline].

cortical cell reduction in epileptic baboons. Submitted to PNAS.

lepsy Behav. Sep 2009;16(1):166-71. [Medline].

Biol. May 2004;142(3):297-312. [Medline].

2009;126(5):677-83. [Medline].

Epilepsia 1989;30(4):389-99.

136 Epilepsy Topics

Scand Suppl. 1989;125:3-149.

2008;49:2022-2029.

line].

line].

Suppl 9:67-72.


**Chapter 7**

**Vagus Nerve Stimulation Therapy for Epilepsy**

The year 2013 marked the 25th year of clinical vagus nerve stimulation (VNS). This chapter will review the preclinical history, the clinical study history, and the long-term outcomes of VNS.

Epilepsy has been treated by many strange remedies over the centuries. In 1883, Corning [1] (Corning, 1883) proposed that vagus nerve stimulation could decrease heart rate and cerebral blood flow, thereby controlling seizures. This early attempt at vagus stimulation for seizure

Bailey and Bremer [2] (Bailey & Bremer, 1938) demonstrated the direct effect of vagus stimu‐ lation on the central nervous system. They found that repetitive electrical stimulation of the central end of the vagus nerve of the cat results in increased amplitude and frequency of the spontaneous potentials of the orbital surface of the frontal lobes of the cerebral cortex. Inhibition of motor activity by activation of visceral vagal afferents was first reported by Schweitzer and Wright [3] (Schweitzer A, 1937) and later confirmed by Paintal [4] (Paintal, 1973). Dell and Olsen [5] (Dell P, 1951) reported that vagus stimulation affected slow wave

These papers and others led Zabara [6] (Zabara, 1992) to investigate vagus stimulation as a potential method to treat epilepsy. Zabara stimulated the cervical vagus nerve in a strychnine dog model of status epilepsy (N=20). He reported that vagus stimulation would interrupt the strychnine-induced seizure, and the amount of seizure interruption was proportional to the length of stimulation, approximately 4 times as long as the stimulation period. Vagal stimu‐ lation terminated seizures within 0.5-5 s. Transection of the vagus distal to the stimulating electrode did not alter the antiseizure effects of vagus stimulation. The optimal stimulus parameters were estimated to be approximately 20 volts (with electrode resistance 1-5 ohms),

> © 2014 The Author(s). Licensee InTech. This chapter is distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/3.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

20-30 Hz stimulation frequency and approximately 0.2 msec pulse duration.

Additional information is available at the end of the chapter

Reese S. Terry Jr

**1. Introduction**

http://dx.doi.org/10.5772/58332

control quickly fell out of favor.

activity in awake cats.

## **Vagus Nerve Stimulation Therapy for Epilepsy**

Reese S. Terry Jr

Additional information is available at the end of the chapter

http://dx.doi.org/10.5772/58332

#### **1. Introduction**

The year 2013 marked the 25th year of clinical vagus nerve stimulation (VNS). This chapter will review the preclinical history, the clinical study history, and the long-term outcomes of VNS.

Epilepsy has been treated by many strange remedies over the centuries. In 1883, Corning [1] (Corning, 1883) proposed that vagus nerve stimulation could decrease heart rate and cerebral blood flow, thereby controlling seizures. This early attempt at vagus stimulation for seizure control quickly fell out of favor.

Bailey and Bremer [2] (Bailey & Bremer, 1938) demonstrated the direct effect of vagus stimu‐ lation on the central nervous system. They found that repetitive electrical stimulation of the central end of the vagus nerve of the cat results in increased amplitude and frequency of the spontaneous potentials of the orbital surface of the frontal lobes of the cerebral cortex. Inhibition of motor activity by activation of visceral vagal afferents was first reported by Schweitzer and Wright [3] (Schweitzer A, 1937) and later confirmed by Paintal [4] (Paintal, 1973). Dell and Olsen [5] (Dell P, 1951) reported that vagus stimulation affected slow wave activity in awake cats.

These papers and others led Zabara [6] (Zabara, 1992) to investigate vagus stimulation as a potential method to treat epilepsy. Zabara stimulated the cervical vagus nerve in a strychnine dog model of status epilepsy (N=20). He reported that vagus stimulation would interrupt the strychnine-induced seizure, and the amount of seizure interruption was proportional to the length of stimulation, approximately 4 times as long as the stimulation period. Vagal stimu‐ lation terminated seizures within 0.5-5 s. Transection of the vagus distal to the stimulating electrode did not alter the antiseizure effects of vagus stimulation. The optimal stimulus parameters were estimated to be approximately 20 volts (with electrode resistance 1-5 ohms), 20-30 Hz stimulation frequency and approximately 0.2 msec pulse duration.

© 2014 The Author(s). Licensee InTech. This chapter is distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/3.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

In 1990 Lockard [7] (Lockard JS, 1990) used an alumnia gel chronic epilepsy model in monkeys. She induced chronic epilepsy by placing alumina gel on the cortex. Previous studies showed seizure rates remained stable for at least 6 months with this model. The instrumentation triggered a 40-second burst whenever it automatically detected a seizure. The stimulation device also provided a stimulation burst once every 3 hours. The stimulation frequency was changed in 2-week intervals from 83 HZ to 143 HZ and then to a random frequency ranging from 50-250 HZ. The stimulation amplitude was increased to tolerance or 5 mA. In 1 animal, investigators reduced seizure rate to near 0, then stopped stimulation and allowed seizures to return to baseline and restarted stimulation, thereby reducing seizure rate to near 0 once more. Seizure frequency was reduced to zero in a second animal, and stability of frequency was affected in the 2 remaining animals. The 3 stimulation patterns seemed to affect seizure rate equally, but the data were not statistically significant. Although Zabara [6] was able to interrupt seizures in the strychnine dog studies, no seizures were interrupted in the Lockard alumnia gel monkey study. The delay between detection and activation was tens of seconds and may have decreased the seizure interruption effect by VNS. Heart rate and blood pressure were unaffected, and stomach ulcers were not noted. The Woodburys [8] (Woodbury, 1991) theorized that stimulation during the seizure disrupts the seizure pathways in the brain, possibly "unlearning" the seizure mechanisms. They hypothesized that stimulation absent a seizure would not have an effect on reducing future seizures, but such a hypothesis has not been tested.

The mean percent seizure reduction after 14 to 35 months was 46.6% for the combined E01 and E02 patients.[9] (Uthman BM, 1993) Two patients, one who previously had 10 to 100 seizures per day before stimulation, had been seizure-free for over 1 year [10]. (Penry JK, 1990)

The stimulation parameters used in the E01 study were quite different from those in current use, and the first study patient had a remarkable response to the high frequency and long OFF-

In the 2 randomized, blinded, active-control trials (E03 and E05), patients were randomly assigned to either of 2 treatment groups: HIGH (believed to be therapeutic) or LOW (adjusted to patient perception but programmed to 1 HZ delivered once every 90 to 180 minutes; parameters believed to be less therapeutic). Patients enrolled in the study were seen every 4 weeks during the baseline period (weeks-12 to 0). Patients meeting eligibility were implanted with the pulse generator and lead are shown in Table 1. Output was adjusted to tolerance at

**Description of Patients**

**Study E01 E02 E04 E03 E05 Total**

Age (range) 32 (20–58) 33 (18–42) 24 (3–63) 33 (13–57) 33 (13–60) 32 (3–63) No. of females (%) 4 (36%) 2 (40%) 57 (46%) 43 (37%) 104 (52%) 210 (46%)

No. of AEDs (avg) 1.0 1.0 2.2 2.1 2.1 2.1

0.6 0.42 0.65 0.70 high/ 0.85

Two weeks after implantation, patients were randomized to the HIGH (frequency/duty cycle) or LOW (frequency/duty cycle) stimulation group, and the Pulse Generator was activated. Patients in the HIGH groups received a higher frequency, greater pulse width, and higher duty cycle of stimulation. The randomized treatment period that followed activation of the Pulse Generator lasted 14 weeks (the last 12 weeks of which were used in the efficacy analysis—the

**Longitudinal Parallel**

11 5 124 115 199 454

10 5 123 115 198 451

22 (13-32) 20 (5-36) 17 (0.8-48) 21 (4-47) 23 (2-52) 21 (0.8-52)

low

0.58 high/ 0.51

Vagus Nerve Stimulation Therapy for Epilepsy

http://dx.doi.org/10.5772/58332

141

low -

each visit for both groups Table 1 provides descriptions of all study patients.

time patterns.

No. of patients implanted

No. of patients stimulated

Years with epilepsy (range)

Median no. of seizures per day at baseline

All patients implanted in all VNS clinical studies, N=454 [13]

first 2 weeks were a treatment ramp-up period).

**Table 1.** Description of Study Patients [13]

**2.2. E03 and E05 pivotal studies**

#### **2. Clinical studies**

#### **2.1. Pilot studies**

The VNS pilot study (E01) included 10 patients and was based on the Zabara [6], Lockard [7] and Woodbury [8] animal studies. Because technology for seizure detection did not exist at that time, the device was programmed to stimulate periodically, and patients received a magnet to self-activate the stimulation when they experienced an aura. As might be expected, most patients did not have auras and so they did not self-activate stimulation during the seizure. Investigators hoped this periodic pattern would occasionally occur during a seizure and eventually reduce the seizure frequency. Initial stimulation parameters were typically 250 µsec pulse width, 60 sec ON, 60 minutes OFF and 50 HZ, although a stimulation frequency of 143 HZ was used on the first patient. This pilot study was a single-blind study consisting of four seizure measurement periods: pre-implant baseline, stimulation period, sham stimulation period and stimulation period. The mean percent seizure reduction was 24.3% (p<0.049) after 4 months of treatment. [9] (Uthman BM W. B., 1993)

The 4-patient E02 pilot study was similar to E01, except pulse width was 500 µsec, ON time 30 sec, OFF time 10 minutes, and frequency 30 HZ. The mean percent reduction after the fourth treatment period was 39.9% (p <0.074).

The mean percent seizure reduction after 14 to 35 months was 46.6% for the combined E01 and E02 patients.[9] (Uthman BM, 1993) Two patients, one who previously had 10 to 100 seizures per day before stimulation, had been seizure-free for over 1 year [10]. (Penry JK, 1990)

The stimulation parameters used in the E01 study were quite different from those in current use, and the first study patient had a remarkable response to the high frequency and long OFFtime patterns.

#### **2.2. E03 and E05 pivotal studies**

In 1990 Lockard [7] (Lockard JS, 1990) used an alumnia gel chronic epilepsy model in monkeys. She induced chronic epilepsy by placing alumina gel on the cortex. Previous studies showed seizure rates remained stable for at least 6 months with this model. The instrumentation triggered a 40-second burst whenever it automatically detected a seizure. The stimulation device also provided a stimulation burst once every 3 hours. The stimulation frequency was changed in 2-week intervals from 83 HZ to 143 HZ and then to a random frequency ranging from 50-250 HZ. The stimulation amplitude was increased to tolerance or 5 mA. In 1 animal, investigators reduced seizure rate to near 0, then stopped stimulation and allowed seizures to return to baseline and restarted stimulation, thereby reducing seizure rate to near 0 once more. Seizure frequency was reduced to zero in a second animal, and stability of frequency was affected in the 2 remaining animals. The 3 stimulation patterns seemed to affect seizure rate equally, but the data were not statistically significant. Although Zabara [6] was able to interrupt seizures in the strychnine dog studies, no seizures were interrupted in the Lockard alumnia gel monkey study. The delay between detection and activation was tens of seconds and may have decreased the seizure interruption effect by VNS. Heart rate and blood pressure were unaffected, and stomach ulcers were not noted. The Woodburys [8] (Woodbury, 1991) theorized that stimulation during the seizure disrupts the seizure pathways in the brain, possibly "unlearning" the seizure mechanisms. They hypothesized that stimulation absent a seizure would not have an effect on reducing future seizures, but such a hypothesis has not

The VNS pilot study (E01) included 10 patients and was based on the Zabara [6], Lockard [7] and Woodbury [8] animal studies. Because technology for seizure detection did not exist at that time, the device was programmed to stimulate periodically, and patients received a magnet to self-activate the stimulation when they experienced an aura. As might be expected, most patients did not have auras and so they did not self-activate stimulation during the seizure. Investigators hoped this periodic pattern would occasionally occur during a seizure and eventually reduce the seizure frequency. Initial stimulation parameters were typically 250 µsec pulse width, 60 sec ON, 60 minutes OFF and 50 HZ, although a stimulation frequency of 143 HZ was used on the first patient. This pilot study was a single-blind study consisting of four seizure measurement periods: pre-implant baseline, stimulation period, sham stimulation period and stimulation period. The mean percent seizure reduction was 24.3% (p<0.049) after

The 4-patient E02 pilot study was similar to E01, except pulse width was 500 µsec, ON time 30 sec, OFF time 10 minutes, and frequency 30 HZ. The mean percent reduction after the fourth

been tested.

140 Epilepsy Topics

**2. Clinical studies**

4 months of treatment. [9] (Uthman BM W. B., 1993)

treatment period was 39.9% (p <0.074).

**2.1. Pilot studies**

In the 2 randomized, blinded, active-control trials (E03 and E05), patients were randomly assigned to either of 2 treatment groups: HIGH (believed to be therapeutic) or LOW (adjusted to patient perception but programmed to 1 HZ delivered once every 90 to 180 minutes; parameters believed to be less therapeutic). Patients enrolled in the study were seen every 4 weeks during the baseline period (weeks-12 to 0). Patients meeting eligibility were implanted with the pulse generator and lead are shown in Table 1. Output was adjusted to tolerance at each visit for both groups Table 1 provides descriptions of all study patients.


All patients implanted in all VNS clinical studies, N=454 [13]

#### **Table 1.** Description of Study Patients [13]

Two weeks after implantation, patients were randomized to the HIGH (frequency/duty cycle) or LOW (frequency/duty cycle) stimulation group, and the Pulse Generator was activated. Patients in the HIGH groups received a higher frequency, greater pulse width, and higher duty cycle of stimulation. The randomized treatment period that followed activation of the Pulse Generator lasted 14 weeks (the last 12 weeks of which were used in the efficacy analysis—the first 2 weeks were a treatment ramp-up period).

For the HIGH Group, a 30 Hz stimulation frequency was chosen based on the Woodbury rat studies [11] (Woodbury DM, 1990) and safety concerns by Agnew [12] (Agnew WF, 1990) that continuous stimulation at frequencies above 50 Hz might induce nerve damage, although the studies had shown that a 4-hour ON and 4-hour OFF at 50 HZ did not cause any damage. The E05 study used 20 Hz instead of 30 Hz. The 30 seconds ON and 5 minutes OFF was chosen as a compromise to extend battery life and achieve a 10% probability of stimulation during a seizure.

**3. Regulatory approvals**

for this indication because of lack of reimbursement.

vary by country.

DEPRESSION

the E03 patients, N=57.

contributed to the change in seizure frequency.

**Table 4.** Stimulation Parameter Ranges [13]

**Table 3.** VNS Regulatory Approval Dates

**3.1. Approved stimulation parameters**

were later revised to those in Table 4 [13] (Cyberonics, 2013).

Signal frequency 1, 2, 5, 10, 15, 20, 25, 30 Hz, ±6% Pulse width 130, 250, 500, 750, 1000 sec ±10%

More than 70 countries have approved vagus nerve stimulation (VNS) for refractory epilepsy. Table 3 lists several countries and regions. Caregivers should be familiar with the regulatory approval status in their own countries when reading this chapter. Reimbursement levels also

Vagus Nerve Stimulation Therapy for Epilepsy

http://dx.doi.org/10.5772/58332

143

The VNS therapy was also approved for depression, but is not being commercially marketed

**EPILEPSY EU USA CANADA CHINA JAPAN** Year Approved 1994 1997 1998 2008 2010 Indicated age All 12 and older All All All Indicated Seizure Type All Partial All All All

Year Approved 2001 2005 2001 N/A N/A

Parameter ranges currently available are provided in Table 4. The originally approved parameters included frequencies up to 143 HZ and output currents up to 12 mA, but these

Signal ON time 7, 14, 21, 30, 60 sec ±15% or + 7 sec, whichever is greater (±15% or ±7 sec in

Signal OFF time 0.2, 0.3, 0.5, 0.8, 1.1, 1.8, 3 min, and 5 to 180 min(5 to 60 in 5-min steps; 60 to 180 in 30-min steps), +4.4 / -8.4 sec

Magnet activation Provided by magnet application (output current, pulse width, and signal ON

\*The acute phase results include seizure frequencies of the E03 Study LOW stimulation group, which included one-half

Patients were permitted to change their AEDs during these long-term follow-up studies, and these changes may have

time may be independently programmed for this purpose)

**Stimulation Parameters Available Parameter Settings**

Output current 0-3.5 mA in 0.25-mA steps\* ±0.25 ≤1 mA, ±10% > 1 mA

Magnet Mode)

Results: The primary efficacy endpoint (percent reduction in seizure rate) was measured over 12 weeks as shown in Table 2. [13] (Cyberonics, 2013) Adverse events were assessed at each patient visit.


All patients in efficacy analyses in all VNS clinical studies, N=441 [13]

*Within group broad analyses:*

\* P ≤ 0.05, by Wilcoxon sign rank.

† P < 0.0001, by anova.

‡ P ≤ 0.05, by Student's *t-test.*

*Between group broad analyses:*

§ P ≤ 0.02, by Wilcoxon rank sum; P ≤ 0.02, by Student's *t-test.*


*Safety information:*

¶ SAEs = serious adverse events.
