TSH Levels Are Predictive of Malignancy in Preoperative Thyroid Nodules: Our Single-Center Experience

Objective: To evaluate the association between preoperative thyroid-stimulating hormone (TSH) levels and postoperative histopathological malignancy in patients undergoing thyroidectomy for thyroid nodules, and to compare the discriminative performance of TSH with that of TI-RADS.Material and Methods: A total of 366 patients operated for thyroid nodules between January 1, 2023 and January 1, 2025 at a tertiary center were retrospectively reviewed. Preoperative TSH, fT3, fT4 and routine biochemical parameters, TI-RADS score, fine-needle aspiration biopsy (FNAB) cytology and type of surgery were recorded. Patients were dichotomized as benign or malignant according to final pathology. ROC analysis was performed for TSH and TI-RADS, with the optimal cut-off determined by the Youden index. Variables associated with malignancy were assessed by univariate and multivariate logistic regression.Results: Final pathology was benign in 223 patients (60.9%) and malignant in 143 (39.1%). Groups were comparable for age and sex (p=0.737 and p=0.477, respectively). TSH was significantly higher in the malignant group (median 1.250 vs 1.115 µIU/mL; p=0.029), as was fT3 (p=0.010). The AUC was low for TSH (0.572; 95% CI: 0.508–0.636) but substantially higher for TI-RADS (0.802; 95% CI: 0.752–0.853). The Youden-derived TSH cut-off was 0.687 µIU/mL (sensitivity 85.8%, specificity 32.8%). In multivariate analysis, TSH ≥0.687 µIU/mL (OR=6.10; 95% CI: 2.28–16.33), TI-RADS score (OR=3.58; 2.25–5.69) and non-benign FNAB cytology (OR=16.04; 6.63–38.79) were independently associated with malignancy (all p