Optimal surgery for papillary thyroid carcinoma.

Abstract

There has been a long, unresolved debate regarding the operation of choice for well differentiated carcinoma of the thyroid. We therefore analyzed the feasibility, scope, sample size, and length of follow-up required to determine the optimal operation for papillary thyroid carcinoma. A statistical approach was used to design a randomized prospective trial comparing the endpoints of complications, recurrence, and cause-specific mortality. A complication comparison trial is prohibitive owing to the large population required--approximately 12,000 randomized patients. A recurrence trial appears feasible based on sample size: approximately 360 to 800 patients with a 6 to 10-year follow-up. However, recurrence detection would be severely compromised in the lobectomy arm, and a unilateral lag-time bias would occur. A cause-specific mortality trial proves to be the least objectionable, although a large sample size (n = 3100) would be required. Such a trial is critically dependent on the ability to select an appropriate endpoint. A cause-specific mortality trial proves to be the most objective and statistically valid endpoint.