FOLLICULAR AND HURTHLE CELL THYROID NEOPLASMS - IS FROZEN-SECTION EVALUATION WORTHWHILE

Objectives: To determine whether (1) frozen-section (FS) evaluation of follicular and Hurthle cell thyroid neoplasms (FHCNs) is accurate, (2 ) FS aids in intraoperative decision-making, and (3) FS is cost-effect ive. Design: Retrospective review of patient histories and FS and para ffin-embedded slides. Permanent histologic sections were considered th e standard criterion. Follow-up was achieved in 92% of patients with a mean follow-up of 5.7 years.Setting: Tertiary care referral center. P atients: All patients undergoing thyroidectomy for a suspected FHCN be tween January 1, 1985, and December 31, 1994. Patients included were t hose whose condition was diagnosed as FHCN, either on FS, permanent hi stologic sections, or both. Main Outcome Measures: Sensitivity, specif icity, positive predictive value, negative predictive value, and accur acy of FS analysis were determined. Total adjusted hospital charges we re compared for those undergoing 1 vs 2 cancer operations. Multivariat e analyses were carried out to determine the optimal predictive model for follicular cancer. Results: The study group included 1023 patients (737 women and 286 men), of whom 83 (8.1%) were diagnosed as having a malignant FHCN on permanent section. The diagnosis of malignant neopl asm was correctly established in 65 (78%) of the 83 patients on FS, th ereby permitting definitive surgical management at the first operation . Sensitivity, specificity, positive predictive value, negative predic tive value, and accuracy for FS diagnosis of malignant FHCN were 78%, 99%, 90%, 98%, and 98%, respectively. In a multivariate analysis, FS w as the most significant variable predictive of malignant neoplasm. App roximately $400 000 was saved in hospital charges by the use of FS as a result of the elimination of many 2-stage operations. Conclusion: Fr ozen-section evaluation of FHCN can be performed with a high degree of accuracy, permitting considerable cost savings.