We read with interest the study by Umapathy et al. (2006), which incorporated a large number of patients with secondary or reinfection cutaneous tuberculosis (CT), and recorded clinical, bacteriological and histopathological observations as well as the response to anti-tubercular therapy (ATT). In Table 1, which depicts the Mantoux test results, some of the percentages may be incorrect, and the computation is unclear (e.g. the total number of patients in the 1st row is 178 rather than 182).

A particularly interesting aspect of the study was the isolation of non-tuberculous mycobacteria (NTM) in 20 patients. We were interested in knowing whether their clinical and histopathological findings, Mantoux test results, drug sensitivity pattern or response to therapy differed from those of Mycobacterium tuberculosis infections. We have observed that patients with such infections show a positivity in the lower range, i.e. around 10 mm, as do those suffering from dermatoses that mimic CT, e.g. lichen planus hypertrophicus and fixed sporotrichosis, thus alerting the clinician to the possibility that the condition may be something other than CT. About 10% of patients with a typical clinical picture of CT show Mantoux readings < 10 mm. Conversely, a high Mantoux reading in a patient without a typical clinical picture of CT does not necessarily indicate a tuberculous aetiology. In unusual situations histopathology, culture for microorganisms and polymerase chain reaction should be done to arrive at the right diagnosis. One of our patients had a verrucous lesion on the dorsum of the hand and a Mantoux <10 mm; he did not respond to ATT and on culture Mycobacterium marinum was isolated. Other patients diagnosed with nocardiosis had taken ATT as the clinical picture strongly suggested tubercular infection, and needed to be reconsidered when they failed to respond.

Irrespective of the drug sensitivity pattern, the authors treated all patients, including those in whom NTM were isolated and those showing resistance to one or more drugs, with rifampicin and isoniazid (INH) for 9 months. Clinical resolution was similar in all; lesions resolved in no more than 9 months. We observed that patients with widespread scrofuloderma required 10 months of therapy with rifampicin and INH, whereas patients with localized disease required 9 months (Ramesh et al. 1991).

It is also unclear why therapy was continued unchanged in patients who were resistant to one of the administered drugs. According to recent CDC guidelines, extrapulmonary tuberculosis should be treated with at least three drugs to prevent the emergence of resistance. Similarly, we were wondering why second-line drugs were not added to the treatment regimen for the two patients whose mycobacteria were resistant to both INH and rifampicin (Nanda et al. 2003).

A therapeutic test of diagnosis is often conducted in difficult cases of suspected CT. Nearly all patients show some improvement within a month, so if they do not get betterment after 6 weeks of ATT the therapeutic trial should be abandoned (Ramam et al. 2005). Did the authors make similar observations and did their findings with the two-drug regimen concur with earlier reports?


  1. Top of page
  2. References