Frozen section was previously required for the management of thyroid nodules; however, because of good results from preoperative fine-needle aspiration, the more invasive frozen section can be reserved for use in cases in which the pathology is ambiguous on fine-needle aspiration. Richards and colleagues conducted a study to determine the optimal role of frozen section in diagnosing thyroid cancer and deciding the extent of surgery.

The researchers studied 231 patients who underwent thyroidectomy at a university medical center between 1995 and 2000, after excluding 10 patients who underwent thyroidectomy for the management of nonthyroid pathology or had unavailable medical records. The retrospective review correlated the final pathologic result from the surgical specimen with the cytologic findings on frozen section or fine-needle aspiration. The results were further analyzed to identify patient subgroups that might be better investigated by frozen section.

The 30 men and 201 women included in the study ranged in age from eight to 86 years (mean age, 44.7 years). The thyroidectomy specimens showed malignancy in 47 cases (20 percent) and benign disease in 184 cases (80 percent). Fine-needle aspiration was performed on 154 patients (67 percent), frozen section on 140 patients (61 percent), and both fine-needle aspiration and frozen section were performed on 103 patients (45 percent). In 27 of the patients (17 percent) who underwent fine-needle aspiration biopsy, the results were nondiagnostic. Ten patients with benign fine-needle aspirations had a false-negative result. The overall sensitivity of fine-needle aspiration for diagnosing thyroid malignancy was 50 percent, and the specificity was 99 percent. Thirteen patients with frozen section had false-negative results. Frozen section had an overall sensitivity for malignancy of 50 percent and a specificity of 100 percent. The extent of thyroid surgery was altered because of frozen-section results in only one case.

Overall, fine-needle aspiration had a 70 percent diagnostic accuracy compared with 90 percent for frozen section; however, the diagnostic accuracy of fine-needle aspiration was lowered by the 27 inadequate biopsies. The final pathology was benign in 89 percent of patients with a nondiagnostic fine-needle aspiration. The lowest diagnostic accuracy occurred in cases with suspicious results on both fine-needle aspiration and frozen section.

The authors conclude that routine frozen section is not supported because its contribution to diagnosis and clinical management is not balanced by the increased operating time, risk to the patient, and requirement for pathologic services. They recommend that frozen section be considered only when the clinical suspicion of malignancy is significant and the fine-needle aspiration results are suspicious or unsatisfactory, and in patients with unexpected findings during surgery.