Abnormal sensorimotor plasticity in CADASIL correlates with neuropsychological impairment
- Francisco J Palomar1,2,
- Aida Suárez3,
- Emilio Franco3,
- Fátima Carrillo1,
- Eulogio Gil-Néciga3,
- Pablo Mir1,2
- 1Unidad de Trastornos del Movimiento, Servicio de Neurología y Neurofisiología Clínica, Instituto de Biomedicina de Sevilla (IBiS), Hospital Universitario Virgen del Rocío/CSIC/Universidad de Sevilla, Seville, Spain
- 2Centro de Investigación Biomédica en Red sobre Enfermedades Neurodegenerativas (CIBERNED), Seville, Spain
- 3Unidad de Demencias, Servicio de Neurología y Neurofisiología Clínica, Hospital Universitario Virgen del Rocío, Seville, Spain
- Correspondence to Dr Pablo Mir, Unidad de Trastornos del Movimiento, Servicio de Neurología y Neurofisiología Clínica, Hospital Universitario Virgen del Rocío, Avda. Manuel Siurot sn., Sevilla 41013, Spain;
- Received 17 August 2012
- Revised 4 December 2012
- Accepted 12 December 2012
- Published Online First 10 January 2013
Objective Cerebral autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopathy (CADASIL) is a small vessel disease of the brain caused by mutations in the NOTCH3 gene. CADASIL progresses, in some cases, to subcortical dementia with a particular cognitive impairment. Different diseases in the dementia spectrum share a central cholinergic and sensorimotor plasticity alteration. We aimed to study different intracortical circuits and sensorimotor plasticity in CADASIL patients using transcranial magnetic stimulation protocols, and to determine whether these characteristics correlated with the results of clinical neuropsychological evaluation.
Methods Ten CADASIL patients and 10 healthy subjects were included in the study. All subjects underwent a transcranial magnetic stimulation study examining different intracortical circuits. Sensorimotor plasticity was also assessed using a paired associative stimulation and extensive neuropsychological tests.
Results CADASIL patients showed a lack of intracortical facilitation, short latency afferent inhibition and sensorimotor plasticity when compared with control subjects. CADASIL patients also showed an altered neuropsychological profile. Correlation between sensorimotor plasticity and neuropsychological alterations was observed in CADASIL patients.
Conclusions These results suggest that acetylcholine and glutamate could be involved in the dementia process in CADASIL and that abnormal sensorimotor plasticity correlates with the neuropsychological profile in CADASIL patients.
Cerebral autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopathy (CADASIL) is a typical hereditary cerebrovascular disease. CADASIL occurs due to mutations in the NOTCH3 gene in chromosome 19. This condition causes small subcortical lacunar infarcts and also neuronal death in the cortex.1 CADASIL is clinically characterised by cerebral ischaemic events (with white matter lesions), migraine with aura, psychiatric disorders and progressive cognitive impairment that leads to dementia.2 ,3 CADASIL is a good model of pure vascular dementia with white matter lesions that may directly affect cholinergic subcortical projections.4 ,5 Gamma-AminoButyric Acid (GABA) and glutamate are also involved in neurotransmission of the cortico-subcortical loop. Glutamate stimulates the release of striatal dopamine6 and basal acetylcholine, whereas GABA is the main neurotransmitter within the basal ganglia.7 The interaction between dopamine, glutamate, acetylcholine and GABA probably underlies the corticostriatal-thalamocortical negative feedback loop in order to limit cortical overstimulation.8
Transcranial magnetic stimulation (TMS) is a non-invasive technique to study primary motor cortex and corticospinal tract. TMS has come to be a useful method for evaluating the involvement of different neurotransmitter systems in the healthy brain and in a variety of neurological diseases. TMS has also been used to study and describe motor changes in patients with cognitive disorders such as AD,9–12 subcortical ischaemic vascular dementia8 ,13 and CADASIL14 where a central cholinergic impairment and an increased cortical excitability has been recently observed, suggesting GABA, glutamate and cholinergic impairment in CADASIL patients as in AD.14
The aim of this study was to better describe different intracortical circuits and sensorimotor plasticity in CADASIL patients by using appropriate TMS protocols. We also aim to correlate possible TMS abnormalities in CADASIL patients with their neuropsychological examination. Our hypothesis was that CADASIL patients could have impairment in these TMS protocols that could confirm and extend, to GABA and glutamate, the previously reported cholinergic impairment.14
Ten patients clinically diagnosed and genetically proven CADASIL (five men, five women, mean age 56.9±9.8 years, range 36–70 years) and 10 healthy subjects with similar age (four men, six women, mean age 57±10.32 years, range 39–71 years) were included in the study. Patients were recruited from the Unidad de Demencias at the Hospital Universitario Virgen del Rocío in Seville. Mean value of the Mini-Mental State Examination (MMSE)15 in CADASIL patients was 24.5±5.02 (ranging from 15 to 30) while mean value of MMSE in healthy subjects was 29.5±0.85 (ranging from 28 to 30). Six CADASIL patients had a MMSE score below 26 while no healthy subject had a MMSE score under 26. All patients’ demographic, clinical, genetic and radiological data are given in table 1. Detailed methods for radiological abnormalities are given in the online supplementary material. No CADASIL patient or healthy subject presented any motor deficit (assessed by a basic neurological examination) or peripheral sensory deficit (assessed by peripheral sensory nerve conduction studies). Detailed clinical data about sensory and motor functions are detailed in table 1S of the online supplementary material. At the time of the study, none of the participants were on any medications that could affect the TMS measurements. Written informed consent was obtained from all participants and the study was approved by the local ethics committee and conducted in accordance with the Declaration of Helsinki.
Electromyographic (EMG) recordings were made from the abductor pollicis brevis (APB) and first dorsal interosseous (FDI) muscles on the side contralateral to the stimulated cortex with Ag-AgCl surface electrodes using a belly-tendon montage. EMG signals were amplified (1000×) and band-pass filtered (bandwidth 20 Hz to 2 kHz) with a Digitimer D360 amplifier (Digitimer, UK), acquired at a sampling rate of 5 kHz through a CED 1401 laboratory interface (Cambridge Electronic Design, Cambridge, UK) and stored on a PC. The EMG traces were analysed using customised SIGNAL software V.4.00.
Transcranial magnetic stimulation procedure and experimental design
Single and paired-pulse TMS of the primary motor cortex were applied using Magstim 200 magnetic stimulators (Magstim Company, Carmarthenshire, Wales, UK). The magnetic stimulators were connected to a standard figure-of-eight coil with an external diameter of 70 mm. A BiStim module (Magstim Company, Carmarthenshire, Wales, UK) was used to interconnect stimulators when paired-pulse TMS protocols were performed. The coil was held tangentially to the skull with the handle pointing backwards and laterally at an angle of ∼45° to the sagittal plane in order to generate a posterior–anterior current in the brain. The ‘hot spot’ was defined as the optimal scalp position for eliciting a motor evoked potential (MEP) of maximal amplitude in the contralateral FDI or APB muscles depending on the performed protocol.
Patients and healthy subjects were tested on left hemisphere in two different TMS sessions, separated by at least a week, one session for FDI targeted protocols and the other one for APB plus FDI targeted protocols. FDI targeted protocols included short intracortical inhibition (SICI), intracortical facilitation (ICF), long intracortical inhibition (LICI) and cortical silent period (CSP). APB plus FDI targeted protocols included short afferent inhibition (SAI) and a paired associative stimulation (PAS) protocol which includes an assessment of resting motor threshold (RMT) and 1 mV MEP before and at 0 (t0), 15 (t15) and 30 (t30) min after the PAS protocol (figure 1). The N20 wave latency and amplitude of cortical somatosensory evoked potentials (SEPs) response, from right median nerve, were recorded and analysed to ensure that there were no delays or amplitude abnormalities in the arrival of peripheral sensory inputs to the left primary somatosensory cortex. Details of different TMS protocols and SEPs performed are more accurately described in the online supplementary material.
Neuropsychological examinations were performed 1 week after the TMS examination. All subjects underwent a neuropsychological assessment with adapted tests for Spanish population to explore different cognitive areas. MMSE15 was used as a screening test for cognitive impairment in both studied groups. Cognitive areas included in the neuropsychological assessment were processing speed, executive function, visual and verbal memory, naming task and constructional praxis. Detailed tests performed for each cognitive area are listed in table 2S of the online supplementary material.
We divided the data into two groups: healthy subjects (CONTROL) and CADASIL patients. The Shapiro-Wilk test was used to assess assumption of normality of the obtained data before the use of parametric tests. Non-normally distributed data were analysed using non-parametric tests between groups. TMS intensities and clinical and demographic data were analysed using χ2 and Student's-t tests for independent measures depending on data type. The result of CSP and N20 latencies and amplitudes of SEPs were evaluated using a t-student comparison between two different groups of study. Results of all other TMS protocols were compared using repeated measures ANOVA with a factor GROUP (CADASIL vs CONTROL) as a between-subject factor. For SICI, ICF, LICI and SAI protocols we used interstimuli intervals (ISI) as a within-subject factor. In the analysis of APB MEP amplitude and variation of APB RMT before and after the PAS protocol we used TIME as a within-subject factor. Mauchley's test assessed sphericity and Greenhouse-Geisser correction was used for non-spherical data. Significant main effects in ANOVA were followed by a post hoc paired t test analysis with Bonferroni correction.
Results of different neuropsychological tests were first compared between groups by using Student's t test for independent measures for those normally distributed in both groups and using Mann-Whitney U test for those not normally distributed. Any statistically significant difference was included in correlation analysis with the most statistically significant TMS finding in CADASIL group using Pearson's or Spearman's correlation analysis depending on normality assumption rising from the initial Shapiro-Wilk test. Afterwards, a multivariate regression test was performed using those neuropsychological tests with a high level of correlation to different TMS findings, sex age and years of formal education.
The significance was preset at p<0.05 for all statistical analyses. All statistical analyses were carried out using PASW 18.0.
Magnetic stimulation intensities
No statistical differences were observed in any of the TMS and electrical intensities used in different TMS protocols as presented in the online supplementary material.
SEPs latencies and amplitudes
Mean values of N20 latency (means±SD) (CADASIL: 20.26 ms±0.98 ms vs CONTROL: 19.61 ms±1.32 ms) were not significantly different between CADASIL patients and healthy subjects (Student's t test p=0.321). Mean values of N20 amplitude (means±SD) (CADASIL: 4.19 μV±0.88 μV vs CONTROL: 4.08 μV±1.44 μV) were also not significantly different between CADASIL patients and healthy subjects (Student's t test p=0.880).
SICI and ICF
We found a significant ISI effect in the amplitude of conditioned MEP compared with unconditioned ones (F4,15=29.287; p<0.001) as well as an ISI×GROUP interaction (F4,15=3.949; p<0.022) in repeated measures ANOVA analysis. Statistically significant inhibition at 2 ms (p<0.001) and 3 ms (p<0.001) and statistically significant facilitation at 10 ms (p<0.001) and 12 ms (p<0.001) were observed in the CONTROL group. Only statistically significant inhibition at 2 ms (p<0.001) and 3 ms (p<0.001) was observed in the CADASIL group while no statistically significant facilitation was present at 10 ms (p>0.05) and 12 ms (p>0.05). Post hoc paired t test analysis with Bonferroni correction in ICF ISIs showed statistically significant differences between CONTROL and CADASIL groups at 10 ms (p=0.003) and 12 ms (p=0.007) (figure 2A).
We found a significant ISI effect in the amplitude of conditioned MEP with respect to unconditioned ones (F3,16=10.210; p=0.001) but no ISI×GROUP interaction (F3,16=0.404; p<0.727) in repeated measures ANOVA analysis. Statistically significant inhibition at 100 ms (p=0.015) and 150 ms (p=0.048) was observed in the CONTROL group, but was not observed at 250 ms (p>0.05). The same was true for the CADASIL group, in which statistically significant inhibition was observed at 100 ms (p=0.043) and at 150 ms (p=0.003) but no inhibition at 250 ms (p>0.05) (figure 2B).
Cortical silent period
Student's t test for analysing independent measures analysis did not show any statistically significant differences between CONTROL and CADASIL groups with regard to CSP duration (Student's t test p=0.732).
Short afferent inhibition
Nine patients as well as nine control subjects participated in this experiment. We found a significant ISI effect (F3,14=14.429; p=0.001) as well as ISI×GROUP interaction (F3,14=2.869; p=0.046) in repeated measures ANOVA analysis. Statistically significant inhibition at 20 ms (p<0.001) and 22 ms (p<0.001) with no statistically significant change in APB MEP amplitude at 30 ms (p>0.05) was observed in the CONTROL group. No changes in conditioned APB MEP were observed at any specific ISI in the CADASIL group (20 ms p=0.115, 22 ms p=0.105 and 30 ms p>0.05). Post hoc paired t test analysis with Bonferroni correction showed statistically significant differences between CONTROL and CADASIL groups at 20 ms (p=0.001) and 22 ms (p=0.011) (figure 3). There were no significant ISI effects (F3,14=1.219; p=0.248) as well as no TIME×GROUP interaction (F3,14=1.327; p=0.256) when FDI MEP amplitudes were analysed.
Nine patients as well as nine control subjects participated in this experiment. All subjects fulfilled the attention criteria and were able to count correctly the number of median nerve electrical stimuli delivered during the PAS protocol. In APB MEP amplitude analysis, we found a significant TIME effect (F3,14=8.190; p=0.002) as well as a TIME×GROUP interaction (F3,14=7.493; p=0.003) in repeated measures ANOVA analysis. Statistically significant APB MEP facilitation was observed at t0 (p=0.014), t15 (p<0.001) and t30 (p=0.012) after PAS in the CONTROL group. No changes in APB MEP were observed at any specific TIME in the CADASIL group (t0 p>0.05, t15 p>0.05 and t30 p>0.05). Post hoc paired t test analysis with Bonferroni correction showed statistically significant differences between CONTROL and CADASIL groups at t0 (p=0.017), t15 (p<0.001) and t30 (p=0.022) after PAS protocol (figure 4A). There was no significant TIME effect (F3,14=1.312; p=0.310) in the amplitude of FDI MEP before the PAS protocol compared with values after PAS as well as no TIME×GROUP interaction (F3,14=0.048; p=0.985).
In the analysis of APB RMT variation, we found a significant TIME effect (F3,14=25.612; p<0.001) and a TIME×GROUP interaction (F3,14=12.233; p<0.001) in repeated measures ANOVA analysis. A statistically significant decrease in APB RMT at t0 (p<0.001), t15 (p<0.001) and t30 (p<0.001) after PAS was observed in the CONTROL group. No changes in APB RMT were observed at any specific TIME in the CADASIL group (t0 p>0.05, t15 p=0.258 and t30 p>0.05). Post hoc paired t test analysis with Bonferroni correction showed statistically significant differences between CONTROL and CADASIL groups at t0 (p=0.001), t15 (p<0.001) and t30 (p<0.001) after PAS protocol (figure 4B).
CADASIL and CONTROL groups showed statistical differences in several neuropsychological areas studied in Student's t test for independent measures analysis. CADASIL patients had worse results in neuropsychological tests assessing executive function and processing speed, verbal memory, visual memory, constructional praxis and naming. Detailed scores and statistical results from the different neuropsychological tests performed are shown in table 2.
Neurophysiological to neuropsychological correlation
TMS values included in the correlation analysis were 10 ms ISI of ICF, 20 ms ISI of SAI and t15 of APB MEP amplitude in the PAS protocol. No correlations were observed between any of the pathological neuropsychological tests and 10 ms ICF or 20 ms SAI protocols. However, several neuropsychological tests were highly correlated with t15 APB MEP amplitude of the PAS protocol. Detailed results of correlation analyses are shown in table 3. Afterwards, all neuropsychological tests correlating with t15 of the PAS protocol were included in a multivariate regression analysis with t15 PAS protocol value as dependent variable. This analysis showed that only letter verbal fluency test remained significant with an r value of 0.941 and p<0.001 (figure 5). Detailed β and p values of this multivariate regression model are detailed in table 3S of the online supplementary material.
We found decreased SAI and ICF as well as a decreased long-term-potentiation-like (LTP-like) induced by a PAS protocol in a group of CADASIL patients with neuropsychological alterations. Novel findings are the absence of ICF, the lack of LTP-like plasticity after the PAS protocol and its correlation with neuropsychological alterations observed in CADASIL patients. To our knowledge this is the first evidence of these abnormalities in CADASIL.
Lack of ICF (at 10 and 12 ms) in our CADASIL partially contradicts a recent study where an enhanced ICF was observed in a sample of subcortical vascular disease (SVD) sample of patients.16 Those different results could be explained by the different clinical situations of two samples (SVD vs CADASIL) and by the experimental procedure used for assessment of ICF. Although CADASIL is proposed as an excellent model for hereditary SVD, there are some different clinical, neuroimaging and cognitive features between them like the appearance of transient ischaemic attacks or stroke,17 the presence of temporal lobe or external capsule leukoencephalopathy and subcortical infarcts17 and the performance of verbal fluency test,18 respectively, that are more affected in CADASIL patients. Regarding the experimental procedure, higher CS and TS TMS intensities used in SVD patients could probably produce a different motor intracortical interaction between CS and TS. Glutamate is the main excitatory neurotransmitter in the human brain by activation of AMPA and N-methyl-D-aspartate (NMDA) receptors and it is thought that ICF in the human motor cortex principally depends on the activity of glutamatergic excitatory circuits.19
An animal study suggested that excitatory postsynaptic potentials (EPSPs) in neurons of motor cortex consist of a fast component mediated by non-NMDA receptors and a slower component mediated by NMDA receptors.20 The latency to onset of the EPSP mediated by NMDA receptor is of the order of 10 ms, which would be consistent with the time course of ICF.21 This idea was supported by several pharmacological studies, demonstrating a decrease of ICF mediated by NMDA antagonists.22–24 Our results in a group of CADASIL patients showing decreased ICF at 10 and 12 ms could suggest a local impairment of glutamatergic function.
Our study also found a reduced SAI in the group of CADASIL patients. The SAI protocol seems to be a good parameter for in vivo determination of central cholinergic function. SAI is reduced by the anticholinergic drug scopolamine,25 and the correlation of SAI and cholinergic system has been confirmed in vivo by an abnormal reduction of SAI in patients with AD and dementia with Lewy bodies.10 ,26 Involvement of a cholinergic impairment in CADASIL patients has been suggested in a study where donazepil administration produced a significant improvement of time to perform the TMT-B test (also impaired in our CADASIL sample) compared with placebo administration.27 Our study showed an abnormal reduction of SAI, thereby reflecting this possible central cholinergic impairment in CADASIL patients.
Finally, our group of CADASIL patients showed a lack of LTP-like plasticity induced by a PAS protocol. This result does not agree with a recent CADASIL study where an enhanced LTP-like sensorimotor plasticity was observed in CADASIL patients after a PAS protocol.28 Several differences are present between our study and this previous one. First of all, CADASIL patients of List's study had a preserved cognitive profile with only slight differences in neuropsychological tests (with respect to their healthy subjects) that could explain the preserved LTP-like PAS plasticity. So, in young CADASIL patients suffering from white matter lesions, the preserved cognitive function could be associated with a LTP-like plasticity increase. With increasing age or longer duration disease, this compensatory mechanism may fail, as resulting to the reduced LTP-like plasticity in our CADASIL group with a substantial cognitive deficit. These cognitive deficits include executive and speed function, and memory functions, thus supported by a recent PAS and cognition study performed in older patients with small vessel disease.29 Secondarily, we have used an enlarged PAS protocol of 240 paired stimuli instead of a short one used by List et al28 and also a different frequency of the paired stimuli. They found a low rate of LTP-like plasticity in their healthy subjects, suggesting their mean age as responsible for this result. These stimulation protocol differences and the older mean age of CADASIL patients in our study could be possible reasons of these different findings. The lack of LTP-like plasticity in our group of CADASIL patients is sustained in the lack of MEP amplitude facilitation and the absence of APB RMT decrease as observed in healthy subjects. The normal decrease of APB RMT reflects an enhanced excitability of the motor cortex after PAS protocol,30 showing the perfect operation of facilitatory mechanisms in healthy humans. PAS-induced LTP-like plasticity seems to be related to GABAB and NMDA receptors as previously reported.31 ,32 Abnormalities found in PAS-induced LTP-like plasticity in our study could suggest, once again, a deficit in the glutamate neurotransmitter system shared with ICF abnormalities also observed in our CADASIL group. The GABAB receptor, as a possible component of PAS-induced plasticity, was ruled out in our study because of the normal results in CSP and LICI protocols in CADASIL patients.
Different imaging studies have shown that recurrent vascular infarcts in different strategic areas including the external capsule, which is typically involved in CADASIL,5 ,33 lead to an interruption of cortico-subcortical cholinergic circuits. Glutamate and GABA circuits receive cholinergic inputs that are able to modulate the efficacy of synaptic transmission.34 The cholinergic impairment observed in our sample of CADASIL patients could therefore influence glutamate systems involved in ICF and LTP-like plasticity induced by the PAS protocol. Disruption of cholinergic circuits could also be responsible for the motor cortex hyperexcitability previously observed.14
Our CADASIL group showed a neuropsychological profile similar to previous studies35 ,36 with subcortical alterations but also with some cortical features. Subcortical impairment includes executive function (verbal fluency tests),18 naming and recall memory domains. No alterations were observed in Symbol Digit Modalities Tests and encoding memory tests; both of them more typical of memory alterations in AD, although processing speed tests could be typically affected in SVD and CADASIL patients.18 In addition, a cortical neuropsychological alteration can be observed in verbal memory, visual memory and constructional praxis areas. Cortical neuropsychological alterations were recently correlated to central cholinergic dysfunction in AD and dementia with Lewy bodies,37 but we could not find any correlation between the CADASIL's neuropsychological profile and SAI protocol performed, as previously reported.14 This lack of correlation could explain why a mixed subcortical and cortical impairment, with a higher subcortical expression, is present in our CADASIL group. Strongest correlation was observed between the lack of LTP-like plasticity in the PAS protocol and verbal fluency (letter) impairment in our CADASIL sample of patients. This correlation did not seem to be affected by sex, age or years of formal education. Ischaemic subcortical lesions observed in CADASIL's MRI studies produce a first subcortical alteration which leads to a cortico-subcortical disconnection and a final cortical and subcortical neuropsychological profile can therefore be observed in our CADASIL group.38 The lack of plasticity is highly correlated with neuropsychological tests that need memory and learning mechanisms to be correctly performed. LTP-like plasticity is known to be related to memory and learning processes and PAS LTP-like plasticity is altered in different neurodegenerative diseases such as Parkinson's disease39 and AD.40 Results of our study agree with these previous ones with the novel finding that this alteration is highly correlated with neuropsychological tests that mainly reflect a subcortical impairment like verbal fluency (letter) which also requires memory and learning mechanisms to be performed in healthy subjects.
We would like to thank Juan Manuel Praena Fernández and Ismael Huertas Fernández for their help with the statistical analysis and all the patients for their kind participation in this study.
Contributors 1A: Substantial contributions to conception and design. 1B: acquisition of data. 1C: analysis and interpretation of data; 2A: Drafting the article. 2B: revising it critically for important intellectual content; 3A: Final approval of the version to be published. FJP: 1A, 1B, 1C, 2A, 2B, 3A. AS: 1B, 2B, 3A. EFM: 1B, 2B, 3A. FC: 1A, 2B, 3A. EG-N: 1B, 2B, 3A. PM: 1A, 1B, 1C, 2A, 2B, 3A.
Funding This work was supported by grants from the Ministerio de Economía y Competitividad de España (SAF2007-60700), the Instituto de Salud Carlos III (PI10/01674), the Consejería de Economía, Innovación, Ciencia y Empleo de la Junta de Andalucía (CVI-02526, CTS-7685), the Consejería de Salud y Bienestar Social de la Junta de Andalucía (PI-0377/2007, PI-0741/2010, PI-0437-2012), the Sociedad Andaluza de Neurología, the Jacques and Gloria Gossweiler Foundation and the Fundación Alicia Koplowitz.
Competing interests None.
Ethics approval Virgen del Rocio University Hospital Local Ethics Committee.
Provenance and peer review Not commissioned; externally peer reviewed