Next Generation Sequencing and Its Clinical Applications: The Growing Pains

Abstract

Short CommunicationAs the list of applications of next generation sequencing (NGS)-based assays continues to grow and the user network continues to expand from academic and pharmaceutical discovery research to clinical decision-making tests, the challenges and controversies continue to persist. Over the past year, the FDA held several workshops on the analytical and clinical validation of NGS tests that resulted in the release of two guidance documents, for which it has requested feedback from NGS stakeholders [1,2]. One key area of debate is defining the best method for establishing analytical validity of NGS tests (use of standards or the use of processes such as quality system regulations, QSRs). Some feel that QSRs are overly burdensome, impractical and cost-prohibitive; others feel the use of standards may be insufficiently rigorous. Our concern is how the standards would be implemented. Another question is related to whether or how clinical labs should confirm novel variant calls. For example, if a whole exome sequencing (WES) of a sample results in 500 somatic mutation calls, it would be essentially impossible (both cost prohibitive and time-consuming) to confirm all 500 variants using an orthogonal method. In some clinical or hospital settings, verifying specific decision-making variant calls for a given patient with special disease condition using orthogonal methods might be feasible or justifiable. However, if the assay is to be used for the determination of patients' hyper-mutation status within a clinical trial, it might be unrealistic to confirm every potential variant call.A third controversy is whether clinical databases could be used for clinical validation of novel mutations not in the literature, especially those mutations for which analytical confirmation was not explicitly performed. Unfortunately, if a public database of genotype-phenotype associations were created using historic data with limited variant confirmation or reproducibility measures in place, even if the data were from reputable labs, using such a database to support clinical validity of an NGS-based in vitro diagnostic might have unintended consequences, and could even increase the risk of getting incorrect diagnoses in the clinic [3,4]. Of course, if the database only contained those specific variants directly supported by the clinical evidence of genotype-phenotype association on a variant-by-variant basis, then that might be acceptable and useful. We can illustrate our concern with an extreme example: suppose a patient's sample that has 500 variants derived from its WES data is confirmed to be a responder for a given treatment, are we saying that now all 500 variants are considered validated and should be deposited into this public database? We would argue against this for several reasons, not the least of which would be the lack of confirmation of each mutation.As for the use of standards in NGS-based in vitro diagnostics for germline diseases, although it should have substantial value in terms of serving as control samples...

A Novel Validation Strategy for Next Generation Sequencing Mutation Profiling in FFPE Tissues

“…Furthermore, one should consider running triplicate library preparations for every single sample in order to optimize the NGS bioinformatics pipeline and generate reproducible mutation profiling clinical data, since we also found that the same mutation in different samples or different mutations in the same sample could all be measured differently in terms of their detection sensitivity [1,[11][12][13]]. …”