The use of biomarkers holds great promise for the development of new therapeutics and the acceleration of clinical research. Yet as Hey et al.^{Reference Hey1} note in their introduction to this thoughtful collection of articles, the realization of these benefits hinges heavily on rigorous biomarker validation — a complex and costly endeavor that demands transparency and high-quality information.

Indeed, all of these articles grapple with the challenges of managing information about biomarkers. How should biomarkers be validated? How can data sharing endeavors and coordination further their discovery and validation?^{Reference Li, Sim and Shrager2} And what opportunities exist to use clinical and regulatory data to systematically study biomarkers associated with therapeutics that are already in use?^{Reference Mavergames and Hey3} These are important questions that demand ongoing discussion by regulators, researchers, and the therapeutics industry.

A related topic that is notably absent from this collection, however, is a discussion of incentives for biomarker development in the first place. As past work has noted, different parties in the health care system have meaningfully different incentives to discover, validate, and use biomarkers in therapeutic development and in health care delivery.^{Reference Stern, Alexander and Chandra4} For example, the types of diagnostic, prognostic, and predictive biomarkers that are highly valued by biopharmaceutical companies pursuing commercialization of targeted therapies may not be those that are most useful for payers or those that are most desired by patient advocacy groups.^{Reference Stern, Alexander and Chandra5}

Some of the initiatives discussed in this set of articles are insulated from direct economic incentives by virtue of their institutional structure (some represent work being done by as not-for-profits), yet further consideration of other incentives (both financial and non-financial) to participate in the programs described will be helpful in maximizing the impact of their implementation.

For example, what incentives exist and — perhaps, most crucially — what incentives are missing for participation in clinical trial data sharing platforms? The International Committee of Medical Journal Editors (ICMJE) has already rolled out thoughtful, forward-looking policies on data sharing, representing an important step towards creating broad incentives for data sharing. Indeed, ICMJE policies have been successful in incentivizing other large-scale information sharing endeavors in the past, such as the increased registration of clinical trials in the early decades of the 21st century, yet plenty of room for improvement in information sharing incentives and platforms for doing so remains.^{Reference Taichman, Sahni, Pinborg, Peiperl, Laine, James, Hong, Haileamlak, Gollogly, Godlee, Frizelle, Califf, Zarin, Kramer, Sherman, Aberle and Tasneem6}

Further, understanding the incentives to use biomarkers in the drug development process — i.e., in the premarket setting — will necessarily shape the set of studies that can be considered in the post-market setting (by organizations such as Cochrane) and shared through clinical trial data sharing platforms (such as Vivli). Thus, greater focus on articulating standards and defining clear incentives for high-quality biomarker development and information sharing early on is of particular importance — not only as a precedent for subsequent therapeutic development, but also for later hypothesis-generation through pooled data resources, and to ensure that evidence-informed decision making related to the use of biomarkers can make its way into clinical practice. Put another way: no matter how well the Cochrane's Linked Data Project is architected and used, it will, by virtue of its post hoc composition, be limited to informing practitioners about a subset of biomarker uses that could have been explored in earlier-stage clinical studies.

This points to an important role for policymakers and the scientific research community in creating incentives to validate biomarkers and suggests a vital role for regulators, such as the U.S. Food and Drug Administration, in helping to clarify evidence required for the use of surrogate endpoints early on in the clinical development process for new therapeutics, since the regulatory process can shape research and clinical practice long beyond new product approval.