Background We measured the incidence and associated factors of urethral and rectal C. trachomatis (CT) and N. gonorrhoeae (NG) among a Kenyan cohort of MSM taking PrEP.

Methods Enrolled MSM underwent audio computer assisted self-interview for behaviour and socio-demographics, with medical examination every 6 months. CT and NG were diagnosed by polymerase chain reaction assay in urine and rectal swabs at baseline, 6- and 12- months. We identified factors associated with incident urethral infection using multivariable Cox regression and report adjusted hazard ratios (aHR).

Results October 2017 - January 2018, 158 participants were enrolled and initiated PrEP. Follow-up was 99% (month 6) and 93% (month 12). At baseline, 10.4% had urethral CT/NG (either or both infections), and 3.3% had rectal CT/NG, with total 11.4% prevalence CT/NG. At 6 months, CT/NG was 7.7% (urethral) and 0% (rectal), and 10.2% (urethral) and 0% (rectal) at 12 months. There were 22 incident infections (19 CT, 2 NG, 1 CT and NG) during 152 person-years (py) follow-up, with no re-infections (13.9 cases/100 py; 95% CI: 9.2–21.1). Risk decreased 12% per one year increase in age (aHR=0.88, p=0.012) and was 73% lower for men reporting last sex partner being female (aHR=0.37, p=0.019). For each one unit increase in social support, men had 3% increased CT and/or NG risk (aHR=1.03, p<0.001), and men reporting having sex in exchange for money at screening were 2.35 times more likely to be infected (p=0.072). Number of sex partners, sexual positioning, condom use, lubricant use, and self-reported PrEP adherence were not associated with infection.

Conclusion The incidence of urethral CT/NG infection was high in our cohort of MSM taking PrEP, despite risk reduction counselling and repeated testing and treatment, supporting need for ongoing etiologic testing and more effective risk reduction intervention. Additional analyses will examine risk compensation.

Disclosure No significant relationships.