POLYMYALGIA RHEUMATICA AND MALIGNANCY: THE QUESTION REMAINS OPEN

Abstract

To the Editor: Recently Martin et al. reported the case of an elderly woman with hematological malignancy associated with polymyalgia rheumatica (PMR).1 They concluded by questioning a true paraneoplastic relationship between both diseases and postulated a chance association. An elderly man who presented with PMR several months before diagnosis of colon carcinoma is described. The PMR symptoms were resolved after colectomy. A 69-year-old man presented with a 5-week history of pain and moderate stiffness in his shoulder and hip girdles, limitation of mobility, and fever. There was no history of headache, visual changes, jaw claudication, anorexia, or weight loss. Medical history was unremarkable. Physical examination revealed a temperature of 37.9°C but no signs of temporal arteritis, synovitis, muscle weakness, or atrophy. Laboratory data showed erythrocyte sedimentation rate (ESR) of 92 mm/h, C-reactive protein of 14 mg/dL, and hemoglobin of 11.9 g/dL. A temporal artery biopsy was negative. The patient, who was diagnosed with PMR, was prescribed 25 mg of prednisone daily. One week later, he became afebrile, and pain and stiffness had significantly decreased, but ESR remained persistently raised and PMR kept recurring with dose reduction. Ten months after discharge, the patient presented with sanguinolent diarrhea. Examination showed moderate pain to pressure on lower left quadrant. Laboratory data showed hemoglobin level of 6.8 g/dL and ESR of 60 mm/h. A colonoscopy and barium enema revealed a neoplastic lesion in the cecum with histological study corresponding to adenocarcinoma. There were no metastases. A colectomy with radiotherapy was performed. Corticosteroid therapy was stopped, and disease remission was achieved. Four years later, he is doing well. Classical PMR is an idiopathic disorder affecting mostly elderly people. Occasionally the diagnosis may be a challenge because PMR appears associated with malignancy.2, 3 This association is known as paraneoplastic PMR, and its reported prevalence is low.4 Some problems may be found in diagnosis of PMR, mainly nonspecific presentation and lack of definitive diagnostic criteria.5 Also, a large variety of tumors could appear as paraneoplastic rheumatic syndromes. Paraneoplastic PMR has to be differentiated from atypical PMR, which complicates metastatic cancer.6 Whether both diseases occur by chance or have a cause-effect relationship has long been a subject of discussion. Some authors postulate that a chance occurrence based on these disorders appears mainly in elderly patients and that no increased risk of malignancy has been found in prospective and retrospective studies,2, 4, 7, 8 but a possible bias of prior selection due to exclusion of disorders presenting with PMR-like symptoms such as malignant neoplasm has to be taken in account.9 This may explain the low prevalence of association. To overcome this problem, a prospective study with no patient selection was begun.9 A highly significant difference between the observed incidence of cancer in the PMR group and the age- and sex-adjusted prevalence of malignancy in the population was seen. In addition, a hazard rate for developing malignancy in patients with positive biopsy temporal arteritis, a disorder commonly related to PMR, was 2.35 times higher than in the controls.7 Strong evidence of a true causal relationship has been suggested in some cases in which PMR resolution was achieved after treatment of neoplasia.3, 10 Thus, it may be hypothesized that, in this reported patient, PMR was a paraneoplastic manifestation of his underlying carcinoma, and it is possible that both diseases were pathogenetically related. Because the long discussion about “chance or causality” appears far from finished, to opt for one of the two hypotheses was unremarkable. What is important is the development of an appropriate index of suspicion in the presence of an incomplete response to treatment, and further evaluation for paraneoplastic PMR should be considered. Although Martin et al. believe that an answer has been found,1 the question is open.