Assessment of PD-L1 expression and oncogenic gene status in patients with small-cell lung cancer: Immuno-oncology biomarker study in LC-SCRUM-Japan.

8558 Background: The clinical impact of PD-L1 expression and oncogenic gene status in patients with small cell lung cancer (SCLC) is not well characterized. We initiated this immuno-oncology biomarker study as part of nationwide genomic screening by LC-SCRUM-Japan (LC-SCRUM-IBIS). Methods: Tumor samples from lung cancer patients enrolled in LC-SCRUM-IBIS were primarily subjected to targeted next-generation sequencing (NGS) with Oncomine™ Comprehensive Assay. The PD-L1 expression was also analyzed by 4 immunohistochemistry (IHC) assays for 22C3, 28-8, SP263 and SP142. At this analysis, 22C3, 28-8, and SP263 were assessed in tumor cells (TC) as positive in > 1%, and SP142 in both TC and tumor-infiltrating immune cells (IC) as positive in > TC1/IC1, as previously reported. The association of PD-L1 expression, oncogenic gene status and clinical outcome was investigated in SCLC patients. Results: Between Feb 2017 and May 2018, 1017 lung cancer patients were enrolled in LC-SCRUM-IBIS. Among them, 933 patients had adequate tumor samples including 101 SCLC and 832 non-small cell lung cancer. Of 101 SCLC patients, the results of PD-L1 expression by 4 IHC assays were 18% in 22C3, 17% in 28-8, 11% in SP263 and 8% in SP142, respectively. Targeted NGS showed that 8 patients had at least one targetable oncogenic alterations, including 3 PIK3CA and 1 KRAS as mutations and 3 PTEN and 1 TSC2 as inactivating mutations. PD-L1 expression by 22C3 was associated with good performance status (P = 0.05) and the presence of oncogenic alterations (P = 0.004). PD-L1 status was not associated with response to cytotoxic chemotherapy and progression-free survival and overall survival in first-line treatment of SCLC patients. Conclusions: The frequency of PD-L1 expression in SCLC patients was relatively lower compared with that reported in other solid tumors. PD-L1 status by TC in 22C3 appears to be not correlated with clinical outcomes for cytotoxic chemotherapy of SCLC patients. Further investigation is needed to explore a predictive biomarker for immune checkpoint inhibitors. Updated results will be presented at the meeting.