Comment on “The lymph node yield in the initial lateral neck dissection predicts recurrence in the lateral neck of papillary thyroid carcinoma: a revision surgery cohort study”
Year: 2026
Type: article
Abstract: Dear Editor, Papillary thyroid carcinoma (PTC), the most prevalent thyroid malignancy, demonstrates a notable tendency for cervical lymph node metastasis[1]. We read with great interest the study by Lei and colleagues, which investigated the association between lateral lymph node yield (LLNY) and the ratio of lateral positive lymph nodes to lymph node yield (LPLR) from the initial surgery and the risk of recurrence in patients undergoing revision lateral neck dissection for PTC[2]. While their findings offer valuable insights for clinical practice, we wish to raise several methodological and interpretive points that we believe are crucial for contextualizing the results. This correspondence was developed following the TITAN Guidelines 2025(2) for the declaration and use of artificial intelligence in scholarly publishing. No AI tools were employed in the conception, design, analysis, interpretation, or writing of this work[3]. First, the biological heterogeneity of PTC, encompassing subtypes with vastly different prognostic implications, presents a potential confounder. Indolent variants (e.g., encapsulated, follicular, cribriform-morular) typically exhibit lower metastatic potential, whereas aggressive subtypes (e.g., tall cell, columnar cell, diffuse sclerosing) are inherently prone to extensive nodal involvement and recurrence[4]. Although the study demonstrates an association between low LLNY and recurrence, the absence of stratification by pathological subtype leaves open the possibility that the observed finding is a confounding effect. A low node yield may simply be an extrinsic marker of the aggressive intrinsic property of high-risk variants, rather than being solely indicative of an incomplete initial dissection. Stratifying the analysis by pathological variant would help clarify whether LLNY retains its prognostic value independently of tumor biology. Second, the statistical robustness of the proposed cutoff values for LLNY (24.5%) and LPLR (32.74%) requires further support. The report omits the area under the curve metrics, leaving readers unable to assess the discriminative power of the ROC analysis. Moreover, the lack of internal validation (e.g., bootstrapping) or sensitivity analysis raises concerns about potential overfitting of these thresholds to the specific study cohort. External validation would be necessary to confirm their generalizability. Third, the variable selection for the multivariate model is insufficiently documented. The authors state that backward elimination was applied, yet the highly significant LPLR from univariate analysis was excluded from the final model, which only included pathological T stage and LLNY. The statistical rationale for this omission, be it collinearity or other criteria, is not provided. This lack of clarity regarding a key step in model refinement undermines the transparency and reproducibility of the findings. Fourth, the number of lymph nodes retrieved is subject to considerable variation, influenced by surgical technique, pathological processing protocols, and patient-specific factors like body habitus or coexistent Hashimoto’s thyroiditis. As most initial surgeries (85.9%) were performed at external institutions, the absence of standardized lymph node examination protocols introduces a potential, unmeasured bias. Furthermore, the analysis did not account for factors known to affect lymph node yield, such as BMI or thyroiditis, which could confound the association between LLNY and recurrence. Notwithstanding these considerations, we commend the authors for addressing an important clinical question. Their work meaningfully contributes to the ongoing effort to establish quantitative metrics for surgical quality in PTC, which could ultimately refine risk stratification and guide postoperative management.
Source: International Journal of Surgery
Authors: YiRong Chen, Shanyan Hong, Yongqiang Chen
Language: English
Subfield: Endocrinology, Diabetes and Metabolism
Field: Medicine
Domain: Health Sciences
Open Access status: hybrid
Estimated APC paid: $3,450
OpenAlex ID: W7124264492