Continued HIV protection from 2:1:1 PrEP dosing in gay and bisexual men in Paris

More than five years ago, the large prospective open-label PREVENIR study provided important additional evidence to support the efficacy of on-demand 2:1:1 event-based dosing using oral TDF/FTC in individuals at high risk of HIV.

At CROI 2026, the final results of PREVENIR were presented by Jean-Michel Molina from the Assistance Publique – Hôpitaux de Paris. [1]

The final dataset included results from more than 3200 people enrolled from May 2017 to May 2024 at 22 sites in Paris.

Baseline characteristics included median age 36 years (IQR: 29 to 43), 98.7% were gay or bisexual men and 0.4% were transgender individuals (n=14). The majority of participants (87%) were born in France, educated (70% to degree level or higher), 14% used chemsex and 42% had not used PrEP before joining the study,

Median follow-up was 45 months (IQR: 25 to 76) and included 13130 person-years of follow-up (PYFU). PrEP was used at least 75% of the time by 38% of people using either daily or on-demand dosing and by 24% of people switching between both dosing options, with roughly 15 people per 100 PY of each dosing group discontinuing PrEP.

Roughly half the cohort used daily and half used 2:1:1 dosing with half of each group later switching.

Roughly 2%, 13% and 7% of participants used neither condoms nor PrEP at last sex in the daily, on-demand and switch groups. The median number of partners dropped by about one-third in all groups but condom use also dropped (presumably due to increased confidence in PrEP).

There were 13 new HIV cases with overall HIV incidence of 0.11 per 100 PY (95% CI: 0.0 to 0.18) with no difference by PrEP dosing (p=0.4). Drug levels from dried blood spot showed very low adherence in 11/13 cases.

The high incidence of bacterial STIs (53 per 100 PY) showed that HIV risk without PrEP would have been significant with incidence rates of 25 and 32 per 100 PYFU for chlamydia and gonorrhoea, respectively.

The primary endpoint for PREVENIR however was whether PrEP could have an impact on local HIV incidence by 2023/24 compared to rates in 2015/16. Although overall incidence didn’t drop to meet this endpoint, there were significant differences in different population groups and these were linked to PrEP access.

So although HIV incidence dropped by 33% in gay men born in France (matching the study population), HIV incidence increased by 73% over the same time in gay men born abroad. This inequity of access was recognised as a weakness in the study and an important challenge for future access to PrEP in France.