Abstract Body

Program evaluation is essential to public health to document success and identify areas for improvement. In British Columbia (BC), Canada, publicly funded pre-exposure prophylaxis (PrEP) has been available since January 2018 at no cost to clients, yet effects on transmission and drug resistance are uncharacterized. To evaluate the BC PrEP program we tested the hypothesis that phylogenetic clustering and drug resistance would differ between previous PrEP users who seroconverted (PUWS) compared to non-PrEP users who seroconverted (NPUWS).

38539 HIV partial pol sequences generated from 10386 participants in the BC Drug Treatment Program were aligned to the HXB2 reference. WHO surveillance drug resistance mutations were extracted from the alignment. Phylogenetic trees were inferred and clusters comprising 5 or more members identified using a patristic distance cutoff of 0.02 substitutions/site, and viral diversification rates calculated for each tip. Date of first detectable viral load was used as a proxy for diagnosis date. Cluster-specific reproduction numbers were estimated using R package EpiEstim.

From 1 January 2018 to 24 June 2021, 7465 persons had ever received PrEP via the BC program, of whom there were 15 (0.20%) PUWS between 23 October 2018 and 20 November 2020. Over this time, there were 314 NPUWS in BC. PUWS were not significantly more likely to cluster than NPUWS (chi-square: p=0.26), nor were viral diversification rates significantly higher (Kruskal test: p=0.09), or cluster-specific median reproduction numbers (Kruskal: p=0.60) (Figure 1). All unique clusters joined by PUWS were also joined by NPUWS. There were no significant differences between proportion of PUWS or NPUWS with baseline resistance to any drug class (chi-square: p=0.88, NNRTI p=0.75, or NRTI p=0.89, Figure 1d). Among 3 PUWS with baseline resistance, 2 had resistance to NNRTI, 2 to NRTI, and 0 to PI or INSTI. PUWS with the most baseline drug resistance to NRTIs had the highest proportion of days not covered by PrEP medication based on dispensing records (0.4 versus median 0.056), the longest gap with no refill (155 days versus median 22), and only 64 days lapse between last prescription and diagnosis date (median 200).

Previous PrEP use among seroconverters in BC was not associated with phylogenetic clustering, viral diversification rates, cluster-specific reproduction numbers, or baseline drug resistance. These results strongly highlight the success of the PrEP program in BC.