Palliative Radiotherapy for Esophageal and Gastric Cancer: Population-Based Patterns of Utilization and Outcomes in Ontario, Canada

Abstract

Objective: Patients with incurable esophageal and gastric cancer may develop local symptoms for which palliative radiotherapy (PRT) may be considered. We sought to evaluate patterns in utilization and outcomes of patients receiving PRT for incurable esophageal and gastric cancer in Ontario, Canada using health administrative data. Methods: Linked health administrative databases were used to identify patients receiving PRT for incurable esophageal and gastric cancer. Primary outcomes were utilization and delivery of PRT, utilization of endoscopic dilation with or without stent insertion after completion of PRT and survival from 1) date of diagnosis and 2) start of PRT. Results: We identified 2500 patients who received PRT. Mean age was 70 ± 13 years and the majority (75%, n = 1873/2500) were male. Over half of the patients had a diagnosis of gastric cancer (58%, n = 1453/2500) and began PRT within 6 months of cancer diagnosis (85%, n = 2125/2500). Of the 2500 patients in the cohort, 2174 patients received EBRT with few receiving brachytherapy (n = 326) or EBRT and brachytherapy combined (n = 88). Over the study period, there was an increase in the number of patients receiving PRT (136 in 2007 to 290 in 2016), as well as in the use of advanced conformal radiotherapy techniques. Only 5% (115/2500) required dilation with or without stent insertion after completion of PRT. Median overall and cancer-specific survival of the cohort was 205 days and 209 days from date of diagnosis and 108 days and 110 days from start of PRT. Conclusions: PRT is an important treatment for patients with incurable esophageal and gastric cancer who present with local symptoms. Utilization of PRT and advanced EBRT techniques increased over the study period. Few patients require endoscopic dilation with or without stent insertion after completion of PRT suggesting that PRT provides favorable symptom control.