Reevaluating the Design of Multicenter Surgical Trials for Esophagogastric Cancer.

[BACKGROUND] Multicenter randomized controlled trials (RCTs) in esophagogastric cancer surgery have repeatedly failed to demonstrate overall survival (OS) benefits for technical interventions. This raises concerns about trial design frameworks, particularly surgical quality assurance (SQA) and their impact on statistical power and sample size requirements.

[METHODS] We conducted a systematic review of MEDLINE, Embase, and Cochrane CENTRAL (January 1990 to June 2024) to identify multicenter RCTs evaluating curative surgical interventions for esophageal or gastric cancer that reported OS outcomes with ≥ 12 months of follow-up. Trials were assessed for SQA strategies, protocol adherence, and sample size assumptions. Post hoc power analyses and statistical simulations modeled the impact of surgical variability on trial outcomes.

[RESULTS] We identified 27 eligible RCTs; 10 were powered for OS superiority, but none showed significant OS benefit. Suboptimal SQA was common, including limited surgeon credentialing, poor adherence monitoring, and inadequate technical standardization. Three trials were underpowered at design, and three failed to meet recruitment targets. Simulations showed that a technical nonadherence rate of 10% per arm could halve statistical power. Sample size requirements to preserve 80% power increased markedly, sometimes by thousands of participants. Most trials lacked internal piloting to assess protocol feasibility and baseline adherence.

[CONCLUSIONS] Inadequate SQA and failure to account for surgical variability may partly explain the lack of OS benefit in multicenter surgical RCTs. Embedding SQA and internal pilot studies into trial design may improve feasibility, ensure protocol fidelity, and strengthen the validity of future trials in complex oncologic surgery.