A Randomized Controlled Clinical Trial of Palliative Therapies for Patients With Inoperable Esophageal Cancer

OBJECTIVES: A dramatic rise in incidence, an aging population, and expensive palliative treatments have led to an escalating burden on clinicians managing inoperable esophageal cancer with only limited evidence of effectiveness. This study compares the clinical effectiveness and cost-effectiveness of self-expanding metal stents (SEMSs) with other palliative therapies to aid clinicians in making an evidence-based treatment choice. METHODS: We conducted a prospective, multicenter, randomized, controlled, clinical trial with 215 patients followed until death or study closure. The primary outcome measures were dysphagia, quality of life (QL) 6 weeks following treatment, and total cost of treatment. Secondary outcome measures included treatment-associated morbidity, mortality, survival, and cost-effectiveness. An intention-to-treat analysis was carried out. RESULTS: There was a significant difference in mean dysphagia grade between treatment arms 6 weeks following treatment (P = 0.046), with worse swallowing reported by rigid stent-treated patients (mean dysphagia score difference = -0.49; 95% confidence interval (CI)-0.10 to -0.89, P = 0.014). Global QL scores were lower at both 1 and 6 weeks following treatment for patients treated by SEMSs (mean difference QL index week 1 = -0.66; 95% CI: -0.02 to -1.30, P = 0.04; mean difference QL index week 6 = -1.01; 95% CI-0.30 to -1.72, P = 0.006). These findings were associated with higher post-procedure pain scores in the SEMS patient group (mean difference of the European Organisation for Research and Treatment of Cancer QLQ C-30 pain symptom score at week 1 = 11.13; 95% CI: 2.89-19.4; P = 0.01). Although mean EQ-5D QL values differed between the treatments (P < 0.001), this difference dissipated following generation of quality-adjusted life year values. Total costs varied between treatment arms but these findings canceled out when SEMSs were compared with non-SEMS therapies (95% CI-845.15-1,332.62). These results were robust to sensitivity analysis. There were no differences in the in-hospital mortality or early complication rates, but late complications were more frequent after rigid stenting (risk ratio = 2.47; 95% CI 1.88-3.04). There was a survival advantage for non-stent-treated patients (log-rank statistic = 4.21, P = 0.04). CONCLUSIONS: The treatment choice for patients with inoperable esophageal cancer should be between a SEMS or a non-stent treatment after consideration has been given to both patient and tumor characteristics and clinician and patient preferences.