CROI 2024 Abstract eBook

Abstract eBook

Poster Abstracts

models to data preceding PrEP to predict growth in the absence of PrEP, with consideration for COVID-19. Predicted cases without PrEP were compared to observed cases to quantify cases averted via PrEP. Results: BC HIV phylogenetic clusters exhibited differential growth and Re since widespread PrEP availability in 2018. Re of most clusters comprised predominantly of men who have sex with men (MSM) have declined since 2018. However, we identified missed opportunities for PrEP among a large cluster of young MSM with Re above 1 between 2018 and 2020, and in a cluster comprised primarily of people who use drugs (PWUD) with elevated Re from 2020-2022. Preliminary models of cluster growth in the absence of PrEP suggest averted HIV cases were heterogeneously distributed across phylogenetic clusters. Conclusion: These results highlight the success of the PrEP program in averting new HIV cases, while emphasizing clusters and sociodemographic groups that could benefit from prioritized access and assistance with adherence to PrEP. 1067 Geospatial and Phylogenetic Clustering of Acute HIV Infections in Lilongwe, Malawi Griffin J Bell 1 , Kimberly Powers 1 , Oliver Ratmann 2 , Ann M. Dennis 1 , Pearson Mmodzi 3 , Mitch Matoga 3 , Edward Jere 3 , David Bonsall 4 , Sharon Weir 1 , Michael Emch 1 , Irving F. Hoffman 1 , Myron S. Cohen 1 , William Miller 1 1 University of North Carolina at Chapel Hill, Chapel Hill, NC, USA, 2 Imperial College London, London, United Kingdom, 3 University of North Carolina Project–Malawi, Lilongwe, Malawi, 4 University of Oxford, Oxford, United Kingdom Background: HIV transmissibility spikes during the first months of infection, especially during acute (pre-seroconversion) HIV infection (AHI). Rapid propagation of HIV during early infection hinders universal test and treat interventions, which typically identify infections after the period of elevated transmission risk. To guide prospective interventions against transmission during this period, we evaluated the geospatial and phylogenetic clustering of acute and early infection in Lilongwe, Malawi. Methods: We identified 144 persons with AHI, 30 people who recently acquired HIV (determined with the Sedia LAg-Avidity EIA), and 652 people with chronic HIV who came to a Lilongwe STI clinic between 2015 and 2019. We mapped the point locations of households with an AHI case and 721 sex-exchange venues identified with the PLACE method. To evaluate the spatial clustering of AHI, we used Tango and Takahashi's flexible scan statistic (α=0.2). Consensus HIV sequences were obtained from blood samples with the veSEQ-HIV protocol and shiver and aligned with MAFFT. Maximum-likelihood trees were built with IQ-TREE. Monophyletic sequences with genetic distances <5.3% were considered phylogenetic clusters. Results: We identified 6 spatial areas (0.2-1.7 km 2 ) in Lilongwe where household locations of people with AHI were overrepresented, comprising 38% of AHI cases in 1% of the populated land area. These spatial areas contained 57 reported sex-exchange venues and were highly connected: the contiguous M1 and S124 roads run directly through 4 spatial clusters. Although viral sequencing failed for 39% of people with AHI and 60% with recent or chronic infections, we still identified 13 nonoverlapping two-person clusters in our phylogenetic analysis. Four pairs (3 acute-acute, 1 acute-recent) attended the clinic 0-95 days apart, suggesting transmission during acute and early infection. A fifth recent recent pair (270 days apart) also suggested early transmission. Acute-acute pairs lived 0.2-4.5 km apart, but not in the same spatial clusters. Conclusion: Spatial clustering of AHI exists in highly connected areas with sex-exchange venues in Lilongwe, suggesting that spatially focused interventions could be key in pursuit of HIV elimination. Evidence of extensive AHI transmission chains was limited: no phylogenetic cluster had >2 members. Mirroring earlier modeling estimates of the role of AHI in Lilongwe, 5 of 13 (38%) phylogenetically linked pairs suggested transmission during acute and early HIV infection.

2-year detection period [Figure]. We first calculated the relative contribution of transmission bursts in a fixed detection period to future transmission in a follow-up period by dividing the number of internal nodes during 2017-2019 descended from bursts by the number of lineages associated with bursts during 2015-2016, compared to non-burst-descended internal nodes divided by non-burst-associated lineages. To characterize PWH diagnosed in the follow-up period who were ever members or descendants of a transmission burst, we then detected bursts within any sliding 2-year period during 2014–2019 and identified populations overrepresented among PWH associated with transmission bursts. Results: The 2,795/86,006 (3.2%) lineages (or persons) associated with a transmission burst during 2015–2016 contributed to 493/3,926 (12.6%) transmissions during 2017–2019 across all jurisdictions. Lineages associated with transmission bursts were 4.3 times as likely as lineages not associated with bursts to contribute to future transmissions. Among PWH diagnosed during 2017–2019, 5,603/43,721 (12.8%) were ever members or descendants of transmission bursts during 2014–2019. Groups overrepresented among members or descendants of transmission bursts (i.e., >12.8%) included PWH aged 13–19 (24.5%) or 20–29 years (16.6%) at HIV diagnosis, diagnosed during acute or early HIV infection (17.9%), who reported male-to-male sexual contact (15.5%), or who were transgender, non-binary, or another gender (15.3%). Conclusion: Lineages associated with transmission bursts contribute disproportionately to future transmission, underscoring the value of detecting and responding to clusters to prevent transmissions. Bursts of rapid HIV transmission likely contribute to disparities in overall HIV incidence for some key populations.

Poster Abstracts

1066 HIV Cases Averted in Phylogenetic Clusters via Preexposure Prophylaxis in British Columbia, Canada Angela McLaughlin , Junine Toy, Paul Sereda, Jason Trigg, Vincent Montoya, Richard H. Liang, Charlotte Beelen, Tim Tang, Alex LaBerge, Chanson J. Brumme, Rolando Barrios, Julio S. Montaner, Jeffrey B. Joy British Columbia Centre for Excellence in HIV/AIDS, Vancouver, Canada Background:Pre-exposure prophylaxis (PrEP) to prevent HIV acquisition is a key pillar in strategies for ending the HIV epidemic. However, PrEP access and adherence have been incomplete across key populations and population-level impacts of PrEP on HIV transmission are not fully understood. We hypothesized that widespread PrEP availability in British Columbia (BC), Canada since 2018 has averted hundreds of new HIV cases heterogeneously distributed across phylogenetic clusters. Methods: Using data from the BC Drug Treatment Program, we aligned 42043 HIV partial pol sequences from 10740 individuals to the HXB2 reference and removed surveillance drug resistance mutations. Phylogenetic trees were inferred to identify clusters with 5 or more members with pairwise tree distance less than 0.02 substitutions/site. New diagnoses within clusters over time were used to estimate cluster-specific reproduction numbers (Re) over time. We tested whether PrEP availability significantly affected Re after 2018 in generalized estimating equations, adjusted for cluster size, median age, risk group composition, COVID-19, and treatment guidelines. We summarized characteristics of clusters without reduced growth following PrEP as missed opportunities. For clusters with significantly reduced Re, we fit counterfactual

CROI 2024 344