**5. Blink reflex abnormalities**

308 Neuroimaging for Clinicians – Combining Research and Practice

Fig. 1. Grand averages and amplitude maps of P300 response in the three groups of subjects. A. Left. Grand averages of P300 responses in the DLB group. Vertical lines mark peak latency. U-shaped bars mark the difference in latency between Fz and Pz leads, same or shorter latency in Pz. Right. Amplitude map of P300 distribution throughout the scalp (at the maximum amplitude recorded) in DLB group. Notice anterior-to-posterior (reversed) amplitude distribution gradient. B. Left. Grand Averages of P300 responses in AD group. Vertical lines mark peak latency. U-shaped bars mark the difference in latency between Fz and Pz leads, longer in Pz. Right. Amplitude map of P300 distribution throughout the scalp (at the maximum amplitude recorded) in the AD group. Notice a posterior to anterior

amplitude distribution gradient. C. Traces and distribution in controls. EOG: electrooculogram; DLB: Dementia with Lewy Bodies; AD: Alzheimer's Disease. Patients with PD exhibit a reduced frequency of blinking leading to a staring appearance [49]. Reduced blink rate can cause an abnormal tear film, dry eyes and reduced vision. A characteristic ocular sign may be the blink reflex, elicited by a light tap on the glabella above the bridge of the nose: successive taps in normal individuals produce less and less response as the reflex habituates but in PD subjects the blink reflex does not disappear on repeated tapping. Habituation may improve after treatment with L-dopa or amantadine. Blink duration and excitability appear to be increased in PD and as in VEP latency may reflect loss of dopamine neurons [50]. The electric Blink Reflex (BR) is a neurophysiological technique exploring pontine structures through a reflex arc connecting nuclei of the 5th to the nuclei of the 7th cranial nerve. The Blink reflex consists of three separate responses: R1, R2, R3. The first one is generated in the trigemino-facial reflex arc, the second and third one are generated in polysynaptic pathways involving the brainstem reticular formation [51]. Clinically, the BR is used to evaluate brainstem lesions and it has been applied in clinical and neurophysiological studies of brainstem lesions and neurodegenerative disorders [52-54].

Is There a Place for Clinical Neurophysiology Assessments in Synucleinopathies? 311

We performed a study of the blink reflex in patients with PD, DLB, MSA, AD and

The subjects were comfortably sitting on an armchair in a quiet room, with eyes gently closed. The recordings took place in a temperature-controlled room (at about 25°C) in halflight. The cathode was placed over the supraorbital foramen and the anode 2cm rostrally. Surface electrodes were placed on the inferior part of the orbicularis oculi muscles on each side, recording ipsilateral R1, and ipsilateral and contralateral R2 and R3. Ground electrode was placed under the chin. Stimuli of 0,1ms of duration with intensity of 5-10 mA elicited stable R1 in repeated trials. Because surface electrodes lay only few centimetres away from the cathode, R1 tended to overlap the stimulus artifact, which could last more than 10ms. A special amplifier with a short blocking time (0.1ms) and low internal noise (0.5 uV at a bandwidth of 2kHz) minimized the problem of stimulus artifact. Signals were amplified and filtered (bandwidth 20-2000Hz), to avoid habituation the interstimuli intervals must be of at least 7 sec, 5-10 responses per site were elicited and stored. BR recording were previously described in MSA, PSP and PD patients: all reports showed R2 latencies inside the 2 SD of the mean and only evidenced enhancement or inhibition of R1-R2 in excitability-duration curve paradigms [52,53,55] in untreated PD. Recently we studied the BR in parkinsonism [56]: in all PD, MSA, PSP and AD patients we found normal R1 and R2 latencies inside the 2SD of the control mean independently of the presence of RBD. Only in DLB patients we found R2 latencies clustering in the upper limits of normality or definitely above the limits (figure 3). All findings were statistically significant. Thus, BR recordings might reveal brainstem dysfunction in DLB, but not in other parkinsonisms where different yet definite brainstem abnormalities are also described. According to the pathophysiological hypothesis [6] our data suggested that in DLB the brainstem is the site of initial lesions, consisting of αsynuclein deposits. Synucleinopathy is ascending from the brainstem, progressively involving the lower brainstem and inducing the appearance of REM Sleep Behaviour Disorder (RBD), then the mesencephalus, inducing the occurrence of parkinsonism and finally involving limbic structures, inducing hallucinations and psychosis, and cortical areas, inducing cognitive disorders. R2 latency delay might be attributed to the ascending synucleinopathy inducing the appearance of RBD, but our findings suggest that this possible correlation is controversial, as normal R2 latencies were observed in PD and MSA patients presenting with RBD, while delayed R2 latencies were recorded in 5 DLB patients who did not present with RBD. Our findings suggest instead that R2 latency delay in DLB is independent of the presence of RBD. The correlation with scores assessing cognitive fluctuations suggests that R2 abnormalities might evidence dysfunction of reticular brain

Hz), corresponding to EEG CSA pattern 2 observed only in DLB patients [5]. Bottom. Example of neurophysiological recordings in an AD patient. Left. P300 recording. P300 response is delayed (410 ms) with posterior to anterior latency distribution gradient. Amplitude is higher in posterior derivations (14.9 uV) compared to anterior derivations (14.5 uV) (inset, I-shaped lines mark peak latencies in the responses to target stimuli). Fz-RF: frontal derivation, Cz-RF: central derivation, Pz-RF: posterior derivation, EOG: ocular derivation. Right. CSA of the same patient, with stable alpha dominant frequency at 9.5 Hz. EOG: electrooculogram; nt: non target stimuli, t: target stimuli; DLB: dementia with Lewy

bodies; AD: Alzheimer's disease.

Progressive Supranuclear Palsy (PSP).

stem pathways involved in vigilance regulation.

Fig. 2. Examples of P300 and CSA traces in one DLB and one AD patient. Top. Example of recordings in a DLB patient. Left. P300 recording. P300 response appears delayed (420 ms) with no latency inter-electrode distribution gradient and has a higher amplitude in frontal derivations (26.7 uV) compared to posterior derivations (21.0 uV) (inset, U-shaped bars mark the anterior-to-posterior (reversed) latency distribution gradient in the responses to target stimuli). Fz: frontal derivation, Cz: central derivation, Pz: posterior derivation, EOG: ocular derivation. Nt: non-target stimuli, t: target stimuli. Right. Quantitative EEG of the same patient represented as Compressed Spectral Array (CSA), i.e. arrays of traces are the representation of EEG power distribution in consecutive 2-second epochs. Peaks of power (amplitude) are found in variable frequencies shifting from alpha (9.6 Hz) to pre-alpha (6.4

Fig. 2. Examples of P300 and CSA traces in one DLB and one AD patient. Top. Example of recordings in a DLB patient. Left. P300 recording. P300 response appears delayed (420 ms) with no latency inter-electrode distribution gradient and has a higher amplitude in frontal derivations (26.7 uV) compared to posterior derivations (21.0 uV) (inset, U-shaped bars mark the anterior-to-posterior (reversed) latency distribution gradient in the responses to target stimuli). Fz: frontal derivation, Cz: central derivation, Pz: posterior derivation, EOG: ocular derivation. Nt: non-target stimuli, t: target stimuli. Right. Quantitative EEG of the same patient represented as Compressed Spectral Array (CSA), i.e. arrays of traces are the representation of EEG power distribution in consecutive 2-second epochs. Peaks of power (amplitude) are found in variable frequencies shifting from alpha (9.6 Hz) to pre-alpha (6.4 Hz), corresponding to EEG CSA pattern 2 observed only in DLB patients [5]. Bottom. Example of neurophysiological recordings in an AD patient. Left. P300 recording. P300 response is delayed (410 ms) with posterior to anterior latency distribution gradient. Amplitude is higher in posterior derivations (14.9 uV) compared to anterior derivations (14.5 uV) (inset, I-shaped lines mark peak latencies in the responses to target stimuli). Fz-RF: frontal derivation, Cz-RF: central derivation, Pz-RF: posterior derivation, EOG: ocular derivation. Right. CSA of the same patient, with stable alpha dominant frequency at 9.5 Hz. EOG: electrooculogram; nt: non target stimuli, t: target stimuli; DLB: dementia with Lewy bodies; AD: Alzheimer's disease.

We performed a study of the blink reflex in patients with PD, DLB, MSA, AD and Progressive Supranuclear Palsy (PSP).

The subjects were comfortably sitting on an armchair in a quiet room, with eyes gently closed. The recordings took place in a temperature-controlled room (at about 25°C) in halflight. The cathode was placed over the supraorbital foramen and the anode 2cm rostrally. Surface electrodes were placed on the inferior part of the orbicularis oculi muscles on each side, recording ipsilateral R1, and ipsilateral and contralateral R2 and R3. Ground electrode was placed under the chin. Stimuli of 0,1ms of duration with intensity of 5-10 mA elicited stable R1 in repeated trials. Because surface electrodes lay only few centimetres away from the cathode, R1 tended to overlap the stimulus artifact, which could last more than 10ms. A special amplifier with a short blocking time (0.1ms) and low internal noise (0.5 uV at a bandwidth of 2kHz) minimized the problem of stimulus artifact. Signals were amplified and filtered (bandwidth 20-2000Hz), to avoid habituation the interstimuli intervals must be of at least 7 sec, 5-10 responses per site were elicited and stored. BR recording were previously described in MSA, PSP and PD patients: all reports showed R2 latencies inside the 2 SD of the mean and only evidenced enhancement or inhibition of R1-R2 in excitability-duration curve paradigms [52,53,55] in untreated PD. Recently we studied the BR in parkinsonism [56]: in all PD, MSA, PSP and AD patients we found normal R1 and R2 latencies inside the 2SD of the control mean independently of the presence of RBD. Only in DLB patients we found R2 latencies clustering in the upper limits of normality or definitely above the limits (figure 3). All findings were statistically significant. Thus, BR recordings might reveal brainstem dysfunction in DLB, but not in other parkinsonisms where different yet definite brainstem abnormalities are also described. According to the pathophysiological hypothesis [6] our data suggested that in DLB the brainstem is the site of initial lesions, consisting of αsynuclein deposits. Synucleinopathy is ascending from the brainstem, progressively involving the lower brainstem and inducing the appearance of REM Sleep Behaviour Disorder (RBD), then the mesencephalus, inducing the occurrence of parkinsonism and finally involving limbic structures, inducing hallucinations and psychosis, and cortical areas, inducing cognitive disorders. R2 latency delay might be attributed to the ascending synucleinopathy inducing the appearance of RBD, but our findings suggest that this possible correlation is controversial, as normal R2 latencies were observed in PD and MSA patients presenting with RBD, while delayed R2 latencies were recorded in 5 DLB patients who did not present with RBD. Our findings suggest instead that R2 latency delay in DLB is independent of the presence of RBD. The correlation with scores assessing cognitive fluctuations suggests that R2 abnormalities might evidence dysfunction of reticular brain stem pathways involved in vigilance regulation.

Is There a Place for Clinical Neurophysiology Assessments in Synucleinopathies? 313

of brain stem reticular pathways involved in vigilance regulation. The administration of donepezil significantly improved BR response in DLB patients, with a mean reduction of 8.2%. R2 mean latency reduction was highly correlated with R2 mean latency delay at baseline, with a 46% of patients showing no difference between R2 mean latency at baseline and after treatment. Thus, the reduction of R2 latency was evident only in patients who had delayed R2 at baseline. A possible explanation is a "bottom effect" of R2 mean latency reduction, meaning that the correction of brain stem dysfunction by ChEI treatment is mostly evident when the alteration of subcortical cholinergic networks is so conspicuous to be evidenced by a BR response alteration. Another possible explanation is that 2 subpopulation of DLB patients can be recognized: responders and not responders to ChEI treatment. No correlation between R2 latency reduction and MMSE scores was found, as expected because of the short test-retest interval and learning effect [65]. However, at baseline, a high correlation between R2 abnormalities and CAF and ODFA scores was found, suggesting that responders are those patients with the worst grade of cognitive fluctuations. In our study, CAF scores were not significantly modified by the 2-week treatment, again as expected, because CAF scores track behaviors, reported by caregivers, of the last month. However, ODFA scores were significantly different after treatment compared with baseline. Our study suggests therefore that ChEI effect is mediated by correction of dysfunction of the brain stem reticular pathways involved in vigilance regulation. A previous study [66] had shown the correlation between improvement of attentional activities and improvement of neuropsychological scores after ChEI therapy and the finding was confirmed by following reports [60,62,67]. The lack of BR response alterations and subsequent absence of R2 latency modification by ChEI in AD patients suggest that due to the lower cholinergic functioning in DLB, a greater potential improvement from these drugs than that seen in AD might be expected, at least in the early phases of DLB pathophysiology, when a prevalent brain stem involvement is called into cause [6]. Furthermore, the presence of fewer neurofibrillary tangles and neuritic plaques and of less neuronal loss in DLB than AD [68,69] suggests that neurons in DLB are more viable than those in AD and could be more responsive to cholinergic stimulation [70-72]. These data suggest that the presence of alterations of neurophysiological responses tracking brain stem reticular formation might also predict the response to ChEI in DLB, as concluded in previous studies [67,73] about the efficacy of ChEI on cognitive impairments and psychiatric symptoms, and foster further studies on the long-term effect of ChEI and

Quantitative Electroencephalography (QEEG) is the measurement, using digital technology, of electrical patterns at the surface of the scalp which primarily reflect cortical activity or "brainwaves". A multi-electrode recording of brain wave activity is recorded and converted into numbers by a computer. These numbers are then statistically analysed and are converted into a colour map of brain functioning. Digital EEG techniques have grown rapidly in both technology and popularity since the early 1980's for recording, reviewing and storing EEG data. Compared to other systems, QEEG is a non-invasive procedure and offers a superior temporal (time) resolution compared with fMRI, SPECT and PET imaging

identification of responders.

**6. Quantitative eeg:qeeg** 

Fig. 3. Example of a blink response in a control subject (top: about 30msec) and in a patient with DLB (bottom: about 45msec). Note the delayed R2 response in the DLB patient in both the ipsilateral and contralateral recordings.

Current interpretations of BR neurophysiology [51,57] suggest that R2 abnormality should be ascribed to disruption of the afferent pathway when it is evident in ipsilateral and contralateral responses to stimuli of one side and efferent when the abnormality is observed in ipsilateral or contralateral responses of only one side, independently of the site of stimulation. Only in 3 of the DLB patients presenting with R2 delays, discrepant latencies on the two sides of stimulation were found [8], yet ipsilateral and contralateral responses were always overlapping, thus it is likely that the afferent pathway is prominently involved in DLB. In a successive study we tested the supposition that BR alterations present in DLB patients are sensitive to cholinergic modulation. It is known indeed, that choline acetyltransferase enzyme levels are lower in DLB compared with AD [58] whereas high muscarinic receptor density has been found in DLB [59]. Alterations of this cortical network are the pathophysiological correlate of cognitive impairment and attention deficit in DLB and are accompanied by abnormal electrocortical arousal [8,60] with alteration of electroencephalogram, event-related potential and choice reaction time. Administration of donepezil has been shown to significantly improve cognitive scores as well as electroencephalogram and event-related potential alterations in patients with fluctuating cognition [12] as a result of improvement of attentional participation in tested activities [60- 63]. So we assessed whether BR alterations present in DLB patients are sensitive to cholinergic modulation [64]. We evaluated 26 patients affected by DLB and 20 patients affected by AD: for each patient, we performed BR recordings before and after 1 and 2 weeks of treatment with donepezil. The correlation between R2 abnormalities and score assessing cognitive fluctuations suggest that R2 latency delay might evidence dysfunctions

Fig. 3. Example of a blink response in a control subject (top: about 30msec) and in a patient with DLB (bottom: about 45msec). Note the delayed R2 response in the DLB patient in both

Current interpretations of BR neurophysiology [51,57] suggest that R2 abnormality should be ascribed to disruption of the afferent pathway when it is evident in ipsilateral and contralateral responses to stimuli of one side and efferent when the abnormality is observed in ipsilateral or contralateral responses of only one side, independently of the site of stimulation. Only in 3 of the DLB patients presenting with R2 delays, discrepant latencies on the two sides of stimulation were found [8], yet ipsilateral and contralateral responses were always overlapping, thus it is likely that the afferent pathway is prominently involved in DLB. In a successive study we tested the supposition that BR alterations present in DLB patients are sensitive to cholinergic modulation. It is known indeed, that choline acetyltransferase enzyme levels are lower in DLB compared with AD [58] whereas high muscarinic receptor density has been found in DLB [59]. Alterations of this cortical network are the pathophysiological correlate of cognitive impairment and attention deficit in DLB and are accompanied by abnormal electrocortical arousal [8,60] with alteration of electroencephalogram, event-related potential and choice reaction time. Administration of donepezil has been shown to significantly improve cognitive scores as well as electroencephalogram and event-related potential alterations in patients with fluctuating cognition [12] as a result of improvement of attentional participation in tested activities [60- 63]. So we assessed whether BR alterations present in DLB patients are sensitive to cholinergic modulation [64]. We evaluated 26 patients affected by DLB and 20 patients affected by AD: for each patient, we performed BR recordings before and after 1 and 2 weeks of treatment with donepezil. The correlation between R2 abnormalities and score assessing cognitive fluctuations suggest that R2 latency delay might evidence dysfunctions

the ipsilateral and contralateral recordings.

of brain stem reticular pathways involved in vigilance regulation. The administration of donepezil significantly improved BR response in DLB patients, with a mean reduction of 8.2%. R2 mean latency reduction was highly correlated with R2 mean latency delay at baseline, with a 46% of patients showing no difference between R2 mean latency at baseline and after treatment. Thus, the reduction of R2 latency was evident only in patients who had delayed R2 at baseline. A possible explanation is a "bottom effect" of R2 mean latency reduction, meaning that the correction of brain stem dysfunction by ChEI treatment is mostly evident when the alteration of subcortical cholinergic networks is so conspicuous to be evidenced by a BR response alteration. Another possible explanation is that 2 subpopulation of DLB patients can be recognized: responders and not responders to ChEI treatment. No correlation between R2 latency reduction and MMSE scores was found, as expected because of the short test-retest interval and learning effect [65]. However, at baseline, a high correlation between R2 abnormalities and CAF and ODFA scores was found, suggesting that responders are those patients with the worst grade of cognitive fluctuations. In our study, CAF scores were not significantly modified by the 2-week treatment, again as expected, because CAF scores track behaviors, reported by caregivers, of the last month. However, ODFA scores were significantly different after treatment compared with baseline. Our study suggests therefore that ChEI effect is mediated by correction of dysfunction of the brain stem reticular pathways involved in vigilance regulation. A previous study [66] had shown the correlation between improvement of attentional activities and improvement of neuropsychological scores after ChEI therapy and the finding was confirmed by following reports [60,62,67]. The lack of BR response alterations and subsequent absence of R2 latency modification by ChEI in AD patients suggest that due to the lower cholinergic functioning in DLB, a greater potential improvement from these drugs than that seen in AD might be expected, at least in the early phases of DLB pathophysiology, when a prevalent brain stem involvement is called into cause [6]. Furthermore, the presence of fewer neurofibrillary tangles and neuritic plaques and of less neuronal loss in DLB than AD [68,69] suggests that neurons in DLB are more viable than those in AD and could be more responsive to cholinergic stimulation [70-72]. These data suggest that the presence of alterations of neurophysiological responses tracking brain stem reticular formation might also predict the response to ChEI in DLB, as concluded in previous studies [67,73] about the efficacy of ChEI on cognitive impairments and psychiatric symptoms, and foster further studies on the long-term effect of ChEI and identification of responders.

#### **6. Quantitative eeg:qeeg**

Quantitative Electroencephalography (QEEG) is the measurement, using digital technology, of electrical patterns at the surface of the scalp which primarily reflect cortical activity or "brainwaves". A multi-electrode recording of brain wave activity is recorded and converted into numbers by a computer. These numbers are then statistically analysed and are converted into a colour map of brain functioning. Digital EEG techniques have grown rapidly in both technology and popularity since the early 1980's for recording, reviewing and storing EEG data. Compared to other systems, QEEG is a non-invasive procedure and offers a superior temporal (time) resolution compared with fMRI, SPECT and PET imaging

Is There a Place for Clinical Neurophysiology Assessments in Synucleinopathies? 315

mPS was divided automatically into 4 frequency bands (1-3.9 Hz [delta], 4-5.5 Hz [theta], 5.6-7.9 Hz [fast theta or pre-alpha], 8-12 Hz [alpha]). These bands were defined after the post hoc analysis with the purpose to facilitate identification of differences, in the description of results, as statistical differences were evidenced when theta band was halved in two parts (4-5.5 Hz, theta and 5.6-7.9 Hz pre-alpha). Fast Fourier transform-QEEG program expressed power values automatically after a log transform (log[x/(1-x)]) and indicated the Dominant Frequency (DF) of the entire power spectrum of each epoch, i.e. the specific frequency where the maximum power for a single epoch or a sum of multiple epochs was contained. Mean Relative Power Spectra (mRPS: percentage of the global mPS of each frequency band) were computed and log transformed [79] to normalize the data, automatically calculated and expressed in numeric percentages for each one of the single epochs obtained from each scalp derivation. EEG power spectra were represented as scalp maps of band amplitudes measured on the 180 sec total analysis (Total Power) and analyzed as Mean Frequency (MF), indicating the average frequency for the 90 epochs, and as mean frequency variability (MFV), representing changes of mean frequency during the 90 epochs. Single channel power spectra were also represented as Compressed Spectral Arrays (CSA) showing the sequences of absolute or relative power spectra in each one of the 90 analysed epochs. mRPS from all the scalp derivations were averaged in order to obtain a single Global Mean RPS representative of the frequency band powers in each patient expressing the average distribution of powers recorded from all the derivations. Eventually, in each epoch, mean band power in each of the three groups of patients (AD, DLB, PDD) as well as in the control

group was then computed by averaging the values of subjects in each group.

Our EEG study [5] completed and detailed results of a preliminary study performed with Magneto-Encephalography (MEG) recordings [80] suggesting that activities in parietal and occipital areas differentiate early DLB from early AD. MEG technique excluded a reference effect, showed differences in reactivity of alpha rhythms between groups of patients, and explored coherence: therefore the present study was focused on waking-closed eyes condition and on methods evidencing differences. The different EEG variables analysed in our study showed some distinct and specific patterns in patients affected by DLB or PDD with cognitive fluctuations (PDDF). When EEGs were interpreted with the classic visual inspection methods, absent alpha in posterior derivations was observed in 63.9% of DLB and in none of AD patients. In PDD absent alpha was observed in 25.7% of patients (all from PDDF group). Intermittent delta and sharp transients, described in previous studies [2,4,9], occurred more frequently in DLB patients compared to AD (13.9% vs 2.5% and 5.6% vs 2.5%) yet these findings were rare, and therefore scarcely useful for diagnostic purposes. Visual inspection was not sufficient to evidence other differences which were observed with QEEG methods: the first relevant finding was the identification of slow activities in posterior derivations, with a frequency of 5.6-7.9 Hz, which were observed in all DLB patients and significantly separated DLB patients from AD. This activity was defined prealpha because it was suppressed by eye opening. Two studies [81,82] quantified EEG characteristics during polisomnography (PSG) in patients with RBD and DLB/PDD and indicated differences with controls in the same EEG frequency band. QEEG was analyzed with different methods. Total power and mRPS showed that pre-alpha activity on posterior derivations expressed the highest statistical difference between AD or PDD without cognitive fluctuations (PDDNF) and DLB or PDDF patients (p<0.01). The study of MFV

techniques. MEG systems, though providing a high temporal and spatial resolution, are a relatively expensive means of monitoring the brain as compared with QEEG. Recently we had designed a QEEG study in a cohort of patients affected by early AD and DLB, whose diagnoses were confirmed by laboratory methods and by a 2-year follow-up, which allowed confirming or discarding earlier diagnoses, and thus reaching the best possible level of certainty on the classification of these two disorders [5]. As specific EEG abnormalities reflecting the presence of cognitive fluctuations (superimposition of pre-alpha/theta activity on alpha dominant frequency, or of theta/delta activity on dominant pre-alpha frequency) were evidenced in early DLB [5,74-76] while alpha dominant activity was more stable in early AD [1], we evaluated possible correlations between P300 and EEG characteristics in AD and DLB. Several electroencephalographic studies on dementia were performed in the years preceding the identification of DLB as a widespread cognitive disorder. Slowing of the rhythms and reduced coherence among brain regions, increased theta and delta activity, in parallel with reduction of alpha and beta rhythms were observed in patients affected by putative AD [76]: computerized EEG spectral analysis showed an increase in delta and theta power in AD patients compared to controls mainly in the left temporal area. EEGs were recorded with Ag/AgCl disk scalp electrodes placed on 19 derivations corresponding to Fp1, Fp2, F3, F4, C3, C4, P3, P4, O1, O2, F7, F8, T3, T4, T5, T6, Cz, Fz, Pz positions of the 10- 20 International System with supplementary A1, A2 derivations. Derivations were grouped in order to define 5 scalp regions: anterior (Fz, Fp2, F7, Fp1, F3, F4, F8), central (T3, C3, Cz, C4, T4), posterior (T5, P3, Pz, P4, T6, O1, O2) and peripheral (Fp1, Fp2, F8, T4, T6, O1, O2, T5, T3, Fz) or internal (F3, F4, Fz, C3, Cz, C4, P3, Pz, P4). Reference was the mean (mean reference) of recordings from all scalp leads, A1, A2 signals were also stored for digitalized derivation reconstruction. Ground was placed at FpZ. Impedance was below 5 KOhm. The patients were seated in a quiet room on a comfortable armchair, awake with closed eyes under continuous control (Video EEG); wakefulness of the patients was verified every 2 min by asking to open eyes and checking block reactions; 2 supplementary derivations monitored electro-oculography (vertical and horizontal), two derivations monitored possible interference of tremor and two pairs of additional bipolar recording channels for the respiration and electrocardiogram were applied. EEG was acquired as a continuous signal for 30 min and visually inspected for current clinical interpretation or detection of artifacts and stored in order to be epoched in off-analysis setting as series of 2 seconds-long epochs. EEGs interpreted with classical visual inspection, corresponding to categories reported in previous literature [77,78] were defined as Classic Interpretation Methods (CIM) in results. The computer collected 10 minutes of EEG recorded with closed eyes, digitized at 1024 Hz with a low filter at 0.5 Hz and high filter at 70 Hz (decay constant 12 dB) with a 50 Hz notch filter in each channel. Blocks of artifact-free 2 seconds-long epochs appearing consecutively for 20-40 sec were selected off-line by visual inspection after pre-programmed automatic blink reduction and muscle and tremor artifact rejection system and were compared with the remaining artifact-free epochs in order to avoid possible discrepancies among acquired sets. A total of 90 epochs per patient were processed by an automatic transforming program present in the NEUROSCAN SynAmps System performing a fast Fourier Transform on each second of EEG acquisition, allowing a frequency sensitivity=0.05 Hz. The obtained spectra values were then processed in order to compute a mean Power Spectrum (mPS) for each epoch and for each channel and expressed in square uV (uV2). The

techniques. MEG systems, though providing a high temporal and spatial resolution, are a relatively expensive means of monitoring the brain as compared with QEEG. Recently we had designed a QEEG study in a cohort of patients affected by early AD and DLB, whose diagnoses were confirmed by laboratory methods and by a 2-year follow-up, which allowed confirming or discarding earlier diagnoses, and thus reaching the best possible level of certainty on the classification of these two disorders [5]. As specific EEG abnormalities reflecting the presence of cognitive fluctuations (superimposition of pre-alpha/theta activity on alpha dominant frequency, or of theta/delta activity on dominant pre-alpha frequency) were evidenced in early DLB [5,74-76] while alpha dominant activity was more stable in early AD [1], we evaluated possible correlations between P300 and EEG characteristics in AD and DLB. Several electroencephalographic studies on dementia were performed in the years preceding the identification of DLB as a widespread cognitive disorder. Slowing of the rhythms and reduced coherence among brain regions, increased theta and delta activity, in parallel with reduction of alpha and beta rhythms were observed in patients affected by putative AD [76]: computerized EEG spectral analysis showed an increase in delta and theta power in AD patients compared to controls mainly in the left temporal area. EEGs were recorded with Ag/AgCl disk scalp electrodes placed on 19 derivations corresponding to Fp1, Fp2, F3, F4, C3, C4, P3, P4, O1, O2, F7, F8, T3, T4, T5, T6, Cz, Fz, Pz positions of the 10- 20 International System with supplementary A1, A2 derivations. Derivations were grouped in order to define 5 scalp regions: anterior (Fz, Fp2, F7, Fp1, F3, F4, F8), central (T3, C3, Cz, C4, T4), posterior (T5, P3, Pz, P4, T6, O1, O2) and peripheral (Fp1, Fp2, F8, T4, T6, O1, O2, T5, T3, Fz) or internal (F3, F4, Fz, C3, Cz, C4, P3, Pz, P4). Reference was the mean (mean reference) of recordings from all scalp leads, A1, A2 signals were also stored for digitalized derivation reconstruction. Ground was placed at FpZ. Impedance was below 5 KOhm. The patients were seated in a quiet room on a comfortable armchair, awake with closed eyes under continuous control (Video EEG); wakefulness of the patients was verified every 2 min by asking to open eyes and checking block reactions; 2 supplementary derivations monitored electro-oculography (vertical and horizontal), two derivations monitored possible interference of tremor and two pairs of additional bipolar recording channels for the respiration and electrocardiogram were applied. EEG was acquired as a continuous signal for 30 min and visually inspected for current clinical interpretation or detection of artifacts and stored in order to be epoched in off-analysis setting as series of 2 seconds-long epochs. EEGs interpreted with classical visual inspection, corresponding to categories reported in previous literature [77,78] were defined as Classic Interpretation Methods (CIM) in results. The computer collected 10 minutes of EEG recorded with closed eyes, digitized at 1024 Hz with a low filter at 0.5 Hz and high filter at 70 Hz (decay constant 12 dB) with a 50 Hz notch filter in each channel. Blocks of artifact-free 2 seconds-long epochs appearing consecutively for 20-40 sec were selected off-line by visual inspection after pre-programmed automatic blink reduction and muscle and tremor artifact rejection system and were compared with the remaining artifact-free epochs in order to avoid possible discrepancies among acquired sets. A total of 90 epochs per patient were processed by an automatic transforming program present in the NEUROSCAN SynAmps System performing a fast Fourier Transform on each second of EEG acquisition, allowing a frequency sensitivity=0.05 Hz. The obtained spectra values were then processed in order to compute a mean Power Spectrum (mPS) for each epoch and for each channel and expressed in square uV (uV2). The mPS was divided automatically into 4 frequency bands (1-3.9 Hz [delta], 4-5.5 Hz [theta], 5.6-7.9 Hz [fast theta or pre-alpha], 8-12 Hz [alpha]). These bands were defined after the post hoc analysis with the purpose to facilitate identification of differences, in the description of results, as statistical differences were evidenced when theta band was halved in two parts (4-5.5 Hz, theta and 5.6-7.9 Hz pre-alpha). Fast Fourier transform-QEEG program expressed power values automatically after a log transform (log[x/(1-x)]) and indicated the Dominant Frequency (DF) of the entire power spectrum of each epoch, i.e. the specific frequency where the maximum power for a single epoch or a sum of multiple epochs was contained. Mean Relative Power Spectra (mRPS: percentage of the global mPS of each frequency band) were computed and log transformed [79] to normalize the data, automatically calculated and expressed in numeric percentages for each one of the single epochs obtained from each scalp derivation. EEG power spectra were represented as scalp maps of band amplitudes measured on the 180 sec total analysis (Total Power) and analyzed as Mean Frequency (MF), indicating the average frequency for the 90 epochs, and as mean frequency variability (MFV), representing changes of mean frequency during the 90 epochs. Single channel power spectra were also represented as Compressed Spectral Arrays (CSA) showing the sequences of absolute or relative power spectra in each one of the 90 analysed epochs. mRPS from all the scalp derivations were averaged in order to obtain a single Global Mean RPS representative of the frequency band powers in each patient expressing the average distribution of powers recorded from all the derivations. Eventually, in each epoch, mean band power in each of the three groups of patients (AD, DLB, PDD) as well as in the control group was then computed by averaging the values of subjects in each group.

Our EEG study [5] completed and detailed results of a preliminary study performed with Magneto-Encephalography (MEG) recordings [80] suggesting that activities in parietal and occipital areas differentiate early DLB from early AD. MEG technique excluded a reference effect, showed differences in reactivity of alpha rhythms between groups of patients, and explored coherence: therefore the present study was focused on waking-closed eyes condition and on methods evidencing differences. The different EEG variables analysed in our study showed some distinct and specific patterns in patients affected by DLB or PDD with cognitive fluctuations (PDDF). When EEGs were interpreted with the classic visual inspection methods, absent alpha in posterior derivations was observed in 63.9% of DLB and in none of AD patients. In PDD absent alpha was observed in 25.7% of patients (all from PDDF group). Intermittent delta and sharp transients, described in previous studies [2,4,9], occurred more frequently in DLB patients compared to AD (13.9% vs 2.5% and 5.6% vs 2.5%) yet these findings were rare, and therefore scarcely useful for diagnostic purposes. Visual inspection was not sufficient to evidence other differences which were observed with QEEG methods: the first relevant finding was the identification of slow activities in posterior derivations, with a frequency of 5.6-7.9 Hz, which were observed in all DLB patients and significantly separated DLB patients from AD. This activity was defined prealpha because it was suppressed by eye opening. Two studies [81,82] quantified EEG characteristics during polisomnography (PSG) in patients with RBD and DLB/PDD and indicated differences with controls in the same EEG frequency band. QEEG was analyzed with different methods. Total power and mRPS showed that pre-alpha activity on posterior derivations expressed the highest statistical difference between AD or PDD without cognitive fluctuations (PDDNF) and DLB or PDDF patients (p<0.01). The study of MFV

Is There a Place for Clinical Neurophysiology Assessments in Synucleinopathies? 317

characterizing DLB [2]: the occurrence of positive CAF scores and of RBD during follow-up supported the categorization of patients. Patients diagnosed as affected by AD, who did not show evidence of fluctuating cognition or RBD, had rare EEG abnormalities. Yet, recent reports [84] showed that power spectra abnormalities in AD patients are characterized only by an increase in theta and a decrease in alpha and beta rhythms at rest and mostly limited to the temporal, and centro-parietal regions, or showed that entropy of EEG, expressing the

In conclusion, the definite presence of EEG abnormalities early in the course of DLB, with a core feature evidencing marked variability of dominant frequencies in posterior derivations, occurring every few seconds, or with the substitution of alpha with frequencies at 5.6-7.9 Hz, suggests that centers regulating EEG rhythms in parieto-occipital areas are affected in

Syncope associated to orthostatic hypotension, urinary incontinence and constipation is common symptoms in demented patients, mainly in DLB and in PDD. AD and FTD show less frequently autonomic dysfunction. There are non invasive tests including standard cardiovascular tests, 123I-MIBG cardiac scintigraphy, urodynamic tests, gastrointestinal motility studies, sweating reflexes and pupillary responses that assess autonomic

123I-MIBG is an analogue of the sympathomimetic amine guanethidine, which is used to determine the location, integrity, and function of postganglionic noradrenergic neurons [85]. Patients with PD can exhibit reduced cardiac 123I-MIBG-derived radioactivity without other evidence of autonomic failure, whereas those with DLB can have reduced cardiac 123I-MIBGderived radioactivity without evidence of parkinsonism [86]. 123I-MIBG may have the potential to differentiate PD from other causes of parkinsonism. For example, MSA and PSP

In PD and DLB, LB are encountered in extracranial tissues, notably in autonomic ganglia [87]. Cardiac sympathetic degeneration can be demonstrated early in the disease process before motor symptoms. In 2005, the DLB Consortium concluded that diminished uptake of 123I-MIBG on cardiac scintigraphy was a ''supportive'' clinical feature that required more

Positron emission tomography (PET) utilizes biologically active molecules in micromolar or nanomolar concentrations that have been labelled with short-lived positron-emitting isotopes. The physical characteristics of the isotopes and the molecular specificity of labeled molecules, combined with the high detection efficacy of modern PET scanners, provide a sensitivity for human in vivo measurement of indicator concentrations that is several orders of magnitude higher than with the other imaging techniques. Whereas the very short halflives of O15 (2 min) and C11 (20 min) limit their use to fully equipped PET centres with a cyclotron and radiopharmaceutical laboratory, F18 labelled tracers (half-life 110 min) can be produced in specialized centres and distributed regionally to hospitals running a PET scanner only. Clinical use of PET is now well established in clinical oncology and it is therefore becoming widely available in major hospitals. In addition to its use in research, brain PET also provides diagnostically relevant information mainly in neurodegenerative disorders, focal epilepsy and brain tumors. In dementia, the measurement of cerebral

irregularity and variability of EEG patterns, is reduced, rather than increased, in AD.

the early course of this disease.

dysfunction in these patients.

pose a difficult diagnostic challenge.

study [2].

**7. Other investigations and future prospectives** 

showed that variability of EEG activity was the second most relevant finding leading to the identification of specific EEG patterns in DLB or PDDF (p<0.001) as reported in a previous study [83]. The variability of the EEG frequencies in relaxed waking conditions was best evidenced by using the CSA method of representation, showing that dominant frequencies (DF) in DLB were either in the pre-alpha band or varied across time with pseudocyclic patterns of delta-theta/pre-alpha or theta-pre-alpha/alpha, differentiating DLB or PDDF patients from AD or PDDNF patients. CSA representation had only been previously used to assess coma or anaesthesia levels [83], and in the present study it allowed to evidence changes of EEG activities in single derivations. It allowed therefore to evaluate local variability, at contrary with total QEEG analyses and MF evaluations, and to evidence that significant differences among groups of patients were prevalent in posterior leads. CSA showed that changes of dominant activity could be separated in five patterns, salient at visual inspection of the sequence of traces: one, with dominant stable alpha, was only observed in early AD and in 54.3% of PDD (PDDNF), while the other patterns, differently grading the dominant frequency variability and pre-alpha presence, were only observed in posterior derivations of early DLB and PDDF. The abnormal patterns consisted either of a stable dominant activity at 5.6-7.9 Hz, encountered in 25% of DLB and 11.4% of PDD, but never in AD, or of unstable activities, all encompassing the presence of the 5.6-7.9 Hz activity and significant variations of the dominant frequency across time. When these EEG abnormalities are observed in a patient with initial signs of cognitive decline, i.e. MMSE<24, they support a diagnosis of DLB. Therefore our study clarifies and quantifies the suggestion that EEG might support the diagnosis [2]. When EEGs were recorded two years later [5], further alterations were observed which differentiated groups of patients, even though the administration of current therapies could have partly marred the results. In DLB patients and in 74.3% of PDD patients EEGs were similar, with a stable pre-alpha activity or unstable DF across time, with variability above 3 Hz, consisting of the presence of unstable alpha, pre-alpha, theta and delta activities. In 72.5% of AD patients and in 25.7% of PDD patients DFV was below 3 Hz and alpha activity was present. At follow-up patterns 2-4 were observed prominently in posterior derivations of DLB and PDDF patients, while only 27.5% of AD patients presented with similar EEG abnormalities. Pattern 5 was observed at followup in patients with severe cognitive deteriorations (5 DLB, 2 AD, 3 PDD), suggesting therefore that this degraded pattern is aspecific. In our experience we recorded pattern 5 activity also in cases of severe Progressive Supranuclear Palsy and Fronto-temporal dementia. In conclusion we would add two further considerations. First, PDD in its early course can be apparently separated in two different groups: one with fluctuating cognition elements and EEG pattern abnormalities akin to the ones observed in early DLB and a group with normal EEG akin to the AD patients and without fluctuating cognition. With followup, however, the majority of PDD patients (74.3%) presents with the same EEG abnormalities characterizing DLB. The presence of two different clusters in early PDD suggests that the distribution of neuropathological abnormalities might have different patterns in different patients, cumulating however across time to show, at follow-up, clinical and EEG patterns similar to the ones observed in DLB. Second, in AD we found less EEG abnormalities than previously reported [76]. We suggest that this finding depends on patients selection methods used in our study. In this study the selection was focused on elements of fluctuating cognition (CAF-ODFA scales) and presence of RBD, prominently characterizing DLB [2]: the occurrence of positive CAF scores and of RBD during follow-up supported the categorization of patients. Patients diagnosed as affected by AD, who did not show evidence of fluctuating cognition or RBD, had rare EEG abnormalities. Yet, recent reports [84] showed that power spectra abnormalities in AD patients are characterized only by an increase in theta and a decrease in alpha and beta rhythms at rest and mostly limited to the temporal, and centro-parietal regions, or showed that entropy of EEG, expressing the irregularity and variability of EEG patterns, is reduced, rather than increased, in AD. In conclusion, the definite presence of EEG abnormalities early in the course of DLB, with a core feature evidencing marked variability of dominant frequencies in posterior derivations, occurring every few seconds, or with the substitution of alpha with frequencies at 5.6-7.9 Hz, suggests that centers regulating EEG rhythms in parieto-occipital areas are affected in

#### the early course of this disease.

316 Neuroimaging for Clinicians – Combining Research and Practice

showed that variability of EEG activity was the second most relevant finding leading to the identification of specific EEG patterns in DLB or PDDF (p<0.001) as reported in a previous study [83]. The variability of the EEG frequencies in relaxed waking conditions was best evidenced by using the CSA method of representation, showing that dominant frequencies (DF) in DLB were either in the pre-alpha band or varied across time with pseudocyclic patterns of delta-theta/pre-alpha or theta-pre-alpha/alpha, differentiating DLB or PDDF patients from AD or PDDNF patients. CSA representation had only been previously used to assess coma or anaesthesia levels [83], and in the present study it allowed to evidence changes of EEG activities in single derivations. It allowed therefore to evaluate local variability, at contrary with total QEEG analyses and MF evaluations, and to evidence that significant differences among groups of patients were prevalent in posterior leads. CSA showed that changes of dominant activity could be separated in five patterns, salient at visual inspection of the sequence of traces: one, with dominant stable alpha, was only observed in early AD and in 54.3% of PDD (PDDNF), while the other patterns, differently grading the dominant frequency variability and pre-alpha presence, were only observed in posterior derivations of early DLB and PDDF. The abnormal patterns consisted either of a stable dominant activity at 5.6-7.9 Hz, encountered in 25% of DLB and 11.4% of PDD, but never in AD, or of unstable activities, all encompassing the presence of the 5.6-7.9 Hz activity and significant variations of the dominant frequency across time. When these EEG abnormalities are observed in a patient with initial signs of cognitive decline, i.e. MMSE<24, they support a diagnosis of DLB. Therefore our study clarifies and quantifies the suggestion that EEG might support the diagnosis [2]. When EEGs were recorded two years later [5], further alterations were observed which differentiated groups of patients, even though the administration of current therapies could have partly marred the results. In DLB patients and in 74.3% of PDD patients EEGs were similar, with a stable pre-alpha activity or unstable DF across time, with variability above 3 Hz, consisting of the presence of unstable alpha, pre-alpha, theta and delta activities. In 72.5% of AD patients and in 25.7% of PDD patients DFV was below 3 Hz and alpha activity was present. At follow-up patterns 2-4 were observed prominently in posterior derivations of DLB and PDDF patients, while only 27.5% of AD patients presented with similar EEG abnormalities. Pattern 5 was observed at followup in patients with severe cognitive deteriorations (5 DLB, 2 AD, 3 PDD), suggesting therefore that this degraded pattern is aspecific. In our experience we recorded pattern 5 activity also in cases of severe Progressive Supranuclear Palsy and Fronto-temporal dementia. In conclusion we would add two further considerations. First, PDD in its early course can be apparently separated in two different groups: one with fluctuating cognition elements and EEG pattern abnormalities akin to the ones observed in early DLB and a group with normal EEG akin to the AD patients and without fluctuating cognition. With followup, however, the majority of PDD patients (74.3%) presents with the same EEG abnormalities characterizing DLB. The presence of two different clusters in early PDD suggests that the distribution of neuropathological abnormalities might have different patterns in different patients, cumulating however across time to show, at follow-up, clinical and EEG patterns similar to the ones observed in DLB. Second, in AD we found less EEG abnormalities than previously reported [76]. We suggest that this finding depends on patients selection methods used in our study. In this study the selection was focused on elements of fluctuating cognition (CAF-ODFA scales) and presence of RBD, prominently
