Annals of the Rheumatic Diseases 1991; So: 854-861

Clinical aspects of systemic sclerosis (scleroderma)

Richard M Silver

Systemic sclerosis (scleroderma) is a disease ofunknown cause, the hallmark of which isinduration of the skin. Although long regardedas a bland fibrotic process, there is now ampleevidence of an active inflammatory processunderlying thepathogenesis ofsystemic sclerosis.In addition, microvascular disease and immuno-logical abnormalities are present in most cases.It remains to be determined just how theimmunological and microvascular changesrelate to the overproduction of collagen andother matrix elements by the fibroblast, butrecent data suggest that products of the immuneresponse may directly affect fibroblasts andendothelial cells in vitro.

This review will focus on recent advances inthe understanding of several clinical aspects ofsystemic sclerosis. The reader is referred toseveral recent chapters and textbooks for a moreextensive review. 1-3

Division of Immunologyand Rheumatology,Medical University ofSouth Carolina,Charleston,South Carolina29425 2229, USAR M Silver

ClassificationScleroderma may exist as a localised or asystemic disease process. The latter has beendelineated 'progressive systemic sclerosis' in thepast, but use of this term has been decried as thedisease is not always progressive and because ofthe emotional burden that this term mayplace on the patient and the patient's family.4

In its localised form scleroderma is confinedto the skin and adjacent tissues, and it can beclassified as linear scleroderma or morphoea(see below). In its systemic form sclerodermamay affect a number of visceral organs. Herethe classification is somewhat controversial andis based on the extent of cutaneous involvement.Table 1 shows a widely accepted classification,where diffuse (particularly truncal) skin diseaseis distinguished from limited skin involvement;the latter encompasses and replaces the CREST(calcinosis, Raynaud's phenomenon, oeso-phageal dysmotility, sclerodactyly, and tel-angiectasia) variant of scleroderma. Limitedcutaneous systemic sclerosis also includes whathas been variously designated acrosclerosis,types I and II scleroderma of Barnett, types Iand II of the German classification, inter-mediate cutaneous systemic sclerosis ofGiordano, and systemic sclerosis withoutscleroderma.4A recent editorial summarises data which

support a conceptual framework of three sub-types of systemic sclerosis: digital, proximalextremity, and truncal,5 but at the present timeserological and microvascular data support thesimpler classification shown in table 1.4 Byclassifying patients into subsets early in the

course of the illness, one hopes to identifypatients at greater or lesser risk of developingcertain visceral complications, as well as providea more homogeneous group of patients forstudies of the pathogenesis, clinical manifesta-tions, and treatment.

Clinical featuresRAYNAUD'S PHENOMENONRaynaud's phenomenon refers to episodic digitalischaemia provoked by cold or emotion.Although classically described as triphasic-that is, pallor followed by cyanosis, and thenhyperaemia accompanied by numbness andpain, such a three colour response does notoccur universally. Pallor seems to be the mostreliable sign and hyperaemia the least reliablesign in subjects who lack the classic triphasicresponse. A recently described questionnaireand colour chart may facilitate the diagnosis ofRaynaud's phenomenon.6

Establishment of the presence or absence ofRaynaud's phenomenon is important whenevaluating a patient with scleroderma-like skin.The absence of Raynaud's phenomenon shouldraise the possibility that one might be dealingwith a disease other than systemic sclerosis(table 2) as 95% of patients with systemicsclerosis have Raynaud's phenomenon. Thisdistinction is illustrated by the recently describedeosinophilia-myalgia syndrome associated withL-tryptophan ingestion, in which scleroderma-like skin lesions commonly accompanied

Table I Subsets of systemic sclerosis. Reproduced, withpermission, from ref 4

Diffuse cutaneous systemic sclerosis*Onset of Raynaud's phenomenon within one year of onset of skin

changes (puffy or hidebound)Truncal and acral skin involvementPresence of tendon friction rubsEarly and significant incidence of interstitial lung disease,

oliguric renal failure, diffuse gastrointestinal disease, andmyocardial disease

Absence of anticentromere antibodiesNailfold capiLary dilatation and capillary destructiontAntitopoisomerase antibodies (30% of patients)

Limited cutaneous systemic sclerosisRaynaud's phenomenon for years (occasionally decades)Skin involvement limited to hands, face, feet, and forearms

(acral) or absentA significant late incidence of pulmonary hypertension, with or

without interstitial lung disease, trigeminal neuralgia, skincalcifications, telangiectasia

A high incidence of anticentromere antibody (70 80%)Dilated nailfold capillary loops, usually without capillary dropout

*Experienced observers note some patients with diffusecutaneous systemic sclerosis who do not develop organ insuf-ficiency and suggest the term chronic diffuse cutaneous systemicsclerosis for these patients.tNailfold capillary dilatation and destruction may also be seenin patients with dermatomyositis, overlap syndromes, and un-differentiated connective tissue disease. These syndromes maybe considered as part of the spectrum of scleroderma associateddisorders.

854

on June 25, 2020 by guest. Protected by copyright.

http://ard.bmj.com

/A

nn Rheum

Dis: first published as 10.1136/ard.50.S

uppl_4.854 on 1 Novem

ber 1991. Dow

nloaded from

Clinical aspects ofsystemic sclerosis (scleroderma)

Table 2 Diseases with cutaneous features resemblingsclroderma

Eosinophilic fasciitisTryptophan associated eosinophilia-myalgia syndromeChemical induced disorders

Bleomycin fibrosisVinyl chloride diseaseTrichloroethylene fibrosisToxic oil syndrome

Graft versus host diseaseDigital sclerosis of diabetes mellitusInfiltrating carcinomasScleroedemaScleromyxoedema (papular mucinosis)Werner's syndromeProgeriaRothmund's syndromeAcrodermatitis chronica atrophicansLichen sclerosis et atrophicusCarcinoid syndromePhenylketonuriaPorphyria cutanea tardaCongenital porphyriaPrimary amyloidosisAcromegaly

myalgia, eosinophilia, and fasciitis (see below).7Few of such cases were accompanied byRaynaud's phenomenon, nor had other diseaseswhich may mimic systemic sclerosis.The duration of Raynaud's phenomenon

before skin involvement is important. Patientswith diffuse cutaneous systemic sclerosis and atendency to early visceral organ damage usuallyhave a brief duration of Raynaud's phenomenonbefore development of skin changes, whereaspatients with limited cutaneous systemicsclerosis usually have many years (often decades)of Raynaud's phenomenon before overt skinand visceral involvement.An important issue for the clinician is the

evaluation of the subject with Raynaud's pheno-menon for the presence of an underlying con-nective tissue disease. Raynaud's phenomenonitself is quite common in the general population,yet only a fraction of such subjects will developa connective tissue disease.8 One survey foundthe prevalence of Raynaud's phenomenon to be4-6% among adults living in South Carolina.9The prevalence may be higher in colder climates.As Raynaud's phenomenon may herald the

development of serious disease, especially con-nective tissue disease, studies have been under-taken to determine which ancillary tests may bepredictive of the evolution to an overt connec-tive tissue disease. Two markers seem to havepredictive value for the subsequent develop-ment of connective tissue disease (usuallysystemic sclerosis): (a) abnormal nailfold capil-laries and (b) antinuclear antibodies.

Definitely abnormal capillary patterns,detected by nailfold capillary microscopy,indicate an increased risk for the presence oreventual transition to a connective tissuedisease.'0 A recent prospective study of patientswith Raynaud's phenomenon confirmed thisfinding and concluded that the presence ofabnormal capillaries was the most significantpredictor of the subsequent developmentof systemic sclerosis (odds ratio, 26-82,p=0 001)." In addition, the degree of capillaryloss assessed by nailfold capillary microscopymay be correlated with clinical and serologicalfeatures and may assist in the subclassificationof systemic sclerosis as diffuse cutaneous or

limited cutaneous systemic sclerosis (see tablei). 12

The other marker which may be useful inpredicting the presence or eventual developmentof connective tissue disease in a subject withRaynaud's phenomenon is the antinuclear anti-body. Here the predictive value depends on themethod and substrate used. Two recent studiesshowed that the immunoblot technique wasmore sensitive than indirect immunofluorescenceand, furthermore, antibody specificity of anti-nuclear antibodies as determined by immuno-blots was predictive of the development ofspecific subtypes of systemic sclerosis. 13 4 In aprospective study of patients with Raynaud'sphenomenon and undifferentiated connectivetissue disease Kallenberg et al found that thepresence of antinuclear antibodies detected byimmunoblots at the time of entry into the studywas associated with the evolution of symptomsof a connective tissue disease, usually systemicsclerosis.'3 Furthermore, anticentromere anti-body was associated with limited cutaneoussystemic sclerosis (sensitivity 60%, specificity98%) and anti-Scl-70 antibody was associatedwith diffuse cutaneous systemic sclerosis (sensi-tivity 38%, specificity 100%). Based on theaforementioned studies, it is recommended thatall patients with Raynaud's phenomenon have acomplete clinical evaluation, including nailfoldcapillary microscopy and antinuclear antibodystudies which, preferably, test for specificantigens associated with subtypes of systemicsclerosis.

SKIN DISEASESclerosis of the skin is the hallmark of systemicsclerosis, though rare patients may have typicalvisceral organ involvement in the absence ofskin disease (systemic sclerosis without sclero-derma). 15 Skin thickening is the definitivediagnostic criterion of systemic sclerosis in thevast majority of cases, and the distribution ofaffected skin serves as the means of classifyingpatients into one or another subset of systemicsclerosis (see above). Such classification mayhave important implications for the risk ofdeveloping visceral organ involvement and formortality (table 3).4Three phases of dermal involvement have

been described. Firstly, there is an oedematousphase, often presenting as stiff and puffy fingersand hands. In this initial phase the condition isoften difficult to distinguish from other connec-tive tissue diseases or carpal tunnel syndrome;nailfold capillary microscopy and determinationof the presence and type of antinuclear antibodyare of greatest diagnostic value in patients inthis disease phase. Secondly, an indurativephase, characterised by tightness of the skin,usually follows the oedematous phase. Here,

Table 3 Survival in systemic sclerosis (%). Reproduced,with permission, from ref 4

Systemic sclerosis I year 6 years 12 yearssubtpe

Limited cutaneous 98 80 50Diffuse cutaneous 80 30 15

855

on June 25, 2020 by guest. Protected by copyright.

http://ard.bmj.com

/A

nn Rheum

Dis: first published as 10.1136/ard.50.S

uppl_4.854 on 1 Novem

ber 1991. Dow

nloaded from

Silver

sclerodactyly and the classic expressionless facemake the diagnosis unmistakable. Ultimately,an atrophic phase occurs, when the skin mayactually soften.

Pitting scars over the fingertips, often accom-panied by loss of substance of the finger pad,are characteristic findings in both limited anddiffuse cutaneous systemic sclerosis. Thepresence of digital pitting scars is one of theminor criteria for systemic sclerosis.'6 Tel-angiectasias are commonly found on the handsand face, especially in the limited cutaneousform of systemic sclerosis. Calcinosis occurscommonly on the volar aspect of the fingertipsand over joints such as the metacarpophalangealand interphalangeal joints, where it mayulcerate.

GASTROINTESTINAL TRACT DISEASEGastrointestinal disease is the most commonlyrecognised visceral manifestation of systemicsclerosis. Dysphagia and heartburn, the mostcommon symptoms, are due to oesophagealdysmotility and oesophageal reflux. Nearly 85%of patients with systemic sclerosis have beenfound to have oesophageal reflux using 24 hourpH monitoring of the distal oesophagus.'7 Theoesophagus is not affected in localised sclero-derma.

Peristaltic abnormalities may delay gastricemptying and may also affect motility of thesmall and large intestines, at times giving rise topseudo-obstruction or malabsorption secondaryto bacterial overgrowth. Malabsorption mayrarely result from pancreatic insufficiency. Arecent study concluded that pancreatic exocrinefunction is often reduced in patients withsystemic sclerosis, but rarely to an extent that isclinically important.18 Primary biliary cirrhosismay occur, usually in association with long-standing limited cutaneous systemic sclerosis.

Palliative treatment of gastrointestinal diseaseis based upon the prevention of gastric refluxand stricture formation by using combinationsof antacids, H2 blockers, sucralfate, withelevation of the head of the bed. Severereflux oesophagitis may respond to treatmentwith omeprazole. Malabsorption and pseudo-obstruction are treated with antibiotics, bowelrest and, at times, parenteral hyperalimentation.

relatively insensitive tests of lung disease.Recent studies using thin section computedtomography have shown abnormalities beforethe onset of dyspnoea or the detection ofabnormalities on routine chest radiograph.'9Tests of pulmonary function often discloserestrictive lung disease with reduced lungvolumes and decreased diffusing capacity. Aslung disease progresses, hypoxia and corpulmonale may occur.

Recently, a number of studies have demon-strated an inflammatory component to the lungdisease in a significant percentage of patientswith systemic sclerosis. Gallium-67 lung scansoften show increased uptake,20 but galliumscans are non-specific and not recommended.Studies using bronchoalveolar lavage haveshown that a significant proportion of patientswith systemic sclerosis have evidence of an'alveolitis', often before the onset of dyspnoeaor abnormal pulmonary function tests or chestradiograph,'9 21 and most studies concur thatthe alveolitis is characterised by an increasednumber of total cells (mostly alveolar macro-phages) and an increased percentage andabsolute number of neutrophils and eosino-phils-similar increases to those described incryptogenic fibrosing alveolitis.22-26 A fewpatients may have a lymphocytic alveolitis,particularly patients with Sjogren's syndrome.

In our study of 43 patients with systemicsclerosis (all non-smokers) an increased bron-choalveolar lavage cell count or an increasedpercentage of neutrophils, or both, was as-sociated with poorer pulmonary function.27 Thecourse of interstitial lung disease in patientswith systemic sclerosis has been shown not to bepredicted by traditional tests, such as initial

4-

3-

>

2-

Initial Follow up Initial Follow up

PULMONARY DISEASEThe prevalence of lung disease ranks second tothat of the gastrointestinal tract. Dyspnoea andhypoxiamay resultfrom interstitial inflammationand fibrosis, or may be the result of pulmonaryhypertension occurring in the absence ofparenchymal lung disease. Severe pulmonaryhypertension is seen most often in patients withlimited cutaneous systemic sclerosis. Withrecent advances in the treatment of sclerodermarenal disease, scleroderma lung disease hasbecome the most common cause of death.Dyspnoea on exertion is the most common

pulmonary complaint and is often accompaniedby a non-productive cough. Chest radiographsoften show diffuse linear and nodular fibrosis inthe lower two thirds of the lung fields, but are

zxIEE

EE

0

up

Changes inforced vital capacity (FVC) and carbonmonoxide transferfactor(TLco) in 10 patients with systemicsclerosis classified according to bronchoalveolar lavagefindings as those without (0) and those with (O) alveolitis.Solid symbols represent means."

v up

I -

856

8-----,O,--..o on June 25, 2020 by guest. Protected by copyright.

http://ard.bmj.com

/A

nn Rheum

Dis: first published as 10.1136/ard.50.S

uppl_4.854 on 1 Novem

ber 1991. Dow

nloaded from

Clinical aspects ofsystemic sclerosis (scleroderma)

chest radiograph or pulmonary function tests,nor by demographic data. In a prospectivestudy of scleroderma lung disease we found thatnormal bronchoalveolar lavage findings wereassociated with a stable course, but abnormalfindings were associated with worseningdyspnoea, worsening chest radiograph, andsignificantly greater decline in forced vitalcapacity and carbon monoxide transfer factor(see figure).27 Thus bronchoalveolar lavage-mayprove to be a useful predictor of the course ofscleroderma lung disease.Treatment of the interstitial lung disease of

systemic sclerosis has been disappointing. Ourpreliminary studies suggest that patients withsystemic sclerosis with alveolitis may benefitfrom treatment with cyclophosphamide, butcontrolled prospective trials are needed.27Perhaps bronchoalveolar lavage and highresolution computed tomography will proveuseful in selecting patients with potentiallyreversible disease, as well as helpful in monitor-ing responses to drug treatment.

Bronchoalveolar lavage does provide infor-mation about cells and proteins obtained fromthe lung, a common site of active disease.Products of alveolar macrophages such asfibronectin, a glycoprotein that may serve asa chemoattractant and growth factor for fibro-blasts, may play a part in the pathogenesis ofscleroderma lung disease. We have found thatscleroderma alveolar macrophages releasesignificantly more fibronectin than controls andthat the concentration of fibronectin correlateswith the degree of alveolitis and correlatesnegatively with the carbon monoxide transferfactor.28 Other growth factors secreted byalveolar macrophages are currently underinvestigation.Pulmonary hypertension (in the absence of

significant interstitial fibrosis) occurs in about5% of patients with limited cutaneous systemicsclerosis but rarely in patients with the diffusecutaneous subtype. A recent case-controlnecropsy study showed that intimal thickeningand luminal narrowing occur in both subsets ofsystemic sclerosis, but that such pulmonaryvascular changes are most pronounced inpatients with limited cutaneous systemicsclerosis.29 By the time that patients presentwith dyspnoea and hypoxia significant pul-monary vascular resistance might have occurredowing to structural luminal narrowing. Thusvasodilator treatment is often unsuccessful atthis point. Perhaps the early use of vasodilatoror other drug treatment in patients identified asbeing at risk for pulmonary hypertension(limited cutaneous systemic sclerosis, dilatednailfold capillaries with minimal capillary loss,anticentromere antibody) will forestall thedevelopment of this often fatal complication ofsystemic sclerosis.

RENAL DISEASEProteinuria, azotaemia, or hypertension occursin 45% of patients with systemic sclerosis.30 Ofall the visceral organs potentially affected bysystemic sclerosis, renal disease has been as-sociated with the highest mortality. Until the

introduction of dialysis and potent antihyper-tensive drugs, renovascular hypertension wasuniformly fatal. The characteristic histopatho-logical finding is that of concentric, subendo-thelial intimal proliferation of small arcuate andinterlobular arteries. A functional correlate ofthis structural abnormality is the reduction inrenal cortical blood flow, which may be furthercompromised by vasospasm associated withRaynaud's phenomenon.30

Significant risk factors for the onset ofscleroderma renal crisis include anaemia, peri-cardial effusion, congestive heart failure, andrapid progression of skin thickening.3' Therenin-angiotensin system plays a major part inthe pathogenesis of malignant hypertensionsecondary to systemic sclerosis. Many suchpatients have increased plasma renin activity,which may precede or coincide with rapiddeterioration in renal function, and many willrespond to angiotensin converting enzymeinhibitor treatment with normalisation of bloodpressure and, sometimes, reversal of renalfailure. The use of captopril and other convert-ing enzyme inhibitors has been associated with amarked reduction in mortality from sclerodermarenal crisis,32 but not all cases will responddespite normalisation of blood pressure.33 3 Arecent report described a small percentage(11%) of patients with systemic sclerosis withrenal crisis not accompanied by hypertension.35When compared with patients with systemicsclerosis with hypertensive renal crisis, suchcases were more likely to have had microangio-pathic haemolytic anaemia and thrombocyto-penia, and were more likely to have receivedhigh doses of corticosteroids during the twomonths immediately preceding the renal crisis.

Renal disease in patients who develop renalfailure despite treatment has been managedwith haemodialysis and transplantation, butexperience is limited and success has beenvariable. Haemodialysis often presents technicalproblems with vascular access. Contrary tosome earlier reports, we and others have foundcontinuous ambulatory peritoneal dialysis to beeffective and well tolerated in patients withsystemic sclerosis with end stage renal disease.36

CARDIAC DISEASEThe landmark paper by Weiss et al establishedthe entity of scleroderma heart disease 48 yearsago.37 Heart disease may present as heartfailure, arrhythmias, conduction disturbances,or chest pain, all of which may be the result ofvascular disease and fibrosis. Weiss et al notedan unusual type of myocardial scarring in thepresence ofnormal extramural coronary arteries.Such scarring is probably antedated by contrac-tion-band necrosis, the classic finding in sclero-derma heart disease, which has been producedexperimentally by transient interruption ofblood flow. Once again, the functional correlateof this structural abnormality may be reversiblevasospasm affecting small coronary arteries.Evidence exists for left ventricular dysfunctioninduced by cold,38 as well as cold inducedregional perfusion defects.39 Structural lesionsof the coronary microcirculation may also be

857

on June 25, 2020 by guest. Protected by copyright.

http://ard.bmj.com

/A

nn Rheum

Dis: first published as 10.1136/ard.50.S

uppl_4.854 on 1 Novem

ber 1991. Dow

nloaded from

Silver

major determinants ofcardiac disease in systemicsclerosis. Some patients with primary sclero-derma heart disease have reduced coronaryblood flow and coronary reserve after maximalcoronary vasodilatation with intravenousdipyridamole.'

Coronary vasospasm seems to be reversiblewith calcium channel blockers such as nife-dipine."' As in the case of pulmonary hyperten-sion, some patients may not respond to vaso-dilator treatment owing to the presence of fixedstructural lesions affecting the coronary micro-circulation. Studies are needed to identifypatients at risk of developing such lesions in thecoronary microcirculation so that treatmentmay be started before irreversible intimalhypertrophy.

PathogenesisGENETIC ASPECTSA number of cases of familial scleroderma havebeen reported.42 Although no common geneticmarkers have been identified, asymptomaticrelatives have a higher incidence of antinuclearantibodies than controls.43 Several populationstudies have shown an increase in the prevalenceof certain HLA types, including DRI, DR3,and DR5, among patients with systemicsclerosis." A recent study showed that theincreased chromosomal breakage rate reportedin patients with systemic sclerosis and their firstdegree relatives is linked to one particular HLAhaplotype, HLA Al, B8, DR3.45 Of interest isthe association of the same haplotype with thesevere form of scleroderma-like disease inducedby vinyl chloride.'

CONNECTIVE TISSUE ASPECTSThe characteristic feature of systemic sclerosisis excessive deposition of collagen and otherconnective tissue matrix proteins. LeRoy notedthat scleroderma dermal fibroblasts synthesiseexcessive collagen in vitro and that this increasedability to synthesise collagen persists in vitro fora number of passages, before declining towardsnormal.47 Increased synthesis of glycosamino-glycans and fibronectin has also been shown.Increased concentrations of mRNA of each ofthese matrix proteins exist in sclerodermadermal fibroblasts. This may result fromincreased rates of transcription, but increasedhalf lives of the mRNAs may also contributeto increased matrix synthesis in systemicsclerosis.48 49

Recent studies have shown that transforminggrowth factor ,B may enhance transcription ofcollagen mRNA through the activation of apromoter region of the collagen gene.50 Wheninjected into experimental animals, transform-ing growth factor c causes a mononuclear cellinflammatory response and fibrosis.5' Also ofpotential relevance to systemic sclerosis, trans-forming growth factor ,B induces the autocrineproduction of a potent fibroblast mitogen,platelet derived growth factor,52 and it inhibitsendothelial cell proliferation in vitro.53 Theseand other studies support a hypothetical role fortransforming growth factor 3 in the induction of

the vascular and interstitial lesions of systemicsclerosis.54

VASCULAR ASPECTSThe physiological (Raynaud's phenomenon)and pathological (in vivo and postmortem)features of vascular disease in systemic sclerosishave been discussed. Raised plasma concentra-tions of factor VIII/von Willebrand factor and aplatelet product, (3 thromboglobulin, furtherattest to microvascular and endothelial injury.55The mediator of endothelial injury in systemicsclerosis remains controversial. Two recentlydefined molecules which are capable of causingendothelial injury are tumour necrosis factor a56and transforming growth factor (.53 Each hasthe ability to stimulate fibroblast proliferation,as well as induce endothelial injury.

IMMUNOLOGICAL ASPECTSThe common occurrence of antinuclear anti-bodies and the presence of mononuclear cellinfiltrates in the dermis suggest a role for alteredimmune responsiveness in the pathogenesis ofsystemic sclerosis. Additional indirect supportfor this notion is the occurrence of scleroderma-like lesions in patients with chronic graft versushost disease. These and other immunologicalfeatures have been reviewed recently.' Newdata implicating the immune response in thepathogenesis of systemic sclerosis are reviewedbelow.As noted above, antinuclear antibodies are

often found in patients with systemic sclerosisand their first degree relatives, up to 95% of theformer and 57% of the latter, when a rapidlydividing, human cell substrate such as HEp-2 isused.43 Two particular antinuclear antibodiesdeserve mention because of their subset specifi-city and potential role in pathogenesis. The firstis the anticentromere antibody, which recognisesthree human chromosomal antigens, and whichis associated with limited cutaneous systemicsclerosis and patients with Raynaud's pheno-menon at risk of developing limited disease.57The association of anticentromere antibodywith certain nailfold capillary abnormalities hasbeen noted'2; both are useful in screeningpatients with Raynaud's phenomenon forunderlying systemic sclerosis.'3 " The second,anti-Scl-70, found in about 30% of patients withdiffuse cutaneous systemic sclerosis, is an auto-antibody directed against the nuclear enzyme,DNA topoisomerase 1.58 Another autoantibody,found in a smaller percentage of patients withdiffuse cutaneous systemic sclerosis, seems tobe directed against a nucleolar enzyme, RNApolymerase V9 Recently it has been suggestedthat topoisomerase I may accelerate collagengene transcription by virtue of its ability to bindto promoter regions and other sites on collagengenes, leading Douvas to propose that in-hibitors of topoisomerase I may ultimatelyprove useful in controlling the excessivecollagen synthesis which characterises systemicsclerosis.'

Activation of the complement pathways hasbeen shown recently in patients with systemicsclerosis.6' Newly developed techniques

858

on June 25, 2020 by guest. Protected by copyright.

http://ard.bmj.com

/A

nn Rheum

Dis: first published as 10.1136/ard.50.S

uppl_4.854 on 1 Novem

ber 1991. Dow

nloaded from

Clinical aspects ofsystemic sclerosis (scleroderma)

showed that concentrations of activated comple-ment components were high in many patientswith systemic sclerosis and these concentrationsreflected the clinical severity.The presence of mononuclear cells in the

dermis of patients with systemic sclerosis hasgenerated interest in the products of such cellsand their potential effects on fibroblasts. Mastcells are present in the dermis of patients withsystemic sclerosis,62 and in a number of experi-mental models and pathological states of fibrosis,including graft versus host disease. Inhibitors ofmast cell degranulation reduce the fibrosiswhich occurs spontaneously in one experimentalmodel of scleroderma, the tight skin mouse.63 6It remains to be determined whether suchinhibitors can modify the course of systemicsclerosis.

Soluble products of lymphocytes and mono-cytes have recently been studied for their effectson vascular endothelium and fibroblasts. Onesuch cytokine, interferon y, can down regulatecollagen synthesis by fibroblasts in vitro. Inter-feron y exerts its action at the level of transcrip-tion by decreasing mRNA for procollagen, andhas a longlasting effect.65 6 In one study a 72hour exposure to interferon y reduced pro-collagen mRNA concentrations in sclerodermafibroblast cell lines to those shown by controlfibroblasts.' Clinical trials of recombinantinterferon y in systemic sclerosis are currentlyunder way.

Other cytokines are also potentially relevantto the pathogenesis of systemic sclerosis. Inter-leukin 1 is mitogenic to fibroblasts, as is aninhibitor to interleukin 1 produced by mono-nuclear cells from patients with systemicsclerosis.67 Lymphotoxin and tumour necrosisfactor a, present in scleroderma serum samples,may also be important mediators.68 Tumournecrosis factor a has been shown in vitro toinjure endothelial cells and to stimulate fibro-blast proliferation.56 68 Future treatments mayuse modifiers of the biological response(s) tosuch cytokines.

Special considerationsLocalised scleroderma is similar to systemicsclerosis histopathologically, but effects of theformer condition are limited to the skin, sub-cutaneous tissue, fascia, and adjacent muscle.Depending upon the location and extent ofaffected skin, localised scleroderma may beclassified as linear or morphoea (plaque orgeneralised). In linear scleroderma scleroticskin occurs in a band-like distribution, oftencrossing joint lines and resulting in contractures.Facial hemiatrophy may be regarded as a formof linear scleroderma affecting the face andscalp (en coup de sabre). Morphoea may occuras circumscribed plaques of sclerotic skin or as amore generalised, symmetric process.

Raynaud's phenomenon is rare among patientswith localised scleroderma, unlike those withsystemic sclerosis. Antinuclear antibodies maybe positive in about 50% of patients withlocalised scleroderma, using the sensitive HEp-2 substrate.69 The nature of the nuclear anti-gen(s) with which serum samples from patients

with localised scleroderma react is unknown,but it differs from topoisomerase I and centro-mere antigens.70 Eosinophilia is more commonthan in systemic sclerosis (31% v 7%), but lesscommon than in eosinophilic fasciitis.7' Thedegree of blood eosinophilia may reflect diseaseactivity.The cause of localised scleroderma is un-

known. Much excitement was generated byreports suggesting a spirochaetal cause (Borreliaburgdorfern),72 but most subsequent investi-gations from other geographical locations havefailed to confirm this association.7374

Diffuse fasciitis with eosinophilia (eosinophilicfasciitis or Shulman's disease) is a syndromecharacterised by scleroderma-like skin changes,fasciitis, and blood eosinophilia. It falls withinthe spectrum of scleroderma, and histologicallyis similar in many respects to what has beendescribed as 'morphoea profunda'.75 Thecharacteristic histopathological features arethickening and hyalinisation of collagen bundlesin the deep dermis and of the septa in thesubcutis; an inflammatory cell infiltrate com-posed mainly of lymphocytes and sometimeseosinophils; and slight deposits of mucin. Suchchanges differ from conventional morphoeabecause of deeper involvement and more pro-nounced inflammatory cell infiltration.

Recently, an epidemic known as the eosino-philia-myalgia syndrome was reported, initiallyand predominantly from the United States,associated with the ingestion of the essentialamino acid L-tryptophan.76 This syndrome issimilar in some respects to sporadic cases ofeosinophilic fasciitis; in fact, retrospectivestudies have shown that a significant proportionof cases of eosinophilic fasciitis were associatedwith the ingestion of L-tryptophan. The eosino-philia-myalgia syndrome also resembles inmany respects the toxic oil syndrome, whichoccurred as an epidemic in Spain in 1981.Thus far the eosinophilia-myalgia syndrome hasaffected over 1500 subjects and has been linkedto more than 20 deaths.Many patients with eosinophilia-myalgia

syndrome have scleroderma-like changes affect-ing the skin and subcutaneous tissue, includingthe fascia and adjacent muscles,7 similar tothose described above in morphoea profunda oreosinophilic fasciitis. Unusual features of thissyndrome include an eosinophilic pneumonitis,hepatitis, and neuropathy in some patients. Thelast of these may be severe and debilitating,resulting in respiratory failure, and has beenresponsible for most of the deaths attributed tothe syndrome. A number of cases have also beendescribed with clinical and histopathologicalevidence of pulmonary hypertension.78 Pneu-monitis, pulmonary hypertension, and neuro-pathy were also prominent features of the toxicoil syndrome.

Several cases of a similar illness were reportedin patients receiving L-5-hydroxytryptophan fortreatment of neurological disorders,79 raisingthe question ofthe role oftryptophan metabolismin these and other fibrosing conditions, includ-ing systemic sclerosis. Biochemical studies oftryptophan metabolites have shown abnor-malities in these patients,7 79 as well as in

859

on June 25, 2020 by guest. Protected by copyright.

http://ard.bmj.com

/A

nn Rheum

Dis: first published as 10.1136/ard.50.S

uppl_4.854 on 1 Novem

ber 1991. Dow

nloaded from

860

patients in the active phase of eosinophilicfasciitis.80 These studies have shown increasedconcentrations of tryptophan metabolites,kynurenine and quinolinic acid, in the plasmaand urine. The pattern of increased kynurenineand quinolnic acid concentrations in patients inthe active phase of their illness is consistent withincreased activity of the rate limiting enzymein the kynurenine pathway, indoleamine-2,3-dioxygenase. Inflammatory mediators, such asendotoxin and interferon y, are known tostimulate the activity of indoleamine-2,3-dioxygenase. It remains to be determinedwhether endotoxin or interferon y, or both, playa part in the biochemical abnormalities andpathogenesis of the current eosinophilia-myalgiasyndrome epidemic. Further studies of trypto-phan metabolism in patients with eosinophilia-myalgia syndrome or eosinophilic fasciitis mayprovide information which is applicable to thepathogenesis of systemic sclerosis.The eosinophilia and oedematous skin

changes, as well as the pneumonitis, associatedwith eosinophilia-myalgia syndrome generallyrespond well to corticosteroid treatment.Neuropathy and pulmonary hypertension maypersist, and chronic skin changes are now beingseen. The natural history of this new epidemicremains to be determined.

Childhood scierodermaScleroderma is rare in childhood, and theclinical presentation is even more varied than inadult life.8" Juvenile onset scleroderma maytake the form of localised scleroderma (linear ormorphoea), including eosinophilic fasciitis,82 orit may occur as systemic sclerosis. In the lattercase Raynaud's phenomenon is commonly thepresenting symptom. Skin involvement may beclassified as localised or diffuse, as with adultonset scleroderma. Visceral organ involvementis similar to the adult form, with oesophagealdysmotility being common, and with cardiacand pulmonary disease primarily responsible formorbidity and mortality. With an HEp-2 cellsubstrate the prevalence of antinuclear anti-bodies has been shown to be similar to that inadult scleroderma: in 100% of children withsystemic sclerosis and in 67% of children withlocalised scleroderma.8' Anticentromere anti-body has been demonstrated in the rare case ofjuvenile onset CREST syndrome.83

I Silver R M, LeRoy E C. Systemic sclerosis (scieroderma). In:Samter M, Talmage D W, Frank M M, Austen K F,Claman H N, eds. Immunological diseases. 4th ed. Boston:Little, Brown, 1988: 1459-500.

2 Jayson M I V, Black CM. Systemic sclerosis: scleroderma. NewYork: Wiley, 1988.

3 Rocco V L, Hurd E R. Scleroderma and scleroderma-likedisorders. Semin Arthritis Rheum 1986; 16: 22-69.

4 LeRoy E C, Black C, Fleischmajer R, et al. Scleroderma(systemic sclerosis): classification, subsets and pathogenesis.Y Rheumatol 1988; 15: 202-5.

Masi A T. Classification of systemic sclerosis (scleroderma):relationship of cutaneous subgroups in early disease tooutcome and serologic reactivity. 7 Rhewnatol 1988; 15:894-8.

6 Maricq H R, Weinrich M C. Diagnosis of Raynaud'sphenomenon assisted by color charts. J Rheumatoi 1988;15: 454-9.

7 Silver R M, Heyes M P, MaizeJ C, Quearry B, Vionnet-Fuasset M, Stemnberg E M. Scleroderma, fasciitis, andeosinophilia associated with the ingestion of tryptophan.N Engl J Med 1990; 322: 874-81.

8 Priollet P, Vayssairat M, Housset E. How to classify

Raynaud's phenomenon. Long-term follow-up study of 73cases. Am J Med 1987; 83: 494-9.

9 Maricq H R, WeinrichM C, Keil J E, LeRoy E C. Prevalenceof Raynaud's phenomenon in the general population. Apreliminary study by questionnaire. J Chronic Dis 1986; 39:423-7.

10 Harper F E, Maricq H R, Turner R E, Lidman R W, LeRoyE C. A prospective study of Raynaud's phenomenon andearly connective tissue disease. A five-year report. Am J7Med 1982; 72: 883-8.

11 Fitzgerald 0, Hess E V, O'Connor G T, Spencer-Green G.Prospective study of the evolution of Raynaud's pheno-menon. AmJ Med 1988; 84: 718-26.

12 Chen Z, Silver R, Ainsworth S K, Dobson R L, Rust P,Maricq H R. Association between fluorescent antinuclearantibodies, capillary patterns and clinical features inscleroderma spectrum disorders. Am J Med 1984; 77:812-22.

13 Kallenberg C G M, Wouda A A, Hoet M H, van VenrooijW J. Development of connective tissue disease in patientspresenting with Raynaud's phenomenon: a six year followup with emphasis on the predictive value of antinuclearantibodies as detected by immunoblotting. Ann Rheum Dis1988; 47: 634-41.

14 Wollersheim H, Thien Th, Hoet M H, van Venrooij W J.The diagnostic value of several immunological tests forantinuclear antibody in predicting the development ofconnective tissue disease in patients presenting withRaynaud's phenomenon. Eur J Clin Invest 1989; 19:535-41.

15 Lomeo R M, Cornella R J, Schabel S I, Silver R M.Progressive systemic sclerosis sine scleroderma presentingas pulmonary interstitial fibrosis. Am J Med 1989; 87:525-7.

16 Masi A T, Rodnan G P, Medsger T A Jr, et al. Preliminarycriteria for the classification of systemic sclerosis (sclero-derma). Arthritis Rhewu 1980; 23: 581-90.

17 Zaninotto G, Peserico A, Costanini M, et al. Oesophagealmotility and lower oesophageal sphincter competence inprogressive systemic sclerosis and localized scleroderma.ScandJ Gastroenterol 1989; 24: 95-102.

18 Hendel L, Worning H. Exocrine pancreatic function inpatients with progressive systemic sclerosis. Scand JGastroenterol 1989; 24: 461-6.

19 Harrison N K, Glanville AR, Strickland B, et al. Pulmonaryinvolvement in systemic sclerosis: the detection of earlychanges by thin section CT scan, bronchoalveolar lavageand 99mTc-DTPA clearance. Respir Med 1989; 83: 403-14.

20 Furst D E, Davis J A, Clements P J, Chopra S K,Theofilopoulos A N, Chia D. Abnormalities of pulmonaryvascular dynamics and inflammation in early progressivesystemic sclerosis. Arthritis Rheum 1981; 24: 1403-8.

21 Wallaaert B, Hatron P Y, Grosbois J M, Tonnel A B,Devulder B, Voison C. Subclinical pulmonary involvementin collagen-vascular diseases assessed by bronchoalveolarlavage. Relationship between alveolitis and subsequentchange in lung function. Am Rev Respir Dis 1986; 133:574-80.

22 Konig G, Lunderschmidt C, Hammer C, Adelmann-GrillB C, Braun-Falco 0, Fruhmann G. Lung involvement inscieroderma. Chest 1984; 85: 318-24.

23 Silver R M, Metcalf J M, Stanley J H, LeRoy E C. Interstitiallung disease in scleroderma. Analysis by bronchoalveolarlavage. Arthritis Rheum 1984; 27: 1254-62.

24 Silver R M, Metacalf J M, LeRoy E C. Interstitial lungdisease in scleroderma: immune complex in sera andbronchoalveolar lavage fluid. Arthritis Rheum 1986; 29:525-31.

25 Owens G R, Paradis I L, Gryzan S, et al. Role otinflammation in the lung disease of systemic sclerosis:comparison with idiopathic pulmonary fibrosis. J Lab ClinMed 1986; 107: 253-60.

26 Rossi G A, Bitterman PB, Rennard S I, Ferrans VJ, CrystalR G. Evidence for chronic inflammation as a component ofthe interstitial lung disease associated with progressivesystemic sclerosis. Am Rev Respir Dis 1985; 131: 612-7.

27 Silver R M, Miller KS, Smith E A, Kinsella MB, Schabel SI. Evaluation and management of scleroderma lung diseaseusing bronchoalveolar lavage. Am J Med 1990; 88: 470-6.

28 Kinsella M B, Smith E A, Miller KS, LeRoy E C, Silver RM. Spontaneous production of fibronectin by sclerodermaalveolar macrophages. Arthritis Rheum 1989; 32: 577-83.

29 Al-Sabbagh M R, Steen V D, Zee B C, et al. Pulmonaryarterial histology and morphometry in systemic sclerosis: acase-control autopsy study.YRheumatol 1989; 16: 1038-42.

30 Cannon P J, Hassar M, Case D B, Casarella W J, Sommers SC, LeRoy E C. The relationship of hypertension and renalfailure in scleroderma (progressive systemic sclerosis) tostructural and functional abnormalities of the renal corticalcirculation. Medicine (Baltimore) 1974; 53: 1-4.

31 Steen V D, Medsger T A Jr, Osial T A, Jr, Ziegler G L,Shapiro A P, Rodnan G P. Factors predicting developmentof renal involvement in progressive systemic sclerosis. AmJMed 1984; 76: 779-86.

32 Steen V D, Medsger T A Jr. Outcome of scleroderma renalcrisis in the pre- and post-captopril eras (abstract). ArthritisRhewn 1988; 31: S21.

33 Waeber B, Schaller M-D, Wauters J-P, Brunner H R.Deterioration of renal function in hypertensive patientswith scleroderma despite blood pressure normalization withcaptopril. Klin Wochenschr 1984; 62: 728-30.

34 Spooner M S, LeRoy E C. The changing face of severescleroderma in five patients. Clin Exp Rheumatol 1990; 8:101-5.

Silver

on June 25, 2020 by guest. Protected by copyright.

http://ard.bmj.com

/A

nn Rheum

Dis: first published as 10.1136/ard.50.S

uppl_4.854 on 1 Novem

ber 1991. Dow

nloaded from

Clinical aspects ofsystemic sclerosis (scleroderma)

35 Helfrich D J, Banner B, Steen V D, Medsger T A Jr.Normotensive renal failure in systemic sclerosis. ArthritisRheum 1989; 32: 1128-34.

36 Jones B, Trevillian P, Lake N. Continuous ambulatoryperitoneal dialysis in scieroderma. Nephron 1990; 54:107-8.

37 Weiss S, Stead E A Jr, Warren J V, Bailey 0 T. Scierodermaheart disease. With a consideration of certain other visceralmanifestations. Arch Intern Med 1943; 71: 749-76.

38 Ellis W W, Baer A N, Robertson R M, Pincus T,Kronenberg M W. Left ventricular dysfunction induced bycold exposure in patients with systemic sclerosis. AmJ Med1986; 80: 385-92.

39 Alexander E L, Firestein G S, Weiss J L, et al. Reversiblecold-induced abnormalities in myocardial perfusion andfunction in systemic sclerosis. Ann Intern Med 1986; 105:661-8.

40 Nitenberg A, Foult J M, Kahan A, et al. Reduced coronaryflow and resistance reserve in primary sclerodermamyocardial disease. Am HeartJf 1986; 112: 309-15.

41 Kahan A, Devaux J Y, Amor B, et al. Nifedipine andthallium-201 myocardial perfusion in progressive systemicsclerosis. N Englj Med 1986; 314: 1397-402.

42 McGregor A R, Watson A, Yunis E, et al. Familial clusteringof scleroderma spectrum disease. Am J Med 1988; 84:1023-32.

43 Takehara K, Moroi Y, Ishibashi Y. Antinuclear antibodies inthe relatives of patients with systemic sclerosis. BrDermatol 1985; 112: 23-33.

44 Black C M, Welsh K I, Maddison P J, Jayson M I V,Bernstein R M. HLA antigens, autoantibodies and clinicalsubsets in scleroderma. BrJ Rheumatol 1984; 23: 267-71.

45 Rittner G, Schwanitz G, Baur M P, et al. Family studies inscleroderma (systemic sclerosis) demonstrating an HLA-linked increased chromosomal breakage rate in culturedlymphocytes. Hum Genet 1988; 81: 64-70.

46 Black C M, Welsh K I, Walker A E, et al. Geneticsusceptibility to scleroderma-like syndrome induced byvinyl chloride. Lancet 1983; i: 53-5.

47 LeRoy E C. Connective tissue synthesis by scleroderma skinfibroblasts in cell culture. Exp Med 1972; 135: 1351-62.

48 Kahari V-M, Multimaki P, Vuorio E. Elevated pro-alpha-2(I) collagen mRNA levels in cultured scleroderma fibro-blasts results from an increased transcription rate of thecorresponding gene. FEBS Letters 1987; 215: 331-4.

49 Kahari V-M, Vuorio E I. Increased half-lives of procollagenmRNAs may contribute to the elevated procoliagen mRNAlevels in cultured scleroderma fibroblasts. MedicalScience Research 1987; 15: 417-8.

50 Rossi P, Karsenty G, Roberts A B, Roche N S,deCrombrugghe B. A nuclear factor 1 binding site mediatesthe transcriptional activation of a type 1 collagen promoterby transforming growth factor B. Cell 1988; 52: 405-14.

51 Roberts A B, Sporn M B, Assoian R K, et al. Transforminggrowth factor type ,: rapid induction of fibrosis andangiogenesis in vivo and stimulation of collagen formationin vitro. Proc Natl Acad Sci USA 1986; 83: 4167-71.

52 Leof E B, Proper J A, Goustin A S, Shipley G D, DiCorletoP E, Moses H L. Induction of c-sis mRNA and activitysimilar to platelet-derived growth factor by transforminggrowth factor-beta: a proposed model for indirect mito-genesis involving autocrine activity. Proc Natl Acad SciUSA 1986; 83: 2453-7.

53 Takehara K, LeRoy E C, Grotendorst G R. TGF-4 inhibitionof endothelial cell proliferation: alteration of EGF bindingand EGF-induced growth-regulatory (competence) geneexpression. Cell 1987; 49: 415-22.

54 LeRoy E C, Smith E A, Kahaleh M B, Trojanowska M,Silver R M. A strategy for scleroderma (systemic sclerosis):Is transforming growth factor beta the answer? ArthritisRheum 1989; 32: 817-25.

55 Kahaleh M B, Osborn I, LeRoy E C. Increased factorVIII/von Willebrand factor antigen and von Willebrandfactor activity in scleroderma and in Raynaud's pheno-menon. Ann Intern Med 1981; 94: 482-4.

56 Sato N, Goto T, Haranaka K, et al. Actions of tumor necrosisfactor on cultured vascular endothelial cells: morphologicmodulation, growth inhibition, and cytotoxicity. NatdCancer Inst 1986; 76: 1113-21.

57 Earnshaw W, Bordwell B, Marino C, Rothfield N. Threehuman chromosomal autoantigens are recognized by serafrom patients with anti-centromere antibodies. Clin Invest1986; 77: 426-30.

58 Shero J H, Bordwell B, Rothfield N G, Earnshaw W C. Hightiters of autoantibodies to topoisomerase I (Scl-70) in serafrom scleroderma patients. Science 1986; 231: 737-40.

59 Reimer G, Rose K M, Scheer L I, Tan E M. Autoantibody toRNA polymerase I in scleroderma sera. Clin Invest 1987;79: 65-72.

60 Douvas A. Does Scl-70 modulate collagen production insystemic sclerosis? Lancet 1988, ii: 475-7.

61 Senaldi G, Lupoli S, Vergani D, Black C M. Activation of thecomplement system in systemic sclerosis. Relationship toclinical severity. Arthritis Rheum 1989; 32: 1262-7.

62 Hawkins R A, Claman H N, Clark R A F, Steigerwald J C.Increased dermal mast cell populations in progressivesystemic sclerosis: A link in chronic fibrosis? Ann InternMed 1985; 102: 182-6.

63 Walker M A, Harley R A, LeRoy E C. Inhibition of fibrosisin TSK mice by blocking mast cell degranulation.J Rheumatol 1987; 14: 299-301.

64 Walker M, Harley R, LeRoy E C. Ketotifen prevents skinfibrosis in the tight skin mouse. Rheumatol 1990; 17:57-9.

65 Duncan M R, Berman B. Persistence of a reduced collagenproducing phenotype in cultured scleroderma fibroblastsafter short term exposure to interferons. Clin Invest 1987;79: 1318-24.

66 Kahari V M, Heino J, Vuorio T, Vuorio E. Interferon andinterferongamma reduce excessive synthesisand procollagenmRNA levels of scleroderma fibroblasts in culture. BiochimBiophys Acta 1988; 968: 4550.

67 Sandborg C I, Berman M A, Andrews B S, Mirick G R, FriouG J. Increased production of an interleukin 1 (ILl)inhibitor with fibroblast stimulating activity by mono-nuclear cells from patients with scleroderma. Clin ExpImmunol 1986; 66: 312-9.

68 Kahaleh M B, Smith E A, Soma Y, LeRoy E C. Effect oflymphotoxin and tumor necrosis factor on endothelial andconnective tissue cell growth and function. Clin ImmunolImnunopathol 1988; 49: 261-72.

69 Falanga V, Medsger TA Jr, Reichlin M, Rodnan G P. Linearscleroderma. Clinical spectrum, prognosis, and laboratoryabnormalities. Ann Intern Med 1986; 104: 849-57.

70 Kikuchi K, Takehara K, Ishibashi Y. Antinuclear antibodiesin localized scleroderma: unique staining in chromosomespreads. Am Acad Dermatol 1989; 21: 1301-3.

71 Falanga V, Medsger T A Jr. Frequency, levels, and signifi-cance of blood eosinophilia in systemic sclerosis, localizedscleroderma, and eosinophilic fasciitis. Am Acad Dermatol1987; 17: 648-56.

72 Aberer E, Stanek G, Ertl M, Neumann R. Evidence forspirochetal origin of circumscribed scleroderma (morphea).Acta Derm Venereol (Stockh) 1987; 67: 225-3 1.

73 Lecerf V, Bagot M, Revuz J, Touraine R, Dournon E.Borrelia burgdorferi and localized scleroderma. ArchDermatol 1989; 125: 297.

74 Garioch J J, Rashid A, Thomson J, Seywright M. Therelevance of elevated Borrelia burgdorferi titres in localizedscleroderma. Clin Exp Dermatol 1989; 14: 439-41.

75 Su W P D, Person J R. Morphea profunda. A new conceptand a histopathologic study of 23 cases. AmJ' Dermatopathol1981; 3: 25140.

76 Anonymous. Eosinophilia myalgia syndrome New Mexico.MMWR 1989; 38: 765-7.

77 Toxic epidemic syndrome study group. Toxic epidemicsyndrome, Spain, 1981. Lancet 1982; ii: 697-702.

78 Tazelaar H D, Myers J L, Drage C W, King Jr T E, AguayoS, Colby T V. Pulmonary disease associated with L-tryptophan-induced eosinophilic myalgia syndrome.Clinical and pathologic features. Chest 1990; 97: 1032-6.

79 Stemnberg E M, Van Woert M H, Young S N, et al.Development of a scleroderma-like illness during therapywith L-5-hydroxytryptophan and carbidopa. N EnglJ Med1980; 303: 782-7.

80 Stachow A, Jablonska S, Kencka D. Tryptophan metabolismin scleroderma and eosinophilic fasciitis. In: Black C M,Myers A R, eds. Systemic sclerosis (scleroderma). New York:Gower Medical, 1985: 1304.

81 Bernstein R M, Pereira R S, Holden A J, Black CM, HowardA, Ansell B M. Autoantibodies in childhood scleroderma.Ann Rheum Dis 1985; 44: 503-6.

82 Grisanti M W, Moore T L, Osborn T G, Haber P L.Eosinophilic fasciitis in children. Semin Arthritis Rheum1989; 19: 151-7.

83 Burge S M, Ryan T J, Dawber R P R. Juvenile onset systemicsclerosis. J7 R Soc Med 1984; 77: 7934.

861

on June 25, 2020 by guest. Protected by copyright.

http://ard.bmj.com

/A

nn Rheum

Dis: first published as 10.1136/ard.50.S

uppl_4.854 on 1 Novem

ber 1991. Dow

nloaded from