Retrospective study of Mycobacterium marinum skin infections

Although infection by Mycobacterium marinum is well reported in the literature, there have been few epidemiologic studies. The purpose of this study was to review the epidemiology of patients with cutaneous M. marinum infection over a 3-year period at the National Skin Centre in Singapore.

Patients with a diagnosis of cutaneous M. marinum infection, confirmed histologically, were collated from computerized data from 1995 to 1997. Thirty-eight patients were diagnosed as having cutaneous M. marinum infection based on history, and clinical and histologic features.

Out of the 38 cases of M. marinum infection, there were 30 men and eight women. The age range was 14-85 years (mean: 44.7 years). The duration of disease ranged from 1 to 132 months (mean: 19 months). Thirteen patients (34.2%) had fish rearing as a hobby and four patients (10.5%) had occupational exposure to fish. Twelve patients (31.5%) gave a history of trauma to the disease site. All patients had biopsies of the lesions. All showed infective granulomas/granulomatous inflammation on histology. Acid-fast bacilli were identified in five out of 38 patients (13.2%) and mycobacteria were isolated in one out of 35 patients (2.9%). Nineteen patients received treatment with cotrimoxazole-trimethoprim alone, three with minocycline alone, five with minocycline and cotrimoxazole-trimethoprim, seven with various combinations of drugs, one with excision, and three defaulted treatment. The duration of treatment ranged from 4 to 38 weeks (mean: 14.9 weeks). Twenty-six patients (68.4%) showed clinical improvement, two (5.3%) had no response, and 10 (26.3%) were lost to follow-up. None of the patients worsened with treatment. The follow-up period ranged from 1 to 20 months (mean: 6.8 months).

The diagnosis of cutaneous M. marinum infection is mainly clinical, with supporting evidence from histologic features and the response to therapy. Risk factors include a history of trauma and water/fish-related hobbies or occupations. There is a poor yield of positive isolates in our experience; however, empirical treatment usually produces a good clinical response. In future, the polymerase chain reaction (PCR) technique may become more widely available as a rapid, sensitive, and specific means of diagnosis.