357 - Clinical and cost impact of long-acting ART in South African youth with HIV: A modeling analysis

Publication Number: 2346.357

Xinwen Xu, Massachusetts General Hospital, Boston, MA, United States; Catherine Orrell, South African Medical Research Council, CAPE TOWN, Western Cape, South Africa; Clare Flanagan, Massachusetts General Hospital, Boston, MA, United States; Lauren Jennings, Desmond Tutu Health Foundation, Cape Town, Western Cape, South Africa; Aaron Wu, Massachsetts General Hospital, Malden, MA, United States; Claire Keene, University of Oxford, Oxford, England, United Kingdom; Sujata Tewari, The Warren Alpert Medical School of Brown University, Providence, RI, United States; John Giardina, Massachusetts General Hospital, Boston, MA, United States; Kenneth Freedberg, Massachusetts General Hospital, Boston, MA, United States; Andrea Ciaranello, Massachusetts General Hospital, Southborough, MA, United States; Anne Neilan, Massachusetts General Hospital, Boston, MA, United States

Background: Youth with HIV (YHIV) aged 15-24 years in South Africa (SA) experience challenges adhering to daily oral antiretroviral therapy (ART) compared to older people with HIV, leading to substantial short- and long-term morbidity and mortality. Long-acting injectable cabotegravir/rilpivirine (CAB/RPV) could reduce adherence barriers but is not available for treatment outside research settings.

Objective: To evaluate the clinical impact, cost and cost-effectiveness of CAB/RPV for YHIV in SA.

Design/Methods: Using the CEPAC model, we simulate YHIV in care and prescribed ART (All_AYA, n=224,000), including those with oral ART adherence challenges (Low_Adh, n=20,000) in SA. We compare two strategies, SOC: standard-of-care oral tenofovir/lamivudine/dolutegravir (TLD) ART versus LA-ART: two-monthly CAB/RPV with transition to TLD if still on CAB/RPV at age 25. Key inputs, informed by published data, for All_AYA/Low_Adh include starting age (mean 20.5 years), sex (70% female), mean CD4 count (447/152 cells/µL), baseline viremia (21/100%), mean ART adherence (% pills taken/month: 74/18%), 24-month retention (91/82%), 1st-line ART efficacy (SOC: 95%, LA-ART: 94-97%) and costs/year (oral ART: $30-280, CAB/RPV: $270, HIV-related care: $130-1,290). Outcomes include life expectancy (LE), costs, and incremental cost-effectiveness ratios (ICERs, $/life-year-saved [LYS], discounted at 3%/year. We vary key inputs and willingness-to-pay thresholds (base case: 50% South Africa's GDP [$3,010/LYS]) in sensitivity analyses.

Results: Over 10 years, LA-ART would yield modest gains in LE for All_AYA (8.8y in LA-ART vs 8.5y for SOC) with larger gains for Low_Adh (6.0y vs 4.6y, table). Over a lifetime, LA-ART would lead to greater LE than SOC in both cohorts (All_AYA: 31.0y vs 29.8y; Low_Adh: 14.9y vs 10.3y), corresponding to 278,000 and 95,000 population-level LYS with ICERs <$3,010/LYS (All_AYA: $1,174/LYS; Low_Adh: $425/LYS). ICERs would remain <$3,010/LYS at yearly drug prices ≤$678 for All_AYA and ≤$3,391 for Low_Adh. Cost-effectiveness conclusions were robust to input variations (Figure). In scenario analyses, we extend CAB/RPV use beyond age 25, which further increases projected LE (All_AYA: 31.9y; Low_Adh: 17.1y) and ICERs (All_AYA: $2,557/LYS; Low_Adh: $586/LYS), with ICERs exceeding $3,010/LYS at yearly CAB/RPV prices ≥$310 for All_AYA, ≥$1,720 for Low_Adh.

Conclusion(s): Prioritizing access to LA ART during the critical developmental period of adolescence and young adulthood could markedly increase long-term survival for YHIV. LA ART could be cost-effective for South African YHIV, particularly those with adherence challenges.

Table. Model-projected outcomes: long-acting cabotegravir/rilpivirine (LA-ART) versus oral tenofovir/lamivudine/dolutegravir (SOC) among youth with HIV in South Africa


One-way sensitivity analyses of LA-CAB/RPV cost-effectiveness for youth with HIV in South Africa, lifetime horizon
This tornado diagram shows results from sensitivity analyses - specifically, the impact of variations in model inputs on projected cost-effectiveness results. The incremental cost-effectiveness ratio (ICER, in $/life-year-saved [LYS]) is displayed on the x-axis. Each row represents an individual model input parameter varied in sensitivity analyses. The range through which each input is varied is shown in parentheses, with the first value correlating to an ICER reduction compared to the base-case ICER and the second value correlating to an ICER increase. Results for the All_AYA cohort are displayed above in panel A; results for the Low_Adh sub-cohort are displayed in panel B. Larger bars indicate parameters to which conclusions are more sensitive. The left-most vertical line indicates the base-case ICER. The second vertical line indicates the base-case willingness-to-pay threshold, set at $3,010/LYS.