J j9 ,9,.,, j. yX - BMJ

Inclusion body myositis: an underdiagnosedcondition?

N D Hopkinson, C Hunt, R J Powell, J Lowe

AbstractInclusion body myositis is an increasinglyrecognised form of inflammatory myopathywith characteristic clinical and histopatho-logical features which has seldom beenreported in the United Kingdom. This paperpresents the clinicopathological features of a

series of patients diagnosed in Nottinghamfrom 1986 to 1990. During this period, 1319muscle biopsy samples were processed by thislaboratory and rimmed vacuoles were seen in17 patients. Eleven patients had definite or

probable inclusion body myositis according topublished criteria. The mean age of the groupwas 69-4 years with a male to female ratio of8:3. Typical clinical features were a slowlyprogressive painless, proximal lower limbweakness, with muscle wasting and early lossofreflexes. The median duration ofillness fromfirst symptom to presentation was five years(range 2-18 years). Fails were a prominentsymptom in six patients and distal weaknessoccurred in nine patients. Creatine kinase wasincreased in 10 patients but only one had a

level >1000 IU/l; the erythrocyte sedimenta-tion rate was normal in five patients. Treat-ment with steroids or cytotoxic drugs, or

both, did not prevent disease progression. Itis confirmed that inclusion body myositis is a

distinct cause ofinflammatory myopathy whichis probably underdiagnosed in the UnitedKingdom. Clinically, it should be suspected inolder patients presenting with muscle weak-ness of insidious onset. Pathologically, a

careful search should be made for rimmedvacuoles and inflammation; ultrastructurally,the presence of inclusions will confirm thediagnosis.

,.,

..... .. 9...........j9_,9,.,, .....j. yX............. :ji-*_..

F Histological features of inclusion body myositis

include fibres containing rimmed vacuoles (arrow) and

atrophic fibres associated with a lymphocytic i'nflammatoryinfiltrate (bottom left). Haematoxylin and eosin frozensection.

Department ofImmunology,University Hospital,Queen's Medical Centre,Nottingham NG7 2UH,United KingdomN D HopkinsonR J PowellDepartment ofHistopathology,University Hospital,Queen's Medical Centre,Nottingham NG7 2UH,United KingdomC HuntJ LoweCorrespondence to:Dr Neil Hopkinson,Department ofRheumatology,Derbyshire Royal Infirmary,Derby DEl 2QY,United Kingdom.Accepted for publication25 July 1992

(Ann Rheum Dis 1993; 52: 147-151)

The term inclusion body myositis was first usedby Yunis and Samaha' in 1971 when theydescribed patients with a slowly progressivemyopathy and characteristic nuclear and cyto-plasmic filamentous inclusions accompanied byvacuoles rimmed by basophilic material in themuscle fibres (fig 1). Electron microscopy ofthese 'rimmed' or 'lined' vacuoles showed thatthey contained cytoplasmic degradation products(fig 2). These inclusions were originally observedby Chou2 in 1967 in a man with 'chronicpolymyositis', and on electron microscopy heobserved aggregates of interwoven filaments inthe sarcoplasmic matrix (fig 3) which resembledmyxovirus nucleocapsids. Since then inclusionbody myositis has been increasingly recognised,

Figure 2 Ultrastructurally, electron microscopy shows thatthe rimmed vacuoles correspond to vacuolation associatedwith autophagic debrisforning membrane-like whorls (largearrow) adjacent to which is a filamentous inclusion bodv(small arrow).

Annals ofthe Rheumatic Diseases 1993; 52: 147-151 147

4Ah..-.

on October 21, 2021 by guest. P

rotected by copyright.http://ard.bm

j.com/

Ann R

heum D

is: first published as 10.1136/ard.52.2.147 on 1 February 1993. D

ownloaded from

Hopkinson, Hunt, Powell, Lowe

Figure 3 At higher magnif cation electron microscopy theinclusion bodies are seen to be composed ofarrays offilaments.

typically having an insidious onset after the ageof 50 years, being more common in men, withthe distal muscles often being affected, and apoor response to steroids.7 Most large series ofpatients with inclusion body myositis have beenreported from the USA where, in one paper, itmade up 28% of all cases of inflammatorymyopathy, the most common cause after poly-myositis (31%), and more common than der-matomyositis (18%).6 A single series of patientswith inclusion body myositis has been reportedin Europe,7 though it is not clear whether thisreflects a genuinely low prevalence or a failureto recognise the disorder. The aim of this studywas to quantify the clinical and laboratoryfindings in a cohort of patients with inclusionbody myositis diagnosed in Nottingham.

Patients and methodsDuring the period 1986-90, 1319 muscle biopsyspecimens were examined in the department ofhistopathology. Because of an interest in musclepathology, the department receives specimensnot only from the whole of Nottingham, butalso from neighbouring district hospitals.

Samples of muscle obtained by either open or

Table 1 Preliminary diagnostic criteria for inclusion body myositis.4 Definite inclusion bodymyosttis requires pathological electron microscopy criterion I and clinical cnrterion 1 plus oneother clinical criterion. Probable inclusion body myositis requires pathological light microscopycriterion I and clinical criterion I plus three other clinical criteria. Posstible inclusion bodymyositis requires pathological light microscopy criterion 2 plus any three clinical cnrtenraPathological criteria

Electron microscopyI Microtubular filaments in the inclusions

Light microscopyI Lined vacuoles2 Intranuclear or intracytoplasmic inclusions, or both

Clinical criteria1 Proximal muscle weakness (insidious onset)2 Distal muscle weakness3 Electromyographic evidence of a generalised myopathy4 Increase in muscle enzyme levels5 Failure of muscle weakness to improve on a high dose regimen of corticosteroids (at least

40-60 mg/day for three to four months)

closed needle biopsy were orientated under adissecting microscope and snap frozen in liquidnitrogen cooled isopentane. Serial 8 ,tm cryostatsections were routinely stained with hae-matoxylin and eosin, NADH Tr, ATPase,periodic acid-Schiff and diastase, or both,myophosphorylase, and acetylcholinesterase.

During this five year period a diagnosis ofinclusion body myositis was pathologically sus-pected in 17 patients because of the presence ofwell defined rimmed vacuoles8 on light micro-scopy. Samples from 15 of these patients wereavailable for conventional transmission electronmicroscopy after fixation in 4% glutaraldehyde.The case records of all 17 patients were

reviewed and specific information sought onage, sex, duration of symptoms, presentingsymptoms, site of muscle disease, clinical signs,associated diseases, drug history, electromyo-gram and laboratory tests, including erythrocytesedimentation rate (ESR), creatine kinase, anti-nuclear antibodies (indirect immunofluore-scence, rat liver substrate), C reactive protein,and blood glucose. All patients were classifiedas having definite, probable, or possible inclu-sion body myositis based on the preliminarydiagnostic criteria of Calabrese et al4 (table 1).

ResultsOf the 17 patients with rimmed vacuoles seen inmuscle biopsy samples, five had typical filamen-tous inclusions on electron microscopy andappropriate clinical criteria to satisfy a diagnosisof definite inclusion body myositis (table 1). Inaddition, an 81 year old man (patient No 4) wasfound to have increased creatine kinase levelsduring investigations for an increased ESR andthrombocytosis (platelet count 707 x 109/1), bothdiscovered during hospital admission for a totalhip replacement for osteoarthritis. A musclebiopsy sample showed rimmed vacuoles andfilamentous inclusions on electron microscopy.He remains well one year after the biopsysample was taken, however, with no evidence ofmuscle weakness. Three patients without fila-mentous inclusions on electron microscopy, andtwo whose muscle was not available for electronmicroscopy, had appropriate criteria to satisfy adiagnosis of probable inclusion body myositis(table 1).

PATIENTS WITH DEFINITE OR PROBABLEINCLUSION BODY MYOSITISTable 2 gives details of these 11 patients. Table3 gives the clinical signs and results of investiga-tions.The mean age of this group was 69-4 years

(range 60-81), with a male to female ratio of8:3. The median duration of illness, from firstsymptom to diagnosis, was five years (range2-18). The most common symptom was progres-sive muscle weakness, which occurred in 10/11patients; the distribution of weakness is shownin table 3. All 10 patients with weakness hadweakness of their legs, usually greater proxi-mally. In contrast, only 8/10 had arm weakness,which in two patients was purely distal. Musclewasting was apparent at presentation in all 10

148

on October 21, 2021 by guest. P

rotected by copyright.http://ard.bm

j.com/

Ann R

heum D

is: first published as 10.1136/ard.52.2.147 on 1 February 1993. D

ownloaded from

Inclusion body myositis: an underdiagnosed condition?

Table 2 Basic clinical information on patients with definite or probable inclusion body myositis

Patient Age Sex Past medical Drug Presenting Length of DiagnosticNo (years) history history svmptoms history (years) crterza

1 73 M Left hemiparesis; Atenolol; Progressive weakness of 6 Definitehypertension nifedipine legs; falls; weak grip;

muscle cramps2 71 M Gout Colchicine Progressive weakness of 7 Definite

legs, falls3 71 M None None Progressive weakness of 3 Definite

legs and arms; myalgia;muscle stiffness

4 81 M Osteoarthritis; hip Sulindac None (increased creatine - Probable'replacement kinase levels on blood

test)5 73 M Osteoarthritis: hip None Progressive weakness of 12 Definite

replacement legs; falls; weight loss6 72 F None None Progressive weakness of 2 Definite

legs; falls7 70 M None None Progressive weakness of 5 Probable

arms and legs; falls8 60 F None None Progressive weakness of 5 Probable

legs; dysarthria9 67 F None None Progressive weakness of 18 Probable

arms and legs10 63 M Discoid lupus None Progressive weakness of 2 Probable

arms and legs11 60 F None None Progressive weakness of 4 Probable

legs; falls

*See text for details.

patients with muscle weakness. Four patientshad a loss of knee reflexes, and in a further threethey were only present on reinforcement. Lossof ankle jerks was more common, these beingcompletely lost in seven patients and presentonly on reinforcement in one patient.Of the nine patients in whom the ESR was

measured, four were abnormal (>20 mm/hour);the mean ESR was 23-9 mm/hour (range 2-74mm/hour). One patient had a normal creatinekinase level and the highest value noted was

1187 IU/l (mean 633 IU/1). C reactive proteinwas normal in the four patients tested: three ofnine patients had positive antinuclear antibodies,none in high titre. No patient had increasedblood glucose.

Table 4 gives light and electron microscopydetails of muscle samples for these patients. Allpatients had a lymphocytic inflammatory infil-trate, which was predominantly perimysial.

Electromyography showed a myopathicpattern in six of seven patients, with fibrillationsuggesting myositis in four of seven patients.One patient (No 2) showed changes suggestiveof chronic denervation with long durationpotentials and without any spontaneous acti-vity; one further patient showed an absent suralsensory nerve action potential.

TREATMENTFive of 11 patients received treatment, all withprednisolone 40 mg/day initially tailing downover at least three months and azathioprine (2 5mg/kg/day). In all patients weakness progressedand azathioprine was changed to cyclophospha-mide (0-5-1 -0 g by mouth or intravenously once

a week for at least six weeks) in three patients,or chlorambucil (4-6 mg/day) in one patient.No improvement, either clinically or in creatinekinase levels, was seen with this change oftreatment. One patient (No 7) died from a

widespread bronchopneumonia associated withgeneralised muscle weakness.

PATIENTS WITHOUT DEFINITE OR PROBABLEINCLUSION BODY MYOSITISSix patients with rimmed vacuoles in a musclebiopsy sample did not satisfy diagnostic criteriafor either definite or probable inclusion bodymyositis according to Calabrese et a14 (table 1).Table 5 details these patients, who had variousclinico-pathological features. Two patientsincluding one with active systemic lupus erythe-matosus (SLE), were considered to have inclu-sion body myositis even though the criteria werenot fulfilled.

Table 3 Clintcal signs and investigations in patients with definite or probable inclusion body myositis

Patient No

1 2 3 4 5 6 7 8 9 10 11

Distribution of weakness*Arms D P=D P None P=D None P=D D P>D P>D NoneLegs P>D P>D P D>P P>D P>D P P>D P>D P>D

Muscle wasting + + + None + + + + + + +ReflexesKnee + (±) + + (+) + (+)Ankle + + (++)

Erythrocyte sedimentation rate(normal <20 mm/hour) N/A 7 9 54 2 74 60 9 N/A 25 10

Creatine kinase (normal <210 IU/l) N/A 760 885 424 745 1187 160 307 444 780 640C reactive protein (normal <20 mg/1) N/A N/A N/A <20 <20 <20 N/A <20 N/A N/A N/AAntinuclear antibodies N/A Negative N/A Negative Weakly Negative Negative Negative Weakly Positive Negative

positive positiveGlucose (normal <7 mmol/l) N/A 5-9 N/A 5 1 N/A 7-1 5 3 N/A 7-0 6 1 5.1

'D=Distal; P=proximal; N/A=not available; +=present; -=absent; (+)=present with reinforcement.

149

on October 21, 2021 by guest. P

rotected by copyright.http://ard.bm

j.com/

Ann R

heum D

is: first published as 10.1136/ard.52.2.147 on 1 February 1993. D

ownloaded from

Hopkinson, Hunt, Powell, Lowe

Table 4 Light and electron microscopy findings in patients with definite or probable inclusion body myositis

Abnormality Patient No Percentageofspecimens

1 2 3 4 5 6 7 8 9 10 Il

Fibres with rimmed vacuoles + + ± ++ ± + + + + + 100Groups of atrophic fibres + + + - - - + + - 45Type I fibre atrophy + - + - + - + + 45Type II fibre atrophy + + - - + + + + + + + 82Necrotic fibres + - - + + + - - 36Eosinophilic inclusions - + 9Lymphocytic inflammatory infiltrate + + + + + + + + + + + 100Filamentous inclusions on electron 55

microscopy + + + + + +

Table S Details of patients with rimmed vacuoles on muscle biopsy who did not fulfil criteria for inclusion body myositis (IBM)

Patient Age Sex Presenting symptoms Creatine kinase Clinical signs Muscle biopsv DiagnosticNo (yrs) (IU/I) findings label

12 87 M Collapse; weakness of legs 21 Quadriceps weak and Atrophic fibres; IBMwasted; knee and ankle lymphocytic infiltrate;reflexes absent rimmed vacuoles;

eosinophilic inclusions13 81 M Collapse due to hemiparesis; 12 Signs of hemiparesis only Atrophic fibres; rimmed Hemiparesis secondary to

muscle biopsy taken vacuoles cerebrovascular accident,because of high not IBMerythrocyte sedimentationrate

14 72 M None; muscle examined 22 No wasting or weakness; Active fibre necrosis; Peripheral vascular disease;after above knee reflexes normal lymphocytic infiltrate; no clinical evidence ofamputation rimmed vacuoles IBM

15 75 F Right footdrop following N/A Right footdrop; absent R No inflammation; atrophic Sciatic nerve lesion possiblyhip replacement for ankle jerk fibres; severe type II resulting from hiposteoarthritis atrophy; rimmed vacuoles replacement; no clinical

evidence ofIBM16 59 M Polyarthralgia; weight loss; 31 Generalised wasting; Lymphocytic infiltrate; Non-Hodgkin's lymphoma

muscle biopsy sample reflexes normal atrophic fibres; rimmed seen on lymph nodetaken to look for vasculitis vacuoles biopsy sample; not IBM

17 49 F Active systemic lupus 47 Abnormal gait; quadriceps Severe type II atrophy; no IBM associated witherythematosus; muscle weakness inflammation (taking systemic lupusbiopsy taken to look for steroids); rimmed erythematosusvasculitis vacuoles; eosinophilic

inclusions

DiscussionThis series of 11 patients with definite or pro-bable inclusion body myositis confirms previousobservations which suggest that inclusion bodymyositis is a distinct cause of inflammatorymyopathy. Characteristic features include:rimmed vacuoles on light microscopy and fila-mentous inclusions on electron microscopy;onset in elderly patients; more common in men;insidious onset of muscle weakness, usuallyproximal but with some distal disease, especiallyin the arms; marked muscle wasting with earlyloss of reflexes and a generally poor response totreatment. The frequency of clinical or probableinclusion body myositis in our laboratory (11/1319 biopsy samples: 0 8%) is similar to thatreported elsewhere in Europe7 (5/850: 0-6%)and the USA3 (7/525: 1-3%).The criteria for definite inclusion body

myositis (see table 1) are microtubular filamentson electron microscopy, proximal muscle weak-ness, and one other clinical criterion. Thesecriteria would be satisfied by three previouslyreported patients who had a hereditary distalmyopathy or oculopharyngeal muscular dys-trophy.9 10 The main differences between thelatter case reports and our own series are thepatient's age and the family history of muscledisease. Our patients all presented when over 60years old, whereas the patients with hereditarymyopathy were 9, 22, and 23 years old at thetime of their first symptom. Moreover, a familyhistory of muscle disease was absent in thepatients with inclusion body myositis. Forprobable inclusion body myositis, Calabrese et

al' propose as criteria the presence of rimmed(lined) vacuoles, proximal muscle weakness,and three of the other four clinical criteria. Inthis study, only 65% of patients with rimmedvacuoles had a diagnosis of inclusion bodymyositis, confirming their non-specificity; pre-viously they have been described in somefamilial distal myopathies,"I denervatedmuscle,'2 and in oculopharyngeal dystrophy.'3For a diagnosis of probable inclusion bodymyositis, it would therefore seem reasonable toinclude further criteria of a myopathy with apoor response to treatment. In the clinicalsetting the diagnosis of inclusion body myositisshould be strongly suspected on clinical groundsin elderly patients with muscle weakness,especially when this is of insidious onset andassociated with muscle wasting and loss ofreflexes. This should then prompt a carefulsearch for rimmed vacuoles on light microscopywhich may, in our experience, require examina-tion of the biopsy sample at many levels.Although a diagnosis of definite inclusion bodymyositis depends on the presence of filamentousinclusions on electron microscopy it is clearfrom this and earlier studies that they are notalways found, possibly because of samplingerrors. Patient No 4 in our series is notable inhaving an increased creatine kinase level, musclebiopsy findings ofrimmed vacuoles, lymphocyticinfiltrates, and filamentous inclusionson electronmicroscopy but remaining symptom free.Whether he develops muscle weakness withtime remains to be seen.

It is likely that inclusion body myositis

150

on October 21, 2021 by guest. P

rotected by copyright.http://ard.bm

j.com/

Ann R

heum D

is: first published as 10.1136/ard.52.2.147 on 1 February 1993. D

ownloaded from

Inclusion body myositis: an underdiagnosed condition?

remains an underdiagnosed disorder. Duringthe five year study period, 25 patients were

diagnosed histologically as having polymyositis,compared with 11 with definite or probableinclusion body myositis. Rimmed vacuoles maybe missed on light microscopy and facilities forelectron microscopy may be limited, leading toa misdiagnosis. We have seen patients with a

diagnosis of polymyositis who remain un-

responsive to treatment; review of their initialmuscle biopsy sample, often performed many

years earlier, has then showed typical rimmedvacuoles leading to a revised diagnosis. Anaccurate diagnosis of inclusion body myositis isimportant because of treatment and prognosticimplications. Unlike the generally good responseof polymyositis to treatment with steroids andimmunosuppressive drugs,'4 the present studyis in agreement with others in showing thattreatment is disappointing. All five patientstreated in this study deteriorated despite cyto-toxic drugs and steroids. Likewise, 25 patientsfollowed up for two or more years in the seriesof Lotz et at6 showed progression of weaknessdespite treatment. Treatment intervention isstill advocated according to some workers,however,'5 because occasionally a response isseen, particularly when inclusion body myositisis in association with a connective tissuedisease. A pragmatic approach to treatmentwould be to stop treatment with cytotoxic drugsif a therapeutic response is not seen shortly afterbeginning steroid treatment.'5

Although inclusion body myositis has beenassociated with a variety of autoimmunediseases,'6 these have usually been single case

reports and may represent a reporting bias dueto the underlying illness being closely monitored.In the largest series of patients with inclusionbody myositis reported to date,6 a single case ofSLE was seen in 40 patients; there were 13other autoimmune diseases associated withinclusion body myositis, including eight patientswith diabetes. No patients with diabetes were

seen in the present series and the relevance ofsuch an association remains unclear.We did not characterise the mononuclear cell

infiltrate in our patients; however, Arahata andEngel have previously reported that the pre-dominantly endomysial inflammation in inclu-sion body myositis is composed of CD8+ Tcells and macrophages in a 2:1 ratio. 17

Historically the involvement of viral agents,such as paramyxovirus, has been implicated.Two studies have provided other possible cluesto the aetiology of this poorly understooddisorder. Mendell et al'8 identified Congo redapple green birefringent deposits in vacuolatedmuscle fibres in inclusion body myositis charac-teristic of amyloid. The distribution of theamyloid collections corresponded to the typicalfilaments of inclusion body myositis. Further-

more, Askanas et al'9 have identified the typeof amyloid as P-(A4) amyloid protein, which iscentral to the pathogenesis of Alzheimer'sdisease. This suggests that these j3-(A4) amyloiddeposits in muscle in inclusion body myositisand the brain deposits in Alzheimer's diseasemay follow similar cellular events.We conclude that inclusion body myositis

does represent a distinct clinical and histopatho-logical entity that is probably underdiagnosed inthe United Kingdom. Evaluation of previousmuscle biopsy or repeat biopsy samples areinvaluable in the confirmation of the diagnosisand thereby targeting therapeutic approaches.Confirmation of the diagnosis relies on identifi-cation of often subtle features within the musclebiopsy sample and electron microscopy remainsessential.

The authors thank Janet Palmer for expert preparation of musclesamples for histology, Trevor Gray for electron microscopy, andBill Brackenbury for the photomicrographs.

1 Yunis E, Samaha F. Inclusion body myositis. Lab Invest1971; 25: 240-8.

2 Chou S-M. Myxovirus-like structures in a case of humanchronic polymyositis. Science 1%7; 158: 1453-5.

3 Danon M, Reyes M, Perurena 0, Masdeu J, Manaligod J.Inclusion body myositis. A corticosteroid resistant idio-pathic inflammatory myopathy. Arch Neurol 1982; 39:760-4.

4 Calabrese L, Mitsumoto H, Chou S. Inclusion body myositispresenting as treatment resistant polymyositis. ArthritisRheum 1987; 30: 397-403.

5 Ringel S, Kenny C, Neville H, Giorno R, Carry M. Spectrumof inclusion body myositis. Arch Neurol 1987; 44: 1154-7.

6 Lotz B, Engel A, Nishino H, Stevens J, Litchy W. Inclusionbody myositis. Observations in 40 patients. Brain 1989;112: 727-47.

7 Mhiri C, Gherardi R. Inclusion body myositis in Frenchpatients. A clinicopathological evaluation. NeuropatholAppl Neurobiol 1990; 16: 333-44.

8 Carpenter S, Karpati G, Heller I, Eisen A. Inclusion bodymyositis: a distinct variety of idiopathic inflammatorymyopathy. Neurology 1978; 28: 8-17.

9 Fukuhara N, Kumamoto T, Tsubaki I, Mayuzumi T,Nitta H. Oculopharyngeal muscular dystrophy and distalmyopathy: intrafamilial difference in the onset and distri-bution of muscular involvement. Acta Neurol Scand 1982;65:458-67.

10 Matsubara S, Tanabe H. Hereditary distal myopathy withfilamentous inclusions. Acta Neurol Scand 1982; 65: 363-8.

11 Nonaka I, Sunohara N, Ishiura S, Satoyoshi E. Familialdistal myopathy with rimmed vacuole and lamella (myeloid)body formation. J7 Neurol Sci 1981; 51: 141-55.

12 Fukuhara N, Kumamoto T, Tsubaki T. Rimmed vacuoles.Acta Neuropathol (Berl) 1980; 51: 229-35.

13 Tome F, Fardeau M. Nuclear inclusions in oculopharyngealdystrophy. Acta Neuropathol (Berl) 1980; 49: 85-7.

14 Henriksson K, Sandstedt P. Polymyositis-treatment andprognosis: a study of 107 patients. Acta Neurol Scand 1982;65:280-300.

15 Richter Cohen M, Sulaiman A, Garancis J, Wortmann R.Clinical heterogeneity and treatment response in inclusionbody myositis. Arthn'tis Rheum 1989; 32: 734-9.

16 Chad D, Good P, Adelman L, Bradley W, Mills J. Inclusionbody myositis associated with Siogren's syndrome. ArchNeurol 1982; 39: 186-8.

17 Arahata K, Engel A. Monoclonal antibody analysis ofmononuclear cells in myopathies, 1: quantitation of subsetsaccording to diagnosis and sites of accumulation anddemonstration and counts of muscle fibres invaded byT cells. Ann Neurol 1984; 16: 193-208.

18 Mendell J, Sahenk Z, Gales T, Paul L. Amyloid filaments ininclusion body myositis: novel findings provide insight intonature of filaments. Arch Neurol 1991; 48: 1229-34.

19 Askanas V, Engel W, Alvarez R, Glenner G. (-Amyloidprotein immunoreactivity in muscle of patients withinclusion-body myositis. Lancet 1992; 339: 560-1.

151

on October 21, 2021 by guest. P

rotected by copyright.http://ard.bm

j.com/

Ann R

heum D

is: first published as 10.1136/ard.52.2.147 on 1 February 1993. D

ownloaded from