Whole-Body Tremulousness Isolated Generalized Polymyoclonus

ORIGINAL CONTRIBUTION Whole-Body Tremulousness Isolated Generalized Polymyoclonus

Andrew McKeon, MB, MRCPI; Sean J. Pittock, MD; Graham A. Glass, MD; Keith A. Josephs, MST, MD; James H. Bower, MD; Vanda A. Lennon, MD, PhD; J. Eric Ahlskog, PhD, MD

Background: Acquired generalized repetitive myoclo- duration was 1.8 years, and mean age at onset was 55 years. nus may be mistaken for tremor. Distinguishing myoc- Referral diagnoses or patient complaints were tremor, lonus has etiologic and therapeutic implications. tremulousness, or shaking in all but 5 patients. All the patients had repetitive myoclonus of all limbs, impairing gait Objective: To describe isolated generalized polymyo- in 14 patients. Surface electromyography confirmed non- clonus and the outcomes of etiologic evaluations at the periodic muscle burst durations of less than 50 millisec- time of diagnosis. onds, typical of myoclonus. Clinical and serologic screen- ing for cancer and autoimmunity revealed metastatic breast Design: Computer search of the Mayo Movement Neu- cancer in 2 patients (1 positive for ganglionic acetylcho- rophysiology Laboratory database and medical records line receptor antibody) and antibody profiles implicating linkage system. neurologic autoimmunity in 3 patients (CRMP-5 IgG or neuronal voltage-gated potassium channel antibodies). Setting: Department of Neurology, Mayo Clinic. Medications known to occasionally trigger myoclonus (opi- oids, selective serotonin reuptake inhibitors, and a serotonin- Patients: Nineteen adults with generalized repetitive myoclonus confirmed using surface electromyography (burst norepinephrine reuptake inhibitor) were being taken by 7 duration Ͻ50 milliseconds), and other neurologic fea- patients. Myoclonus resolved after discontinuation of se- tures minimal or absent. lective serotonin reuptake inhibitor therapy in 1 patient; drug discontinuation was declined and follow-up was in- Interventions: Treatment of myoclonus and underly- adequate in the other 6. ing causes. Conclusions: Isolated whole-body tremulousness should Main Outcome Measures: Clinical presentation and raise the suspicion of generalized polymyoclonus, con- underlying etiologies. firmed using routine surface electromyography. Recog- nition is important because the differential diagnosis in- Results: We identified 19 patients with isolated general- cludes autoimmunity and drug-induced myoclonus. ized polymyoclonus resembling whole-body tremor. On- set was most often subacute (12 patients), mean symptom Arch Neurol. 2007;64(9):1318-1322

N OUR ROUTINE MOVEMENT DISOR- However, when generalized polymyoclo- ders clinical practice, occasion- nus develops as an isolated syndrome, the ally patients have been referred absence of clinical context and the resem- with a diagnosis of an atypical gen- blance to tremor may be diagnostically chal- eralized tremor syndrome that is lenging. The distinction is important be- ultimatelyI recognized as repetitive low- cause the symptomatic treatments differ and amplitude myoclonus (polymyoclonus). adult-onset generalized myoclonus is typi- Author Affiliations: Generalized polymyoclonus may re- cally acquired and may have an autoim- Departments of Neurology semble tremor but is distinguished by the mune basis, including paraneoplastic, or (Drs McKeon, Pittock, Glass, absence of true periodicity, which can be may be drug induced.1 Herein we describe Josephs, Bower, Lennon, and confirmed using surface electromyogra- a series of adult patients with a syndrome Ahlskog), Immunology phy (EMG). Although this resemblance may of generalized repetitive myoclonus, most (Dr Lennon), and Laboratory Յ Medicine and Pathology occur in the context of a toxic-metabolic en- with recent onset (duration 2 years), con- (Drs Pittock and Lennon), cephalopathy or a neurodegenerative confirmed by means of surface EMG. Tremor, Mayo Clinic, Rochester, dition, the accompanying clinical features tremulousness, or gait disorder was the typi- Minnesota. of these disorders facilitate recognition. cal referral diagnosis.

(REPRINTED) ARCH NEUROL / VOL 64 (NO. 9), SEP 2007 WWW.ARCHNEUROL.COM 1318

Patient No./ Age at Symptom Onset of Diagnosis at Sex Symptom Onset, y Duration, mo Symptoms Referral Diagnosis Chief Complaint Initial Examinationa 1/F 73 18 Insidious Tremor Jerks Myoclonus 2/M 55 120 Insidious Involuntary movements Jerks Myoclonus 3/M 51 24 Insidious Gait disorder Jerks Myoclonus or tremor 4/M 66 4 Subacute Tremulousness Tremulousness Myoclonus or tremor 5/M 34 6 Subacute Involuntary movements Shaking Myoclonus or tremor 6/M 44 24 Subacute Tremor Tremulousness Tremor 7/F 69 6 Insidious Parkinsonism Balance problems Myoclonus 8/F 63 6 Subacute Dystonia, possibly nonorganic Loss of limb dexterity Myoclonus 9/M 33 18 Insidious Tremulousness Tremulousness Myoclonus 10/M 77 12 Subacute General functional decline Tremors Myoclonus 11/M 64 6 Subacute Gait disorder Tremors Tremor and myoclonus 12/F 23 12 Subacute Involuntary movements Shaking Myoclonus 13/F 30 6 Subacute Tremor Tremor Myoclonus 14/F 67 36 Insidious Tremulousness Tremulousness Myoclonus 15/M 77 36 Subacute Tremor Tremors Tremor 16/M 76 15 Acute Gait disorder Shaking Myoclonus 17/F 72 12 Subacute Tremor Tremors Myoclonus 18/F 48 Uncertain Subacute Tremulousness Tremulousness Myoclonus 19/F 22 24 Insidious Involuntary movements Balance problems Myoclonus

a By a staff neurologist.

METHODS tion of symptoms at the time of presentation was 21.3 months (range, 4-120 months). Myoclonus was not a re- We sought to capture data on adult patients with a syndrome ferral diagnosis in any of the 19 patients, all of whom had of acquired myoclonus affecting the whole body and develop- been seen by a neurologist before review at Mayo Clinic. ing in the absence of obvious systemic illness or other neuro- The chief complaints from the patients were tremors or logic disease. The specific clinical inclusion criteria were as fol- tremulousness (10 patients), shaking (3 patients), jerks lows: (1) a syndrome of generalized repetitive myoclonus (3 patients), and balance or dexterity problems (3 pa- confirmed by surface EMG; (2) other neurologic features mini- tients). A family history of myoclonus or dystonia or ame- mal or absent, except for impaired gait, balance, and dexterity lioration of symptoms with alcohol use was not re- directly attributable to the myoclonus; (3) absence of major or- corded in any case. gan failure; and (4) age at onset older than 21 years. The surface EMG criteria were 2-fold: (1) documentation of muscle All 19 patients were examined by a Mayo staff neu- burst durations less than 50 milliseconds and (2) absence of rologist. A movement disorders specialist examined all periodicity (tremor is, by definition, periodic). but 2 patients before the neurophysiologic assessment; We ascertained these cases by searching the Mayo Move- 13 had their condition evaluated by at least 1 of us. My- ment Neurophysiology Laboratory computerized database and oclonus (sudden, brief, shocklike, involuntary move- medical records linkage system (1998-2006). Patients with fo- ments) was identified on clinical examination in 13 pa- cal, segmental, epileptic, or propriospinal myoclonus were ex- tients, with 4 others being characterized as having cluded, as were those with coexisting movement disorders (dys- myoclonus and/or tremor; however, tremor was the di- tonia, tremor, or chorea). The neurophysiologic features were described and myoclonus was documented by movement dis- agnosis in 2 patients. In all of the patients, the observed orders neurophysiology staff in all cases. From this original co- tremulous movements affected all 4 limbs and were re- hort the medical records were then reviewed to exclude pa- petitive (generalized polymyoclonus), with upper limb tients not meeting the previously described criteria. Thus, predominance found in 2 patients. Cranial involvement patients with myoclonus occurring in the context of neurode- was documented in 3 patients, and it was sufficient enough generative disease, encephalopathy, epilepsy, or obvious toxic- to cause dysarthria in 2. Opsoclonus or palatal myoclonus/ metabolic disease were excluded. We retained patients with mi- palatal tremor was not present in any of the patients. In nor or incidental neurologic findings in the study. 14 of the 19 patients, gait and stance were impaired by the ongoing generalized polymyoclonus. This ranged from RESULTS subjective unsteadiness alone to severe gait instability; 1 patient experienced myoclonic knee buckling, ob- PATIENT DEMOGRAPHICS served by the physician. In 2 patients, the physician com- AND CLINICAL FEATURES mented that the myoclonic jerks in the lower limbs during standing were better appreciated by palpation than We identified 19 patients who met the previously men- inspection. Minor or incidental features noted on exami- tioned criteria of isolated generalized polymyoclonus; 10 nation were findings suggestive of small-fiber neuropa- were men (Table 1). The mean age at symptom onset thy (patient 3), mild cognitive impairment (patient 7), was 55 years (range, 22-77 years), and the mean dura- and reduced arm swing on ambulation (patient 15).

(REPRINTED) ARCH NEUROL / VOL 64 (NO. 9), SEP 2007 WWW.ARCHNEUROL.COM 1319

Cancer or Patient Autoantibodies No. Detected Treatment Follow-up 1 None Fluoxetine Myoclonus resolved hydrochloride B cessation 6 None Clonazepam Marked improvement 7 CRMP-5 IgG None NA 8 Metastatic Clonazepam No improvement with breast clonazepam noted, cancer, immunomodulatory ganglionic therapy 200 µV AChR recommended 330 ms autoantibody 9 None Clonazepam and Marked improvement gabapentin Figure. Multichannel surface electromyographic recordings. A, Myoclonus is shown in patient 1, with repetitive, nonrhythmic discharges of short duration 11 None Methylprednisolone No improvement (Ͻ50 milliseconds). B, By way of contrast, a recording from a patient with sodium succinate, essential tremor shows bursts of rhythmic (6-Hz) muscle activity. 1000 mg IV ϫ 5d 14 Metastatic Levetiracetam In remission from breast cancer cancer and myoclonus free 5 y ELECTROPHYSIOLOGIC ANALYSIS after surgery and radiotherapy for Surface EMG recordings revealed electrophysiologic evi- chest wall dence of generalized polymyoclonus in all of the pa- metastatic lesion tients characterized by repetitive muscle burst dura- 15 None Levetiracetam Marked improvement 16 CRMP-5 IgG Clonazepam NA tions of less than 50 milliseconds and without evidence 17 Voltage-gated Plasma exchange Good symptomatic of periodicity (unlike tremor) (Figure). Myoclonus was potassium response but observed at rest in 11 patients and with posture and ac- channel subsequent relapse tion in all of the patients. Routine electroencephalogra- autoantibody phy was performed in 12 patients, demonstrating epi- Abbreviations: AChR, acetylcholine receptor; IV, intravenously; NA, not leptiform discharges in patients 7 and 17, neither of whom available. had a history of seizures.

ETIOLOGY OF MYOCLONUS mative or advised to be undertaken. However, follow-up was 1 month or less in all but 7 of the 19 patients. The avail- A definite cause was established for the generalized poly- able follow-up data are given in Table 2. myoclonus in 3 patients. Metastatic breast cancer was No definite etiology was identified in the 13 remain- found in patients 8 and 14, discovered only after the can- ing patients. However, 1 or more medications known to cer search was triggered by recognition of myoclonus. cause myoclonus were being administered long-term to Patient 1 experienced resolution of generalized myoclo- 6 of these patients. The potentially offending drugs were nus after stopping use of the selective serotonin reup- narcotics (patients 4, 5, 6, 13, and 18), SSRIs (patients take inhibitor (SSRI) fluoxetine hydrochloride, which im- 3, 4, and 6), and a serotonin-norepinephrine reuptake plicated this as the cause. inhibitor (patient 18). Patients 4 and 6 (both with chronic Serologic evidence of neurologic autoimmunity was as- pain syndrome) and 18 (with bipolar affective disorder) sessed in 15 patients and was detected and etiologically im- chose to continue taking the potentially causative medi- plicated in 4 patients (Table 2): patient 8 with breast can- cations. The potential for a drug-induced disorder was cer had antibodies to ganglionic acetylcholine receptor not recognized in patients 3, 5, or 13 at the time of evalu- (AChR) (0.04 nmol/L; reference value, Յ 0.02 nmol/L), pa- ation, and, therefore, drug cessation was not consid- tients 7 and 16 had CRMP-5 autoantibodies (recombi- ered. Because these 6 patients continued taking these nant Western blot assay), and patient 17 had neuronal volt- medications, causation could not be proved. age-gated potassium channel antibodies (0.79 nmol/L; reference value, Յ0.02 nmol/L). Cancer was not found in TREATMENT the latter 3 patients, but because of the known association with small cell lung carcinoma (patient 17 was a smoker), Patient 14 had surgical treatment and radiotherapy to her continued cancer surveillance was advised. Of the 9 pa- chest wall metastatic lesion. Her myoclonus gradually re- tients whose cerebrospinal fluid was tested, only patient solved during a few months with effective treatment of 13 had an abnormality (protein level, 12 g/dL). In all but her breast cancer, and 5 years later she remains in can- 1 patient (patient 13), the possibility of neurologic auto- cer remission without myoclonus. Patient 17 experi- immunity related to an occult cancer was considered, and enced initial symptomatic improvement after a course of a serologic and oncologic evaluation was either noninfor- plasma exchange but subsequently relapsed. An empiri-

(REPRINTED) ARCH NEUROL / VOL 64 (NO. 9), SEP 2007 WWW.ARCHNEUROL.COM 1320

©2007 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 09/30/2021 cal 5-day course of intravenous methylprednisolone so- central nervous system by definition (although it may be dium succinate did not improve symptoms in patient 11. triggered by peripheral stimuli). Exact localization in the Symptomatic treatment for myoclonus was undertaken central nervous system requires special studies and equip- in 6 patients, 5 of whom had adequate follow-up. The ment that are not routinely available in most clinical prac- myoclonus was markedly improved with clonazepam use tices (eg, jerk-locked, back-averaged electroencephalo- alone (patient 6), treatment with clonazepam combined graphic-EMG polygraphic recordings), unlike the surface with gabapentin (patient 9), or levetiracetam mono- EMG used to confirm myoclonus. However, the crucial therapy (patients 14 and 15). Patient 8 did not improve step is recognition of myoclonus; hence, we emphasize with clonazepam treatment. that the electrophysiologic techniques available in routine neurologic practices are sufficient for that purpose. COMMENT In contrast to these patients, myoclonus is often a com- ponent of a broader syndrome rather than an isolated phe- The phenotype of adult-onset, isolated, diffuse (whole- nomenon. Thus, generalized myoclonus may occasion- body) tremulousness should raise the suspicion of gen- ally occur in neurodegenerative disorders, such as eralized polymyoclonus. If confirmed (and especially if Alzheimer disease,2 Parkinson disease,3 inborn errors of subacute or acute), this should generate a differential di- metabolism, primary epilepsy syndromes,4 and toxic- agnosis to include 3 potentially treatable conditions: oc- metabolic encephalopathies. The myoclonus seen in some cult cancer, idiopathic neurologic autoimmunity, and ad- of these disorders has been primarily limited to small- verse effect of medication. amplitude jerks of the distal limbs (predominantly fin- Although this syndrome seems uncommon (19 pa- gers), termed minipolymyoclonus.2,5 Similar to the pres- tients in 9 years), we included only patients in whom my- ent patients, these nonperiodic repetitive twitches of oclonus had been assessed and confirmed in the Mayo minipolymyoclonus may resemble tremor. However, our Movement Neurophysiology Laboratory. Given the dif- patients’ phenotypes differ in terms of the prominence ficulty recognizing this clinical entity, numerous other and generalized distribution of the myoclonus. cases may have been overlooked by the staff, similar to The known association of myoclonus with malig- the physicians who referred these patients (as exempli- nancy prompted a search for cancer, which led to the iden- fied by patients 6 and 15). tification of occult metastatic breast carcinoma in 2 pa- In this series, generalized polymyoclonus was consis- tients, 1 of whom (patient 8) had ganglionic AChR tently unrecognized and mistaken for tremor. Clinical antibodies, but without adequate follow-up to know the clues include the following. First, tremor syndromes rarely outcome of treatment. Ganglionic AChR antibodies have involve the entire body, and certainly not subacutely or previously been reported with solid tumors and dysau- acutely. An exception is tremor related to toxic- tonomia.6 In patient 14, the myoclonus resolved after ef- metabolic or drug etiologies; however, such patients typi- fective treatment of the cancer with surgery and radio- cally have systemic clues, are encephalopathic, or are tak- therapy. Three patients in whom cancer was not found ing drugs known to cause generalized tremor, such as had neuronal antibodies specific for CRMP-5 or voltage- lithium or valproic acid. Second, it is rare for tremor to gated potassium channels. CRMP-5 IgG is most com- degrade gait and stance, with the primary exception being monly associated with small cell lung carcinoma.7 Pa- orthostatic tremor. Third, the nonperiodicity of the move- tient 17 had neuron-specific voltage-gated potassium ments on close inspection would exclude tremor be- channel autoantibodies and initially responded well to cause, by definition, tremor is perfectly periodic (rhyth- plasma exchange. This antibody has previously been de- mic). Such nonperiodicity may not be obvious in the scribed in association with peripheral nerve hyperexcit- present context because the myoclonic jerks were of low ability, epilepsy, and limbic encephalitis and may be as- amplitude and rapidly repetitive. Thus, clinical suspi- sociated with thymoma, small cell lung carcinoma, or cion and focused observation of an outstretched arm with other solid tumors.8 fingers extended will often reveal the irregular twitches Certain commonly prescribed drugs may cause this that raise the suspicion of myoclonus. myoclonic phenotype in the absence of other signs of tox- The physicians who ultimately made the diagnosis in icity; these drugs include tricyclic antidepressants9, this series relied on neurophysiologic analyses to con- SSRIs and a serotonin-norepinephrine reuptake inhibi- firm myoclonus. The neurophysiologic technique we used tor,10 narcotics, and a combination of these drugs.11 is relatively straightforward using equipment available Whereas a drug-induced disorder was definitively proved in all EMG laboratories, placing pairs of surface elec- only in patient 1, 6 other patients were taking 1 or a com- trodes approximately 3 cm apart on the skin overlying bination of these medications. Three patients elected to the most active muscles. Thus, surface EMG recordings continue taking the potentially causative agents, whereas documented muscle burst durations of less than 50 mil- follow-up was inadequate in the other 3 patients. Thus, liseconds, which is consistent with myoclonus (Figure). in patients with this myoclonic syndrome, review of medi- The surface EMG recordings also displayed the irregu- cations should be one of the first diagnostic maneuvers. larly repeating pattern, unlike the perfect periodicity of Other drugs have been implicated as rare causes of my- tremor. Hence, neurophysiologic confirmation can be oclonus, including lithium,12 certain anticonvulsants (la- made by practicing neurologists in the absence of so- motrigine,13 carbamazepine,14 and topiramate15), ceph- phisticated equipment. alosporin antibiotics,16 amantadine,17 and verapamil.18 The exact neuroanatomical origin of myoclonus in Myoclonus is also important to recognize because these patients is unknown. Myoclonus originates in the symptomatic drug treatment is often effective and dif-

(REPRINTED) ARCH NEUROL / VOL 64 (NO. 9), SEP 2007 WWW.ARCHNEUROL.COM 1321

©2007 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 09/30/2021 fers from that used for tremor. Thus, 4 of the 6 patients 2. Hallett M, Wilkins DE. Myoclonus in Alzheimer’s disease and minipolymyoclonus. improved markedly after treatment with clonazepam, le- Adv Neurol. 1986;43:399-405. 3. Caviness JN, Adler CH, Beach TG, Wetjen KL, Caselli RJ. Small-amplitude cor- vetiracetam, or gabapentin. Although gabapentin may be tical myoclonus in Parkinson’s disease: physiology and clinical observations. Mov used to treat myoclonus, it has paradoxically caused my- Disord. 2002;17(4):657-662. oclonus in the setting of renal failure.19 4. Hallett M. Myoclonus: relation to epilepsy. Epilepsia. 1985;26(suppl 1):S67- In summary, the phenotype of whole-body tremu- S77. lousness should suggest consideration of the syndrome 5. Wilkins DE, Hallett M, Erba G. Primary generalised epileptic myoclonus: a fre- quent manifestation of minipolymyoclonus of central origin. J Neurol Neuro- of generalized polymyoclonus, which can be easily con- surg Psychiatry. 1985;48(6):506-516. firmed using surface EMG available in routine neuro- 6. Vernino S, Lennon VA. Neuronal ganglionic acetylcholine receptor autoimmunity. logic practice. Recognition is important because it is symp- Ann N Y Acad Sci. 2003;998:211-214. tomatically treatable and may be a clue to a treatable 7. Yu Z, Kryzer TJ, Griesmann GE, Kim K, Benarroch EE, Lennon VA. CRMP-5 neu- autoimmune etiology (paraneoplastic or idiopathic) or ronal autoantibody: marker of lung cancer and thymoma-related autoimmunity. a drug effect. Ann Neurol. 2001;49(2):146-154. 8. Thieben MJ, Lennon VA, Boeve BF, Aksamit AJ, Keegan M, Vernino S. Poten- tially reversible autoimmune limbic encephalitis with neuronal potassium chan- Accepted for Publication: January 27, 2007. nel antibody. Neurology. 2004;62(7):1177-1182. Correspondence: Andrew McKeon, MB, MRCPI, De- 9. Garvey MJ, Tollefson GD. Occurrence of myoclonus in patients treated with cy- partment of Neurology, Gonda 8 South, Mayo Clinic, 200 clic antidepressants. Arch Gen Psychiatry. 1987;44(3):269-272. First St SW, Rochester, MN 55902 (mckeon.andrew 10. Ghika-Schmid F, Ghika J, Vuadens P, Bogousslavsky J, Regli F, Despland PA. Acute reversible myoclonic encephalopathy associated with fluoxetine therapy. @mayo.edu). Mov Disord. 1997;12(4):622-623. Author Contributions: All the authors had full access to 11. Potter JM, Reid DB, Shaw RJ, Hackett P, Hickman PE. Myoclonus associated all the data in the study and take responsibility for the with treatment with high doses of morphine: the role of supplemental drugs. BMJ. integrity of the data and the accuracy of the data analy- 1989;299(6692):150-153. sis. Study concept and design: McKeon, Pittock, Glass, Len- 12. Caviness JN, Evidente VG. Cortical myoclonus during lithium exposure. Arch Neurol. non, and Ahlskog. Acquisition of data: McKeon, Pittock, 2003;60(3):401-404. 13. Crespel A, Genton P, Berramdane M, et al. Lamotrigine associated with exacer- Glass, Josephs, Bower, Lennon, and Ahlskog. Analysis and bation or de novo myoclonus in idiopathic generalized epilepsies. Neurology. 2005; interpretation of data: McKeon, Pittock, Glass, and Ahl- 65(5):762-764. skog. Drafting of the manuscript: McKeon, Lennon, and 14. Sullivan JB Jr, Rumack BH, Peterson RG. Acute carbamazepine toxicity result- Ahlskog. Critical revision of the manuscript for important ing from overdose. Neurology. 1981;31(5):621-624. intellectual content: Pittock, Glass, Josephs, Bower, Len- 15. Alonso-Navarro H, Jimenez-Jimenez FJ. Reversible tremor, myoclonus, and fas- ciculations associated with topiramate use for migraine. Clin Neuropharmacol. non, and Ahlskog. Study supervision: Pittock, Josephs, 2006;29(3):157-159. Bower, Lennon, and Ahlskog. 16. Chan S, Turner MR, Young L, Gregory R. Cephalosporin-induced myoclonus. Financial Disclosure: None reported. Neurology. 2006;66(6):E20. 17. Matsunaga K, Uozumi T, Qingrui L, Hashimoto T, Tsuji S. Amantadine-induced cortical myoclonus. Neurology. 2001;56(2):279-280. REFERENCES 18. Vadlamudi L, Wijdicks EF. Multifocal myoclonus due to verapamil overdose. Neurology. 2002;58(6):984. 1. Caviness JN, Brown P. Myoclonus: current concepts and recent advances. Lan- 19. Zhang C, Glenn DG, Bell WL, O’Donovan CA. Gabapentin-induced myoclonus in cet Neurol. 2004;3(10):598-607. end-stage renal disease. Epilepsia. 2005;46(1):156-158.

Announcement

Sign Up for Alerts—It’s Free! Archives of Neurology of- fers the ability to automatically receive the table of con- tents of Archives when it is published online. This also allows you to link to individual articles and view the ab- stract. It makes keeping up-to-date even easier! Go to http://pubs.ama-assn.org/misc/alerts.dtl to sign up for this free service.

(REPRINTED) ARCH NEUROL / VOL 64 (NO. 9), SEP 2007 WWW.ARCHNEUROL.COM 1322