A national cross-sectional analysis of surveillance drug resistance mutations among recently diagnosed antiretroviral naïve Brazilian people with HIV

Abstract

Background: Brazil pioneered free and universal HIV antiretroviral therapy (ART) access in 1996. This study evaluated the prevalence of surveillance drug resistance mutations (SDRM) among individuals newly diagnosed with HIV and initiating ART.

Methods: Using WHO HIV Threshold Survey methodology, participants were recruited from seven major cities, representing all Brazilian macro-regions. Dried blood spots collected between August 2021 and November 2023 were transported at room temperature to a central laboratory for genotyping.

Findings: A total of 327 samples were analyzed, 271 (82.9%) were males. The self-reported ethnic distribution was obtained from 257 people, including 92 white (35.8%), 115 brown (44,7%), 46 black (17.9%), 2 indigenous (0.8%), and 2 Asian (0.8%) people. Mean/median CD4+ T-cell counts were 377/325 cells/μL, and mean/median viral load was 468,000/71,800 copies/mL. Overall SDRM prevalence was 11.6%, with regional values of 16.0% (Rio de Janeiro), 11.5% (Santos), 13.0% (Belém), 11.6% (Porto Alegre), 10.6% (Itajaí), 10.6% (Brasília), and 9.0% (Salvador). Resistance prevalence by class was 3.4% for nucleoside reverse transcriptase inhibitors, 5.2% for non-nucleoside reverse transcriptase inhibitors, 1.5% for protease inhibitors, and 1.8% for integrase strand transfer inhibitors (INSTI). One individual harbored the dolutegravir-associated R263K mutation. Compared to a prior survey conducted a decade earlier, subtype C prevalence increased markedly, and a probable new circulating recombinant form was detected.

Interpretation: Nationwide SDRM prevalence in Brazil remains stable relative to data from 10 years ago. However, the emergence of low-level INSTI SDRM, including dolutegravir-specific resistance, is noteworthy. These results emphasize the importance of continuous SDRM and HIV genetic diversity surveillance. Sustained molecular monitoring is essential for optimizing prevention strategies and preserving long-term treatment effectiveness.

Funding: FAPESP and CNPq.

Keywords: Antiretroviral therapy; Brazil; Dried blood spots; Genotyping; HIV-subtype; Recent HIV diagnosis; Surveillance drug resistance mutations.

Fig. 1

Study Flow Chart. RT, reverse transcriptase.

Fig. 2

Prevalence of Resistance-Associated Mutations to Nucleoside Reverse Transcriptase Inhibitors (NRTI, panel a), Non-nucleoside Reverse Transcriptase Inhibitors (NNRTI, panel b), Protease Inhibitors (PI, panel c), and Integrase Strand Transfer Inhibitors (INSTI, panel d). In red are SDRM mutations, and in blue are the other mutations that can be either selected by antiretrovirals or HIV clade-related polymorphisms. ∗Mutations not counted in the prevalence of SDRM due to hypermutation. ∗∗ Only one of the sequences harboring this mutation was hypermutated and not counted in the prevalence of SDRM. #Mutation not counted in the prevalence of SDRM due to lack of confirmation of its impact on resistance.

Fig. 3

Brazilian map showing the location of the cities and the Brazilian macro-regions where samples were collected. The prevalence of the subtype profile is indicated for each city. Regions are N, North; NE, Northeast; CW, Centralwest; SE, South-east; and S, South. At the bottom, subtype profiles are presented in percentages, where the first letter describing the recombinant subtype refers to the 5′ region of the pol gene, whereas the second letter refers to the 3′ region in a concatemer combining the protease + RT (first letter) and integrase regions (second letter: BC, BF, CB, etc.).

Fig. 4

Maximum-likelihood phylogenetic trees based on HIV-1 protease (294 bp) (a), reverse transcriptase (603 bp) (b), and integrase (594 bp) (c) protein-encoding genetic regions of the pol gene. Colored tips correspond to distinct HIV-1 reference subtypes. Blacktip codes represent sequences from this study. Asterisks on the main branches assume SH-like support values, with only values equal to or greater than 70% shown. Branch lengths are drawn to scale and measured in nucleotide substitutions per site, as indicated by the scale bar in the figure.