Objective: Currently, immune checkpoint inhibitors (ICIs) play a central role in the treatment of lung cancer. However, ICI re-administration is still uncommon, and its utility should be evaluated as early as possible.Patients and Methods: Twenty-five patients who received ICIs twice or more in any of the drug treatment lines for advanced/relapsed non-small cell lung cancer were included. OS, PFS, ORR, and DCR were examined, and factors such as age, sex, histopathological type, PD-L1 expression, whether radical surgery was performed, driver gene mutations, and immune-related adverse events (irAEs), were evaluated for their relevance and as prognostic factors.Results: Of the 25 patients, 17 were men and 8 were women, with an average age of 68 ± 8.4 (range, 48–85 years), and histology was non-squamous cell carcinoma/squamous cell carcinoma in 19/6 cases. One driver gene mutation positive case was included. PD-L1 TPS was ≥50%/1–49%/0–1%/ unknown in 7/8/5/5 cases. The first ICI administered was pembrolizumab/nivolumab/atezolizumab in 5/13/7 cases. The median number of courses was 9 (range, 1–52) months, and the median PFS was 9 (95% CI, 6.0–12.0) months. Cytotoxic chemotherapy or radiation therapy was administered to 6 patients during the interval up to re-administration. The second ICI administered was pembrolizumab/nivolumab/atezolizumab in 5/8/12 cases, and all patients received antibody drugs different from those given as the first ICI. The median number of courses was 5 (range, 1–24), and the median PFS was 3 months (95% CI, 1.0–5.0) months. In 5 of the 6 patients (24%) who achieved PFS of 6 months or longer after re-administration, the order of administration was anti-PD-1 antibody to anti-PD-L1 antibody.Conclusion: The effect of re-administration is limited, but it may be effective depending on the type of cases and the order of ICI administration. Further studies are required to verify its effectiveness.

Objective: In recent years, an association between serum soluble immune checkpoint molecules (sICMs) and malignant tumors has been reported, which may become important biomarkers in the future. Although several reports have suggested a correlation between sICMs and prognosis, their origin is unclear. In this study, changes in serum soluble PD-L1 (sPD-L1) during the perioperative period and its origin were analyzed in patients with lung cancer.Patients and Methods: Patients with lung tumors (n=39) were included. Samples for sPD-L1 measurements were collected at five time points before and after surgery, and their changes over time were analyzed. ELISA was used to measure sPD-L1 levels.Results: Thirty-nine patients with lung tumors (31, males; 8, females; age, 74 (years) ± 7.7 (range: 51–89) years; malignancy/benign, 33/6) were enrolled. Eight cases of driver gene mutation-positive tumors were included. Twenty-eight (72%) patients were smokers, and their performance status was 0-1 in all 39 patients. PD-L1 TPS was ≥50%/1–49%/<1% in 8/10/14 patients. Stage I/II/III/IV/postoperative recurrence of lung cancer was observed in 21/0/6/5/1 patients, respectively. There were no significant correlations between sPD-L1 levels and clinicopathological features and no correlation with PD-L1 TPS. Comparing localized lesions (stages I–III) with advanced lesions (stage IV and postoperative recurrence), the distribution of sPD-L1 was slightly higher in advanced lesions, although the difference was not significant. No obvious changes in sPD-L1 expression were observed before and after surgery.Conclusion: sPD-L1 levels tended to be high in stage III and above lung cancer. There was no change in sPD-L1 levels before and after surgery. sPD-L1 levels did not correlate with the PD-L1 TPS.