Inter-Institutional Survival Heterogeneity Between Surgery and Chemoradiotherapy for Clinical Stage I Esophageal Cancer: Exploratory Analysis of Jcog0502

Background Surgery and chemoradiotherapy (CRT) are treatment options for clinical stage I esophageal squamous cell carcinoma (ESCC). However, the distribution of these treatments varies among institutions, potentially contributing to differences in outcomes between high- and low-volume centers. JCOG0502, a parallel-group controlled trial, showed comparable efficacy between surgery and CRT for stage I ESCC, though institutional preferences may have influenced the patient’s treatment selection. While previous reports suggest physician recommendations impact patient decision-making, variations in outcomes and adverse events (AEs) between institutions remain insufficiently explored. Methods Of the 379 cases enrolled in JCOG0502, 364 were analyzed, excluding 15 cases due to ineligibility or non-protocol treatments. Institutions that enrolled ≥5 cases were categorized as high-enrollment centers (High), while those with <5 were classified as low-enrollment centers (Low). The primary endpoint was overall survival (OS), and secondary endpoints included progression-free survival (PFS), AEs, and surgical complications. Results A total of 202 cases (n = 186: High, n = 16: Low) in the surgery group and 162 cases (n = 150: High, n = 12: Low) in the CRT group were analyzed. The 5-year PFS (High vs. Low) was 81% vs. 93% in the surgery group (p = 0.23) and 71% vs. 66% in the CRT group (p = 0.91). The 5-year OS (High vs. Low) was 87% vs. 93% in the surgery group (p = 0.44) and 85% vs. 92% in the CRT group (p = 1.00). ≥ Grade 3 neutropenia and myocardial ischemia in Low was significantly higher than in High in the CRT group (33% vs. 9%, 17% vs. 2%, p < 0.05). Conclusion In stage I ESCC, no significant difference in survival outcomes was observed between the high-enrollment and low-enrollment centers in either the surgery or CRT groups. Several AEs of early and late phase in CRT were significantly different between high-enrollment and low-enrollment centers. However, due to the imbalance in case numbers, caution is required when interpreting these results.