Clinical and biological factors predictive of malignancy in dermatomyositis

Editor Although the paraneoplastic nature of a subset of dermatomyositis (DM) cases is unquestionable, the factors that indicate the coexisting cancer still remain unclear. Several predictive signs have been postulated as a marker of underlying malignancy: older age, male gender, rapid onset of the disease, presence of cutaneous necrosis, increased erythrocyte sedimentation rate (ESR), and increase or normality of creatine phosphokinase (CPK).1–8 In order to identify potential risk factors for associated cancer in patients with DM, we reviewed clinical and laboratory data of 32 patients with DM (17 females and 15 males, aged 11–78 years), who have been treated during a 23-year period (1985–2007). Diagnosis of DM was based on the criteria of Bohan and Peter.9 Typical cutaneous signs and muscular involvement (proximal muscle weakness and/or elevated muscle enzymes and/or electromyography findings and/or muscle histology) were observed in all patients. The main recorded data included an association with a cancer, age at the time of the diagnosis, gender, a rapid onset of symptoms (considered if the diagnosis was made within 3 months after the appearance of initial symptoms), signs of severity (presence of dysphonia, dyspnoea and/or dysphagia), some clinical features such as cutaneous necrosis (defined as cutaneous and/or mucosal necrotic lesions or ulcerations) and periungual erythema, evaluated ESR (superior to 40 mm during the first hour), serum muscle enzymes levels – CPK, lactate dehydrogenase (LDH), aspartate aminotransferase (AST) and alanine aminotransferase (ALT) – as well as presence of antinuclear autoantibodies. Statistical analysis using Fisher’s exact test for qualitative data and Mann-Whitney’s test for quantitative data were performed to investigate differences between patients with and without associated malignancy. The difference was considered significant at P < 0.05. Malignant tumours were detected in 8 patients (25%), with equal number of female and male patients. Malignancies included colon cancer (2 patients), ovarian cancer (2 patients) and the remaining cancers were those of lung, breast, pancreas and prostate. The prevalence of predicting factors of malignancy is listed in Table 1. Cutaneous necrosis was presenting sign in 76% of our patients with cancer and in only 8.3% of the patients without cancer. Some previously published studies have pointed out that cutaneous necrosis is highly predictive of an associated cancer.3–6 Including our trial, cutaneous necrosis is thought to increase the probability of occult malignancies in 27 (61%) of cases associated with cancer, opposite to 7 (8%) cases of DM without cancer. Since there is no clear definition for the term ‘cutaneous necrosis’, as this skin sign comprises a wide range of symptoms, from small digital necrosis to extensive skin and mucosal necrosis, these results, as pointed out by Burnouf et al.,4 must be analysed with reserve. In our study, elevated muscle enzyme levels were also associated with underlying malignancy, especially elevated level of CPK. The validity of this criterion has been confirmed by most formerly published trials1,4 but is in contrary to some studies that normal muscle enzyme levels tend to be a risk factor in developing cancer.7,8 The observations that the same antigens are expressed at high levels in several cancers known to be associated with the development of inflammatory myopathy, but not in corresponding normal tissue, suggest that the link between malignancy and myositis relates to the expression of common autoantigens between cancer tissue and muscle tissue in patients with DM.10 We can confirm that factors predictive of concomitant malignancy are the presence of cutaneous necrosis and elevation of the muscular enzymes. These parameters, which are easy to evaluate by clinicians, justify the inclusion of an extensive malignancy search that should be orientated by the most frequent localization of cancer.