Paper # 986
Risk Determinants of Men-Who-Have-Sex-with-Men Transmission Cascades
Bluma Brenner*1, M Roger2, R Turgel1, J-P Routy3, D Stephens4, D Moisi1, F Ohnona1, S Gagnon2, J Koopman5, M Wainberg1, and Montreal PHI Cohort and SPOT Study Group
1McGill AIDS Ctr, Montreal, Canada; 2Ctr Hosp Univ Montreal, Canada; 3McGill Univ Hlth Ctr, Montreal, Canada; 4McGill Univ, Montreal, Canada; and 5Univ of Michigan, Ann Arbor, US
Background: Our past findings, using population-based
genotypic surveillance, estimated acute/recent infection to be a critical
period for onward spread of the men-who-have-sex-with-men (MSM) epidemic in Quebec. Infections appear highly transmissible and individuals unaware of their status may engage
in high-risk behaviors. This study combined molecular phylogenetic and
epidemiological approaches to model MSM transmission dynamics and prevention
interventions.
Methods: HIV sequence data were obtained from
primary HIV infections (PHI, <6 months post-infection) collected from the Quebec genotyping program (2002 to 2009, n = 1127). Phylogenetic analysis of male only,
subtype B infections inferred MSM transmission dynamics (n = 922). Bayesian analysis
using the BEAST package showed partial sampling is not detrimental to inference
and time stability of infection dynamics. Behavioral determinants of clustering
was assessed using risk profiles reported by MSM participants in the primary
HIV-infection (PHI) cohort (n = 126). Changes in risk behaviors were determined
at 5 time points (0, 6, 12, 18, and 24 months) following diagnosis.
Results: Clustering of PHI represented 50% and 66%
of transmissions in 2005 and 2009, respectively. Phylogenetic analysis stratified
MSM into non-clustered (n = 321); small clustered (2 to 4 PHI/cluster, 2.4 ± 0.8
PHI, n = 203) and large clustered (≥5 PHI/cluster, 9.9 ± 1.2
PHI/cluster, n = 398) transmission events. Large clusters rose from of 30% of
transmissions in 2005 to 43% of PHI in 2009. Cohort participants reporting
>10 partners, 3 months prior to diagnosis were more frequent in large
clusters as compared to the other 2 groups (39% vs. 13%, odds ratio 4.67, P <0.001).
Higher frequencies of insertive sex were reported by participants in the small
cluster as compared to the other 2 groups (54% vs. 16%, odds ratio 9.3,
p <0.01). Significant declines in anal and oral receptive risk
behaviors, coupled with increased condom use were observed in all groups and sustained
over >12 months. Little change in insertive sex was reported over time. In Montreal, 1/3 persons were unaware of their status at diagnosis and <30% had not had an
HIV test one year prior to diagnosis.
Conclusions: Molecular epidemiology findings have
revitalized efforts to avert PHI-associated HIV transmissions. Rapid anonymous
HIV testing with structured counseling administered by nurses and counselors,
trained on MSM dynamics, at a new community site has been created to improve
testing access with structured behavioral interventions.
|