Spatiotemporal dynamics of HIV-1 transmission in France (1999-2014) and impact of targeted prevention strategies

Abstract

Background: Characterizing HIV-1 transmission networks can be important in understanding the evolutionary patterns and geospatial spread of the epidemic. We reconstructed the broad molecular epidemiology of HIV from individuals with primary HIV-1 infection (PHI) enrolled in France in the ANRS PRIMO C06 cohort over 15 years.

Results: Sociodemographic, geographic, clinical, biological and pol sequence data from 1356 patients were collected between 1999 and 2014. Network analysis was performed to infer genetic relationships, i.e. clusters of transmission, between HIV-1 sequences. Bayesian coalescent-based methods were used to examine the temporal and spatial dynamics of identified clusters from different regions in France. We also evaluated the use of network information to target prevention efforts. Participants were mostly Caucasian (85.9%) and men (86.7%) who reported sex with men (MSM, 71.4%). Overall, 387 individuals (28.5%) were involved in clusters: 156 patients (11.5%) in 78 dyads and 231 participants (17%) in 42 larger clusters (median size: 4, range 3-41). Compared to individuals with single PHI (n = 969), those in clusters were more frequently men (95.9 vs 83%, p < 0.01), MSM (85.8 vs 65.6%, p < 0.01) and infected with CRF02_AG (20.4 vs 13.4%, p < 0.01). Reconstruction of viral migrations across time suggests that Paris area was the major hub of dissemination of both subtype B and CRF02_AG epidemics. By targeting clustering individuals belonging to the identified active transmission network before 2010, 60 of the 143 onward transmissions could have been prevented.

Conclusion: These analyses support the hypothesis of a recent and rapid rise of CRF02_AG within the French HIV-1 epidemic among MSM. Combined with a short turnaround time for sample processing, targeting prevention efforts based on phylogenetic monitoring may be an efficient way to deliver prevention interventions but would require near real time targeted interventions on the identified index cases and their partners.

Keywords: HIV-1; Phylogeography; Primary infection; Transmission network; Treatment as prevention.

Fig. 1

Inferred HIV transmission clusters. HIV-1 transmission cluster diagrams illustrating the structure and demographics of the putative transmission clusters identified in the PRIMO ANRS CO6 cohort. A total of 387 of the 1356 (28.5%) individuals were connected with at least one other individual. Color indicates the reported transmission risk [red MSM; green heterosexual (HTS), purple others]; and shape denotes gender (ellipse male, square women). All edges represent a genetic distance of ≤1.5% separating nodes. All shapes are labeled according to the HIV-1 subtype. NA not available. White and unfilled dots correspond to missing informations

Fig. 2

Characteristics of the 3 larger clusters B1, AG1 and AG2. a Transmission network of the three larger cluster AG1 (n = 14), AG2 (N = 41) and B1 (n = 9) and evolution of the main clusters over the study period. b Map representing the number of clustering individuals by location of residence. c Ancestral root state probabilities. The root state probabilities are presented with the color codes corresponding to the 11 equally populated regions

Fig. 3

Bayesian time-scaled tree of the HIV transmission network of subtype B (a) and CRF02_AG (b) pol sequences in clusters from the participants enrolled in the PRIMO ANRS cohort between 1999 and 2014. Time scaled in year. Nodes and branches are colored according to the most probable location state of their descendent nodes. Tips are colored according to the recorded location of sampling