Population-level impact of expanding PrEP coverage by offering long-acting injectable PrEP to MSM in three high-resource settings: a model comparison analysis

Abstract

Introduction: Long-acting injectable cabotegravir (CAB-LA) demonstrated superiority to daily tenofovir disoproxil fumarate/emtricitabine (TDF/FTC) for HIV pre-exposure prophylaxis (PrEP) in the HPTN 083/084 trials. We compared the potential impact of expanding PrEP coverage by offering CAB-LA to men who have sex with men (MSM) in Atlanta (US), Montreal (Canada) and the Netherlands, settings with different HIV epidemics.

Methods: Three risk-stratified HIV transmission models were independently parameterized and calibrated to local data. In Atlanta, Montreal and the Netherlands, the models, respectively, estimated mean TDF/FTC coverage starting at 29%, 7% and 4% in 2022, and projected HIV incidence per 100 person-years (PY), respectively, decreasing from 2.06 to 1.62, 0.08 to 0.03 and 0.07 to 0.001 by 2042. Expansion of PrEP coverage was simulated by recruiting new CAB-LA users and by switching different proportions of TDF/FTC users to CAB-LA. Population effectiveness and efficiency of PrEP expansions were evaluated over 20 years in comparison to baseline scenarios with TDF/FTC only.

Results: Increasing PrEP coverage by 11 percentage points (pp) from 29% to 40% by 2032 was expected to avert a median 36% of new HIV acquisitions in Atlanta. Substantially larger increases (by 33 or 26 pp) in PrEP coverage (to 40% or 30%) were needed to achieve comparable reductions in Montreal and the Netherlands, respectively. A median 17 additional PYs on PrEP were needed to prevent one acquisition in Atlanta with 40% PrEP coverage, compared to 1000+ in Montreal and 4000+ in the Netherlands. Reaching 50% PrEP coverage by 2032 by recruiting CAB-LA users among PrEP-eligible MSM could avert >45% of new HIV acquisitions in all settings. Achieving targeted coverage 5 years earlier increased the impact by 5-10 pp. In the Atlanta model, PrEP expansions achieving 40% and 50% coverage reduced differences in PrEP access between PrEP-indicated White and Black MSM from 23 to 9 pp and 4 pp, respectively.

Conclusions: Achieving high PrEP coverage by offering CAB-LA can impact the HIV epidemic substantially if rolled out without delays. These PrEP expansions may be efficient in settings with high HIV incidence (like Atlanta) but not in settings with low HIV incidence (like Montreal and the Netherlands).

Keywords: Europe; HIV prevention; North America; PrEP; men who have sex with men; modelling.

Figure 1

Baseline model characteristics in Atlanta, Montreal and the Netherlands. (A) HIV incidence. (B) Proportion of the population and estimated HIV incidence in each sexual activity group in 2022. (C) Antiretroviral therapy (ART) and viral suppression in the PLHIV population (viral suppression was not simulated in the Netherlands model). (D) Projected tenofovir disoproxil and emtricitabine (TDF/FTC) use among MSM not living with HIV. Bold lines show means of all scenario replicates. Thin lines show individual simulations. Variation shown in the range of simulation lines comes from parameter variability in the deterministic compartmental models and stochastic variability in the agent‐based model. Abbreviations: PLHIV, people living with HIV; MSM, men who have sey with men

Figure 2

Pre‐exposure prophylaxis (PrEP) expansion scenario structure. (A) The baseline (counterfactual) scenario projected that tenofovir disoproxil and emtricitabine (TDF/FTC) usage will increase slightly over time from its current level in 2022 (29% Atlanta, 6% Montreal and 4% Netherlands) to 2042 (32% Atlanta, 10% Montreal and 5% Netherlands, black lines). Reaching a targeted coverage (40%) in 10 years (by 2032) required initiating new long‐acting cabotegravir (CAB‐LA) users (coloured lines). (B) The distribution of CAB‐LA and TDF/FTC use in each scenario depended on the prevalence of TDF/FTC use in the baseline scenario. Expansion with 0% CAB‐LA use caused only those in the expansion group to use CAB‐LA. Expansion with 50% CAB‐LA caused 50% of the baseline TDF/FTC users to instead use CAB‐LA, in addition to those in the expansion group who always use CAB‐LA.

Figure 3

Population effectiveness. (A) Acquisitions averted over 20 years (2022–2024) with pre‐exposure prophylaxis (PrEP) coverage achieved by 2032. Fifteen percent and 30% PrEP coverage levels were not modelled in Atlanta as baseline PrEP coverage was 29%. (B) Acquisitions averted with 40% PrEP coverage achieved by 2032 for different proportions of tenofovir disoproxil and emtricitabine (TDF/FTC) users switching to long‐acting cabotegravir (CAB‐LA). (C) Projected reductions in HIV incidence in 2032 relative to 2022 incidence in the baseline (white) and expanded PrEP coverage (colours) scenarios. A and C cumulate all TDF/FTC to long‐acting cabotegravir (CAB‐LA) switching scenarios. (D) Proportion of risk covered for different PrEP expansions, estimated by the proportion of PrEP coverage of each sexual activity group weighted by the proportion of new acquisitions expected to occur in that group. Notches in boxplot show 95% credible interval for the median. Dotted lines show full results range without outliers.

Figure 4

Population efficiency. Additional years on pre‐exposure prophylaxis (PrEP) needed to prevent one HIV acquisition with PrEP coverage achieved by 2032 in (A) Atlanta, (B) Montreal and (C) the Netherlands. Includes all tenofovir disoproxil and emtricitabine (TDF/FTC) to long‐acting cabotegravir (CAB‐LA) switching scenarios. Note different y‐axes. Notches in boxplot show 95% credible interval for the median. Dotted lines show full results range without outliers.

Figure 5

Effects of pre‐exposure prophylaxis (PrEP) expansion on racial disparities in the Atlanta model. (A) Racial differences in the proportion of men who have sex with men (MSM) with PrEP indications on PrEP with baseline and expanded PrEP coverage. Median PrEP coverage of MSM with no PrEP indication was 0% in all scenarios. (B) Acquisitions averted over 20 years in Black and White MSM with and without a PrEP indication. Notches in boxplot show 95% credible interval for the median. Dotted lines show full results range without outliers.

Figure 6

Additional pre‐exposure prophylaxis (PrEP) expansion scenarios. (A) PrEP expansion among PrEP‐eligible men who have sex with men (MSM) versus proportionally among all MSM not living with HIV. Acquisitions averted when PrEP expansion is only to those PrEP‐eligible (colours) or proportionally among all MSM not living with HIV (grey) with 40% PrEP coverage, including all tenofovir disoproxil and emtricitabine (TDF/FTC) to long‐acting cabotegravir (CAB‐LA) switching scenarios. (B) Prioritized interventions based on adherence to TDF/FTC in the Atlanta model. Poor TDF/FTC adherers switching to CAB‐LA (darker colours) and good TDF/FTC adherers switching to CAB‐LA (lighter colours) with 40% and 50% PrEP coverage achieved by 2032. Notches in boxplot show 95% credible interval for the median. Dotted lines show full results range without outliers.