Recent HIV Risk Behavior and Partnership Type Predict PrEP Adherence in Men Who Have Sex with Men
Published Web Locationhttps://doi.org/10.1093/ofid/ofx162.038
Abstract: Background: Individuals engaging in higher risk behavior are often more adherent to PrEP but it is unclear if partnership type itself affects PrEP adherence. We examined the effect of recent HIV risk behaviors and partnership type on PrEP adherence in men who have sex with men (MSM) taking PrEP. Methods: CCTG 595 is a 48-week PrEP demonstration study of 398 HIV− at-risk MSM. At baseline and week 48, HIV risk score was estimated as the probability of seroconversion over the next year based on number of condomless anal sex acts with HIV+/unknown partners in the last month and any STI diagnosed at study visit. HIV risk score was categorized as low (<0.12), moderate (0.12−0.59) and high (>0.59) risk based on population seroconversion probabilities. Partnership type was assigned as no/single HIV− partner, single HIV+ partner, or multiple partners of any serostatus in the past 3 months. PrEP adherence was estimated by intracellular tenofovir-diphosphate (TFV-DP) levels as a continuous variable at week 48. Statistical methods included McNemar’s test, Wilcoxon rank-sum test, and linear regression model where appropriate. Results: Of 313 MSM who completed week 48, there was no significant change in HIV risk category from baseline to week 48 (low: 44 to 42%; moderate: 27 to 24%; high: 28 to 34%; P = 0.25). There was a significant change in partnership type, with the proportion of those with no or single HIV− partnerships increasing (1 to 9%, P < 0.001). In univariate analysis, moderate and high-risk groups had higher TFV-DP levels than the low-risk group at week 48 (P = 0.018). Participants with no/single HIV− partner had significantly lower TFV-DP levels than those with one HIV+ partner or multiple partners (P = 0.007). In a multivariable linear regression model, only low-risk partnerships remained significant where no/single HIV− partnerships were associated with lower TFV-DP levels (mean difference = −344fmol/punch [−617, −71], P = 0.014). Conclusion: Although there was a shift in partnership type towards lower risk partnerships, objective HIV risk behavior remained stable over time. Individuals with higher HIV risk behaviors and risk partnerships had higher TFV-DP levels suggesting maintained strong motivation for PrEP adherence. Thus, recent sexual risk behavior and partnership type may be important predictors of PrEP adherence in MSM. Disclosures: All authors: No reported disclosures.