ORALTETRAHYDROAMINOACRIDINE IN LONG-TERM TREATMENT OF SENILE DEMENTIA, ALZHEIMER TYPE

'vVn.LlAM KOOPMANS Sl;MMERS, i'I'f.D., LAWRENCE VICTOR I'vIAJOVSKI, PH.D., GARY ~1ART!;-J :'v1ARSH, PH.D., KE:'JNETH TACHIKI, PH.D., AND ARTH R KLING, ]'>/1.0.

Abstract We treated 17 patients who had moderate to severe Alzheimer's disease with oral \etrahydroamino­acridine (THA), a centrally active anticholinesterase, in a three-phase study. In the nonblinded first phase of the study, significant improvement occurred in subjects who received the drug. as compared with their pretreatment status, on the global assessment (P ~ 0.001), the Orienta­tion Test (P = 0.001), and the more sophisticated Names Learning Test (P = 0.001). During the second phase, the SUbjects served as their own controls in a double-blind, placebo-controlled, cross-over study in which the order of administration of the drug and placebo was randomly assigned. Among the 14 subjects completing Phase II, THA treatment produced significantly better results than

DErVIE0ITIA of' tht: .... lzheiIDer type is a slowly pro­o-ressivc neuropsychialric condition Ihat is prin­

cipally ll1anilcslec! by memory deficits and that results in death from debilitating disease in 6 to 12 yea)". I Although its cause is unknown. ome specific neuro­chemical and analomicallesions have been observr:d. In 1976 Davies and ~'laloney demonstrated a specific deficit ill choline acetyltransferase in autopsy material (i'om patients with Alzheimer',; disease 2 This f'nzyme is responsible for the formation of' acetylcholine from choline and acetyl-coenzy'me A. The findings of Da­vies and Maloney have been confirmed by several oth­cr investigators J -6 Coyle et al. demonstrated a selec­live loss of cholinergic neurons in the septal-diagonal band of' the Broca-nu leus basalis system in patients \\lith Alzheimer's disease,7

Once these specific transmitter lesions were identi­fied, the development of a treatment strategy for Alz­heimer's disease became possible. In 1gSI the results of a study of the short-term use of in trav.;nous 1,2,3,4­tetrahydro-9-amilloacridine (THA), a potenl central­ly acting anticholine. terase, for therapy of memory deficits in patients Wilh Alzheimer's disease \-IITI pub­lished B The 75 percent positive response to THA not

From the Dcpanment of Psychiatry. University of California, La Angeles: the. Department of Rehabilitation. Huminglon Memori,,1 Hospital. P"sadena. Calif.: the Depar1ment of Biostatistics. niversity of Pitt,burgh. Pinsbu'·gh. Pa.: and the SepulVeda Velerans Administratiun Hospital, SepUlveda. Calif. Address reprint rcquesls 10 Dr. Summers al 624 W. Duane Rd .. Arcadia, CA 91006.

placebo on the global assessment (P = 0.003), the Ori­entation Test (P ~ 0.004). the Alzheimer's Deficit Scale (P 0.003). and the Names Learning Test (P = 0,001)_c

Twelve subjects have entered Phase 111, which involves long-term administration of oral THA. The average duration of treatment in these subjects at present is 12.6 months; symptomatic improvements have occurred, and no seri­ous side effects attributable to THA have been observed.

These encouraging initial results suggest that THA may be at least temporarily useful in the long-term palliative treatment of patients with Alzheimer's disease. We stress that further observations will be required before a clear assessment of the role of this agent can be made. (N Engl J Med 1986; 315:1241-5.)

only supported the cholinergic hypothesis but was en­couraging with regard to possible palliative treatment of Alzheimer's disease.

111 this papet· we report our r\:perienee with long­term administration of oral THA to patients in the middle and late stages or suspected Alzhcimn's disease.

METHODS

Twenty-Ihree sul~jec{s wilh a clinical diagnosis of undifferentiat­ed demenlia werc considered for Ihe study.~' The criteria for inclu­sion \Ver~ a clinical historv consistent with demcllIia, an age of S5 or older, and rhc ahilil, to speak English. The criteria for exclusion were evidence t,f abnormal liver function and moderalely severe obslructi"e pulmonary disease. No exclusions were made because of concomitant use of other medications. All subjects were asked to take nine or more capsules of commercial lecithin () 100 mg per capsule) or its eqnivalenl per day. The subjects were inpatients al lhe Huntington :Vlemorial Hospilal or the Las Encinas Hospital in Pasadena, California, for 2 to 14 days of Ihe sludy. The remainder of the investigation was conducted in "n outpatielll sClling.

Thc study was divided into Ihree phases. Various questions werc asked in the different phases of the slUd\'. I n Phase I, Ihe questions were (I) Is Ihe diagnosis of Alzheimer's disease accurate' (2) Does THA improve mentation in a nonblinded e\'aJuation:' (3) What dose of THA provides optimal benefit without side effects' In Phase II, we asked, At the optimal dose of THA for each subject in Phase I, is the active agent, as compared with the placebo, producing a difference in mentation? Two quc tions were a 'ked in Phase III: Al the optimal Jose of THA, is the improvement seen in Phases I and [[ maintained? Is there long-term unsuspected toxicity due to THA'

In Ihe first phase, extcns;ve diagnostic evaluation ""as done, and oral THA was administered in a non blinded Irial. The diagnostic evaluation included Slandard procedures. ' ·

IO In addilion, all sub-

Reprinted from the New England Journal of Medicine 315:1241-1245 (November 13), J986

jects were sc)"(,ened b\' a I-rng' elexarnethasu"e-wpprc,sion lc,t with semm cortisol rnea,urements ;It 17 ;mel 24 hours, 24-holll" Hoher monitoring, a carotid duplex scan, ;lnd a cranial nuclear magnetic resonance scan. Th~ screenin~ psychometric tests wcre the Luria­Nebraska Neurop vcllOlogica! Test, the ;Vlinnesota \lultiphasic Personalit), Inventory, the ;'\ie\\' York University :'v!cmor" Test, Wechsler's Aelult Intellig.-nee Scale, the Names Learning Test, the Alzheimer's Deficit cale. tilt· Orientation Test, and the ;'vlini :\·len­tal Status Test. Adelitionalt~sts,,'ere ordered if clinieallv indicated.

During Phase T, a 2.'J-mg dose of oral THA \vas given in the first 24 hours. The dosage of" TH!\ was then increased in ineremeJ1ls of :25 to 50 mg- per day to a peak dose of ISO to 200 mg pn day, over 7 to 10 days. The peak dose was ideillitied bv the observation of clinical improvem.-nt without side effects or of treatment with 200 mg per day without dinieal improvement. Clini al improvement was measured hy a daily global assessment, the Names Learning Test, and the Orientation Test"·11

Phase II was conducted in an outpatient setting. The patients were randomly assigncdlO receive either THA at the optimal dose determined in Phase 1 ur a similar numbel' or placebo pills fur three weeks. The family. the patient, the family physician, and the dinical technicians wcre blinded to which medication was being used. The patients served as their own controls: in the next illI'.-e weeks of Phase II, th.-y received th.- other a!<~nt. Clinical assessment was done with the psychometric leSts Ii ted earlier. Toxicit~, was as­sessed in Phase II by weekly omp1ete blood counts and chemistry profiles and monthly dectroeardiograms.

Phase I I I was a long-term slLlely of oral THA treatment. Clinical improvement during- therapy with THA (as compared with pla­cebo) in Phase II was mandatory before a subject was accepted into Phase Ill. The initial dose of THA in Phase III was the optimal dose established in Phases I and 11. ~-[onlhly assessments for long­term toxicit\' and clinical effects were done as in Phase II.

THA was ohtained from Aldrich Chemical (Milwaukee) and then formulated in 2.~-mg ablets (Solo Research). The placebo tablets ,,'cre identical in appearance to the THA tablets.

The statistical anal\'sis of the data focused on the pairwise com­parison of median test scores as measured by ['Iur scales, the glohal assessment. thc Orientation Test, the Names Learning Test, ilnd the Alzheimer's Defil'it Scale. Three IreatmeJ1l conditions were compared: pre-THA versus THA, prc-THe\ \"(rsu, plae bo, and THA "crsus placebo. Pail"\\'isc comparisons were "Iso performed on two subgroups of the data defined by the pretreatment global a, ­sessrnent score; Group I included patients in Stages :1 and 4 of Alzheimer's disease. and Group 2 included those in Stages 5 and 6. [n all paired comparisons, the Wilcoxonmalched-pair, signed-rank, nonpal'am~tric statistical test was used [0 perform a two-tailed test of the null hypothesis that the median test scores in each sample were equal. I"

RESULTS

Of the 23 subjects considered for this stuely, 6 were considered not to have Alzheimer's disease on the ba­sis of laboratory tests. There were no significanr differ­ences in age, sex, or race between the subjects with and without Alzheimer's disease: The mean (±SE) age of the 17 subjects with Alzheimer's disease was 70.65± 1.92 years; that of tbe ubjects without Alz­heimer's diseas was 72.5± 1.86. Of the six subj,xts without Alzheimer's disease, one was found at autopsy to have had amyotrophic lateral sclerosis. One sub­ject had severe hypothyroidism (the level of thyroid­stimulating hormone was 470 J.Lg per milliliter). Two had extensive multiple small cerebral infarcts on a nuclear magnetic re onance scan. Two had normal­pressure hydrocephalus, as assessed by cisternogra­phy. Four of the six subjects without Alzheimer's dis­ease received a trial of oral THA at doses of 1I p to 200 mg per day, without any ,'pparent benefit.

~)eventeen subjects were considered to have Alz­heimer's disease_ five of the 17 had additional diag­noses that were thought to have contribul,~d to the dementia but not 10 have caused it. One had a low B I2

lev I hut a minimally positiv<" Shilling t Sl. Supple­mental intramuscular- Bt2 was given for two months, but there was no positive response. This patient did jmprov~ on the THA regimen. One- subject had Par­kinson's rlisease, a mild clini~al depression, and a posi­tive dexamethasone-suppression test. Another patient with Alzheimer's disease had three small areas of foet! ('r:rebral infar-ction, with no history of hypertension. One patient with 'ever dementia also had cerebral anoxia without focal infarction, sr:condary to pulmo­nary atelectasi on the second day 01 tbe study; however, this appeared to be independent of 'iHA

GLOBAL STAGES

6 5 4 3 2 a ---'----'-c;::::::===='::;;;--L.....--.L.I...J1

SD-01 0 .. 9r- A...·>---------'­

80-03 ~~.---------·...J9

o------.....'r·SD-05 t;... ­

0-------.e..,9SD-06 t;.4 ­

O,.------_e9SD-07 t;.. ­

80-08 [)----------.0

(J~----------------_.. 980-09 t;.4 ­

[)--------oo.,O

SO-13

50-11

g.. "9

0...-----.....,98D-14 A.....e-------' ­SO-IS

80-16 or---------.....9A .J_

oe-----~e..,930-20 t;. • ­

50-21

80-22

23 0----080- 1

50-24 0--0 'I___'--__...L.--__-'--__....... --__..L-.II I

6 5 4 23 o1

STAGE

Figure 1. Change in Stages of Alzheimer's Disease during Oral THA Treatment.

The squares denote the initial pretreatment slage of Alzheimer's disease, the circles the eHecl of THA treatment on global staging, and the triangles Ihe placebo effect The palient numbers are shown on the ordinate. Tile assessments were based on 0 10 6

global stages of Alzheimer's disease.s

100 each test were compared in the lo­** 90

80

70

60

50

40

30

20

10

tal sample, the subjects with mod­***

I erately severe Alzheimer's disease ~Stage 3-4) at the beginning of the

I [ . ·;tudY, and the subjects with severe \Jzheimer's disease (Stage 5-6) at base line. Comparisons were made between base line ,pretreatment) and oral THA treatment, oral THAI I ~ 1 trealment (both blinded and non­} 1 t blinded) and placebo treatment (the blinded data collection of Phase II), and base line and placebo treatment. [i I The results of all four tests showed BASELINE THA PLACEBO BASELINE THA PLACEBO BASELINE THA PLACEBO

significant improvement with the ALL SUBJECTS STAOE 3-4 SUBJECTS STAGE 5-6 SUBJECTS

THA treatment, as compared with Figure 2. Results of Assessments with the Alzheimer's Deficit Scale. base line, in the entire group. In the

The median values and the 25th to 75th percentile limits are shown for three sub­ group of patients with severe Alz­groups: all subjects, subjects in Stage 3 to 4. and subjects in Stage 5 to 6. For each heimer's disease, the more ,'ophisti­subgroup, base-line values, values during oral THA treatment, and values during cated i\ames Learning Test did not

placebo treatment are shown (: denotes P<O.05...* P<O.01, and *** P<O.001).

administration. One 94-year-old .subject had insulin­dependent diabetes mellitus. The remaining 12 sub­jects had uncomplicated Alzheimer's disease.

All 17 subjects who were considered to have Alz­heimer's disease completed Phase 101' the study. Fif­teen subjects entered Phase II and 12 subjects entered Phase I I I. Two subjects were exd ucled from the study after Phase I because of noncompliance. One of these suusequentiy died, and the diagnosis of Alzheimer's disease was confirmed at autopsy.

Figure I shows the change in global assessment be­tween base line and the period of THA treatment in 17 patients with Alzheirnn's disease. The median glob­al stage before THA was 4.0, which indicated severe deficits due to Alzheimer's disease. During the admin­istration 01" oral THA, the median global stage was 2.0. During the administration of placebo, the average global stage was 4.0.

The Alzheimer's Deficit Scale is a psvchometric in­strument that assigns numerical values from a (worst) to 100 (asymptomatic) to reAect the severity of Alz­heimer's disease. Figure 2 shows the median scores on the Alzheimer's Deficit Scale be­fore treatment, eluring oral THA treatment, and during placebo ad­ministration. The results on the Alz­ :::: [ *1** heimer's Deficit Scale were consis­

8.0 ~ tent with the global assessment in that they showed significant im­provement during oral THA ther­ ::f I Iapy, but not during placebo ad­ 2.0~ministration or before treatment. Figure 3 shows the median Orien­ BASELINE THA PLACEBO

show a significant change. Twelve subjects have encered

Phase HI, in which the effects of long-term adminis­tration of oral THA are being evaluated. As of this writing, the first subject has been taking oral THA Iljr 26 months; the average duration of oral THA adminis­tration is 12.6 months per patient (range, 3 to :/(i!. The degree of improvement has often been dramatic. One Sll bject was able to resume most of her homelllaking tasks, one was able to resume employment on a part­time basis, and one retired subject \vas able to resume playing golf daily. In other cases, there were improve­ments in activities of daily living, such as self-reeding at the family table, where (Otal care had previously been reg uired.

The toxic eDects of the study drug were also consid­ered. THA is a monoamine acridine, which may be associated with hepatic, hematologic, and neurologic toxicity. Transient side effects in Phase I included nausea (six subjects), belching (two subjects), emesis (two subjects), excessive micturition (one subject), ab­dominal discomfort (two subjects), diarrhea (two sub­jects), and diaphoresis (three subjects). Thesc side ef~

fects were seen in only eight subjects, all of whom were

I ** ..

I I I I 1­

BASELINE THA PLACEBO BASEL~E THA PLACEBO

ta tion Te t scores for the various ALL SUBJECTS ST AGE 3-4 SU8JECTS STAGE 5-6 SUBJECTStreatment conditions.

Table I gives the summary of Figure 3. Orientation Test Results. statistical comparisons for all four Median values and the 25th to 75th percentile limits are shown for three subgroups assessments used. The result of (~ denotes P<O.05, ** P<O,01, and ., P<O.001).

Table 1. Summary of Statistical Comparisons.' mental anticholinergic ~yndromes.IFl

ASSES$ME f AND PRE11l:E4TMEl'-T THA THERAPY

SUBGROUP VERSlI\ T'HA THERAPY VERSUS Pl.ACt:BO

Global assessment All slages P = 0.001 (17) P = 0.003 (13) Slage 3-4 P = 0.009 (10) P = 0.022 (8)

Stage - 6 P ~ 0.022 (7) ,S (5)

OrientAtion Test All slage, P = 0001 (17) P = 0004 (14) Slage 3-4 P = 0.006 (10) P = 0024 (9)

Stage 5-6 P = 0.022 (7) NS (5)

Names Learning Test All "ages P = 0.001 ( 17) I' = 0.001 (14) Stage 3-4 P = 0.006 (10) P = 0.033 (9)

Siage 5-6 NS (7) NS (5)

Alzheimer Deficit Scale All stages P = 0.001 (17) P = 0.003 (12) Siage 3-4 P = 0.006 (10) P = 0.022 (7) Slage 5-6 P = 0.022 (7) NS (5)

Synaptic cholincrgic enhallcemen t PRETREJITME r

hy use of muscarinic agonists has also resulted in improvements in

VERSUS PLACI;BO

learning and memory. 19,20 Recently, NS (8) Harbaugh et al. reponed the re~ults NS (5)

NS (13)

of a treatment in which bcthanc­ehol, a quaternary ammonium com­

NS (14) NS (9i pound that is unable to penetrate NS (5) the central nervous system \'liht'll

given systemically, was injected di­NS (14) rectly into the cerebrospinal Auid by NS (9) NS (5\ a neurosurgical tt'chnique.:lO Betha­

nechol, like other muscarinic ago­NS (12) nists, has a low therapeutic index,NS (7)

NS (5) and as an unsubstituted carbamyl ester, it is totally resi~tant to hy­

*Based on the Wilcoxon ma!ched·pair-.;. signed-rank tesl. NS denOrt5 n\.)( :'"lgmtic3nf (a [wo-railed P value >0.05), The drolysis by ei ther acetylcholine ornumbers of pairs compared are given ill par<:nthesC5,

receiving 150 to 200 mg of THA per day. In all cases, reduction of the dost, administration of oral glycopyr­rolate (Robinul), or both, re ulted in disappearance of the side eITects. Four adverse events occurred durill,-" Phase III of the protocol, but none appeared to be caused by THA.

DISCUSSION

This was a large study of long-term palliative phar­macologic treatment of subjects with presumptive Alzheimer's di~ease. Although only 12 subjects have been enrolled in the long-term phase of this study, the results are encouraging, All four of the a~sessments

that were employed showed improvernenl on a regi­men of oral THA, which is unusual in an illness that is considered to be progressive, with only rare plateaus at best. I The placebo data showed essentially a return to pretreatment status.

The action of THA is thought to be due principally to potent central nervous system anticholinestera~e

activity.13 However, because of the structural simi­larity of THA to 4-aminopyridillf" it is probable that THA also selectively blocks potassium channels in the central nervous system, which results in an increased release of acetylcholine and a prolongation of the ac­tion potential of the presynaptic cholinergic neurons. H

Other investigators have reported variable success with cholinergic enhancement in the central nervous system in subjects with Alzheimer's disease. The ad­ministration or more than 8 g of choline chloride or its equivalent for more than three weeks has produced moderate improvement in ubjects with carly Alz­heimer's disease. 15,J6 Synaptic enhancement by intra­venous or oral administration of the anticholines­terase physostigmine has shown some promise in the treatment of Alzheimer's disease. 17 However, physo­stigmine has a lower therapeutic incJcx and a shorter duration of action than THA. 1:

s,;;.; Physostigmine is also Ie s eITective than THA in reversing exp ri­

nonspecific cholinesterase. All (our patients in this single-blind study

had subjective improvement. The safety of such a technique remain to be demonstrated.

In the pilot study of intravenou' THA in 1981,9 of 12 subjects (75 percent) had improvement as as~essed

by global staging, a subjective measure8 Objcctive psychometric tests showed improvement in 7 of 9 and dramatic impro ement in 4 of 9. Tn the present study, 16 of the 17 subjects had subjective global improve­ment, and 10 of the 17 had dramatic objective im­provement on psychometric testing.

The apparent improvements may have been due to better diagnostic techniques, specifically routine nu­cleLlr magnetic resonance scanning. Tbrec of the six SlI bj ects "vi thou t Alzheimer's disease in th is stud y would have been included as "probable Alzheimer's disease subjects" in the 1981 study. All three received THA in Phase I of the protocol, and none had im­provement. A second possi ble basis for the better re­sponse in this study was the availability of a serum THA assay that was sensitive to 2 ng per milliliter. The earlier assays were sensitive only to 50 ng per milliliter, and because the therapeutic range of THA appears to be between 5 and 70 ng per milliliter, such assays were not clinically useful. Indeed, in Pha~e III, three subjects v\lith moderate to minimal improvement vvere found to have;' serum THA levels of less than 7 ng per milliliter. When the dose of THA was increased, the clinical response improved dramatically. Before serum THA as .. ays were available, a limit or 150 mg of oral THA per day was Sf:t arbitrarily. However, these thret subjects required 200 mg per day.

Prudence injudging these results is advised. THA is no more a cure for Alzheimer's disea~e thall levodopa is a cure for Parkinson's disease. Just as levodopa ceases to have efTects in patients in the fmal stages of Parkinson's disease, we anticipate that oral THA will cease to have effect· as Alzheimer'~ di ease pro­lrrc,ses 21 Indeed, two of our subjects have had some dClcrioration despite maintenance therapy with oral

TH A, olle aCter 17 months and lhe other after 7 mOllths of therapy. However, THA is .1 potential pallialive t r<:a lmen [.

REFERENCES

I. Cumming' JI.. Benson DF. Dementia. a clinical appruoch. Boston: Buller­worth. 1983:36·7.

2. Dov;es P. Malone} AJF. Selective 10" of ecntml cholinergic neuroos in Alzheimer's di-ea·e. Lancet 1976; 1:1403.

3. Bowen DM. Smith CB. White P. I)awson AN. Neurotransmitter-relaled enzyme and indices of hypox ia in senile dementia and other abiolrophies. Brain 1976: 99:459-96.

4. Perry EK. Perry RH. Illessed G. Tomlinson BE. ecropsy evidence of central cholinergic deftcils in senile dementia. Lancel 1977; 1:189. Whi,e P. Kiley CR. Goodhardt MJ. et al. Neocortical cholinergic neurons in elderly people. Lancet 1977: 1:668-71.

6. 01' Den Velde W. Slam FC. Some cerebral pmtein' and enzyme YSlems in Alzheimer's presenile and senilG dementia. J Am Geria" Soc 1976; 14: 11-6.

7. Coyle JT. Price DT. Dc.Long MR. Alzheimer's disease: a disorder of coni­cal cholinergic innervation. Science 1983: 219: 1184-90.

8. Summers WK. Viesselman JO. Mar~h GM. Candelora K. Usc of THA in trealment of Alzhcimer·likl.: dcmemia: pilot study in twelve patients. Bioi Psyehiotry 1981; 16: 145-53.

9. Feighner JP. Robin~ E. uuze SB. Woodruff RA Jr. Winokur G. Munoz R. DiagnostiC criteria for use in psychiatric research. Arch G"n Psychiatry 1972: 26:57-63.

10. Wells CEo Diagno is 01 dementia. Psychosomatics 1979: 20:517-12. II. Irvlllg G. Robinson RA. McAdam W. The validily of some cognitive tests in

Ihe diagnosis of dementia. Br J Psychialry 1970: 117:J49-5b. 12. Hollander M. Wolfe DA. Nonparametric statistical tllelhods. New York:

John Wiley. 1973. 13. Summers WK. Kaulman KR. Altman F Jr, Fischer JM. TI-IA -a review of

the literature ::md ilS use in treillment of llve ovcrdos~ patients. Clin Toxicol 1980; 16:169-81.

14. Son; N. Kam P. 4-Aminopyrid;ne - a revicw Anae.th (ntensive Care 1982; 10:120-6.

15. Boyd WD. Graham-White J. Blackwood G. Glen 1. McQueen J. Clinical effeclS or choline in Alzheimer ,elllie ,Iementia. Lancct 1977: 2:71 J.

16. Etienne P. Gaulh,er S. John,on G. et al. Climeal effeelS of choline in Alzheimer's disease. Lancet 1978: 1:508-9.

17. Brinkman SD, Gcrshon S. Measurement of cholincrie drug effeets on memo ory in Alzheimer's disease. Neurobiol Agmg 1983: 4: 139-45

18. lIil T, Fink M. Anticholinergic drug-induced delirium: experimentalmodif,· calion. quantitative EEG and behavioraJ correlations. J Nerv Ment Dis 1966: 143:492-507.

19. Christie JE. Shering A, Ferguson 1. Glen AiM. Phy 0 tigmine and areco­line: effects ofintravenous infusions in Alzheimer's presenile dementia. Br J Psychiatry 1981; J38:46-50.

20 Harbaugh RE. Robens DW. Coo01b DW. Saunders RL, Reeder TM. Pre­liminary report: intracranial cholinergic drug infusion in patients with Alz­heimer's disease. Neurosurgery 1984: 15:514-8.

21. McDowell FH. Lee JE, Sweet RD. EKll'.pyramid,1! disease. In: Baker A. Baker LH. cds. ClinieaJ neurology. ew York: Harper & Row. 1978: 1-67.

tfJCopyright. 1986. by the Massachusells Medical Society Prinled in the U_S.A.

THE NEW ENGLAND JOURNAL OF MEDICINE is published weekly in the English language by the Massachusetts Medical Society (Waltham, MA, USA). Material printed in the Journal is covered by copyright. No part of this reprint may be reproduced or transmitted in any form withoul written permission. All rights reserved. Direct permission requests to the Permissions Dept. at the USA subscription oHice. Editorial oHices: 10 Shattuck Street, Boston, MA 02115, USA. SUBSCRIPTIONS: Subscription of­fices: 1440 Main Street, Waltham, MA 02254, USA; 13 Colina Road, Tottenham, London, England N15 3JA. U.K. Subscription prices: U.S.A.'. $60 per year (interns, residents $35 per year; stu­

dents $30 per year). Outside U.S.A. (US funds on a US bank): $75 per year (interns, residents $55 per year; students $45 per year): pounds sterling drawn on a U.K. bank: £52 per year (in­terns, residents £38 per year; students £31 per year). Rates are SUbject to change without notice. Sample copies available upon request. International copies are delivered by airlift, at no addi­tional charge, to all countries except Canada and the U.K. [)ay­ments and correspondence for all subscriptions delivered to Ja­pan should be sent directly to: Nankodo Co., Ltd., 42-6, Hongo 3­chome, Bunkyo-ku, Tokyo 113, Japan.