Use of viability PCR for detection of live Chlamydia trachomatis in clinical specimens.

Background: The current testing approach to diagnose Chlamydia trachomatis (CT) infection relies on nucleic acid amplification tests (NAATs). These tests are highly sensitive, but do not distinguish between active infection and residual bacterial nucleic acid which may remain after resolution of infection, or via cross-contamination. Better methods to assess the viability of CT detected in clinical samples would be useful in determining the relevance of CT detection in a variety of clinical settings. The goal of this study was to test viability PCR (vPCR) as a method to distinguish viable bacteria from non-viable CT.

Methods: The vPCR relies on a propidium monoazide dye (PMAxx), which intercalates into accessible DNA from dead organisms and prevents their detection in a PCR assay for the CT ompA gene. We used digital PCR to quantify absolute genome copy numbers from samples. We validated the vPCR approach using laboratory stocks of CT with known viability. Then, we tested total DNA, viable CT DNA, and culture results from 18 clinical vaginal specimens and 25 rectal clinical specimens, all of which had tested positive by NAAT.

Results: In laboratory stocks of CT, vPCR using defined ratios of heat-killed to live bacteria tracked closely with expected results. In vaginal clinical specimens, vPCR and total DNA results were correlated, though total DNA genomes outnumbered viable genomes by 2.2-52.6-fold more copies. As expected, vPCR detected more total genomes than culture results. Both vPCR and total DNA correlated with culture results (Spearman correlation R = 0.8425 for total DNA and 0.8056 for vPCR). Ten rectal NAAT positive specimens were negative by total DNA PCR, vPCR, and were negative or inconclusive by culture. Of the 6 rectal specimens that were culture positive, all were total DNA and vPCR positive. vPCR additionally detected viable bacterial DNA in 8 specimens which were NAAT + and culture negative, though levels were very low (mean 1,357 copies/ml).

Conclusions: vPCR is a fast and easy method to assess viability in clinical specimens and is more correlated with culture results than total DNA PCR. Inconsistent ratios between total DNA and vPCR results suggest that the amount of dead bacteria varies considerably in clinical specimens. Results from rectal specimens suggest that many NAAT positive specimens do not in fact represent live replicating bacteria, and likely result in significant overuse of unnecessary antibiotics.