194 Understanding use of reflex biomarker testing at diagnosis in NSCLC: Results from an online survey

Abstract Introduction/Objective Clinical guidelines in cancer diagnostics recommend comprehensive cancer biomarker testing to guide therapy and improve patient outcomes. However, limited data exist on biomarker testing practices, protocols, and policy barriers to timely implementation. This study aimed to understand ordering patterns, opportunities to incorporate testing into diagnostic workflows, and potential policy barriers. Methods/Case Report Members of LUNGevity Foundation, Association for Molecular Pathology, and American Society for Clinical Pathology were sent an online survey to obtain information on current cancer biomarker ordering practices. Criteria assessed included testing performed at initial diagnosis, reflex testing to multi-gene panels, and result availability within 21 days and prior to the patient’s oncology consult. Respondents (n = 115) were primarily employed by an academic/university medical center-owned institution (53%) and 61% (n = 70) were pathologists or molecular pathologists. Results Fifty-six percent of respondents (n = 64) reported either no standard biomarker ordering protocol or protocols that didn’t meet the criteria outlined in the study. Among the respondents employing protocols or considering them, 58% (n = 38) specified the pathologist as the ordering physician, indicating a broadly supported practice. Reimbursement and the “14-day rule” regulation were significant barriers identified to establishing effective biomarker testing standard protocols. Future research informing biomarker testing implementation and regulatory alignment is needed. Conclusion Findings from this survey highlight persistent gaps in standardized reflex biomarker testing at NSCLC diagnosis. While many institutions support pathologist-led ordering, inconsistent protocols and systemic barriers—particularly reimbursement challenges and the CMS 14-day rule—continue to impede timely, comprehensive testing.