Subscribe to RSS
DOI: 10.1055/s-0040-1704483
PREDICTIVE VALUE OF ENDOSCOPIC ESOPHAGEAL ABNORMALITIES FOR RESIDUAL ESOPHAGEAL CANCER AFTER NEOADJUVANT CHEMORADIOTHERAPY
Publication History
Publication Date:
23 April 2020 (online)
Aims Endoscopic evaluation of the esophageal mucosa may play a role in an active surveillance strategy after neoadjuvant chemoradiotherapy (nCRT) for esophageal cancer. This study aimed to investigate the yield of endoscopic abnormalities of the esophageal mucosa for detection of residual disease (RD).
Methods A retrospective chart review of 156 patients was performed. All patients underwent nCRT followed by surgery for esophageal cancer. Upper endoscopy was performed six and twelve weeks after nCRT. Endoscopic records were reviewed for presence of non-passable strictures, relative strictures, residual tumor, scar tissue, or ulceration. Presence and type of endoscopic esophageal abnormalities at 6 and 12 weeks were compared between patients with RD and patients with a complete response (CR) based on the resection specimen.
Results 118 of 156 (76%) patients had RD. A non-passable stricture was present in eleven patients at six weeks (RD vs CR, 5% vs 13%, P=0.09), preventing full examination of the esophagus. In the remaining 145 patients, ulceration was the most prevalent endoscopic feature (RD vs CR at 6 weeks: 55% vs 61%, P=0.59; 12 weeks: 40% vs 43%, P=0.77). Comparable outcomes were found for presence of a relative stricture (18% vs 24%, P=0.41; 18% vs 14%, P=0.67) and scar tissue (10% vs 15%, P=0.39; 19% vs 29%, P=0.31). Endoscopic suspicion of residual tumor was significantly associated with RD; at 6 weeks 40/44 patients had RD (36% vs 12%, P=0.01), while at 12 weeks all sixteen patients had RD (22% vs 0%, P< 0.01).
Conclusions Endoscopic esophageal abnormalities after nCRT were found to be of limited value for detection of residual esophageal cancer. Endoscopic suspicion of residual tumor was the only finding that was associated with RD. Data from prospective studies are needed to confirm its predictive value before this parameter can be implemented in clinical practice.
#