We read with great interest the paper by Salazar and colleagues, analyzing postural and action myoclonus in patients with parkinsonian-type multiple system atrophy (pMSA). 1 We agree with their observation that abnormal involuntary movements (AIMs) involving fingers and hands are frequently seen in pMSA, and they are myoclonus rather than tremor. 1 However, their conclusion that the AIMs can be described as minipolymyoclonus 2 needs more evidence, since they could not obtain any premyoclonic cerebral potentials by jerk-locked averaging. Minipolymyoclonus of central origin has been reported in certain types of epilepsy, cerebral palsy, and Alzheimer's disease, accompanied by characteristic premyoclonic bifrontal negative slow waves. 2,3 Hallett et al. 3 postulated a subcortical generator rather than a cortical generator. Demonstration of such cerebral potentials would be necessary to describe this particular type of myoclonus. 3 Recently, we had a chance to investigate the myoclonic tremor in patients with pMSA, and disclosed premyoclonic cerebral potentials, compatible with cortical reflex myoclonus, by jerk-locked averaging in five out of six patients (Okuma et al., unpublished observations). Therefore, the tremulous myoclonic AIM in pMSA appears to be better described as a form of cortical reflex myoclonus or cortical reflex myoclonic tremor rather than minipolymyoclonus. The reason why Salazar et al. could not show premyoclonic cerebral potentials is uncertain, but it may be possible that the potentials were so small that they could not be differentiated from background electroencephalographic activity (EEG) even after 50 times of back averaging.
Although we recommend the term "cortical reflex myoclonic tremor," the features of myoclonic tremor in pMSA are slightly different from typical cortical tremor, which was first reported by Ikeda et al. 4 Subsequently, cortical tremor has been described mainly in Japanese familial cases. [5][6][7][8] Familial cortical tremor is relatively more rhythmic, and therefore is often misdiagnosed as essential tremor. 4,7 On the other hand, the AIM in pMSA tends to be irregular, small amplitude, and sometimes confined to the fingers as reported by Salazar et al. 1 Moreover, familial cortical tremor usually accompanies giant somatosensory evoked potentials (SEP), whereas that of pMSA often does not. 1,9 We suggest that the myoclonic AIM in pMSA and cortical tremor are both in the spectrum of cortical reflex myoclonus, but the mechanism may differ.
Finally, we do not exclude the possibility of existence of minipolymyoclonus in pMSA. In fact, Kofler et al. 10 reported two different kinds of myoclonus, cortical and brainstem in origin, in the same patient with pathologically confirmed MSA. Therefore, further studies are necessary to determine the nautre of AIMs in MSA.