A cross-sectional study was carried out between March and August 2023 in 30 screening and care centers for the treatment of AIDS, viral hepatitis and sexually transmitted infections spread across six health regions in Togo (Figure 1).

The study population is consisted of HIV-1-positive, antiretroviral-naive adults aged 18 and over and all of whom signed a free and informed consent form.

Information on socio-demographic characteristics such as age, gender, marital status, date of diagnosis and ART initiation, HIV-1 viral load and planned antiretroviral combination for the person were collected using a questionnaire. ART history was also collected from persons’ medical records.

The 2-ml plasma aliquots were stored at −20 °C at the collection sites and then at −80 °C at the national reference laboratory for HIV, viral hepatitis and sexually transmitted infections (LNR-VIH-HV-IST) until used for laboratory testing. Two rapid tests, SD Bioline™ HIV/SYPHILIS Duo (Abbott, Santa Clara, CA, United States) and First Response HIV 1–2. O Card Test version 2.0 (A1-302, GIDC, Sarigam 396,155. Dist. Valsad, Gujarat, India) were used to confirm HIV serology at the sampling sites. A participant was considered HIV-infected when both tests used were positive, in accordance with the national algorithm for HIV testing in Togo (27). Viral RNA was extracted using the GXT NA VER 1.0 extraction kit (BRUKER, Hain Lifescience GmbH) by mixing 250 μL plasma and 10 μL proteinase K. After incubation at room temperature for 10 min, the mixture is transferred to NorDIAG® equipment for extraction. The extraction protocol of 250 μL of plasma to 50 μL of eluate, comprising the steps of lysis, coupling of magnetic beads and RNA, washing and elution, was chosen (28). At the end of extraction, 10 μL of RNA were used to measure HIV-1 viral load using the GENERIC VIH-1 kit (Biocentric, Bandol, France) at the LNR-VIH/IST (29). Under these conditions, the quantification threshold was 390 RNA copies/ml plasma. Samples with viral load value greater than or equal to 6500.00 copies/mL (N = 336) were sent to the TransVIHMI—UMI233 laboratory (IRD, Montpellier, France) for resistance genotyping testing. The threshold of 6500.00 copies/mL was selected because it falls within the range of 5000.00 to 10000.00 copies/mL, from which NGS sequencing techniques show significantly higher sensitivity (30). HIV-1 nucleic acids were extracted from 250 μL of eluted in 60 μL of buffer using the magnetic silica Boom method on the Nuclisens EasyMAG automatic extractor (BioMerieux). Nested RT-PCRs using HotStarTaq DNA Polymerase Qiagen GmbH, Germany were performed to separately amplify fragments of the pol gene, coding, respectively, for protease (PR, covering amino acids from position 1 to position 99), part of reverse transcriptase (RT, covering amino acids from position 1 to 257) and part of integrase (IN, covering amino acids from position 37 to position 288) according to the ANRS protocol.¹ The genes encoding protease, part of reverse transcriptase and part of integrase were amplified by PCR according to the modified ANRS, 2023 protocol, including primers and PCR program. Briefly,

a first-round reverse transcriptase PCR was performed using Invitrogen’s SuperScript IV enzyme (Invitrogen, Carlsbad, CA), PR and RT were amplified using the same PCR program. IN amplification was performed using a different program. Five microliters of the amplicons from the first PCR run were used as template for nested PCR using HotStarTaq DNA Polymerase (Qiagen GmbH, Germany), including the primers and PCR program applied.

Amplification products were visualized by 1% agarose gel electrophoresis with ethidium bromide (BET). The three amplicons (PR, RT, IN) confirmed positive by electrophoresis were grouped inequal quantity for each sample, in a final volume of 60 μL then purified using a 0.6X ratio of Agencourt AMpure XP beads (BECKMAN COULTER, Life Sciences). Approximately 500 ng of product was used for library preparation using the Illumina DNA Prep kit according to the manufacturer’s instructions with average size measured using the Agilent High Sensitivity DNA kit on a Bioanalyzer 2,100 (Agilent Technologies, Santa Clara, CA). Once prepared and verified, the products were sequenced using next generation sequencing (NGS) technology on Illumina (Miseq) platform using the Miseq R v2 500 Cycle Reagent Kit. FASTQ files extracted from the Miseq platform were analyzed using the HIVdb-NGS pipeline,² applying a mutation analysis threshold of 20% to generate consensus sequences in FASTA format. The quality of the sequences was assessed with the Stanford tool. Sequences with a median read depth less than 10,000X were excluded from analysis to ensure sufficient depth to confirm the reliability of the existence of the observed minor mutation (31).

We assessed HIVDR mutations patterns and genotypes’ sensitivity to protease inhibitor (PIs), nucleoside and non-nucleoside reverse transcriptase inhibitors (NRTIs and NNRTIs) and integrase inhibitors (INI) using the Stanford HIVDB algorithm, version HIVDB 9.5.1 (last updated 05/11/2023, available at https://hivdb.stanford.edu/hivdb/by-mutations/).

For analysis of results and study conclusions, mutations were interpreted using the National Agency for Research on AIDS and Viral Hepatitis (ANRS) algorithm (version 35, [April 8,2024]) and classified into three categories: susceptible, confirmed resistance (presence of at least one major mutation) and possible resistance (accessory, polymorphic, or mixture mutations in the absence of major mutations). In cases of coexistence, the classification followed ANRS rules, with priority given to “confirmed resistance” (32). Prevalence was calculated as the proportion of sequences in each category relative to the total number of sequences analyzed (24). Identification of subtypes and circulating recombinant forms (CRF) was carried out using the Stanford site program.

Data were entered into Excel and analyzed using R software version 4.2.3 (R Foundation for Statistical Computing, Vienna, Austria). Qualitative variables were summarized as absolute and relative frequencies, along with their 95% confidence intervals (95% CIs) where appropriate. Quantitative variables were presented as medians with their interquartile ranges (IQRs). No comparative or inferential analyses were performed in this study according to the objectives.

This study was conducted in accordance with the ethical principles of the Declaration of Helsinki. The study protocol was approved by the National Bioethics Committee for Health Research (CBRS; Number 013/2023/CBRS; Supplementary File 1). Participation in the study was entirely voluntary. Written informed consent was obtained from all enrolled participants prior to data collection. All data were collected and analyzed anonymously to ensure confidentiality and privacy.

Among the 376 persons enrolled into the study, 40 (10.6%) were excluded for a viral load below 6500.00 copies/mL. At the end, of the 336 selected for PCR amplification of the protease, reverse transcriptase and integrase genes, nine samples were not selected for sequencing (six due to the limited number of places per run on Miseq and three due to failure of the three genes following amplification), i.e. 327 sequenced by NGS. At the end of NGS sequencing and after analysis of the quality of the runs, 321 samples with at least 1 sequenced gene and only 306 samples with the 3 validated PR/RT/IN sequences were selected for resistance analysis (Figure 2).

The median age was 39 years (IQR = 30–47 years), with 64.9% women. At the time of HIV diagnosis, 62.1% were in stage 1 of the WHO classification (Table 1). In this study, the viral loads of the 336 retained samples ranged from 6500.00 to 1000000.00 copies/mL, with a median of 174000.00 (IQR = [60100.00–621500.00]). The sociodemographic and clinical characteristics are presented in Table 1.

This study addresses the lack of information on the prevalence of primary HIV-1 resistance mutations among persons initiating antiretroviral therapy (ART) in Togo in the context of widespread use of dolutegravir (DTG). The main finding is the absence of primary resistance to protease inhibitors (PIs) and dolutegravir.

Our results are consistent with resistance data from other low-resource countries. The absence of PI resistance is comparable to that observed in Vietnam (0%) (33), China (0.2%) (34), Zimbabwe (1%) (35), and Madagascar (0%) (36) likely due to the limited use of PIs as first-line therapy and their high genetic barrier. The observed NRTI resistance rate (0.6%) is lower than those observed in China (3.5%) (34), Vietnam (1.2%) (33), and Zimbabwe (4%) (35). The two mutations detected (T215TS and K70KE) have already been documented in Togo by Dagnra et al. (11), which reinforces the need for continued surveillance given the widespread use of tenofovir in treatment regimens.

In contrast, the prevalence of NNRTI resistance of 6.3% found in our study corresponds to a moderate level according to the WHO classification. This level is above the 5% threshold that justifies increased vigilance in the choice of treatment regimens, but remains below the 15% threshold considered high and requiring a systematic change of treatment lines at the programmatic level (34). These data are consistent with results reported in other countries like in Vietnam (6.3%) (33), in China (3.4%) (34) and in Zimbabwe (15%) (35). The predominance of the K103N mutation, also reported in Madagascar (3.6%) (24), may reflect historical exposure to nevirapine during Prevention of Mother-to-child Transmission of HIV/AIDS (PMTCT) programs, particularly in the predominantly female population of our cohort (33, 34).

It is important to note that no dolutegravir resistance mutations were detected in our study, confirming findings from Zimbabwe and Madagascar (24, 32, 36). However, the detection of the T97A and E157Q mutations, associated with possible resistance against first-generation NNRTIs (e.g., raltegravir, elvitegravir), in treatment-naïve individuals, signals the possibility of transmitted resistance and underscores the need for integrase resistance surveillance in West Africa. Our results are in contrast to those from a pediatric cohort in Togo, reporting high levels of NRTI and NNRTI resistance in children and adolescents receiving DTG-based regimens who did not achieve virological suppression (37).

This discrepancy could be explained in particular by population differences: our study targeted ART-naïve adults, while the cited cohort included HIV-positive adolescents from Prevention of Mother-to-child Transmission of HIV/AIDS (PMTCT) programs who had already received treatment and were experiencing virological failure, a factor that can favor the selection of resistance, particularly in cases of suboptimal adherence or drug exposure.

The high proportion of the CRF02_AG and CRF06_cpx subtypes is consistent with previous findings from Togo (11, 34) and Mali (37, 38). The recently identified CRF45_cpx subtype expands the known molecular epidemiology of HIV in the region and could have implications for diagnostic test performance, vaccine development, and the evolution of drug resistance.

These results have several implications. The absence of resistance to dolutegravir (DTG) is reassuring and justifies its continued use as first-line treatment.

However, moderate resistance to NNRTIs and the presence of INSTI mutations warrant systematic resistance monitoring, particularly before initiating antiretroviral therapy, when resources permit.

The integration of basic resistance testing, particularly in high-risk or previously treated populations, could help adapt treatment regimens and reduce the risk of treatment failure (38).

Monitoring genetic diversity is essential to monitor the evolution of circulating strains, anticipate diagnostic challenges, and inform vaccine design (39).

In terms of limitations of our study, although we used a high viral load threshold (≥ 6500.00 copies/ml), we were unable to measure antiretroviral drug levels to exclude prior ART exposure, which could overestimate the prevalence of resistance. Second, the lack of phylogenetic analysis limits our understanding of potential transmission chains or resistance clusters. Finally, the cross-sectional nature of the study does not allow conclusions to be drawn on clinical or virological outcomes on the efficacy of dolutegravir-based regimens.