The utility of preoperative serum thyroid-stimulating hormone level for predicting malignant nodular thyroid disease

IntroductionThe aim of this study was to assess whether serum thyroid-stimulating hormone (TSH) levels are of value in predicting malignancy in patients with nodular thyroid disease (NTD).MethodsPatients with NTD and a preoperative TSH level who underwent thyroidectomy between 1990 and 2008 were identified from a prospective database. Age, sex, TSH concentration, nodule size, and pathology were evaluated. Logistic regression analysis was used to determine which factors were predictive of malignancy.ResultsSix hundred fifty-three patients were analyzed. The overall rate of malignancy was 20%; the rate was highest in patients <30 years (32%). The mean TSH level was higher in the malignant group (5.5 μIU/mL vs 1.4 μIU/mL, P < .0001). The rate of malignancy was 65% in patients with TSH levels >5.5 μIU/mL. Logistic regression analysis revealed that TSH level was the only significant risk factor for malignancy.ConclusionThe serum TSH level may be useful in predicting the probability of cancer and optimizing the extent of thyroidectomy in patients with NTD.