Development and validation of the biomarker toolkit: a tool aiming to quantifiably assess biomarker utility and guide development.

Abstract

Introduction: Increased resources have been spent on cancer biomarker discovery, but very few of these biomarkers have been clinically adopted. To help bridge the gap between biomarker discovery and clinical use, this PhD thesis will develop the Biomarker Toolkit, a tool aiming to evaluate biomarker candidates and identify most clinically promising biomarkers worthy of additional research. Methods and Results: To generate the Biomarker Toolkit, mixed methodology was used including systematic literature search and qualitative techniques (semi-structured interviews two-stage Delphi Survey). The Biomarker Toolkit checklist was created following PRISMA guidelines. Upon article selection and screening, 129 characteristics were identified to be linked with a clinically useful biomarker. These characteristics were grouped in four main categories including: Rationale, Analytical Validity, Clinical Validity and Clinical Utility. This checklist was subsequently developed using semi-structured interviews of experts in the field. Thematic analysis was conducted, and additional details reported by participants (n=34) were added under already existing characteristic categories. Moreover, 88.23%, 75% consensus was achieved regarding the identified characteristics, via a two-stage online Delphi Survey (n=51). We then aimed to quantitively validate the Biomarker Toolkit. Quantitative validation was achieved utilising clinically implemented and non-clinically implemented biomarkers developed for two cancer types (breast and colorectal cancer). Systematic literature searches were conducted to identify relevant publications. Each publication was scored positively for the presence of characteristics included in the Biomarker Toolkit. Cox-regression analysis suggested that total score is a significant driver of biomarker success in breast and colorectal cancer publications (BC: P>0.0001, 95% CI: 0.869-0.935, CRC: P>0.0001, 95%CI: 0.918-0.954). Conclusion: This novel study generated a validated checklist with literature-reported attributes linked with successful biomarker implementation. Upon future work, this toolkit could be used i) to detect biomarkers with the highest potential of being clinically implemented and ii) to shape how biomarker studies are designed/performed.