Clinician perspectives on digital and computational pathology in oncology: Opportunities and challenges.

Abstract

e15172 Background: Digital and computation pathology (DCP) is a rapidly emerging testing technology seeking to address both workflow constraints in the pathology lab and support clinician decisions. As oncology targets become more complex and nuanced, more sophisticated technologies like DCP will be increasingly required to accurately identify patients eligible for specific targeted therapies. DCP technologies historically focused on improving pathology lab workflows and centered on pathologists as the primary stakeholder to drive adoption. However, novel DCP algorithms to support diagnoses and inform specific treatment decisions are on the horizon. With the emergence of these algorithms, clinicians’ perspectives on the benefits and concerns surrounding these technologies are increasingly important to understand if widespread adoption is to become a reality. Methods: We conducted a double-blinded survey in December 2024 of oncologists (n = 101) to evaluate current awareness, perceived benefits, concerns, and willingness to order DCP tests. This evaluation was assessed both generally and with a specific theoretical FDA approved DCP-based companion diagnostic test. Ultimately, this data would help identify what concerns need to be addressed to drive future adoption of DCP algorithms for supporting clinical decisions. Results: Awareness of DCP was highly varied across clinicians surveyed, with 17% knowingly using a DCP-based test today (e.g., Paige Prostate, uPATH HER2, etc.) 26% reporting little to no knowledge of DCP, and 58% having a modest or moderate awareness of the technology. However, when introductory level information on DCP was provided, consensus formed among the majority of clinicians around the potential benefits. As such, 62% of clinicians reported being comfortable or very comfortable ordering an FDA-approved DCP-based test included in NCCN guidelines. Anticipated barriers for DCP-based tests match traditional systemic barriers to novel testing (e.g., insurance coverage concerns, patient out-of-pocket costs, etc.). However, full adoption will also require addressing skepticism around the perceived “black box” of DCP algorithms causing 10-12% of clinicians to be unwilling to adopt DCP-based tests without additional oversight or regulations. Conclusions: Our survey results revealed a positive outlook for DCP tests among clinicians with potential for consensus around the benefits. However, ensuring widespread adoption will require additional education efforts to both broaden awareness and alleviate the concerns of “black box” approaches in a subset of clinicians. These findings help further elucidate the necessary steps for successful adoption of DCP tests into the clinical paradigm.