Real-life occurrence of bacterial sexually transmitted infections among PrEP users: improving the diagnosis of Chlamydia trachomatis and Neisseria gonorrhoeae with multisite screening

Abstract

PrEP users are under high risk for bacterial sexually transmitted infections (STI), including those caused by Treponema pallidum (Tp), Chlamydia trachomatis (Ct) and Neisseria gonorrhoeae (Ng). Ct and Ng screening at multiple anatomic sites may improve the diagnostic sensitivity among high-risk populations. We analyzed the prevalence and incidence of Ct, Ng, and Tp and investigated predictors of bacterial STI occurrence between January 2018 and November 2019 in a retrospective cohort of PrEP users in Sao Paulo, Brazil. We describe the frequency and percentage of Ct/Ng per anatomical site and calculate the percentage of missed diagnosis if molecular testing were applied only in symptomatic patients, or only in urine samples. Patients underwent syphilis testing every 3-4 months and Ct/Ng testing every 6 months. We included 413 PrEP users with a median age of 31 years. At baseline, 25% had a positive treponemal test and 7% had active syphilis; Ct and Ng were more frequently detected in the oropharynx and anus (6.4-6.9%) than in urine samples (0.7-2.6%). Twelve months after the onset of PrEP, the incidence of Tp, Ct and Ng was, respectively, 13.4%, 11.4% and 8.9%. During follow-up, 23 out of 33 Ct/Ng cases (69%, 95% CI 51-84) would have been missed if oropharynx and anus samples had not been tested. In addition, if only symptomatic cases had been tested, 30 out of 33 Ct/Ng cases (90%, 95% CI 75-98) would have been missed. Participants with incident STI had a higher baseline number of sexual partners and a longer follow-up. Our study reinforces that active and frequent screening for STI is a powerful strategy to improve the diagnostic sensitivity.

Figure 1. Cumulative incidence of syphilis (A), Ct (B) and Ng (C) infections during the follow-up of the participants in the study.

Figure 2. Percentage of missed Ct/Ng diagnosis at baseline and follow-up if molecular testing had been applied only in urine samples or only to symptomatic patients. Whiskers represent 95% confidence intervals.