MOAD0305
Share
 
Title
A competing risks model for the use of condom in the open-label extension of the ANRS Ipergay study
Presenter
Luis Sagaon Teyssier
Authors
L. Sagaon Teyssier1,2, M. Mimi1,2, D. Rojas Castro3,4, N. Hall5, C. Capitant6, L. Meyer6,7, C. Chidiac8, C. Tremblay9, G. Pialoux10, C. Pintado11, M. Préau12, J.-M. Molina13, B. Spire1,2, ANRS IPERGAY Study Group
Institutions
1SESSTIM UMR 1252, Marseille, France, 2ORS PACA, Observatoire Régional de la Santé Provence-Alpes-Côte d'Azur, Marseille, France, 3AIDES, Paris, France, 4Coalition PLUS, Community-based Research Laboratory, Paris, France, 5CHU de Nantes, Department of Infectious Diseases, Nantes, France, 6INSERM SC10 US19, Villejuif, France, 7Université Paris Sud, Paris Saclay, France, 8Hôpital de la Croix Rousse, INSERM U1052, Department of Infectious Diseases, Lyon, France, 9Research Center of the Centre Hospitalier de l'Université de Montréal, Montréal, Canada, 10Hôpital Tenon, Assistance Publique Hôpitaux de Paris, Maladies Infectieuses, Paris, France, 11Hospital Saint-Louis, Assistance Publique Hôpitaux de Paris, Department of Infectious diseases, Paris, France, 12Groupe de Recherche en Psychologie Sociale (EA 4163), Université Lumière Lyon 2, Université de Lyon, Lyon, France, 13Hospital Saint-Louis, Assistance Publique Hôpitaux de Paris, Department of Infectious Diseases, Paris, France

Background: During the open-label extension (OLE) phase of the ANRS Ipergay trial, the overall use of condoms decreased when participants knew about the PrEP efficacy with TFDF/FTC vs. placebo. Still some participants resumed condom use during OLE. We aimed investigating the factors associated with using condoms back again combined or not with PrEP.
Methods: ANRS Ipergay OLE started in November 2014. Follow-up between M0-M18 included bimonthly HIV/STI testing and online questionnaires collecting sexual behavior, PrEP (correct/sub-optimal) and condom use at the most recent anal intercourse (MRAI). This analysis focused on participants using only PrEP at OLE enrolment. The outcome: time elapsed until condom resumption alone or combined with PrEP. Competing risks survival model estimated hazard-ratios (HR). Individual characteristics, sexual behavior and STI onset before condom resumption during OLE were specified as time-dependent variables.
Results: Among 361 OLE participants, 146 (40.4%) used only PrEP for their MRAI at enrolment. Compared to other participants, those using only PrEP were less educated (63.3% vs 36.7%, p=0.03), and had higher HIV-risk perception (scale 0-10; median[IQR]:4[2-7] versus 2[1-4], p< 0.001). No difference was found for age, number of sexual partners and number of sexual intercourses. 70% of these 146 participants were low-level condom users during the double-blind phase with: on average, 2/10 MRAI with condom (6/10 for the remaining 30%). Among these 146 participants, 8.2% resumed condoms instead of PrEP on average after 5.5 months (sd.±3.4); and 59.6% PrEP+condoms on average after 6 months (sd.±4.4). On average, condoms resumption lasted 7.6 months (sd.±6.1). Condom use probability increased with the number of sexual partners (previous 2 months) (HR:1.01, p=0.035); and the number of sexual intercourses (previous 4 weeks) (HR:1.05, p< 0.001). Sensations seeking increased condom resumption probability in combination with PrEP (HR:1.04, p=0.038). However, chemsex reduced condom resumption alone (HR:0.157, p=0.014) or combined with PrEP (HR:0.712, p=0.003). Finally, the onset of STIs during OLE increased condom resumption probability alone (HR:3.69, p< 0.001) and combined with PrEP (HR:1.42, p=0.013);
Conclusions: Condom resumption during the OLE phase of the ANRS Ipergay study was strongly associated to the onset of STIs. In contrast, chemsexers were less likely to resume condom while using PrEP.