Definitive radiotherapy in patients with clinical T1N0M0 esophageal squamous cell carcinoma: A multicenter retrospective study (KROG 21–10)

•cT1a ESCC patients who cannot receive endoscopic resection showed equivalent failure rates as cT1b patients.•There was no difference in survival and failure rates between ENI and IFI in cT1b ESCC, treated with definitive RT.•In cT1b ESCC, IFI with concurrent chemotherapy is reasonable. To assess the failure pattern and analyze the treatment scheme of definitive radiation therapy (RT) for T1N0M0 esophageal squamous cell carcinoma (ESCC). We performed a multi-institutional retrospective analysis in T1N0M0 ESCC patients who underwent definitive RT from 2010 to 2019. Patterns of failure were demonstrated as in-, and out-field locoregional, and distant metastasis. In the analysis, freedom-from locoregional recurrence (FFLRR) and their association with clinicopathologic factors were evaluated. Propensity score matching in cT1b patients was done. 168 patients were included with a median follow-up of 34.0 months, and 26 cT1a, 116 cT1b disease. The rates of 3-year all and locoregional recurrence for cT1a were 30.5% and 24.1% and those for cT1b were 27.1% and 25.9%, respectively. Among 116 cT1b patients, 69 patients received elective nodal irradiation (ENI) and 47 received involved field irradiation (IFI). After propensity score matching, the 3-year FFLRR rate was 84.5%. There was no difference between ENI and IFI in FFLRR (P = 0.831) and OS (P = 0.525). The 3-year FFLRR was 83.8% (95% Confidence interval (CI), 61.8–93.8%) in IFI group and 85.3% (95% CI, 65.1–94.3%) in ENI group. In multivariate analysis, concurrent chemotherapy use was marginally associated with FFLRR (Hazard ratio, 0.16; P = 0.064). cT1a patients who cannot receive endoscopic resection showed similar failure rates as cT1b patients, questioning the staging accuracy and raised the need for thorough treatment like chemoradiotherapy. In cT1b patients, IFI with 50 to 60 Gy and concurrent chemotherapy could be reasonable.