Determinants of a positive resection margin (R1 status) in esophageal squamous cell carcinoma: a retrospective analysis of 148,121 cases from China

To investigate risk factors for positive proximal and distal resection margins after surgery for esophageal squamous cell carcinoma (ESCC). Total initially enrolled esophageal squamous cell carcinoma patients were 148,121 cases. After strict case screening and quality control, a total of 80,881 complete cases were included in the primary analytical cohort, and 29,546 patients from the post-2011 subgroup were further analyzed for sensitivity analysis. Statistical analyses were performed using R4.4.3. Chi-square test was used for intergroup comparisons, and the false discovery rate (FDR) correction was applied for multiple comparisons. Firth penalized logistic regression was adopted for all univariable and multivariable analyses of both proximal and distal margin status across the whole manuscript. Results of the Primary Analysis: Significantly higher positive proximal margin rates were found in patients from low-tier hospitals, low-volume hospitals, males, those aged <60 years, HIA residents, patients with positive family history, cervical+upper tumors, poorly differentiated tumors, T3/T4 lesions, positive lymph node metastasis, advanced TNM stage, multifocal lesions, patients receiving non-minimally invasive surgery, and earlier treatment periods (all P < 0.05). Multivariate adjusted regression confirmed independent risk factors: low hospital volume (20–80: OR=2.158, 95%CI 1.858-2.494, P<0.001; <20: OR=2.804, 95%CI 2.133-3.619, P<0.001), males (OR=1.174, 95%CI 1.070–1.289, P=0.001), poorly differentiated tumors (OR=1.205, 95%CI 1.045-1.392, P=0.010), TNM stage (III: OR=1.315, 95%CI 1.128-1.539, P<0.001; IV: OR=1.754, 95%CI 1.400-2.189, P<0.001) and multifocal lesions (OR=2.459, 95%CI 1.398-4.011, P=0.003). Independent protective factors comprised LIA residence (OR=0.730, 95%CI 0.659-0.809, P<0.001), middle/lower esophageal location (Middle: OR=0.539, 95%CI 0.486-0.600, P<0.001; Lower: OR=0.654, 95%CI 0.572-0.746, P<0.001), and modern treatment eras (2001–2010: OR=0.679, 95%CI 0.569-0.816, P<0.001; 2011–2024: OR=0.509, 95%CI 0.421-0.620, P<0.001); Higher positive distal margin rates were significantly observed in patients from low-tier hospitals, cancer specialized hospitals, HIA regions, rural areas, patients with positive family history, patients receiving non-minimally invasive surgery, and cases diagnosed in earlier time periods (all P < 0.05). Multivariate adjusted analysis identified two independent risk factors: cancer specialized hospitals (OR=1.592, 95%CI 1.269-2.016, P<0.001) and positive family history (OR=1.531, 95%CI 1.260-1.858, P<0.001). Independent protective factors included LIA residence (OR=0.614, 95%CI 0.482–0.776, P<0.001), and the 2011–2024 period (OR=0.298, 95%CI 0.183-0.496, P<0.001). Clinical characteristics differed significantly between patients with positive versus negative proximal and distal surgical margins. Sensitivity analysis verified the robustness of core predictors for positive proximal resection margins. Male sex, TNM stage IV and multifocal lesions remained independent risk factors, while LIA residence and mid/lower thoracic location acted as consistent independent protective factors. Multivariate regression revealed cancer specialized hospitals and positive family history as risk factors for distal resection margin positivity, with LIA residency and 2011–2024 treatment as protective factors, yet this hospital correlation was driven by baseline bias instead of poor surgical quality, and tumor pathology rather than hospital type governed margin status.