**3. Link between impaired sensory gating and AD pathology**

by the constant repetition of both short- and long-term patterns (S1–500 ms; S2–8 s) could result in long-term repetition suppression produced by the dlPFC. On the contrary, an oddball paradigm, characterized by the varying occurrence probability of a novel stimulus between a series of repeated stimuli, could put the neural network into a state of perceptual expectation (gating-in) while simultaneously suppressing redundant stimuli (gating-out). This phenomenon could be interpreted as the bottom-up, stimulus-driven initiation of attention during the

**Figure 1.** Auditory sensory gating network evoked by the oddball paradigm (panel A) and paired-click paradigm (panel B). The best-fitting source locations are superimposed on volumetric MRI head data to achieve a spatial (3D) rendering of the auditory gating topology. In addition to anticipated bilateral generators in primary auditory areas (blue and green dots), both paradigms evoke a prefrontal gating generator (red dot). While oddball tones provoke medial PFC activation,

In addition to the spatial localization of auditory gating networks, we have disclosed the functional relation within network generators providing strong evidence of a modulatory role for the mPFC generator on bilateral superior temporal gyri (STG) sources dynamics during gating processing (**Figure 2**). This result discloses the long-sought mechanism underlying the auditory gating effect [60]. We demonstrated the complex form of estimated cortical morphology of the neural responses produced by the gating generators. In particular, gating response of STG generators is found to be composed of two consecutive cortical subcomponent, Mb1 peaking at 35–53 ms and Mb2 peaking at 75–95 ms post-stimulus, for both oddball paradigm tones [60]. We also provide the first estimates of cortical gating response produced by the

mPFC generator, which has an analogous tandem form [60], as shown in **Figure 2**.

very early sensory processing [100] executed by the mPFC region.

*2.2.7. Functional mechanisms underlying auditory sensory gating*

the tones of a paired-click paradigm activate dorsolateral PFC areas.

178 Biomarker - Indicator of Abnormal Physiological Process

The initial symptoms of AD include subtle decline of the ability to learn new information along with diverse amnestic disorders without present brain injuries. Clinically indicated AD signs are pointing to the existence of functional impairment of synapses that are involved in converting and forming new declarative memory [32]. Whereas longer retention of sensory memory traces derive more successful memory encoding [101], sensory gating process, conceptualized as the ability of the neural system to modulate its responses to subsequent stimuli, has a fundamental role in guiding successful encoding of new information. Impaired auditory gating processing may reduce pre-attentive signal-to-noise ratio and desynchronize synaptic consolidation in the initial phases of memory formation [58]. Augmentation of dysfunctional sensory gating process could yield to the first amnestic symptoms seen in AD neuropathology.

A range of EEG/MEG studies, measuring evoked sensory responses, have reported differences in early processing of auditory [102–104] stimuli in symptomatic AD patients, affirming the possibility of impaired inhibition of redundant information (gating-out) and processing of novel information (gating-in) in the initial phase of disease. Also, alerted connectivity among different brain regions and decline of long-distance synchronization are found to be responsible for some of the earliest cognitive changes in early phases of symptomatic AD [105]. Although the study of the extracranial neurophysiological (EEG/MEG) signals provides valuable information regarding the pathology-related changes in the amplitudes, latencies, frequency bands, spectral densities, and coherence of oscillatory brain dynamics, identified relations have received limited attention in the search for a biomarker of AD. The main reason is that classification based on the difference between group means of sensor-level measures generally cannot provide a clear individual boundary value between healthy and pathological responses and thus result in lower clinical significance. However, MEG spatiotemporal localization of cortical sources underlying extracranial magnetic field shows internal consistency and provides highly reliable and stable results of both cortical dynamics and topology of the activated network [60, 82, 84], enabling a search for an AD biomarker at the individual level.

#### **3.1. Localization of auditory gating network generators: a topological biomarker of AD**

Using MEG spatiotemporal source calculations, we have demonstrated [64] the potential of topological localization of mPFC generator within an auditory gating network as a discrete, binary, noninvasive tool for detection of AD at the individual subject level. We found three types of gating network topologies evoked by a simple auditory oddball paradigm across the research sample of elderly individuals, ranging from 63 to 87 years of age, which comprised patients with clinical diagnosis of symptomatic AD (MCI and moderate AD) and non-symptomatic elderly controls. Discrete localization/non-localization of mPFC gating generator absolutely discriminate symptomatic AD from controls confirm the indiscernibility between amnestic MCI (aMCI) and AD patients and differentiate two distinct gating network types within the elderly controls, one of which is suggested as preclinical AD. The lack of mPFC generator localization within an auditory gating network as a biomarker of symptomatic AD shows a large effect size (>0.9) and high accuracy, sensitivity, and specificity (100%) in respect of clinical diagnosis [64].

statistical review of group differences on neuropsychological tests confirmed the lowmagnitude but significant differences in Mini Mental State Exam (MMSE) and delayed Rey-Osterreith Complex Figure Test (dROCFT) scores across subjects in different cluster groups [64]. The first group of controls, characterized by sustained gating activation of mPFC generator for both oddball tones, had the highest MMSE scores and the highest performance on the

**Figure 3.** Dendrogram based on the neuropsychological and neurophysiological variables. Subjects were assigned numbers from 1 to 20 and grouped into clusters. Variables were normalized before using a clustering algorithm with Ward's minimum variance method as amalgamation rule and with Euclidean distance as a measure of dissimilarity. Cluster grouping disclosed three groups across two clinical categories (H-healthy subjects, MCI/AD—symptomatic AD subjects) suggesting the potential of the proposed approach to detect subgroup of control ("healthy") subjects who are

Topological Biomarker of Alzheimer's Disease http://dx.doi.org/10.5772/intechopen.76633 181

The second cluster group of controls was characterized by the first signs of neurophysiological gating alternation, which emerged as a partial activation of the mPFC gating generator to the deviant tone only, as shown in **Table 1**. The standard tone did not evoke mPFC activation, suggesting that a very early stage of impairment in sensory gating processing is manifested by an absence of mPFC gating transmission that corrupts habituation to redundant stimuli [64]. We confirmed the presence of insidious cognitive decline in this subgroup of controls, demonstrating the low-amplitude but significantly lower both MMSE and dROCFT scores in comparison with the high-functioning subgroup of controls [64]. Overall, both neuropsychological and neurophysiological impairments characteristic of an AD type of dementia found in lowfunctioning control group, although they did not yet meet clinical criteria for aMCI, indicate a possible preclinical AD phase. The additional weight to our speculation gave Takayama,

dROCFT, thus considered to be cognitively healthy elderly group.

in the possible preclinical phase of AD (cluster category 2).

**3.3. Partial activation of mPFC gating generator: a biomarker of preclinical AD**

#### **3.2. Sustained activation of mPFC gating generator: a sign of healthy cognitive aging**

We have applied spectra of the multivariate analyses to disclose potentially hidden structure in complete set of our data, i.e., scores of wide neuropsychological screening and neurophysiological gating network topology results [64]. Clustering based on principal variables indicated the existence of three stable clusters across subjects as shown in **Figure 3**. Subsequent

as the ability of the neural system to modulate its responses to subsequent stimuli, has a fundamental role in guiding successful encoding of new information. Impaired auditory gating processing may reduce pre-attentive signal-to-noise ratio and desynchronize synaptic consolidation in the initial phases of memory formation [58]. Augmentation of dysfunctional sensory

A range of EEG/MEG studies, measuring evoked sensory responses, have reported differences in early processing of auditory [102–104] stimuli in symptomatic AD patients, affirming the possibility of impaired inhibition of redundant information (gating-out) and processing of novel information (gating-in) in the initial phase of disease. Also, alerted connectivity among different brain regions and decline of long-distance synchronization are found to be responsible for some of the earliest cognitive changes in early phases of symptomatic AD [105]. Although the study of the extracranial neurophysiological (EEG/MEG) signals provides valuable information regarding the pathology-related changes in the amplitudes, latencies, frequency bands, spectral densities, and coherence of oscillatory brain dynamics, identified relations have received limited attention in the search for a biomarker of AD. The main reason is that classification based on the difference between group means of sensor-level measures generally cannot provide a clear individual boundary value between healthy and pathological responses and thus result in lower clinical significance. However, MEG spatiotemporal localization of cortical sources underlying extracranial magnetic field shows internal consistency and provides highly reliable and stable results of both cortical dynamics and topology of the activated network [60, 82, 84], enabling a search for an AD biomarker at the individual level.

gating process could yield to the first amnestic symptoms seen in AD neuropathology.

180 Biomarker - Indicator of Abnormal Physiological Process

**3.1. Localization of auditory gating network generators: a topological biomarker** 

Using MEG spatiotemporal source calculations, we have demonstrated [64] the potential of topological localization of mPFC generator within an auditory gating network as a discrete, binary, noninvasive tool for detection of AD at the individual subject level. We found three types of gating network topologies evoked by a simple auditory oddball paradigm across the research sample of elderly individuals, ranging from 63 to 87 years of age, which comprised patients with clinical diagnosis of symptomatic AD (MCI and moderate AD) and non-symptomatic elderly controls. Discrete localization/non-localization of mPFC gating generator absolutely discriminate symptomatic AD from controls confirm the indiscernibility between amnestic MCI (aMCI) and AD patients and differentiate two distinct gating network types within the elderly controls, one of which is suggested as preclinical AD. The lack of mPFC generator localization within an auditory gating network as a biomarker of symptomatic AD shows a large effect size (>0.9) and high accuracy, sensitivity, and specificity (100%) in respect of clinical diagnosis [64].

**3.2. Sustained activation of mPFC gating generator: a sign of healthy cognitive aging**

We have applied spectra of the multivariate analyses to disclose potentially hidden structure in complete set of our data, i.e., scores of wide neuropsychological screening and neurophysiological gating network topology results [64]. Clustering based on principal variables indicated the existence of three stable clusters across subjects as shown in **Figure 3**. Subsequent

**of AD**

**Figure 3.** Dendrogram based on the neuropsychological and neurophysiological variables. Subjects were assigned numbers from 1 to 20 and grouped into clusters. Variables were normalized before using a clustering algorithm with Ward's minimum variance method as amalgamation rule and with Euclidean distance as a measure of dissimilarity. Cluster grouping disclosed three groups across two clinical categories (H-healthy subjects, MCI/AD—symptomatic AD subjects) suggesting the potential of the proposed approach to detect subgroup of control ("healthy") subjects who are in the possible preclinical phase of AD (cluster category 2).

statistical review of group differences on neuropsychological tests confirmed the lowmagnitude but significant differences in Mini Mental State Exam (MMSE) and delayed Rey-Osterreith Complex Figure Test (dROCFT) scores across subjects in different cluster groups [64]. The first group of controls, characterized by sustained gating activation of mPFC generator for both oddball tones, had the highest MMSE scores and the highest performance on the dROCFT, thus considered to be cognitively healthy elderly group.

#### **3.3. Partial activation of mPFC gating generator: a biomarker of preclinical AD**

The second cluster group of controls was characterized by the first signs of neurophysiological gating alternation, which emerged as a partial activation of the mPFC gating generator to the deviant tone only, as shown in **Table 1**. The standard tone did not evoke mPFC activation, suggesting that a very early stage of impairment in sensory gating processing is manifested by an absence of mPFC gating transmission that corrupts habituation to redundant stimuli [64]. We confirmed the presence of insidious cognitive decline in this subgroup of controls, demonstrating the low-amplitude but significantly lower both MMSE and dROCFT scores in comparison with the high-functioning subgroup of controls [64]. Overall, both neuropsychological and neurophysiological impairments characteristic of an AD type of dementia found in lowfunctioning control group, although they did not yet meet clinical criteria for aMCI, indicate a possible preclinical AD phase. The additional weight to our speculation gave Takayama,


**3.4. Complete absence of mPFC gating generator activation: a biomarker of** 

Subjects within the cluster group characterized by the complete absence of mPFC gating generator activation regardless of the tone condition (i.e., standard and deviant tone) had the lowest neuropsychological test results and belonged to the symptomatic AD patients with clinical diagnosis [64], as shown in **Figure 3**. Clearly, a topological deficit of mPFC gating generator activation is augmented in symptomatic AD phase, taking place during both gating out (inhibition of standard tone) and gating in (enhanced processing of novel tone), suggesting complete disruption of sensory gating process. This absolute break of a fundamental sensory process is associated with the onset of the amnestic AD symptoms. It is possible that progressive failure in sensory gating-out, found in possible preclinical phase of disease, is likely to lead to an overload of working memory due to noise accretion and consequent to the first

Topological Biomarker of Alzheimer's Disease http://dx.doi.org/10.5772/intechopen.76633 183

There are numerous evidences of association between symptomatic AD pathology and alerted physiology of PFC region. Decreased frontal-parietal correlations of glucose metabolism [110], prefrontal glucose hypometabolism on 18F-fluorodeoxyglucose PET scans [111], and frontal retention of 11C-Pittsburgh compound [112] are found in AD patients. Klupp and colleagues [113] found a significant hypometabolism in PFC regions that are not loaded with amyloid plaques, suggesting that the deficit in mPFC gating activation may be related to longitudinal amyloid deposition in different but functionally connected brain regions. For instance, absence or desynchronization of mPFC gating activity could induce atrophy across fast auditory sensory gating loop, which may result in increased amyloid vulnerability of synapses terminated within primary auditory areas involved in the gating process. Inactivity of the mPFC generator during processing of repetitive stimuli (gating-out) likely impairs the ability to distinguish novel from repetitive information, which is critical for long-term memory encoding. It has been demonstrated that activation of presynaptic α-7-nAChRs, involved in gating transmission, induces long-term potentiation of the excitatory input [114]. Therefore, a topological deficit of mPFC generator activation during gating transmission could be a consequence of impaired nAChR levels found in PFC regions of AD patients. Alerted function of α-7-nAChRs could induce lower levels of intracellular calcium, consequently impairing the calcium cascade in producing synchronized postsynaptic signals required for effective both

Identification of a novel biomarker of AD (**Figure 4**) with the potential to detect both putative preclinical and clinical stages at the individual subject level represents significant progress toward improving diagnosis of AD and accelerating the field toward the neurobiological advantage of earlier intervention. Although the study engaged only a research sample, the very large effect size (>0.98) of proposed test, thanks to its binary nature, provides high relevance to the finding. Such a large effect size enables this study with the research sample size

**symptomatic AD**

symptoms of memory impairment.

gating processing and long-term memory encoding.

**4. Conclusion**

to yield power greater than 85%.

**Table 1.** Subject scores on the MMSE, dROCFT, mPFC gating generator activation (1-activated, 0-non-activated), clinical diagnosis, and cluster group.

whose 10-year longitudinal study also discerns the dROCF test as highly assertive indicator of conversion to symptomatic phase of Alzheimer's disease, i.e., found significant sensitivity of dROCFT scores to early, possible preclinical AD pathology [106].

The possibility that partial gating activation of the mPFC generator in lower-functioning controls may be associated with preclinical AD phase is also confirmed by several recent findings. There are evidences that reduced functional connectivity affecting the PFC is associated with amyloid-β-related hypersynchronization [107] and p-tau pathology [108] in a very early phase of AD-type memory impairment. The impaired mPFC activity during endogenous brain activity or memory tasks is found in cognitively normal individuals who were AD APOE ε 4 carriers [109]. Also, evidence of decreased extracranial gating dynamics as a predictor of cerebrospinal amyloid-β reduction is demonstrated in MCI patients [65]. Alerted synaptic function along with subsequent synaptic loss and transneuronal spread of pathological tau forms [108] through PFC regions could result in the topological gating deficit that we found in a low-functioning subgroup of controls. This topological gating deficit could reflect a possible preclinical phase of AD pathology before widespread of cognitive symptoms.

### **3.4. Complete absence of mPFC gating generator activation: a biomarker of symptomatic AD**

Subjects within the cluster group characterized by the complete absence of mPFC gating generator activation regardless of the tone condition (i.e., standard and deviant tone) had the lowest neuropsychological test results and belonged to the symptomatic AD patients with clinical diagnosis [64], as shown in **Figure 3**. Clearly, a topological deficit of mPFC gating generator activation is augmented in symptomatic AD phase, taking place during both gating out (inhibition of standard tone) and gating in (enhanced processing of novel tone), suggesting complete disruption of sensory gating process. This absolute break of a fundamental sensory process is associated with the onset of the amnestic AD symptoms. It is possible that progressive failure in sensory gating-out, found in possible preclinical phase of disease, is likely to lead to an overload of working memory due to noise accretion and consequent to the first symptoms of memory impairment.

There are numerous evidences of association between symptomatic AD pathology and alerted physiology of PFC region. Decreased frontal-parietal correlations of glucose metabolism [110], prefrontal glucose hypometabolism on 18F-fluorodeoxyglucose PET scans [111], and frontal retention of 11C-Pittsburgh compound [112] are found in AD patients. Klupp and colleagues [113] found a significant hypometabolism in PFC regions that are not loaded with amyloid plaques, suggesting that the deficit in mPFC gating activation may be related to longitudinal amyloid deposition in different but functionally connected brain regions. For instance, absence or desynchronization of mPFC gating activity could induce atrophy across fast auditory sensory gating loop, which may result in increased amyloid vulnerability of synapses terminated within primary auditory areas involved in the gating process. Inactivity of the mPFC generator during processing of repetitive stimuli (gating-out) likely impairs the ability to distinguish novel from repetitive information, which is critical for long-term memory encoding. It has been demonstrated that activation of presynaptic α-7-nAChRs, involved in gating transmission, induces long-term potentiation of the excitatory input [114]. Therefore, a topological deficit of mPFC generator activation during gating transmission could be a consequence of impaired nAChR levels found in PFC regions of AD patients. Alerted function of α-7-nAChRs could induce lower levels of intracellular calcium, consequently impairing the calcium cascade in producing synchronized postsynaptic signals required for effective both gating processing and long-term memory encoding.
