Palliation of dysphagia in patients with non-curable esophageal cancer - a retrospective Danish study from a highly specialized center.

Purpose: A majority of the patients with esophageal cancer (EC) suffer from dysphagia. Several endoscopic treatment options are available such as stent placement, argon plasma coagulation, and esophageal dilatation. This study aimed to map the use of endoscopic dysphagia relieving interventions and secondly investigate possible impact on survival.

Methods: Data was collected at the Dept. of Surgery & Transplantation, Rigshospitalet, Denmark. Patients with non-curable EC referred from 2016 to 2019 were included. Type of dysphagia treatment, complications and the need for repeated treatments, and survival were registered.

Results: In the study, 601 patients were included. Forty-five percent were treated with an endoscopic procedure due to dysphagia (82% had a stent placed). The median time from diagnosis to intervention was 24 days. The overall complication rate was 35% (38% in the stent group and 20% in the non-stent group, p = 0.03) and 13% of the patients were readmitted due to a complication. After 26% of the procedures, a repeated treatment was required. Patients having an endoscopic intervention had a worsened survival prognosis compared with the patients in the non-intervention group (HR: 2.17, 95% CI: 1.80-2.61, p < 0.001). In the sub analysis where only patients who had an intervention was included, a survival difference in favor of the non-stent group was found (HR: 0.61, 95% CI: 0.43-0.86, p = 0.005).

Conclusion: In this cohort, the incidence of endoscopic procedures was high, complication rates were considerable, and many the patients required a second treatment. A survival difference was seen, where the patients who had a stent placed seemed to have the worst survival outcomes. However, the causal relationship is yet to be determined why the results must be interpreted carefully. New interventions and tailored approaches that may positively affect functional and long-term oncological outcomes are highly warranted and this should preferably be investigated in randomized clinical trials.