Abstract
PURPOSE: The survival of esophageal and gastric cancer patients treated with chemotherapy is rarely assessed outside of clinical trials. Therefore, we compared the effectiveness of various curative or palliative chemotherapy regimens on the survival of esophageal and gastric cancer patients in a "real world" clinical setting. METHODS: We identified a cohort of 966 incident esophageal and gastric cancer patients in Stockholm/Gotland County (a low-risk Western population) during 2008-2013. Patients who received chemotherapy with curative intention (n = 279) and palliative intention (n = 182) were analyzed separately. Using Cox proportional hazards regression models, we estimated hazard ratios (HRs) with 95% confidence intervals (CIs) and adjusted for the potential confounding factors: age, sex, TNM stage, radiotherapy, comorbidity, marital status, education, income, and country of birth. RESULTS: In esophageal cancer patients with curative treatment intention, we observed a higher hazard for death among patients who received carboplatin-fluorouracil compared to patients who received cisplatin-fluorouracil, corresponding to a HR of 2.18 (95% CI 1.09-4.37). Conversely, in patients with cancer in the gastroesophageal junction who had a curative treatment intention at diagnosis, we observed a reduced hazard for death among those who received fluorouracil-oxaliplatin, compared to patients who received cisplatin-fluorouracil (HR 0.28; 95% CI 0.08-0.96). CONCLUSION: Among patients with esophageal cancer who received treatment with curative intention, cisplatin-fluorouracil was associated with better survival compared to carboplatin-fluorouracil, while patients with gastroesophageal junction cancer who were treated with cisplatin-fluorouracil had worse survival compared to fluorouracil-oxaliplatin.