Long-acting injectable cabotegravir accounted for only a small share of HIV pre-exposure prophylaxis use in the US through 2024, and persistence declined from about one-half of patients at 1 year to fewer than one-quarter at 2 years.
In a retrospective cohort study, researchers analyzed national pharmacy and medical claims from the Symphony Health PatientSource database from January 2022 through December 2024. The analysis included 781,040 patients with at least one claim for pre-exposure prophylaxis (PrEP), including oral medications or long-acting injectable cabotegravir. The primary outcomes were uptake of long-acting injectable cabotegravir (LAI-CAB) and persistence with injectable therapy. A secondary outcome assessed overall PrEP persistence among LAI-CAB users, including patients who transitioned to oral PrEP.
During the study period, 24,194 patients received LAI-CAB, representing 3% of all PrEP users. By July through December 2024, LAI-CAB accounted for 4% of patients receiving PrEP nationally. Among patients with at least 1 year of follow-up, 50% remained persistent with LAI-CAB, while 57% remained engaged in PrEP overall after accounting for transitions to oral therapy. Among patients with at least 2 years of follow-up, LAI-CAB persistence declined to 23%, and overall PrEP persistence declined to 30%.
Researchers found that LAI-CAB users differed from oral PrEP users in several respects. A greater proportion of LAI-CAB users were covered by Medicaid (26% vs 14%), and female patients accounted for a larger share of LAI-CAB users than oral PrEP users (15% vs 11%). However, long-term persistence was lower among female patients, with 13% remaining on LAI-CAB at 2 years compared with 25% of male patients. Persistence was generally similar across racial and ethnic groups. Older patients were more likely to remain on therapy than younger patients, and Medicare beneficiaries demonstrated higher long-term persistence than patients covered by commercial insurance or Medicaid.
Use of LAI-CAB increased steadily during the study period, rising from 704 users in the first half of 2022 to 16,557 users in the second half of 2024. Despite that growth, injectable therapy remained a small proportion of overall PrEP use. The researchers noted that persistence observed in routine clinical practice was substantially lower than rates reported in clinical trials and smaller observational studies. They suggested that structural barriers—including insurance requirements, medication acquisition, and clinic logistics—may contribute to lower uptake and persistence, although the study was not designed to determine why patients discontinued or switched therapy.
The study had several limitations. It relied on administrative claims data and could not capture reasons for discontinued or switched therapies. The database excluded some closed US health care systems, nearly one-half of race and ethnicity data required statistical imputation, and information on gender identity and sexual risk behaviors was unavailable.
The findings suggest that wider availability of long-acting PrEP alone may not substantially increase PrEP use without improving long-term persistence and addressing barriers to implementation.
"The implementation of more effective new PrEP modalities alone is unlikely to substantially increase PrEP use in the US, and future research should examine structural supports and behavioral interventions to facilitate scale-up of new, highly efficacious modalities," wrote lead study author Shi Hao Ernest Koh, MPH, of the Rollins School of Public Health at Emory University, and colleagues.
Disclosures: Mr Corbin-Gutierrez reported grants from Gilead Sciences and grants and personal fees from ViiV Healthcare outside the submitted work. Dr Sullivan reported grants and personal fees from Gilead Sciences and grants from Merck outside the submitted work. Dr Siegler reported grants from Gilead Sciences, ViiV, and Merck (paid to Emory University) outside the submitted work. The study was funded by the National Institute of Allergy and Infectious Diseases.
Source: JAMA Network Open
