Characteristics of human immunodeficiency virus (HIV) seroconversions in a large prospective implementation cohort study of oral HIV preexposure prophylaxis in men who have sex with men (EPIC-NSW)

Background: Most human immunodeficiency virus (HIV) seroconversions in people who have initiated preexposure prophylaxis (PrEP) occur in the context of insufficient adherence. We describe participants who seroconverted after being dispensed PrEP in a large PrEP implementation study in Australia. Methods: Expanded PrEP Implementation in Communities in New South Wales was an implementation study of daily oral PrEP in individuals aged ≥18 years at high risk for acquiring HIV. HIV seroconversions were defined as a positive HIV test by either antigen, antibody, or detectable HIV viral load after enrollment. Insufficient adherence, measured by dispensing logs or participant self-report, was defined as <4 PrEP doses per week. Results: A total of 9596 participants were enrolled and dispensed PrEP between 1 March 2016 and 30 April 2018; 30 were diagnosed with HIV by 31 March 2019. The median (interquartile range [IQR]) age was 31 (25-38) years, all identified as male, 29 (97%) identified as gay or homosexual, and 20 (69%) lived in a postcode with a low concentration of gay male residents. The median (IQR) days from first PrEP dispensing to diagnosis was 409 (347-656). There was no evidence that participants who seroconverted had been sufficiently adherent to PrEP. Nineteen (63%) participants who seroconverted were diagnosed with chlamydia, gonorrhoea, syphilis, or new hepatitis C infection. One participant had resistance to emtricitabine (M184V mutation) at diagnosis. Conclusions: Participants who seroconverted were insufficiently adherent to PrEP despite being at high risk for acquiring HIV. Understanding the reasons for poor PrEP adherence in individuals who subsequently acquire HIV is critical to improving PrEP effectiveness.