HIV-1 genotypic resistance testing using single molecule real-time sequencing

• DOI: 10.1016/j.jcv.2024.105717
• Publication Date: 2024-07-24T19:59:10Z
• Authors: Stéphanie Raymond; Nicolas Jeanne; Camille Vellas; Florence Nicot; Karine Saune; Noémie Ranger; Justine Latour; Romain Carcenac; Agnès Harter; Pierre Delobel; Jacques Izopet
• Keywords: Genotype; Genotyping Techniques; Anti-HIV Agents; High-Throughput Nucleotide Sequencing; HIV Infections; Microbial Sensitivity Tests; Sequence Analysis, DNA; Viral Load; Sensitivity and Specificity; Drug Resistance, Viral; Mutation; HIV-1; Humans; RNA, Viral; Prospective Studies

HIV-1 resistance testing is recommended in clinical management and next-generation sequencing (NGS) methods are now available in many virology laboratories.To evaluate the diagnostic performance of Long-Read Single Molecule Real-time (SMRT) sequencing (Sequel, PacBio) for HIV-1 polymerase genotyping.111 prospective clinical samples (83 plasma and 28 leukocyte-enriched blood fraction) were analyzed for routine HIV-1 resistance genotyping using Sanger sequencing, Vela NGS, and SMRT sequencing. We developed a SMRT sequencing protocol and a bio-informatics pipeline to infer antiretroviral resistance on both haplotype and variant calling approaches.The polymerase was successfully sequenced by the three platforms in 98 % of plasma RNA samples for viral loads above 4 log copies/mL. The success rate decreased to 83 % using Sanger or Vela sequencing and to 67 % using SMRT sequencing for viral loads of 3 to 4 log copies/mL. Sensitivities of 50 %, 54 % and 61 % were obtained using SMRT, Vela, and Sanger sequencing, respectively, in cellular DNA from patients with prolonged undetectable plasma HIV-1 RNA. Ninety-eight percent of resistance-associated mutations (RAMs) identified with Sanger sequencing were detected using SMRT sequencing. Furthermore, 91 % of RAMs (> 5 % threshold) identified with Vela NGS were detected using SMRT sequencing. RAM quantification using Vela and SMRT sequencing was well correlated (Spearman correlation ρ = 0.82; P < 0.0001).SMRT sequencing of the full-length HIV-1 polymerase appeared performant for characterizing HIV-1 genotypic resistance on both RNA and DNA clinical samples. Long-read sequencing is a new tool for mutation haplotyping and resistance analysis.