Development and evaluation of a next-generation digital PCR diagnostic assay for ocular Chlamydia trachomatis infections

Abstract

Droplet digital PCR (ddPCR) is an emulsion PCR process that performs absolute quantitation of nucleic acids. We developed a ddPCR assay for Chlamydia trachomatis infections and found it to be accurate and precise. Using PCR mixtures containing plasmids engineered to include the PCR target sequences, we were able to quantify with a dynamic range between 0.07 and 3,160 targets/μl (r(2) = 0.9927) with >95% confidence. Using 1,509 clinical conjunctival swab samples from a population in which trachoma is endemic in Guinea Bissau, we evaluated the specificity and sensitivity of the quantitative ddPCR assay in diagnosing ocular C. trachomatis infections by comparing the performances of ddPCR and the Roche Amplicor CT/NG test. We defined ddPCR tests as positive when we had ≥95% confidence in a nonzero estimate of target load. The sensitivity of ddPCR against Amplicor was 73.3% (95% confidence interval [CI], 67.9 to 78.7%), and specificity was 99.1% (95% CI, 98.6 to 99.6%). Negative and positive predictive values were 94.6% (95% CI, 93.4 to 95.8%) and 94.5% (95% CI, 91.3 to 97.7%), respectively. Based on Amplicor CT/NG testing, the estimated population prevalence of C. trachomatis ocular infection was ∼17.5%. Receiver-operator curve analysis was used to select critical cutoff values for use in clinical settings in which a balance between higher sensitivity and specificity is required. We concluded that ddPCR is an effective diagnostic technology suitable for both research and clinical use in diagnosing ocular C. trachomatis infections.

Fig 1

Confocal photomicrograph of ddPCR droplets from a representative C. trachomatis positive sample post-PCR. A bright-field image of droplets is shown at the bottom left. C. trachomatis plasmid PCR-positive droplets are shown on the 6-carboxyfluorescein (FAM) (green) channel (top left), and human RNase P/MRP 30-kDa subunit gene-positive droplets are shown on the HEX (red) channel (top right). A composite of the bright-field, FAM, and HEX channels is shown at the bottom right. All droplets have noticeable baseline fluorescence on both channels. PCR-positive droplets fluoresce with much greater intensity than template-negative droplets. The majority of droplets are PCR negative.

Fig 2

(A) Standard calibration curve of ddPCR concentrations (plasmids/μl) against dilution factors, r² = 0.9927, intercept = 0.037 plasmids/μl, slope = 1.01. (B) Standard calibration curve of ddPCR concentrations (human RPP30 copies/μl) against dilution factors, r² = 0.02896, intercept = 1,478.1 plasmids/μl, slope = −0.003. All plots are drawn on logarithmic scales.

Fig 3

Flow diagram of participant samples in the retrospective validation of ddPCR against the standard Amplicor CT/NG test. Numbers in parentheses refer to Amplicor CT/NG results.

Fig 4

(A) Distribution of Amplicor CT/NG test OD₄₅₀ values, showing (i) the most-recent Amplicor CT equivocal zone, (ii) the earlier Amplicor equivocal zone, and (iii) the broad equivocal zone identified using the observations of this study. (B) Amplicor CT test OD₄₅₀ (x axis) against C. trachomatis plasmid quantity (y axis), as determined by ddPCR. OD₄₅₀ values of 0.1, 0.8, and 3.1 are indicated. ddPCR values of zero are assigned an arbitrarily low (0.001 plasmids/μl) value. Positive ddPCR with high estimates of the infectious load are infrequent when the Amplicor CT/NG OD₄₅₀ is <3.1.