BACKGROUND: Cutaneous lymph node tuberculosis (CLTb) represents 25.4% of all cases of extra-pulmonary Tb in the state of Amazonas. The current methods of diagnose including bacteriological and histopathological assays involve some technical difficulties, and the polymerase chain reaction - PCR arise as an alternative method allowing specific results faster than the others. In this context the accuracy of PCR4 Marchetti et al. protocol was compared with traditional methods. MATERIAL AND METHOD: Nested-PCR for IS6110 (123 pb) were applied on 83 CLTb suspicious formalin fixed and paraffin embedded samples of tissues (52 cutaneous and 31 lymph node), obtained from 1997 to 2002. All cases were evaluated by bacteriological and histopathological methods. Accuracy analyses were carried out between the PCR amplification results and those related on bacteriological and histopathological methods. RESULTS AND DISCUSSIONS: Positive results of PCR4 were about 50.6% (59.6% in cutaneous samples and of 35.5% in lymph nodes samples). In both groups were observed false-negative and false-positive results. Some hypotheses that explain those results are related to the presence of IS6110 in environmental mycobacterias in the Amazon region and the absence of standardized DNA concentration to amplification assays. CONCLUSIONS: The proposed protocol was as positive as others ones available in the literature. Definitive Tb diagnostic can be obtained on lymph node paraffin embedded PCR in association with bacteriological or histopathological method. A better accuracy of an amplification assay applied to cutaneous Tb suspicious lesions has to be still under research.
Cutaneous tuberculosis; formalin-fixed and paraffin-embedded tissues; Mycobacterium tuberculosis; PCR; IS6110
