Background
Tuberculosis is still a significant diagnostic and therapeutic challenge with high proportion of smear- and culture- negative incidences worldwide. The conventional diagnostic tests are time-consuming and have a low sensitivity. Digital PCR is a novel technology which can detect target sequences with relatively low abundance and obtain the absolute copy numbers of the targets.
Methods
We assessed the accuracy of dPCR in TB diagnosis using more than 250 specimens, and for the first time, we selected M.tuberculosis-specific IS1081 in addition to widely used IS6110 as the amplification targets for dPCR. The quantification of target DNA was calculated using QuantaSoft Version 1.7.4.0917 (BioRad), and SPSS version 13.0 software (SPSS Inc., Chicago, IL, USA) was used for statistical analyses.
Results
IS6110-dPCR was more sensitive than IS1081, with the sensitivity and specificity accounting for 40.6 and 93.4% respectively. When we classified the TB patients by personal factors for high copy number of M.tuberculosis derived DNA in plasma: bilateral TB, extrapulmonary TB and disseminated TB, the sensitivity of both IS6110- and IS1081- dPCR was the highest in patients with disseminated TB (IS6110, 100%; IS1081, 68.8%), while their sensitivity was a bit higher in patients with extrapulmonary TB (IS6110, 50.0%; IS1081, 39.3%) than that in bilateral TB (IS6110, 43.3%; IS1081, 33.3%). Compared with traditional TB diagnostic tests, joint detection IS6110 & IS1081-dPCR was not as sensitive as smear microscope or mycobacterial culture, but it was higher than IS6110 qPCR (p < 0.05) and was able to detect 47.4% of smear-negative TB patients.
Conclusion
Our study suggested that plasma IS6110-dPCR is a rapid, moderate accurate and less-invasive method to detect M.tuberculosis DNA in plasma of TB patients and IS6110 & IS1081-dPCR has a potential to aid diagnosis of smear-negative TB.