Pretreatment HIV-1 drug resistance, genetic diversity, and molecular transmission networks among individuals initiating antiretroviral therapy in Luzhou, China

Pretreatment drug resistance (PDR) may reduce the effectiveness of initial antiretroviral therapy (ART), but contemporary data from Luzhou, southwestern China, are limited. This study characterized HIV-1 genetic diversity, PDR, and molecular clustering among individuals initiating ART in Luzhou. In this cross-sectional study, 550 individuals with HIV-1 infection who were initiating ART and were diagnosed from March 2024 through February 2025 were included. Protease, reverse-transcriptase, and integrase regions of the pol gene were sequenced by Sanger sequencing. Drug resistance was interpreted using the Stanford HIV Drug Resistance Database; resistance was defined as low-level resistance or higher for any antiretroviral drug. A molecular network was constructed using Tamura-Nei 93 pairwise distances at a 1.5% threshold. Correlates of PDR were assessed using binary logistic regression. Pretreatment drug resistance (PDR) was the primary outcome, defined as low-level resistance or higher (Stanford score ≥ 15). Transmitted drug resistance (TDR) could not be directly measured because prior antiretroviral exposure could not be excluded; the WHO 2009 surveillance drug resistance mutation (SDRM) list was therefore used only to derive SDRM-based prevalence for comparison with WHO surveillance thresholds, not as a measure of TDR. The predominant genetic forms were CRF08_BC (179/550, 32.55%), CRF07_BC (178/550, 32.36%), and CRF01_AE (150/550, 27.27%). Overall PDR prevalence was 16.73% (92/550); class-specific prevalence was 13.09% for non-nucleoside reverse-transcriptase inhibitors (NNRTIs), 2.00% for nucleoside reverse-transcriptase inhibitors (NRTIs), 1.64% for protease inhibitors (PIs), and 0.91% for integrase strand-transfer inhibitors (INSTIs). By the WHO 2009 SDRM definition, the SDRM-based prevalence was 11.27% (62/550) overall and 8.91% (49/550) for NNRTIs. The network contained 224 participants, 302 links, and 71 connected components. Forty-five(45) resistant participants occurred in 22 components, and five (5) PI-resistant participants formed two (2) complete components: a three-participant CRF07_BC component in Longmatan District and a two-participant CRF08_BC component in Xuyong County. In the multivariable model, residence in Gulin County (adjusted odds ratio [aOR], 3.137; 95% confidence interval [CI], 1.502–6.552), residence in Naxi District (aOR, 2.576; 95% CI, 1.060–6.262), CD4 + T-cell count ≥ 200 cells/µL (aOR, 1.656; 95% CI, 1.003–2.734), and CRF08_BC infection (aOR, 2.251; 95% CI, 1.397–3.626) were associated with higher adjusted odds of PDR; these associations were exploratory. PDR, particularly NNRTI resistance, was common in this municipal cohort. The genetically linked PI-resistant components warrant focused epidemiological investigation, but the sequence data alone do not establish direct transmission. Continued local resistance surveillance and access to baseline genotyping where feasible may support selection of effective initial ART.