OBJECTIVES: Recent advances in lung cancer drugs have been tremendous, and it leads to succeed prolonging the lives of inoperable stage lung cancer patients. Epidemiological and health economic studies using real world data (RWD) are increasing; however, obtaining the overall survival (OS) data and progression-free survival (PFS) data from RWD is not always possible. Therefore, generating the surrogate variables for OS and PFS is considered valuable to design database studies. METHODS: In this study, we defined time to the last claim (TLC) as a surrogate variable for OS, and utilize time to treatment failure (TTF) as a surrogate variable for PFS. We calculated TLC and TTF of non-small cell lung cancer (NSCLC) population selected from medical claim database (provided by Medical Data Vision Co., Ltd.), who had been prescribed pembrolizumab as first-line mono-therapy. The results were compared with OS and PFS of the clinical trial result to evaluate validity of the surrogate variables. As exploratory work, we also calculated TLC and TTF of NSCLC population prescribed pembrolizumab and nivolumab as second-line or later treatment. In addition, QALYs were also calculated using the surrogate variables. RESULTS: Medians of TLC and TTF in the first-line pembrolizumab population were 16.1 months (95%CI; 14.8-17.1) and 3.7 months (95%CI; 3.5-4.5) respectively, and those were in concordance with OS and PFS of the clinical trial result. Median of TLC and TTF in the second or later line pembrolizumab population were 13.5 months (95%CI; 12.1-15.0) and 2.8 months (95%CI; 2.8-3.5) respectively. Median of TLC and TTF in nivolumab population were 8.6 months (95%CI; 8.0-9.5) and 1.9 months (95%CI; 1.8-2.1) respectively. Estimated QALYs for the second or later line pembrolizumab and nivolumab population were 0.985 and 0.804 respectively. CONCLUSIONS: The results showed possibility to utilize TLC and TTF as surrogate variables of OS and PFS.