Sexual behaviour and viral load in MSM HIV-diagnosed at primary infection stage and followed in the French ANRS - PRIMO cohort

Background: From 2008, the diffusion of the concept of treatment as prevention (TasP) in the HIV-infected population may lead to an increase in at-risk sexual behaviour. We aimed to assess the evolution over time of sexual behaviour in HIV-infected MSM followed in the French ANRS-PRIMO cohort.
Methods: In 1996-2014, 1,226 MSM have been enrolled during HIV primary infection. At each visit
(every 6 months), a clinical questionnaire including lab measurements is completed and a self-administered questionnaire collects data on sexual partners, and condom use during the past six months (reasons of inconsistent condom use are available since 2013). The evolution of sexual behaviour in 2000-2014 and the association with the viral load at the preceding visit (undetectable: yes/no) were assessed using GEE models taking into account longitudinal data.
Results: We analyzed 8,395 follow-up visits with completed data on partners and condom use. The proportion of visits where ≥1 casual partner was reported has significantly increased of 3%/year (p=0.03) from 2001 to 2010, and of 11%/year (p< 0.0001) since 2011 to reach 60% of visits in 2013.
In MSM reporting ≥1 casual partner, the frequency of inconsistent condom use with partners of negative or unknown HIV status has risen of 5%/ year (p=0.02) from 18% of visits in 2001 to 23% in 2010, and then of 18%/year (p< 0.0001) up to 36% of visits in 2013. There was no difference in condom use according to the VL (p=0.52).
Since 2013, among the 94 MSM who did not use condom consistently with casual partners despite a detectable VL, 45% thought that they could not transmit HIV because they were treated or thought (wrongly) they had an undetectable VL.
Conclusions: In these MSM, inconsistent condom use has increased since the beginning of 2000s and more dramatically after 2010. We can assume that the concept of TasP is widespread in recent years and assign to it this kind of disinhibition in HIV-infected MSM, but no association was observed with VL. Efforts should be done to inform patients about the impact of treatment on HIV-transmission and the importance to know their VL.

K. Champenois1, R. Seng1,2,3, A. Essat1, S. Orr1, A. Persoz1, C. Goujard1,2,4, L. Meyer1,2,3, ANRS PRIMO Cohort Study Group
1Inserm UMR 1018, CESP-4: Epidemiology of HIV and STIs, Le Kremlin Bicêtre, France, 2South-Paris University, South-Paris Medical School, Le Kremlin Bicêtre, France, 3AP-HP, Bicêtre Hospital, Epidemiology and Public Health Service, Le Kremlin Bicêtre, France, 4AP-HP, Bicêtre Hospital, Internal Medicine Department, Le Kremlin Bicêtre, France