Parallelism: the foundation of biomarker assay development and validation

In recent years, there has been growing appreciation that biomarker assay development and validation must be considered in their own context and not incongruously fitted into concepts/approaches developed to evaluate PK assay performance. This opinion, succinctly coined as ‘Biomarker assays are not PK assays’ in the LBA sessions at the Crystal City VI Workshop on bioanalytical method validation for biomarkers [1,2] has since become the basis for driving discussions on the appropriate scientific approaches for biomarker assay development and validation, with characterization of parallelism being a foundational component. Over the last ten years, there has been an increasing amount of literature in the ligand-binding assay field discussing the relative importance of evaluating parallelism in the context of both pharmacokinetic and biomarker assay development and validation [3–6]. Importantly, in June 2017 a public workshop entitled ‘Regulatory considerations for the analytical validation of assays used in the qualification of biomarkers in biological matrices’ was held in Washington DC, USA with the goal of creating alignment on the evidentiary considerations for the analytical validation of biomarker assays within their specific context(s) of use [7]. Discussions at that workshop highlighted that there remain gaps in both breadth and depth of conceptual understanding of parallelism among professionals across industry, academia and regulatory agencies. Here, we summarize the critical concepts associated with parallelism assessments in LBA biomarker assay development and validation, including key discussion points raised at the public workshop.