Virological failure and HIV drug resistance among adults living with HIV on second-line antiretroviral therapy in the Asia-Pacific

• Main Authors: Ross, Jeremy, Jiamsakul, Awachana, Kumarasamy, Nagalingeswaran, Azwa, Raja Iskandar Shah Raja, Merati, Tuti Parwati, Do, Cuong Dat, Lee, Man-Po, Ly, Penh S., Yunihastuti, Evy, Nguyen, Kinh Van, Ditangco, Rossana, Ng, Oon Tek, Choi, Jun Yong, Oka, S., Sohn, Annette H, Law, Matthew G.
• Format: Article
• Published: Wiley 2021
• Subjects: R Medicine; RB Pathology; Theories of disease. Etiology. Pathogenesis
• Online Access: http://eprints.um.edu.my/34654/

To assess second-line antiretroviral therapy (ART) virological failure and HIV drug resistance-associated mutations (RAMs), in support of third-line regimen planning in Asia. Methods Adults > 18 years of age on second-line ART for >= 6 months were eligible. Cross-sectional data on HIV viral load (VL) and genotypic resistance testing were collected or testing was conducted between July 2015 and May 2017 at 12 Asia-Pacific sites. Virological failure (VF) was defined as VL > 1000 copies/mL with a second VL > 1000 copies/mL within 3-6 months. FASTA files were submitted to Stanford University HIV Drug Resistance Database and RAMs were compared against the IAS-USA 2019 mutations list. VF risk factors were analysed using logistic regression. Results Of 1378 patients, 74% were male and 70% acquired HIV through heterosexual exposure. At second-line switch, median interquartile range (IQR)] age was 37 (32-42) years and median (IQR) CD4 count was 103 (43.5-229.5) cells/mu L; 93% received regimens with boosted protease inhibitors (PIs). Median duration on second line was 3 years. Among 101 patients (7%) with VF, CD4 count > 200 cells/mu L at switch odds ratio (OR) = 0.36, 95% confidence interval (CI): 0.17-0.77 vs. CD4 <= 50) and HIV exposure through male-male sex (OR = 0.32, 95% CI: 0.17-0.64 vs. heterosexual) or injecting drug use (OR = 0.24, 95% CI: 0.12-0.49) were associated with reduced VF. Of 41 (41%) patients with resistance data, 80% had at least one RAM to nonnucleoside reverse transcriptase inhibitors (NNRTIs), 63% to NRTIs, and 35% to PIs. Of those with PI RAMs, 71% had two or more. Conclusions There were low proportions with VF and significant RAMs in our cohort, reflecting the durability of current second-line regimens.