Beyond median overall survival: Estimating multiple survival scenarios in patients with metastatic esophagogastric cancer.

Abstract

261 Background: Recent clinical trials of novel systemic therapies showed improved survival of patients with metastatic esophageal cancer (EC) and gastric cancer (GC). Survival improvements observed in clinical trials might be unrepresentative for the total population as the percentage of patients whom participate in clinical trials is limited and more than half of all patients receive best supportive care (BSC). The aim of our study is to assess the best-case, typical and worst-case survival scenarios in patients with metastatic esophagogastric cancer. Methods: We selected patients with metastatic EC (including junction) or GC diagnosed in 2006-2019 from the Netherlands Cancer Registry. Survival scenarios were calculated based on percentiles of the survival curve stratified by tumor location and treatment (tumor-directed therapy or BSC). Survival scenarios were calculated for the 10th (best-case), 25th (upper-typical), 75th (lower-typical) and 90th (worst-case) percentiles. Linear trend analysis was performed to test if changes in survival over the diagnosis years were significant. Results: We identified 12739 patients with EC and 6833 patients with GC. Percentage of patients receiving tumor-directed therapy increased from 34% to 47% and 30% to 45% for patients with EC and GC, respectively. The median survival remained unchanged for patients with EC (5.0 months) and improved slightly for patients with GC (3.1 to 3.7 months; P=0.006). For patients with EC survival of the best-case scenario improved (17.4 to 22.8 months; P=0.001), whereas the other scenarios remained unchanged: upper-typical 11.2 to 11.7 (P=0.11), lower-typical 2.1 to 2.0 (P=0.10) and worst-case 0.9 to 0.8 months (P=0.22). For patients with GC survival improved for the best-case (13.1 to 19.5; P=0.005) and upper-typical scenario (6.7 to 10.6 months; P=0.002), whereas the lower-typical (1.2 to 1.4 months; P=0.87) and worst-case (0.6 to 0.6 months; P=0.60) remained unchanged. For patients with EC receiving tumor-directed therapy survival in all scenarios remained unchanged while for patients receiving BSC survival decreased: best-case 11.8 to 9.8 (P=0.005), upper-typical 6.0 to 5.0 (P=0.002), lower-typical 1.4 to 1.0 (P=0.003) and worst-case 0.7 to 0.5 months (P=0.03). For patients with GC receiving tumor-directed therapy survival improved for all scenarios: best-case 19.8 to 30.4 (P=0.005), upper-typical 6.4 to 10.3 (P=0.002), lower-typical 3.6 to 5.4 (P<0.001) and worst-case 1.4 to 2.6 months (P<0.001), and for patients receiving BSC survival for all scenarios remained unchanged. Conclusions: The proportion of patients with EC and GC receiving tumor-directed therapy increased over time. Despite the fact that survival improvements were not observed across all scenarios, at least an increase in survival was observed in certain subgroups of patients.