Impact of smoking on immunotherapy outcomes in gastric, gastro-esophageal junction, and esophageal cancers (G/GEJ/EC): A real-world analysis.

Abstract

320 Background: Outcomes of G/GEJ/EC remain poor despite the recent addition of immunotherapy (IO) to a chemotherapy backbone. The impact of smoking on the efficacy of IO remains unknown. Given the known association between smoking, tumor mutational burden (TMB) and tumor microenvironment, we explored the impact of smoking history on outcomes in patients (pts) with G/GEJ/EC treated with IO. Methods: We conducted a single-center retrospective study. Smoking history (never smokers vs. ever smokers), treatment intent (curative vs. palliative) and outcomes were collected. Differences in relapse-free survival (RFS) and overall survival (OS) by smoking status were analyzed using Kaplan-Meier curves adjusting for age, performance status, body mass index, Charlson Comorbidity Index and stage. Hazard ratios (HR) were estimated using Cox proportional hazards and Fine and Gray regression models. Results: 204 pts were included (median age 61.9 years); 93 (46%) were never smokers and 111 (54%) were ever smokers. Adenocarcinoma was the most common histology in 156 pts (78%). Tumor location included gastric (38%), gastro-esophageal junction (34%), and esophagus (28%). IO was given with curative intent in 31 pts (15%) and palliative intent in 173 pts (85%). Nivolumab was the most common IO used (67%). In the overall cohort, ever smokers had numerically longer median RFS (17.2 vs. 13.6 months; HR 0.80, 95% CI 0.49–1.32, p=0.39) and OS (14.1 vs. 12.4 months; HR 0.92, 95% CI 0.56–1.51, p=0.75) compared to never smokers. In those treated as curative intent, ever smokers had significantly higher risk of relapse (HR 4.07, 95% CI 1.26–13.1, p=0.019) and a non-significant increased risk of death (HR 3.38, 95% CI 0.40–28.41, p=0.26) than never smokers. In the palliative cohort, 52 pts (30%) had recurrent disease and 121 pts (70%) had de novo metastatic disease. Among those with recurrent disease, ever smokers had significantly worse median OS compared to never smokers (9.1 vs. 25.8 months; HR 4.62, 95% CI 1.15–18.56, p=0.031). Among pts with de novo metastatic disease, ever smokers had numerically longer median RFS (15.8 vs. 7.7 months; HR 0.66, 95% CI 0.36–1.18, p=0.16) and OS (15.7 vs. 9.7 months; HR 0.63, 95% CI 0.33–1.20, p=0.16), though not statistically significant. Conclusions: Smoking history was associated with different IO outcomes based on disease and treatment setting. Ever smokers had higher relapse risk in the curative setting and worse survival with recurrent disease. In contrast, ever smokers with de novo metastatic disease showed a trend toward improved outcomes. Future analyses should incorporate biomarkers such as PD-L1 combined positive score (CPS) and TMB to better evaluate the association between smoking and IO outcomes.