Role of immunotherapy in stage IV non-small cell lung cancer with liver metastasis: A NCDB analysis.

person Takefumi Komiya Penn State Cancer Institute, Hershey, PA info_outline Takefumi Komiya, Shinkichi Takamori

Authors person Takefumi Komiya Penn State Cancer Institute, Hershey, PA info_outline Takefumi Komiya, Shinkichi Takamori Organizations Penn State Cancer Institute, Hershey, PA, Kyushu University Hospital, Fukuoka, Japan Abstract Disclosures Research Funding No funding received None. Background: Despite approvals of immune checkpoint inhibitors in stage IV non-small cell lung cancer (NSCLC), survival benefit of immunotherapy (IO) in those with liver metastasis is controversial. Randomized controlled trials demonstrated conflicting results. Methods: Using National Cancer Database (NCDB), Stage IV NSCLC cases diagnosed in 2016-2017 with at least 30-day follow up were analyzed. Clinical demographics included age (20-69 vs. 70+), sex (male vs. female), race (whites vs. others), institution (academic vs. others), Charlson-Deyo score (0-1 vs. 2-3), year of diagnosis (2016 vs. 2017), Histology (adenocarcinoma NOS vs. other), brain metastasis (Yes vs. No), bone metastasis (Yes vs. No), liver metastasis (Yes vs. No). They were divided into liver-metastasis positive (LM+) vs. negative (LM-). Information regarding cancer treatment was limited to the first course of therapy, including surgery for primary lesion (Yes vs. No), radiation (Yes vs. No), multi-agent chemotherapy (Yes vs. No), and IO (Yes vs. No). Overall Survival (OS) analysis was performed using Kaplan-Meier curves and Log-rank tests. Cox proportional hazard model was used for multivariate analyses. A two-sided p-value < 0.05 was considered as significant. Results: A total of 13,594 LM+ and 69,885 LM- cases met eligibility and further analyzed. 2,944 and 15,553 patients were treated with IO, respectively. No significant association between use of IO and status of liver metastasis was observed (p = 0.124). The median OS in LM+ cases was significantly shorter than that in LM- cases (5.0 vs 8.8 months, respectively, p < 0.0001). Use of IO was associated with improved OS in both LM+ and LM- cohorts with similar HRs/p-values (0.62/ < 0.0001 and 0.64/ < 0.0001, respectively). Multivariate analysis demonstrated that use of IO had a significantly improved OS in both LM+ (HR = 0.69, p < 0.0001) and LM- (HR = 0.64, p < 0.0001) cohorts. The OS benefit of IO in LM+ cohort remained with propensity score matching analysis (mOS 9.0 vs. 4.4 months, HR = 0.62, p < 0.0001). Further exploratory analysis focusing on those treated with multiagent chemotherapy showed that the addition of IO improved OS in both LM+ (HR = 0.77, p < 0.0001) and LM- (HR = 0.79, p < 0.0001) cohorts. Conclusions: This retrospective study using one of the largest cancer databases suggests that use of IO improves OS of NSCLC patients with LM regardless of chemotherapy status, and its magnitude of OS benefit is similar to that in LM- patients. LM+ with IO LM+ without IO LM- with IO LM- without IO Median OS (month) 9.0 4.1 15.6 7.2 12-month OS (%) 41.7 24.1 57.8 37.9 24-month OS (%) 24.8 13.1 37.6 23.7