Evaluation of two bioinformatic algorithms for the interpretation of HIV-1 drug resistance and subtyping in Cameroon: Translational application for ART optimization in low-and middle-income countries.

BACKGROUND

Efficient monitoring of HIV drug resistance (HIVDR) relies on standardized bioinformatics tools for accurate identification of drug resistance mutations (DRMs). Thus, we sought to compare the concordance of HIV-1 genotypic profiling from sequences analyzed with two commonly-used editing algorithms in low- and middle-income countries (LMICs).

METHODS

A laboratory-based comparative study was conducted among treatment-experienced people living with HIV attending the Chantal BIYA International Reference Centre in Yaoundé-Cameroon from October-2022 through July-2023. For each individual, raw data of HIV-1 sequences were analyzed simultaneously using RECall (semi-automated) vs. Exatype (automated) algorithms. Outputs were compared for DRMs, polymorphisms and subtyping between the two algorithms, with significance at p<0.05.

RESULTS

Overall, 221 participants were included (mean-age 32±15 years, 52.5% female). Validation of sequence quality was 70.1% (155/221) by RECall vs. 60.2% (133/221) by Exatype, Ka=0.78 (p<0.0001), indicating a good agreement between both algorithms. Importantly, a perfect concordance (100%) was found in HIV-1 clade inference (CRF02_AG [82/82], A1 [29/29], G [5/5], F2 [5/5] and others [12/12]). Similarly, high concordances were found for the identification of DRMs to protease-inhibitors (99.0%), nucleoside reverse-transcriptase inhibitors (98.0%), non-nucleoside reverse-transcriptase inhibitors (98.6%) and integrase-inhibitors (100.0%). The average turn-around-time was two-folds longer with RECall (5.5±1.7 min) vs. Exatype (2.5±1.1 min); giving a lower efficiency (i.e. validation rate/time) with RECall (12.7) vs. Exatype (24.1).

CONCLUSIONS

Semi-automated (RECall) and automated (Exatype) tools showed excellent agreement in detecting HIV-1 clades and DRMs, supporting their interoperability in clinical practice. Following efficiency, Exatype can be considered preferential, while RECall remains a quite suitable alternative for LMICs.