Next Generation Sequencingadmin
Next Generation Sequencing Provides Comparable Data to Sanger Sequencing for the Detection of HIV Protease Inhibitor Mutations
Mukhlid Yousif1,2, Sergio Carmona3,4, Johanna Ledwaba1, Kim Steegen3,4, Lynn Morris1,2 and Gillian Hunt1,2
National Institute for Communicable Diseases, Centre for HIV and STI, Johannesburg, South
Africa1, University of the Witwatersrand, Virology, Johannesburg, South Africa2, University of the Witwatersrand, Molecular Medicine and Haematology, Johannesburg, South Africa3, National Health Laboratory Service, Johannesburg, South Africa4
South African national guidelines recommend HIV drug resistance (HIVDR) testing for all HIV-infected patients failing a protease inhibitor (PI) based regimen. Currently, Sanger sequencing is used for HIVDR testing, however the implementation of next generation sequencing (NGS) could improve sensitivity, reduce cost and increase specimen throughput. The aims of this study were to compare Sanger and Illumina NGS methods for their ability to detect PI drug resistance mutations among patients failing a PI regimen. A total of 162 specimens with PI mutations detected by Sanger sequencing were selected for this study. A 1.7kb fragment spanning protease and reverse transcriptase genes was amplified using an in-house PCR assay. Library preparation was performed using a Nextera XT kit. Paired-end libraries were indexed in a single MiSeq run (96 specimens) and sequenced using MiSeq Reagent kit V3. FastQ files were analysed using DeepChek®. A consensus sequence was generated for each specimen to compare with Sanger sequences. The validation process of the NGS and Sanger sequences was done successfully using phylogenetic analysis and pairwise analysis. Using a 15% cut-off for generating a consensus sequence, 155/162 (95.7%) of specimens showed similar HIVdb resistance mutation scores between Sanger and NGS, while 7 specimens (4.3%) showed discordance between the two techniques. When using 5-15% consensus cut-off, an additional 13 specimens (8%) showed discordance. In conclusions, detection of DRMs using MiSeq and Sanger sequencing showed high concordance (95.7%). The difference in the HIVdb resistance mutation scores was due to discrepancies in mutations detected. The discrepancies had minor impact on clinical interpretation as all scores were >15 for LPV/r and ATZ/r, which is the cut-off for switching patients to 3rd line therapy. The use of NGS for HIVDR testing is therefore reliable and allows for large specimen numbers to be tested in a more efficient workflow.