OBJECTIVE: Miliary tuberculosis (MTB) is difficult to diagnose. When prompt diagnosis is necessary, the polymerase chain reaction (PCR) to detect mycobacterial DNA may be valuable. SETTING: Tuberculosis clinic in an academic tertiary-level hospital in Mexico. DESIGN: Bone marrow (BM) aspiration samples from 30 consecutive clinically suspected MTB patients and 58 non-tuberculosis hematologic patients were evaluated by in-house PCR using a fragment of the insertion sequence IS6110; results were compared with those obtained by acid-fast-stained smears, culture in Löwenstein-Jensen medium, histology, and serology. RESULTS: Tuberculosis diagnosis was confirmed in all MTB suspects, 28 by microscopy and culture in pulmonary or extra-pulmonary samples other than BM, and two by clinical and radiologic improvement after antituberculosis treatment. In fresh BM specimens, in-house PCR was positive in 21/30 (70%) suspects, contrasting with only one positive (3.3%) in staining and culture, and four with compatible histologic findings (13.3%). BM samples from the control group showed negative results in bacteriologic and histologic studies, except in nine who had positive PCR results. These nine control cases had malignant processes. CONCLUSION: PCR in aspirates of BM is a useful diagnostic assay in cases of MTB, mainly when bacteriological results are negative.
OBJECTIVE:Miliary tuberculosis (MTB) is difficult to diagnose. When prompt diagnosis is necessary, the polymerase chain reaction (PCR) to detect mycobacterial DNA may be valuable. SETTING:Tuberculosis clinic in an academic tertiary-level hospital in Mexico. DESIGN: Bone marrow (BM) aspiration samples from 30 consecutive clinically suspected MTB patients and 58 non-tuberculosis hematologicpatients were evaluated by in-house PCR using a fragment of the insertion sequence IS6110; results were compared with those obtained by acid-fast-stained smears, culture in Löwenstein-Jensen medium, histology, and serology. RESULTS:Tuberculosis diagnosis was confirmed in all MTB suspects, 28 by microscopy and culture in pulmonary or extra-pulmonary samples other than BM, and two by clinical and radiologic improvement after antituberculosis treatment. In fresh BM specimens, in-house PCR was positive in 21/30 (70%) suspects, contrasting with only one positive (3.3%) in staining and culture, and four with compatible histologic findings (13.3%). BM samples from the control group showed negative results in bacteriologic and histologic studies, except in nine who had positive PCR results. These nine control cases had malignant processes. CONCLUSION: PCR in aspirates of BM is a useful diagnostic assay in cases of MTB, mainly when bacteriological results are negative.