OBJECTIVES: Estimate the cost-effectiveness of durvalumab following chemoradiotherapy (CRT) versus CRT alone for unresectable stage III NSCLC using 5-year follow-up data from the PACIFIC trial.

METHODS: 5-year progression-free survival (PFS), time to progression (TTP), post-progression survival (PPS), and overall survival (OS) were used with updated costs and subsequent therapy data to generate results from an existing state-transition model (progression-free, progressed disease, and death) that had previously used PACIFIC data of shorter maturities. Utilities and adverse event inputs remained constant since more recent data were not available. Modelling was conducted from the Medicare payer perspective over a 30-year time horizon.

RESULTS: 5-year time-to-event data for both PACIFIC arms closely matched the model’s previous OS and PFS extrapolations. For the durvalumab arm, there were 0.2% and -1.8% differences between OS and PFS values predicted using 3-year PPS and 2-year TTP and PFS data and the actual 5-year data, and 1.1% and a -1.8% differences between values predicted using 4-year data and actual 5-year data. For the placebo arm, there were OS and PFS differences of -7.0% and -7.7%, respectively, between estimates from the model with 2- and 3-year data and the actual 5-year data, and -3.4% and -5.0% differences between estimates from the model with 4-year data and actual 5-year data. In the models based on 2- and 3-year, 4-year, and 5-year PACIFIC data, respectively, durvalumab following CRT was associated with incremental quality-adjusted life years (QALY) of 1.65, 1.49 and 1.51 and corresponding incremental cost-effectiveness ratio (ICER) values of $55,285, $60,692 and $63,891 per QALY.

CONCLUSIONS: The model was highly accurate at predicting 5-year OS and PFS. ICER estimates based on 5-year time-to-event data from PACIFIC were less than a $100,000 willingness-to-pay threshold, indicating that durvalumab remains a cost-effective treatment option for patients with unresectable stage III NSCLC following CRT.