RESUMEN
The value of routine frozen section examination for intraoperative diagnosis of thyroid cancer is controversial and needs to be evaluated on an institution to institution basis. This paper highlights the local experience with frozen section and analyses its role, accuracy and limitation in diagnosing thyroid cancer. A retrospective study of 61 patients who underwent thyroidectomy with routine thyroid frozen section examination at King Fahd Hospital of the University, Al-Khobar, Kingdom of Saudi Arabia. Results were compared with the final diagnosis to evaluate its effectiveness in predicting malignancy. Frozen section diagnosis, proved to be benign in 44 [72%] specimens, malignant in 10 [16%] and the diagnosis deferred in 7 [11.5%] patients. After examination of the permanent sections of the specimens, 15 [28%] were found to be malignant including 3 [43%] out of 7 deferred cases. With an overall accuracy of >90% and a specificity of 100%, the sensitivity of frozen section diagnosis remains low. Only 60% of papillary carcinoma and 25% of follicular carcinoma were correctly diagnosed on frozen section. The low rate of diagnosis of follicular carcinoma is due to its encapsulated and minimally invasive nature. The inability to diagnose follicular carcinoma intraoperatively with frozen section is the most significant factor accounting for the relatively low sensitivity of frozen section diagnosis of thyroid malignant neoplasm. In our institution, frozen section is considered a complementary investigation to emerging fine needle cytopathology in the region