Background  It is difficult to demonstrate Mycobacterium tuberculosis in smears or biopsies and to grow it in culture in cutaneous tuberculosis because most cases are paucibacillary. A therapeutic trial of antitubercular drugs is frequently used to confirm the diagnosis in difficult cases. Information is lacking on the response to antitubercular therapy in cutaneous tuberculosis; consequently there are no clear guidelines on when to expect a response and also when to abandon a therapeutic trial.

Methods  We studied the records of 60 patients treated for cutaneous tuberculosis at our hospital to study the time course of the therapeutic response. All patients were treated with a short-course antitubercular regimen consisting of isoniazid 300 mg daily, rifampicin 450 mg daily, ethambutol 800 mg daily and pyrazinamide 1500 mg daily for 2 months followed by isoniazid and rifampicin in the same doses for 4 months. At follow-up visits, each patient was assessed by a dermatologist who recorded the presence or absence of clinical improvement in the skin lesions.

Results  Of the 60 patients seen, eight patients did not follow up after the initial consultation, 48 patients improved with treatment and four patients were classified as treatment failures. The timing of the first visit varied from 3 days to 15 months (median 27.5 days, mean 58.96 ± 94.50) after initiation of treatment. Twenty-one patients were recorded to have improved within the first month of therapy. Twenty-seven patients who first reported more than 30 days after initiation of treatment were found to have improved. Four patients failed to respond during follow up ranging from 3 to 17 months.

Conclusion  When a therapeutic trial is undertaken in cutaneous tuberculosis, 6 weeks of therapy with four drugs appears adequate to prove (or disprove) the diagnosis.