As next-generation sequencing (NGS) costs gradually decline, the Food and Drug Administration’s (FDA) Division of Antiviral Products (DAVP) anticipates that more companies will make the switch to NGS for future antiviral drug resistance analyses and other additional uses.
The push to more NGS use comes as DAVP independently assessed NGS resistance data for three new drug applications (NDAs), which taught the agency a few important lessons and will help it prepare for the next wave of applications, DAVP’s Eric Donaldson, Damon Deming, Julian O’Rear and Lisa Naeger note in an editorial in Biomarkers in Medicine.
Historically, genotypic antiviral resistance data were generated by Sanger sequencing and submitted by companies to the FDA’s DAVP using standardized datasets.
The emergence of less-costly NGS technologies, however, has allowed for “deeper sequencing of viral populations,” though because NGS is still emerging, it also “presents many data analysis and data integrity issues that must be considered when conducting a regulatory review,” the authors say.
Some of these issues include:
- Multiple sequencing platforms with different sequencing lengths and error rates require bioinformatics analysis using a platform-agnostic analysis pipeline.
- No standardized NGS analysis pipelines exist despite the existence of hundreds of algorithms for performing assemblies and alignments, calling variants and identifying issues with genomic structure.
- Different algorithms often require proprietary scripts or programs to tie them into an analysis pipeline and the results of different programs “can vary greatly depending upon the algorithm used.”
As far as how companies can submit NGS data for resistance analysis in support of an NDA, the authors say that in general, data “can be submitted along with the appropriate details for the sequencing platform, the protocols to be used for sample preparation, the raw NGS data and a detailed description of the methods used to analyze the data.”
DAVP also recommends that companies communicate early in the NDA process and provide DAVP with these details prior to submitting the NGS data.
More specifically, the authors note that the independent analyses of the resistance data for the three NDAs already conducted was used to determine if: “treatment-emergent amino acid substitutions can be correlated with treatment failure (emergence of resistance), resistant variants persist after treatment failure, shifts in cell culture susceptibility can be associated with treatment failure, baseline polymorphisms lead to reduced efficacy and resistance-associated substitutions confer cross resistance to other antiviral drugs.”
Those analyses also taught the agency a few lessons, particularly on the cross contamination of barcoded samples, the sensitivity of NGS compared with Sanger sequencing, how NGS can provide a more accurate picture of resistance pathways, how reverse transcription polymerase chain reaction (RT-PCR) artifacts can introduce false positives, and -- what they deemed to be “the most important lesson” – “that low frequency treatment-emergent substitutions of interest have been identified.”
In addition to the resistance analyses, the authors predict there will be a number of additional uses of NGS data in the future.
“Several other submissions of NGS data have been received or are anticipated in the coming months,” they write. “In addition to resistance analyses, we expect to see additional uses of NGS, including for viral tropism assessments, characterization of virus stocks for studies supporting drugs seeking approval using the Animal Rule, metagenomics studies looking at the impact of antivirals and antimicrobials on the normal flora and eventually for use in precision medicine where the host exome or genome may be used to identify subjects who will have the best chance of responding to a specific antiviral treatment.”
Biomarkers in Medicine Study