Cost-effectiveness of first-line pembrolizumab monotherapy in PD-L1–high metastatic non-small cell lung cancer in Australia: evidence from trials and real-world practice

INTRODUCTION: Pembrolizumab monotherapy improves survival in advanced non-small cell lung cancer (NSCLC) with programmed death ligand 1 (PD-L1) tumour proportion score (TPS) ≥ 50%. However, survival in randomised trials may not fully reflect outcomes in routine practice, with implications for economic evaluation. AIM: To evaluate the cost-effectiveness of pembrolizumab monotherapy compared with platinum-based chemotherapy from an Australian health-payer perspective, integrating both randomised controlled trial (RCT) and real-world evidence (RWE). METHOD: A semi-Markov model with three health states (progression-free, progressed disease, death) was developed over a 3-year horizon. Clinical inputs were derived from KEYNOTE-024 for the RCT scenario and from nationwide Australian real-world data for the RWE scenario, with chemotherapy outcomes retained from the pivotal trials in both scenarios. Direct medical costs (2024 AU$) and quality-adjusted life years (QALYs) were estimated. Incremental cost-effectiveness ratios (ICERs) were calculated, probabilistic and deterministic sensitivity analyses conducted. RESULTS: Compared with chemotherapy, pembrolizumab monotherapy was associated with incremental costs of AU$87,399 (RCT-based) and AU$75,935 (RWE-based), and incremental QALY gains of 0.23 and 0.11, respectively. Corresponding ICERs were AU$385,561/QALY (RCT) and AU$705,729/QALY (RWE). At a AU$75,000/QALY willingness-to-pay (WTP) threshold, the probability of cost-effectiveness was < 1% in both scenarios. Differences in overall survival between trial and real-world cohorts likely contributed to the variation in ICERs. CONCLUSION: At current pricing and Australian WTP thresholds, pembrolizumab monotherapy is unlikely to be cost-effective, with RWE-informed scenarios yielding less favourable cost-effectiveness estimates than from trial data. These findings highlight that cost-effectiveness estimates may be sensitive to differences between trial-based and real-world effectiveness inputs.