Abstract / Summary
Abstract Objective Thyroid nodules with indeterminate cytology often undergo diagnostic surgery despite a relatively low risk of malignancy, and molecular testing is limited by cost and availability. Because the Warburg effect is independent of cytological category, we evaluated intranodular lactate across all Bethesda categories, with emphasis on indeterminate cytology. Methods Prospective, cross-sectional study of 338 surgically treated thyroid nodules at a single tertiary referral center, with histopathology as the reference standard. Nodules were classified according to the Bethesda System and TI-RADS. Lactate was measured with a portable analyzer from the solid component of each nodule and from extranodular thyroid parenchyma, with capillary lactate as a systemic control. Performance was assessed by ROC analysis, sensitivity, specificity, likelihood ratios, and Bayesian post-test probabilities. Results Lactate increased progressively from extranodular parenchyma (1.85 ± 0.49 mmol/L) to benign (2.30 ± 0.96) and malignant nodules (4.69 ± 1.92 mmol/L; p < 0.0001). The optimal cutoff was 3.0 mmol/L (AUC 0.90 overall; 0.83 in Bethesda III–IV), with 91% sensitivity and 83% specificity. Negative predictive value was 92% for Bethesda I/III/IV/V and 95.6% for Bethesda III/IV. Lactate did not differ between classic and follicular papillary carcinoma subtypes (p = 0.34), and no extranodular sample affected by lymphocytic thyroiditis exceeded the cutoff. Estimated relative reduction in diagnostic surgery was 61.7% (Bethesda III–IV) and 48.5% (Bethesda III–V). Conclusions Intranodular lactate is a feasible, low-cost metabolic biomarker that performs consistently across cytological categories, with greatest clinical value in Bethesda I, III, and IV nodules, where it may substantially reduce unnecessary diagnostic surgery.