Long-term follow-up of patients with papillary and follicular thyroid cancer: a prospective study on 715 patients.

PURPOSE: This prospective study evaluated the recurrence rate in 715 patients with differentiated thyroid cancer who had no evidence of persistent disease after total thyroidectomy and lymph node dissection in 94% of them followed up by radioiodine ablation (30-100 mCi) and assessed the predictive value of the initial thyroglobulin (Tg) levels for detecting recurrence, both during levothyroxine (LT4) treatment and after TSH stimulation. PATIENTS AND METHODS: Patients had Tg determinations performed at 3 months on LT4 treatment (Tg1) and at 9-12 months after stimulation by either thyroid hormone withdrawal or recombinant human TSH (Tg2); the Access kit was used (functional sensitivity of 0.11 ng/ml); they had undetectable anti-Tg antibodies. Patients were followed up annually. Predictive values were calculated by comparing Tg levels (Tg1 and Tg2) and the outcome in terms of recurrence. RESULTS: During the median follow-up of 6.2 yr, 32 patients had a recurrence. Assuming a cutoff level for Tg1 at 0.27 ng/ml, Tg1 sensitivity and specificity reached 72 and 86%, respectively, whereas predictive positive and negative values were 20 and 99%, respectively. With a cutoff level for Tg2 at 1.4 ng/ml, sensitivity and specificity reached 78 and 90%, respectively, whereas positive and negative predictive values were 26 and 99%, respectively. CONCLUSION: This large prospective cohort of patients presented a low rate of recurrence. Initial Tg measurements allow to predict long-term recurrence with an excellent specificity. Stimulated Tg determination presented a slightly higher sensitivity than Tg determination on LT4. TSH stimulation may be avoided when Tg measured 3 months after ablation is less than 0.27 ng/ml during LT4 treatment.