Association of HIV Preexposure Prophylaxis Use With HIV Incidence Among Men Who Have Sex With Men in China A Nonrandomized Controlled Trial

IMPORTANCE Evidence on HIV preexposure prophylaxis (PrEP) among Chinese men who have sex with men (MSM) is critical to guide its large-scale implementation in low- and middle-income countries. OBJECTIVE To evaluate incident HIV infection, adherence, safety, and changes in sexual behaviors among MSM using daily PrEP (D-PrEP) and event-driven PrEP (ED-PrEP) in 4 cities in China. DESIGN, SETTING, AND PARTICIPANTS This nonrandomized controlled trial was conducted among HIV-seronegative MSM from December 11, 2018, to November 30, 2020, in Beijing, Shenyang, Chongqing, and Shenzhen. Participants self-chose D-PrEP or ED-PrEP regimens at baseline and could switch regimens during the 12-month study period. HIV-negative MSM who declined to initiate PrEP (nonusers) in the same cities joined a separate parallel prospective cohort and served as control individuals. INTERVENTIONS PrEP consisted of coformulated tenofovir disoproxil fumarate, 300 mg, and emtricitabine, 200 mg. MAIN OUTCOMES AND MEASURES The main outcome was incident HIV infection. Poisson regression was used to obtain the HIV incidence rate ratio (IRR). RESULTS A total of 1530 MSM were included in the analysis (median age, 30 [IQR, 25-37] years). At baseline, 520 MSM chose D-PrEP (median age, 29 [IQR, 25-35] years) and 503 chose ED-PrEP (median age, 29 [IQR, 25-36] years). The median HIV Risk Index score was 18 (IQR, 12-22) among D-PrEP users and 18 (IQR, 11-22) among ED-PrEP users. Among 507 PrEP nonusers, the median age was 33 (IQR, 27-43) years, and the median HIV Risk Index score was 12 (IQR, 7-18). Although PrEP users had more baseline behaviors associated with HIV risk, the HIV incidence was lower among all PrEP users (adjusted IRR, 0.09 [95% CI, 0.04-0.21]), ED-PrEP users (adjusted IRR, 0.05 [95% CI, 0.01-0.22]), and D-PrEP users (adjusted IRR, 0.12 [95% CI, 0.04-0.33]) compared with PrEP nonusers. There was no difference in HIV incidence between D-PrEP users and ED-PrEP users (IRR, 0.33 [95% CI, 0.06-2.04]). Event-driven PrEP users consumed 40% fewer tablets than D-PrEP users during the study period. Adherence, defined as the proportion of self-reported days with sexual intercourse in which PrEP was taken according to prescription of at least 90%, increased over time among ED-PrEP users (from 57.4% to 77.8%; P < .001 for trend) and decreased over time among D-PrEP users (from 75.1% to 72.1%; P = .02 for trend). Daily PrEP users reported fewer adverse events than ED-PrEP users (193 of 520 [37.1%] vs 241 of 503 [47.9%]). CONCLUSIONS AND RELEVANCE The findings of this study suggest that D-PrEP and ED-PrEP regimens are associated with lower incidence of HIV and a good safety profile among high-risk MSM in China.