Patients and methods

Ascertainment

We retrospectively analysed data from SSc patients with myopathy followed between 1995 and 2005 in four French centres (three internal medicine and one rheumatology departments). All patients fulfilled the American College of Rheumatology10 and/or Leroy and Medsger11 criteria for the diagnosis of SSc and had a history of myopathy explored by muscle biopsy. Myopathy was defined as the presence of muscle weakness, myalgia, or creatine kinase (CK) greater than 5 N (upper normal range), together with evidence of muscle involvement on electromyography (low voltage and/or short duration potential during maximal contraction, fibrillation or sharp wave) or on muscle biopsy.

SSc-related features

The main SSc-related features were recorded together with antinuclear, anti-DNA, anticentromere, antitopoisomerase 1, antipolymyositis/scleroderma, antiribonucleoprotein, anti-Sjogrens’ syndrome A and Sjogrens’ syndrome B, anti-Jo1 and anti-Ku antibodies.

Myopathy evaluation

The time from SSc diagnosis to the onset of myopathy, the location and intensity of myalgias, Medical Research Council manual muscle testing, levels of serum muscle enzymes (CK and aldolase) and results of any electromyographic examination or muscle magnetic resonance imaging (MRI) were recorded. The following myopathological criteria were assessed: myofibre atrophy and necrosis/regeneration; inflammation; fibrosis; microangiopathy (defined as capillary loss; lumen dilatation; thickening of capillary wall; and/or small patches of necrosis); vasculitis involving small arteries; mitochondrial abnormalities and neuropathic changes. The type, dose and duration of treatments were recorded.

Favourable muscle outcome was defined as the improvement of muscle grading (normalisation or improvement >1 point in at least two muscle groups) and myalgia intensity (normalisation or subjective significant decrease); or, if no muscle weakness or myalgia (two cases), normalisation or greater than 50% decrease in CK levels. Unfavourable outcome included all other conditions. Complete remission corresponded to the disappearance of all clinical and biological muscle symptoms. Favourable response to corticosteroid therapy was defined as the achievement of favourable muscle outcome within 3 months following the initiation of corticosteroids.

Statistical methods

Associations between myopathological features were tested by Fisher’s exact test. Factors associated with favourable muscle outcome and response to corticosteroids were assessed by logistic regression. Explanatory variables were the presence/absence of myalgia, muscle weakness, CK level greater than 5 N, abnormal electromyography and muscle MRI at myopathy diagnosis, presence/absence of the myopathological criteria described above. Multivariate analyses included all explanatory variables with a p value less than 0.1 on univariate analysis, using a stepwise procedure. Analyses were performed using R software (http://cran.r-project.org/, version 2.4.1).

Results

Thirty-five SSc patients with defined myopathy were enrolled (see supplemental table available online only). Thirty (86%) patients were women, 27 (77%) were white, 4 (11%) were Arabic and four (11%) were black. The median age at diagnosis of myopathy was 44 years (interquartile range (IQR) 30–52), the median time from SSc diagnosis was 5 years (IQR 2–8, range 0–23).

SSc-related features

Twenty-six (74%) patients had diffuse SSc; the median Rodnan skin score was 18 (IQR 12–34). SSc-related complications were: heart involvement (left ventricular ejection fraction <60% eight (23%); arrhythmia four (11%) or conductive abnormalities six (17%)); interstitial lung disease 23 (66%); pulmonary arterial hypertension nine (26%); oesophagus 23 (66%) or bowel four (11%) involvement; history of pericarditis four (11%), scleroderma renal crisis three (9%) and digital ulcers 17 (48%). The prevalences of autoantibodies were antinuclear 28 (82%), anti-DNA two (6%), anticentromere two (6%), antitopoisomerase 1 six (18%), antipolymyositis/scleroderma four (12%), antiribonucleoprotein five (14%), anti-Sjogrens’ syndrome A and Sjogrens’ syndrome B seven (20%), anti-Jo1 0/27 and anti-Ku one (9%).

Muscle features and evolution

Twenty-seven (77%) patients had muscle weakness, with proximal distribution in all; and 30 (86%) had myalgia. Serum CK and aldolase levels were elevated in 27 (82%) and 16 (76%) cases, respectively. Among patients with electromyography results, 13 (93%) had a myopathic pattern and four (29%) a combined peripheral neuropathy. Twelve patients underwent muscle MRI; eight (67%) showed a high intensity signal of girdle muscles in T2 or short T1 inversion–recovery, with muscle atrophy and fatty infiltration in three (none having previously received corticosteroids); the others had no detectable abnormality.

Myopathological features were: mononuclear inflammation 22 (63%), localised in septal/perimysia (n  =  7), endomysial (n  =  5) and/or perivascular (n  =  11) areas; atrophy and irregular muscle fibre diameter 22 (63%); fibre necrosis 22 (63%); fibre regeneration 15 (44%); central nuclei eight (24%); endomysial and perimysial fibrosis eight (24%); microangiopathy 10 (29%); vasculitis involving small arteries three (9%) and mitochondrial abnormalities one (3%). No neuropathic change was observed. Hyperexpression of class I human leucocyte antigen molecules on myofibres surface was observed in 11/17 cases. Complement deposits were seen on the vascular walls of four of seven cases. Four inflammatory myopathies had available extensive immunostaining. In three, the infiltrate was mostly composed of CD4 T cells, B cells and complement deposits on vascular walls and in one, macrophages and CD8 T cells predominated. Except for a tight association between fibre regeneration and necrosis, no specific association was found between these pathological features. Four biopsies were normal; none of the corresponding patients underwent immunosuppressive therapy at the time of biopsy.

Treatment for myopathy was corticosteroids for 28 (80%) patients, at a high dose (1 mg/kg per day) for 24. Eleven (31%) patients received intravenous cyclophosphamide because of worsening interstitial lung disease. Nine patients received intravenous immunoglobulin, mostly because of failure (n  =  1) or relapse (n  =  7) after corticosteroids. Several other immunosuppressive drugs were used as maintenance therapy: methotrexate (n  =  8); azathioprine (n  =  8); mycophenolate mofetil (n  =  3) and cyclosporine (n  =  3). Overall, after a median follow-up of 4.5 years (IQR 2–5.5) from the onset of myopathy, favourable muscle outcome was achieved in 24 (69%) patients, including 14 complete remissions. Two patients developed renal crisis after high dose corticosteroids. Three SSc patients with myopathy died during follow-up, none of the deaths was due to the skeletal myopathy.

Factors associated with muscle outcome

No clinical and biological features at diagnosis could predict muscle outcome. Patients with abnormal muscle MRI results had a more favourable outcome, but the small sample size precludes further interpretation. Conversely, two histological parameters—inflammation and, to a lesser extent, necrosis—were significantly associated with muscle outcome (table 1). The percentage of favourable outcomes was 13% in patients without evidence of inflammation or necrosis, 89% with either inflammation or necrosis and 100% with both inflammation and necrosis.

Response to corticosteroid therapy was associated with no bioclinical parameter but a CK level greater than 5 N—on univariate analysis only, whereas myopathological inflammation was strongly associated with a favourable response in multivariate analysis (table 2).

Discussion

As reported in the literature,7 we observed a broad spectrum of histological patterns in SSc-associated myopathies. Approximately two-thirds of biopsies had at least one classic feature of inflammatory myopathy (mononuclear inflammation, necrosis, regeneration or atrophy),12 associated or not with more specific features (fibrosis, microangiopathy). No association was found between the different variables, except for muscle fibre necrosis and regeneration. Inflammatory infiltrates were largely composed of mononuclear cells and were more prominent in the perivascular area, as previously described.3,13 Extensive characterisation of these infiltrates has seldom been reported. In one study, the inflammatory cells were predominantly CD8 cells, with few CD4 and B cells and no complement deposits, consistent with a cell-mediated immune response.14 Conversely, in three-quarters of our patients, as in Evans et al,15 the immunopathological pattern was evocative of a humoral immune process. Further studies are therefore needed to characterise better the immunopathological processes involved in SSc-related myopathy.

Bioclinical and electromyographic features of SSc-related myopathies are indistinguishable from those of polymyositis,6 although severe muscle weakness5 and high muscle enzyme levels3,4 seem less frequent. We did not observe any significant association between the bioclinical presentation of myopathy and muscle outcome upon multivariate analysis. In contrast, we found a strong impact of myopathological inflammation and necrosis on muscle prognosis: 100% of patients with both features had a favourable outcome versus 14% only in patients with neither inflammation nor necrosis.

There are no therapeutic guidelines for SSc-related myopathy, but in our study corticosteroids was the most prescribed treatment, usually at a high dose, regardless of muscle pathology. In particular, eight out of 13 patients without histological evidence of inflammatory myopathy received corticosteroids and responded poorly (38% favourable response). By contrast, a 90% favourable response was observed in patients with myopathological inflammation. Unfortunately, the absence of treatment randomisation did not allow us to assess the benefits of high-dose corticosteroids in inflammatory myopathies associated with SSc.

Because of the retrospective study design and small sample size, our results need to be confirmed by a prospective assay with systematic screening of muscle involvement in SSc patients. Nevertheless, we suggest that muscle biopsy is critical for the therapeutic management of SSc-related myopathy. In these patients, at risk of the development of renal crisis, we recommend not to use high-dose corticosteroids in the absence of histologically documented inflammatory myopathies, because non-inflammatory myopathies are unlikely to improve with this treatment.