Background: NGS testing for genomic alterations has been found to expedite time to appropriate targeted therapy initiation compared to PCR testing in mNSCLC. This study assessed US payers’ costs of testing, delayed care, and suboptimal treatment initiation for PCR versus NGS testing in patients with mNSCLC. Methods: A decision tree model evaluated testing for guideline-recommended alterations (EGFR, ALK, ROS1, BRAF, KRAS, MET, HER2, RET, NTRK1/2/3) in patients with newly-diagnosed mNSCLC from first test until appropriate therapy initiation using 1) liquid or tissue NGS (60% [90% tissue, 10% liquid] of patients), 2) exclusionary KRAS test followed by sequential PCR tests (5%), 3) sequential PCR tests (5%), or 4) hotspot panel (30%) PCR tests. The proportion of patients testing positive for a targetable alteration and time to appropriate therapy initiation were evaluated. Cost components included genetic testing, testing-related medical services, delayed care (i.e., costs for medical services while waiting for test results), immunotherapy [IO]/chemotherapy [CTX] initiation prior to test results, and suboptimal treatment initiation after test results (i.e., costs of first-line IO/CTX in patients falsely testing negative for an actionable mutation). Results: In a modeled cohort of 1 million members (75% commercial, 25% Medicare), an estimated 1,119 patients had mNSCLC and were tested. The proportion of patients testing positive for a targetable alteration was 45.9% for NGS and 39.4% for PCR testing (Table). Total mean per patient costs were lower for NGS ($8,866) than for PCR ($18,252), driven by more rapid treatment with appropriate therapy (2.1 versus 5.2 weeks), resulting in lower delayed care costs and IO/CTX costs prior to test results. Conclusions: Results suggest that NGS testing may lead to better patient care and reduced costs to the health system relative to PCR testing. Costs of inappropriate IO/CTX initiation are an important component of increased costs of PCR-based approaches.

*Sum of components differs slightly due to rounding.