The Predictive Value of PD-L1 Expression Level in Evaluating the Cost-Effectiveness of Atezolizumab/Pembrolizumab.

Objective: Recently, the significant improvement of atezolizumab and pembrolizumab over chemotherapy for treatment-naïve stage IV non-small cell lung cancer (NSCLC) has been demonstrated, but the cost-effectiveness of these regimens remains unknown. Methods: A Markov model was adapted from the US healthcare perspective to assess the cost-effectiveness of atezolizumab, pembrolizumab, and chemotherapy in treatment-naïve NSCLC. Pseudo-individual patient data were generated from digitized Kaplan-Meier curves. Direct medical costs and utility values were sourced from the database and literature. Quality-adjusted life-years (QALYs), total costs, and incremental cost-effectiveness ratios (ICERs) were computed. Sensitivity analyses and budgetary impact analyses were calculated. Results: In any and high programmed cell death 1-ligand 1 (PD-L1) expression populations, with chemotherapy, atezolizumab provided ICERs of $234,990 and $130,804 per QALY, while pembrolizumab yielded ICERs of $424,797 and $140,873 per QALY. The ICER of atezolizumab vs. pembrolizumab was $56,635 and $115,511.82 in any and high PD-L1 expression population, respectively. The critical drivers of ICERs included the cost of atezolizumab and pembrolizumab. The accumulated incremental budgetary impact of atezolizumab vs. chemotherapy increased to approximately $39.1 million in high PD-L1 expression patients over 5 years. Conclusions: In the high PD-L1 expression population, both atezolizumab and pembrolizumab were cost-effective for stage IV NSCLC compared to chemotherapy, which is contrary to that in any PD-L1 expression population. Atezolizumab shows a higher acceptability in both populations. Treating with immune checkpoint inhibitors (ICIs) has a substantial budgetary impact on the medical burden. The PD-L1 expression level has the potential to be a predictor for the economics of ICIs.