Preoperative Prediction of the Risk of Malignancy in Thyroid Nodules

Objective: This study aimed to propose a formula for predicting the biological nature of thyroid nodules, thus helping in clinical decision making and avoiding unnecessary surgery. Patients and methods: Prospective, longitudinal and observational study was conducted on 400 patients submitted to primary thyroid surgery at the Department of Surgery, University of Pisa, Italy, between April 1st and June 30th 2008, whose clinical charts indicated gender, age, ultrasonographic findings, TSH concentrations, while not receiving thyroid hormones and whose definite histopathology report was available. Other variables recorded were: lymph node characteristics, estimated thyroid volume, thyroid functional tests, thyroglobulin, anti thyroid antibodies, calcitonin, fine needle aspiration cytology, frozen sections and vocal cord assessment through laryngoscopy.Statistical analysis was performed using Pearson correlation test and logistic regression backward elimination. Results: We designed two formulas which included thyroid functional tests, calcitonin and fine needle aspiration with and without anti thyroid antibodies and demographic variables. These formulas could predict the biological nature of thyroid nodules with 92.3% sensitivity and 91.67% specificity, positive and negative predicting values of 88.5 % and 94.49%, respectively. Both formulas showed a better performance compared to fine-needle aspiration biopsy (FNA) alone. Conclusion: We developed a formula that can predict the biological behavior of thyroid nodules in Italian patients, their applicability in other countries requires further study.