Background: The diagnosis of pulmonary tuberculosis (PTB) is conventionally established by examination of three Ziehl-Neelsen stained smears; however, negative results do not preclude active TB. Since tubercle bacilli or their nucleic acids are also expected to be excreted through the kidneys, we assessed spot urine as a supplementary specimen for diagnosing PTB.
Methods: A total of 164 respiratory specimens (147 sputum, 15 bronchoalveolar lavage, and two gastric lavage) from 81 suspected PTB cases were prospectively collected and processed. A total of 112 non-TB controls were also included in the study. For three consecutive days, morning urine specimens were collected from all patients and controls, and were processed for culture by BACTEC MGIT 960 (mycobacteria growth indicator tube) and Lowenstein-Jensen methods and for PCR by amplifying a 441-bp fragment of the hsp65 gene (Mycobacterium genus-specific) and a 786-bp fragment of the cfp32 gene (TB complex-specific).
Results: Of the 81 patients suspected of having PTB, 46 (56.8%) were sputum culture-positive. Of these, 12 (26.1%) were also urine culture-positive for Mycobacterium tuberculosis. Of the 35 sputum culture-negative cases, three (8.6%) were urine culture-positive. The TB complex specific PCR (cfp32) was positive in 52.2% (24/46) of the bacteriologically-confirmed and 28.6% (10/35) of the bacteriologically-negative PTB patients. In none of the control subjects were urine culture or PCR found to be positive for M. tuberculosis.
Conclusions: Specific PCR and culture examination of spot urine samples from suspected PTB patients significantly improved the detection rate of PTB and should be encouraged in resource-limited settings and where multiple pulmonary specimens are not feasible.