Indirect clinical validation for predictive biomarkers in oncology: International Quality Network for Pathology (IQN Path) Position Paper
Emina Torlakovic, Raed Al Dieri, Tony Badrick, Zongming Eric Chen, Carol C. Cheung, Zandra C. Deans, Andrew Dodson, Francesca Fenizia, Hiroshi Kijima, Joerg Maas, Antonio Martı́nez, Søren Nielsen, Simon Patton, Etienne Rouleau, Peter Schirmacher, Tanuja Shet, Tracy Stockley, Nicola Normanno
Abstract
Validation of biomarker assays is mandatory not only for their applications in clinical trials but also for their subsequent transfer to clinical laboratories in routine clinical care. There are two critical components relevant to their transfer to clinical practice: regulatory oversight and methodology transfer. Both aspects are simplified where companion diagnostic (CDx) assays relevant to a given indication are being implemented in clinical laboratories. However, when laboratory developed tests (LDTs) are being used either because CDx is not available or because LDT is preferred, both aspects need special consideration from regulatory agencies as well as clinical laboratories. The key component that links these two aspects is evidence of validation of the new LDTs. For predictive and prognostic biomarkers in oncology, clinical validation is feasible only in clinical trials. This approach is not available or feasible to clinical laboratories that develop LDTs. While clinical laboratories routinely perform technical/analytical validation, depending on the type of biomarker, this may not be sufficient to provide evidence of the LDT's clinical relevance. Laboratories must perform and document their assessment for the need for indirect clinical validation. When indirect clinical validation is required, it must be performed according to existing guidelines for this purpose. This paper provides expert consensus guidance and recommendations on how to assess for the need for indirect clinical validation and how to perform indirect clinical validation where required. This paper also provides a conceptual framework to regulatory agencies for determining requirements for validation of predictive and prognostic biomarkers in oncology.