Extent of lymphadenectomy dictates adjuvant chemotherapy benefit in primarily resected esophageal squamous cell carcinoma: a multi-cohort validation

Background While primary surgery remains a crucial treatment for resectable esophageal squamous cell carcinoma (ESCC), the uniform application of postoperative adjuvant chemotherapy for pathologically node-positive patients exposes many to severe systemic toxicity. This study hypothesizes the existence of a specific lymph node dissection (LND) threshold that stratifies the survival benefit of adjuvant chemotherapy and aims to define this cutpoint to enable safe therapeutic de-escalation. Methods We conducted a multi-cohort retrospective study using data from the Surveillance, Epidemiology, and End Results (SEER) database (n = 753, training cohort) and a Chinese single-center repository (n = 447, validation cohort) of patients with ESCC who underwent primary radical esophagectomy without radiotherapy. The optimal prognostic cutpoint for LND was mathematically determined using X-tile software. An interactive Cox proportional hazards nomogram incorporating an LND-by-chemotherapy interaction term was developed and validated. Subsequently, layer-specific 1:1 propensity score matching (PSM) was performed within separate surgical strata of the validation cohort to assess chemotherapy efficacy. Results X-tile analysis established 21 lymph nodes as the optimal LND threshold. Multivariable analysis confirmed that adequate LND (≥ 21 nodes) and adjuvant chemotherapy were independent protective prognostic factors. The interactive nomogram demonstrated good discrimination and calibration across both populations. Crucially, layer-specific PSM in the external validation cohort revealed divergent therapeutic responses dependent on surgical quality. For patients with suboptimal LND (< 21 nodes), adjuvant chemotherapy provided a significant survival benefit over surgery alone (P = 0.048). In contrast, for patients achieving an adequate LND (≥21 nodes), the addition of adjuvant chemotherapy yielded no further survival advantage (P = 0.085). Conclusions Achieving the removal of ≥ 21 lymph nodes serves as a robust anatomical biomarker that fundamentally alters the postoperative therapeutic landscape for ESCC. Patients with suboptimal LND require adjuvant chemotherapy to compensate for potential residual disease, whereas those achieving the 21-node clearance threshold derive no additional survival benefit from systemic interventions and can be safely exempted from cytotoxic over-treatment.