Survival outcomes for post-endoscopy upper gastrointestinal cancer in England

Background: Post-endoscopy upper gastrointestinal cancer (PEUGIC) is defined as cancer diagnosed 6-36 months after a prior index endoscopy during which cancer was not diagnosed. Few studies have examined PEUGIC survival. We describe survival in PEUGIC patients compared with upper gastrointestinal cancer (UGIC) patients who did not undergo endoscopy without a cancer diagnosis prior to their diagnosis. Methods: This was a population-based retrospective survival analysis of UGIC patients diagnosed between January 2009 and December 2018 in England. Patients were categorized into "detected UGIC" (without prior endoscopy) and "PEUGIC" diagnosed 6-12, 12-18, 18-24, or 24-36 months after index endoscopy. PEUGIC survival was analyzed according to pre-existing Barrett's esophagus (BE), esophageal cancer histological, and gastric cancer. Kaplan-Meier all-cause survival curves investigated survival from both diagnosis and index endoscopy date. Results: 98710 UGIC patients (9068 PEUGIC [9.2%]) were studied. Pre-existing BE was recorded for 9975 UGIC patients (10.1%), including 2583 PEUGIC (25.9%). PEUGIC survival was better than detected UGIC survival but the difference decreased as time between index endoscopy and diagnosis increased. Patients with pre-existing BE and PEUGIC had much better survival, with more stage 1 cancer (39.5% versus 8.4% for PEUGIC without BE; P < 0.001). Survival in PEUGIC and detected UGIC patients was similar when BE patients were excluded. Survival in PEUGIC patients with squamous cell carcinoma and gastric cancer was the same as for detected UGIC. Conclusions: Better survival in patients with PEUGIC compared with detected UGIC appeared to be related to BE surveillance. Epidemiological biases, including selection, lead time, and immortal time, should be considered when reporting PEUGIC survival outcomes.