Sirs: Paroxysmal dyskinesias are a heterogeneous group of movement disorders characterized by recurrent brief episodes of abnormal involuntary movements [1–3] Among them, paroxysmal dystonic choreoathetosis (PDC) is associated with dystonic or choreic attacks unrelated to movement. It may occur spontaneously at rest without any triggering factors but attacks are frequently precipitated by caffeine or alcohol consumption, stress and fatigue. In contrast to paroxysmal kinesigenic choreathetosis (PKC), PDC attacks are more dystonic, usually last longer, and are less frequent with long attack-free intervals [3, 4]. Although most cases are familial, sporadic and secondary forms of PDC have been reported [4, 5]. We report a patient with a sporadic severe PDC which has been dramatically improved by acetazolamide over a period of three years. This 17-year-old boy who had a normal birth and no familial history developed his first symptoms at the age of 12. He presented recurrent episodes of left arm mild dystonia, along with attacks of paroxysmal non-kinesigenic dyskinesia involving the whole body. Short attacks (20 to 120 s) occurred 3 times a week to several times per day. They manifested as dystonia frequently beginning in the lower limbs, either on the left or right side of the body, spreading to the trunk and upper limbs and finally leading to falls. No aura or changes in consciousness were noted and no precipitating factors were reported. Clinical examination was normal. Occasionally, a mild left hand dystonic posture was observed when walking. Blood chemistry was normal. Wilson’s disease was excluded. Antibasal ganglia antibodies were negative on two examinations. Brain MRI, interictal electroencephalograms and iodobenzamide SPECT were normal. A videoencephalogram was performed for three consecutive days during which the patient exhibited several attacks without any epileptic activity on the EEG during attacks. Psychiatric evaluation was unremarkable. Owing to the analogy with paroxysmal ataxia (EA2) [6], screening was performed for CACNA1A gene. Screening by direct DNA sequence analysis of the 47 exons of the CACNA1A gene did not reveal any mutation. Analysis of triplets (involved in SCA6) was also normal. We also sought a mutation in the MR-1 gene encoding for the Myofrillogenesis Receptor 1 [7, 8]. Sequencing of the 10 exons of the transcript specifically expressed in the brain did not detect any mutation in the MR-1 gene. Tiapride chlorydrate, valproate sodium, carbamazepine and clonazepam at usual dosage were ineffective. Owing to similarities between attacks of PDC and periodic ataxia, acetazolamide (10 mg/ kg/day) was used off-label and the attacks resolved within two weeks. During the subsequent three years of follow up, the patient reported only two episodes of attacks during holidays. They clearly occurred when the patient neglected to take his treatment in the morning and they disappeared with acetazolamide, showing the strict correlation between relapse of attacks and drug intake. Treatment with acetazolamide was clinically well tolerated and regular examinationof blood samples was unremarkable. The clinical presentation of our patient points to a diagnosis of PDC because the paroxysmal dystonic attacks were not triggered by any particular movement, unlike in paroxysmal kinesigenic dyskinesia (PKD). Attacks were of short duration but a wide variability of symptoms is frequently noted in PDC, particularly concerning attack duration [2–4]. For example, one patient (P7) in the series of Bressman et al. [4] had attacks lasting only 10 sec to 5min, close to those observed in our patient. The dystonic nature of the attacks, the normal consciousness, and above all the absence of EEG V. Michel, MD AE D. Guehl, MD, PhD B. Bioulac, MD, PhD P. Burbaud, MD, PhD (&) Service de Neurophysiologie clinique Hopital Pellegrin Avenue Amelie-Rabat Leon 33076 Bordeaux, France Tel.: + 33/5 57 57 15 86 Fax: + 33/5 56 89 20 61 E-Mail: Pierre.Burbaud@umr5543. u-bordeaux2.fr