Frozen Section-Guided Selective Ipsilateral Central Neck Dissection in Bethesda V-VI Thyroid Nodules

This observational retrospective cohort study was conducted in a tertiary referral endocrine surgery unit and was based on prospectively collected institutional data. The study included consecutive adult patients undergoing primary thyroid surgery for cytologically suspicious or malignant thyroid nodules, classified as Bethesda V or VI according to the Bethesda System for Reporting Thyroid Cytopathology.

The study was designed to evaluate whether an intraoperative strategy based on ipsilateral central neck dissection (ipsiCND) with frozen section analysis (FSA) could improve the adequacy of surgical treatment in selected patients with low-risk thyroid cancer, potentially reducing both overtreatment and undertreatment. The institutional protocol incorporating ipsiCND-FSA was introduced in September 2023. Patients treated before implementation of this protocol generally underwent total thyroidectomy according to the standard institutional practice in place at that time. After implementation of the protocol, hemithyroidectomy combined with ipsilateral central neck lymph node sampling and intraoperative frozen section analysis was systematically proposed as the preferred surgical strategy in appropriately selected patients. Total thyroidectomy remained an option for patients who declined the conservative approach after counselling regarding the malignant or suspicious nature of the thyroid nodule and the available surgical alternatives.

Eligible patients were adults aged 18 years or older with a solitary thyroid nodule measuring less than 2.5 cm in maximum diameter, classified as Bethesda V or VI on preoperative cytology, with a clinically node-negative neck documented by preoperative cervical ultrasound. Patients were considered only if they underwent primary thyroid surgery and had complete histopathological data available for the thyroid specimen and, when removed, central compartment lymph nodes. Patients with previous thyroid or neck surgery, known distant metastases at the time of surgery, incomplete clinical or pathological data relevant to the study endpoints, or reoperations such as completion thyroidectomy or secondary lymph node dissection were excluded.

All patients received detailed preoperative counselling and provided written informed consent before surgery. The information provided included the planned procedure, the possible need for intraoperative modification of the surgical strategy according to real-time findings and frozen section results, and the potential risks and benefits of the available surgical options. Patients also consented to the inclusion of their clinical and pathological data in the institutional database and to their use for research purposes. The study was conducted in accordance with the principles of the Declaration of Helsinki, followed good clinical practice standards, and was approved by the local Ethics Committee under protocol PA-1-RETRO-2026-TBD. Reporting was planned in accordance with the STROCCS recommendations for observational studies.

All surgical procedures were performed or supervised by experienced endocrine surgeons. Intraoperative nerve monitoring was routinely used in all cases as part of the standard institutional practice. The extent of thyroid surgery, either total thyroidectomy or lobectomy/isthmectomy, and the strategy for management of the central neck compartment were planned preoperatively according to cytological diagnosis, imaging findings, clinical risk profile, and institutional practice. In patients managed according to the ipsiCND-FSA protocol, the operative plan could be modified intraoperatively on the basis of frozen section findings.

Patients were stratified according to intraoperative management of the central compartment. In the FSA group, one or more ipsilateral level VI lymph nodes were removed during hemithyroidectomy and immediately submitted for frozen section analysis. Lymph nodes were processed using standard cryostat techniques, stained with hematoxylin and eosin, and evaluated intraoperatively by an experienced pathologist. The intraoperative nodal status was reported as positive or negative for metastatic disease. Definitive paraffin-embedded histopathology served as the reference standard for final assessment. In the non-FSA group, no intraoperative frozen section analysis of central compartment lymph nodes was performed. In this cohort, patients underwent total thyroidectomy, reflecting the inability to obtain sufficiently reliable preoperative risk stratification according to the standard approach used before implementation of the new protocol.

The primary objective of the study was to assess treatment adequacy. Treatment adequacy was defined as the concordance between the surgical procedure actually performed and the minimum surgical treatment considered necessary to achieve appropriate locoregional disease control on the basis of final histopathological findings. The minimum necessary extent of surgery was determined retrospectively, with particular attention to definitive tumor characteristics, including tumor focality and laterality, extrathyroidal extension, vascular invasion, aggressive histological variants, and nodal involvement.

Overtreatment was defined as performance of total thyroidectomy in patients who were ultimately found to have a solitary unilateral carcinoma without microscopic or macroscopic extrathyroidal extension, vascular invasion, or aggressive histological variants. Undertreatment was defined as an insufficient extent of surgery in the presence of adverse pathological features requiring a more comprehensive oncological procedure. These definitions were applied irrespective of study group allocation. For comparative analyses, overtreatment and undertreatment were also combined into a composite endpoint of surgical mis-treatment.

Secondary objectives included evaluation of the diagnostic performance of frozen section analysis for detecting central compartment lymph node metastases. Diagnostic performance was assessed by calculating sensitivity, specificity, positive predictive value, negative predictive value, overall accuracy, and the area under the receiver operating characteristic curve, using definitive histopathology as the reference standard. Additional secondary outcomes included postoperative complications and length of hospital stay. Postoperative complications included major complications such as recurrent laryngeal nerve injury, hypoparathyroidism, postoperative hemorrhage, and minor complications such as seroma, hematoma, and wound infection.

Clinical, surgical, pathological, and postoperative data were extracted from the prospectively maintained institutional database and cross-checked against source documents, including operative reports, pathology reports, ultrasound reports, discharge summaries, and postoperative follow-up records. Data quality was supported by predefined data fields, standardized definitions of variables and endpoints, internal consistency checks, and verification of missing or inconsistent values against the original clinical documentation. Missing data were handled by excluding patients with incomplete information relevant to the primary endpoint or key secondary endpoints. No formal sample size calculation was performed because of the retrospective observational design; all consecutive eligible patients treated between January 2023 and June 2025 were included.

Statistical analyses were planned to compare baseline characteristics, surgical management, treatment adequacy, diagnostic performance of frozen section analysis between study groups. Diagnostic performance of frozen section analysis was evaluated against definitive histopathology as the reference standard. Additional analyses were planned to account for potential confounding and to assess the robustness of the primary findings.