Thyroid cancer: a multivariate analysis on influence of treatment on long-term survival.

Contemporary management of patients with follicular pattern thyroid cancer remains uncertain. This retrospective multivariate analysis studied 410 such patients treated in a cancer hospital in the years 1932-72 and providing a follow-up of 20 years. In papillary carcinoma thyroxine administration (P < 0.005) and surgery (P < 0.001) improved survival together with youth (P < 0.001) and being female (P < 0.05). In follicular carcinoma, thyroxine therapy (P < 0.001) increased survival as did surgery but it failed to reach significance (P = 0.19); increasing age (P < 0.001), stage M1 (P < 0.05) and 'complete' radiotherapy (P < 0.05) decreased survival. In anaplastic carcinoma survival was improved by thyroxine therapy (P < 0.001), a new finding, but decreased by stages T3 (P < 0.001) and M1 (P < 0.05); however, radiotherapy, the mainstay in control of local disease, did not increase survival. Overall, total thyroidectomy reduced local recurrence with an increase in complications and no operation gained a significant increase in survival. Radio-iodine achieved no benefit in survival which raises the possibility of confounding. Thyroxine therapy is indicated as initial treatment of follicular cell thyroid cancer.