Association of recurrent laryngeal nerve lymph node retrieval with survival in early-stage resectable esophageal squamous cell carcinoma: a retrospective cohort study

Purpose The clinical relevance of recurrent laryngeal nerve lymph node (RLN LN) management in early-stage resectable esophageal squamous cell carcinoma (ESCC) remains controversial. We evaluated the association of RLN LN retrieval with overall survival (OS) and disease-free survival (DFS) in patients with pT1bN0/pT2N0 thoracic ESCC and examined whether total lymph node (LN) yield was associated with outcomes. Methods We retrospectively analyzed 212 patients with pT1bN0 or pT2N0 ESCC who underwent esophagectomy between January 2015 and September 2018. Patients were categorized by RLN LN retrieval status (no RLN LN retrieved vs ≥1 RLN LN retrieved (sampling or dissection)) and by total LN yield (<15 vs ≥15). RLN LN procedures were further classified as sampling vs formal dissection and unilateral vs bilateral retrieval. OS and DFS were estimated using Kaplan–Meier methods and compared with log-rank tests. To reduce confounding, we applied stabilized inverse probability of treatment weighting (sIPTW) based on prespecified preoperative covariates (excluding postoperative variables), with weight trimming at the 1st–99th percentiles, and fitted weighted Cox models with robust standard errors. Sensitivity analyses additionally adjusted for postoperative factors and included an approach-restricted analysis within the McKeown subgroup. Results RLN LN retrieval (≥1) was associated with improved OS and DFS vs no RLN LN retrieved in both unweighted and sIPTW-weighted analyses (sIPTW-weighted HR 2.75, 95% CI [1.53–4.96] for OS; HR 2.84, 95% CI [1.61–5.01] for DFS; both p < 0.001). In sIPTW-adjusted multivariable models, absence of RLN LN retrieval remained associated with higher risks of death (adjusted HR 3.18, 95% CI [1.60–6.30]; p < 0.001) and recurrence/death (adjusted HR 3.51, 95% CI [1.86–6.63]; p < 0.001). Recurrent laryngeal nerve palsy occurred more frequently in the RLN LN retrieval group (21.9% vs 9.3%, p = 0.021). In exploratory procedure-focused analyses among RLN LN–retrieved patients, formal dissection ( vs sampling) was associated with improved DFS (HR 0.40, 95% CI [0.18–0.90]; p = 0.027) with a borderline OS signal, whereas bilateral retrieval showed no additional benefit. Total LN yield (<15 vs ≥15) was not materially associated with OS or DFS overall or within pT1b/pT2 strata. Conclusions In patients with pT1bN0/pT2N0 thoracic ESCC, retrieval of at least one RLN LN was associated with improved OS and DFS, whereas higher total LN yield was not clearly associated with survival. Because RLN-chain nodal retrieval was closely related to surgical approach and lymphadenectomy field, these findings should be interpreted as observational associations rather than proof of an independent causal benefit of RLN LN dissection itself. Larger, ideally multicenter studies are needed to clarify the respective contributions of RLN-chain nodal management, staging adequacy, and operative approach to long-term outcomes.