Real-world prevalence of integrase inhibitor resistance and virological failure since adoption as guideline-preferred therapy

• Published in: Drugs in Context Vol. 13; pp. 1 - 10
• Main Authors: Januszka, Jenna, Drwiega, Emily, Burgos, Rodrigo, Smith, Renata, Badowski, Melissa
• Format: Journal Article
• Language: English
• Published: England BioExcel Publishing Ltd 16.05.2024
• Subjects: antiretroviral therapy; hiv/aids; integrase inhibitors; Original Research; resistance mutations; HIV/AIDS; integrase inhibitors; antiretroviral therapy; resistance mutations
• ISSN: 1745-1981; 1740-4398
• DOI: 10.7573/dic.2023-12-4

Limited data reporting real-world prevalence of integrase strand transfer inhibitor resistance (INSTI-R) in the USA are available because their recommendation as first-line treatment in 2017. Reported national surveillance data in the USA estimated INSTI-R to be 6.3% as of 2018. This article aims to describe estimated prevalence of INSTI-R within a single clinic network in Chicago, IL, USA, and identify risk factors for resistance and virological failure (VF). This was a retrospective, single-centre study of adults with HIV starting an INSTI-containing regimen between September 2017 and 2020. The primary endpoint was the difference in INSTI-R of the sample population compared with the national prevalence. Other outcomes included VF and documented INSTI-R mutations. Of 948 participants screened, 321 were included. Eight people had baseline INSTI-R testing results available, of which five had INSTI-R at baseline for an estimated prevalence of 1.6%. This estimation was significantly less than the national estimated prevalence of 6.3% ( <0.001). VF occurred in 26 (7.8%) individuals. Because no participants acquired INSTI-R during the study period, investigators were unable to identify risk factors associated with the development of INSTI-R. People with high pre-treatment viral loads had 1.21 (95% CI 1.05-1.39) higher odds of VF. Amongst participants on INSTI-containing regimens, INSTI-R rates were estimated to be lower than the estimated national prevalence. Detectable pre-switch viral loads were more associated with VF than undetectable viral loads.