Comparative Analysis of Molecular and Serologic Testing for Primary Syphilis: A Population-Based Cohort Study

Abstract

Rising rates of syphilis (T. pallidum; Tp) requires rapid diagnosis and treatment to manage the growing epidemic. Syphilis serology is imperfect and requires interpretation of multiple tests while molecular diagnostics allows for potential yes-no identification of highly infective, primary anogenital lesions. Accuracy of this testing modality has thus far been limited to small, highly selective studies. Therefore, we retrospectively assessed a large, adult population of patients with anogenital lesions seen at Sexually Transmitted Infection (STI) clinics in Alberta, Canada who were screened for syphilis and herpes simplex (HSV) 1/2 using PCR to evaluate Tp-PCR versus serology to diagnose primary syphilis. 114 (3.1%) of the 3,600 adult patients had at least one Tp-PCR+ anogenital lesion with 99 (2.8%) patients having newly positive syphilis serology (new INNO-LIA positive or 4-fold RPR increase). Tp-PCR had a sensitivity of 49.3% (95% CI 42.6-56.1) and specificity of 99.9% (99.7-100.0). Positive predictive values and negative predictive values in the study population or when corrected for provincial prevalence were 97.4% (92.5-99.5) or 0.4% (0.4-1.2) and 96.7% (96.1-97.3) or 100.0% (100.0-100.0), respectively. Positive and negative likelihood ratios were estimated at 555 (178-1733) and 0.5 (0.4-0.6), respectively. Review of all Tp-PCR performed with or without exclusion of HSV-positive lesions resulted in no significant change in Tp-PCR characteristics. Interestingly, 12 of the Tp-PCR+ samples had negative serology at time of lesion sampling but became positive within our 28-day testing window. Overall, this study further supports the use of Tp-PCR as an accurate assay to rapidly identify, treat, and prevent the spread of primary syphilis.