Abstract

Next‐generation sequencing (NGS) is beginning to show its full potential for diagnostic and therapeutic applications. In particular, it is enunciating its capacity to contribute to a molecular taxonomy of cancer, to be used as a standard approach for diagnostic mutation detection, and to open new treatment options that are not exclusively organ‐specific. If this is the case, how much validation is necessary and what should be the validation strategy, when bringing NGS into the diagnostic/clinical practice? This validation strategy should address key issues such as: what is the overall extent of the validation? Should essential indicators of test performance such as sensitivity of specificity be calculated for every target or sample type? Should bioinformatic interpretation approaches be validated with the same rigour? What is a competitive clinical turnaround time for a NGS‐based test, and when does it become a cost‐effective testing proposition? While we address these and other related topics in this commentary, we also suggest that a single set of international guidelines for the validation and use of NGS technology in routine diagnostics may allow us all to make a much more effective use of resources. Copyright © ۲۰۱۴ Pathological Society of Great Britain and Ireland. Published by John Wiley & Sons, Ltd.

Introduction

The recent paper by Tothill et al in one of the recent issues of Journal of Pathology 1 illustrates an intelligent application of complex genomic information in a very specific clinical problem. Both fresh‐frozen and formalin‐fixed, paraffin‐embedded (FFPE) samples of patients with cancers of unknown primary (CUP) were analysed with next‐generation sequencing (NGS) technology. In 75% of the patients tested, the results revealed new therapeutic options, as well as certain signatures that are ‘aetiological’ in nature and, as such, are indicative of a likely tissue/organ of origin. Based on this and other prior studies, the authors suggest a pathway to integrate high‐throughput genomic information in the overall therapeutic decision‐making. This type of studies are taking us to a new scenario, already predicted when these technologies were made available in their first instance, beyond their clear role in the discovery of new molecular mechanisms of disease. Indeed, the work by Tothill et al, as well as other contemporary papers, is beginning to illustrate the diagnostic and therapeutic application of massively parallel sequencing (MPS) approaches for the reclassification of diagnosis 2, the therapeutic decision‐making beyond the classic organ‐specific treatment options 1 or the detection of standard‐of‐care mutations in comparison with the single‐gene analysis gold‐standard 3.