Milken Institute School of Public Health Poster Presentations (Marvin Center & Video)

Cross Sectional Analyses of HIV-1 Drug Resistance Mutations and Transmission Networks Present in People Living with HIV-1 in Washington, DC by Next Generation Sequencing: Comparison of Drug Resistance Mutation Reports Generated using a Publicly-Available Pipeline to that of a Commercial Pipeline

Poster Number

63

Document Type

Poster

Status

Staff

Abstract Category

Epidemiology and Biostatistics

Keywords

Next Generation Sequencing, HIV, Genotypic Drug Resistance Mutations, DC Cohort

Publication Date

Spring 2018

Abstract

Background: Detecting clinically relevant drug resistance mutations (DRMs), especially those conferring “high-level” resistance is important when making treatment decisions for people living with HIV (PLWH) and for public health surveillance efforts alike. Next Generation Sequencing (NGS) platforms are making possible detection of lower frequency DRMs that are not detectable using Sanger-based sequencing platforms. Study objectives included: 1) Describe our NGS approach using publicly-available software for data analysis and compare these results to those obtained using a commercial pipeline, 2) Determine what impact the selection in frequency cutoffs has on the number of DRMs detected for protease, reverse transcriptase (RT) and integrase gene targets from NGS data, and 3) Generate transmission networks to better understand potential linkages among an urban cohort of viremic PLWH. Methods: Plasma from 79 viremic participants was extracted and used to generate libraries for targeted NGS using Illumina® MiSeq platform. Sequences were uploaded into HyDRA Web developed by the Public Health Agency of Canada where minimum amino acid frequency cutoffs of ≥20%, ≥15%, ≥10% and ≥1% were selected to create separate DRM reports employing both the publicly available pipeline and SmartGene IDNSÒ 5 for HIV-1 Deep-Sequencing, along with the Stanford HIV Drug Resistance Database using the Genotypic Resistance Interpretation Algorithm. Consensus RT sequences at 20% were used to generate transmission networks using HIV-TRACE software developed by UCSD and Temple University. Results: There was 99.5% agreement (2172/2184 data points) seen between the DRM frequency reports obtained using the publicly-available pipeline and those obtained using the commercial pipeline. Highest rates of DRMs among these participants were seen for RT Inhibitors, with the next highest being for Integrase Strand Transfer Inhibitors. In all, 40% of DRMs conferring “high-level” resistance would have been missed had a >20% cutoff been used rather than a >1% cutoff. Transmission networks were constructed in 14.1% (11/78) of participants; most were Non-Hispanic Black males who reported being MSM or high-risk heterosexuals. Conclusions: DRM reports generated from NGS data analyzed using publicly-available software was validated against a commercial pipeline from SmartGene®. There is a risk-benefit issue in detecting low frequency DRMs; they may be an early indicator of emerging resistance leading to virologic failure, or they may not be clinically relevant. Future longitudinal studies are needed to assess this.

Creative Commons License

Creative Commons License
This work is licensed under a Creative Commons Attribution-Noncommercial-No Derivative Works 4.0 License.

Open Access

1

This document is currently not available here.

Share

COinS
 

Cross Sectional Analyses of HIV-1 Drug Resistance Mutations and Transmission Networks Present in People Living with HIV-1 in Washington, DC by Next Generation Sequencing: Comparison of Drug Resistance Mutation Reports Generated using a Publicly-Available Pipeline to that of a Commercial Pipeline

Background: Detecting clinically relevant drug resistance mutations (DRMs), especially those conferring “high-level” resistance is important when making treatment decisions for people living with HIV (PLWH) and for public health surveillance efforts alike. Next Generation Sequencing (NGS) platforms are making possible detection of lower frequency DRMs that are not detectable using Sanger-based sequencing platforms. Study objectives included: 1) Describe our NGS approach using publicly-available software for data analysis and compare these results to those obtained using a commercial pipeline, 2) Determine what impact the selection in frequency cutoffs has on the number of DRMs detected for protease, reverse transcriptase (RT) and integrase gene targets from NGS data, and 3) Generate transmission networks to better understand potential linkages among an urban cohort of viremic PLWH. Methods: Plasma from 79 viremic participants was extracted and used to generate libraries for targeted NGS using Illumina® MiSeq platform. Sequences were uploaded into HyDRA Web developed by the Public Health Agency of Canada where minimum amino acid frequency cutoffs of ≥20%, ≥15%, ≥10% and ≥1% were selected to create separate DRM reports employing both the publicly available pipeline and SmartGene IDNSÒ 5 for HIV-1 Deep-Sequencing, along with the Stanford HIV Drug Resistance Database using the Genotypic Resistance Interpretation Algorithm. Consensus RT sequences at 20% were used to generate transmission networks using HIV-TRACE software developed by UCSD and Temple University. Results: There was 99.5% agreement (2172/2184 data points) seen between the DRM frequency reports obtained using the publicly-available pipeline and those obtained using the commercial pipeline. Highest rates of DRMs among these participants were seen for RT Inhibitors, with the next highest being for Integrase Strand Transfer Inhibitors. In all, 40% of DRMs conferring “high-level” resistance would have been missed had a >20% cutoff been used rather than a >1% cutoff. Transmission networks were constructed in 14.1% (11/78) of participants; most were Non-Hispanic Black males who reported being MSM or high-risk heterosexuals. Conclusions: DRM reports generated from NGS data analyzed using publicly-available software was validated against a commercial pipeline from SmartGene®. There is a risk-benefit issue in detecting low frequency DRMs; they may be an early indicator of emerging resistance leading to virologic failure, or they may not be clinically relevant. Future longitudinal studies are needed to assess this.