Serial post-surgical stimulated and unstimulated highly sensitive thyroglobulin measurements in low- and intermediate-risk papillary thyroid carcinoma patients not receiving radioactive iodine.

The purpose of this study was to determine the natural temporal trends of serial thyroglobulin (Tg) among low/intermediate-risk PTC patients not receiving radioactive iodine (RAI) using TSH-stimulated Tg (Stim-Tg) and unstimulated highly sensitive Tg (u-hsTg). We prospectively analyzed serial Stim-Tg measurements after total thyroidectomy ± therapeutic central neck dissection among 121 consecutive low/intermediate-risk PTC patients who did not receive RAI, of whom 104 also had serial u-hsTg measurements available. Median follow-up was 6.5 years with Stim-Tg measurements commencing 3 months after surgery and u-hsTg commencing 1.8 years after surgery (when the assay became available). TSH stimulation was performed with 9-day T3 withdrawal, 22-day T4 withdrawal, or using recombinant human TSH (rhTSH). To account for within-patient correlations of repeated Tg measurements, temporal trends in Stim-Tg and u-hsTg were assessed using Generalized Estimating Equations. Stim-Tg models were adjusted for the method of TSH stimulation, whereas the u-hsTg models were adjusted for concurrent TSH level. Linear regression modeling was used to assess the trend in serial Stim-Tg and u-hsTg measurements as a function time from time of surgery throughout the duration of follow-up. The main outcome measured was the change in u-hsTg and Stim-Tg measurements over time. A total of 337 Stim-Tg (2.8/patient) and 602 u-hsTg (5.8/patient) measurements were analyzed. Among the 337 Stim-Tg measurements, Stim-Tg was assessed using rhTSH in 202 (60 %), T4 withdrawal in 41 (12 %), and T3 withdrawal in 94 (28 %) measurements. The overall mean ± 1SD for Stim-Tg and u-hsTg measured was 1.0 ± 1.2 and 0.2 ± 0.1 μg/L, respectively. When adjusted for method of TSH stimulation, serial Stim-Tg measurements did not significantly change over time (all p = NS). The estimated changes in Stim-Tg per year for rhTSH, T4 withdrawal, and T3 withdrawal were 0.01, -0.08, and 0.04 μg/L, respectively. Upon exclusion of 73 patients with an initial undetectable Stim-Tg (n = 48), serial Stim-Tg measurements did not change significantly over time (all p = NS). For these patients, the estimated changes in Stim-Tg per year for rhTSH, T4 withdrawal, and T3 withdrawal were -0.09, -0.10, and 0.01 μg/L, respectively. Serial u-hsTg measurements did not significantly change over time after adjusting for TSH level (p = NS). The estimated change in u-hsTg per year was -0.003 μg/L. No patients had any clinical or imaging evidence of a recurrence during the duration of their follow-up. Among low/intermediate-risk PTC patients not treated with RAI, serial post-surgical Stim-Tg and u-hsTg measurements do not change significantly over a median follow-up of 6.5 years.