Survival discriminants for differentiated thyroid cancer.

Since 1975, the American Cancer Society, Illinois Division, has published end results of major cancer sites drawn from patient data contributed voluntarily by hospital cancer registries throughout the state. The current study was undertaken, in part, to apprehend information regarding contested areas in the management of patients having differentiated (papillary/follicular) thyroid cancer. A total of 2,282 patients with either papillary or follicular carcinoma of the thyroid from 76 different Illinois hospitals and providing 10 years of follow-up information (life-table analysis) were retrospectively analyzed for demographic, disease, and treatment-related predictors of survival. Multivariate analysis using the Cox proportional hazards method was made for stage, age, race, sex, morphology, history of radiation exposure, presence of positive lymph nodes, initial surgical treatment, postoperative iodine 131 therapy, and replacement/suppressive thyroid hormone treatment. Statistically significant (p less than or equal to 0.05) predictors of favorable survival after thyroid cancer were low stage (I and II), young age (less than 50 years), white race, female sex, and the administration, postoperatively, of either thyroid hormone or radioactive iodine. Factors that had no influence on survival were lymph node status, choice of initial surgical treatment, and a history of prior irradiation. We suggest that where a prospective clinical trial is impracticable, a retrospective analysis of a large and detailed database, such as that available from cooperating hospital-based tumor registries, may yet provide useful insights to solutions of cancer management problems.