The circulating Tg concentration reflects the net sum of: (a) the mass of Tg-secreting thyroid tissue present (normal tissue and/or tumor), (b) any inflammation of, or injury to, thyroid tissue secondary to fine needle aspiration biopsy (FNAB), surgery, radioactive iodine (RAI) therapy or thyroiditis, and (c) the degree of TSH receptor stimulation (from endogenous TSH, rhTSH, high hCG, or thyroid stimulating immunoglobulins). In patients with DTC, thyroid mass and TSH status are the major factors that influence the interpretation of serum Tg concentrations, as schematically shown in Fig. 13B.1.

As shown in Table 13B.1, some methods are still not directly standardized against the Certified Reference Material (CRM-457). Between-method variability remains high (∼30%)—two-fold higher than within-person variability or the coefficients of variation of measurements made with the same method in different laboratories (17,24,25,26,27). The between-method variability of IMA methods appears higher than for RIA (24). This likely reflects standardization and assay matrix issues, coupled with the employment of monoclonal antibody IMA reagents with narrow epitope specificity for different Tg isoforms, in contrast with the polyclonal antibodies with broader epitope specificities used for RIA (Table 13B.1) (72,84,85,86,87). Tg epitopes are conformational (88,89). There is potential for tumors to secrete abnormal immature Tg isoforms with low iodine and altered carbohydrate content resulting in an abnormal tertiary molecular structure that interacts differently with the antibody reagents of different assays (24,85,86,90,91,92,93,94,95). The wide between-method variability of current assays precludes changing the Tg method during long-term postoperative Tg monitoring of DTC patients. A change in method has the potential to result in a two-fold difference in Tg values that could be interpreted as a clinically significant change [more than 1.5 μg/L (ng/mL)] (76) and either mask disease or prompt unnecessary concern for recurrence. When practical, the archiving of unused specimens for concurrent re-measurement of a past with a current specimen could eliminate run-to-run variability and facilitate re-baselining, should a change in Tg method become necessary.

As seen in Table 13B.1, methods have Tg reference ranges that vary and functional sensitivities (76) that are often close to the lower reference limit (17,24,75). The variability in upper reference limits likely reflects assay specificity differences coupled with the rigor of selecting subjects without thyroid pathology. The reference ranges established for normal control subjects with intact thyroid glands are not relevant to managing thyroidectomized DTC patients unless adjusted for the influence of TSH and the degree of thyroid surgery (76). Specifically, as shown in Fig. 13B.1, lobectomy plus TSH suppression each reduce serum Tg by approximately 50% (62,76). When iodine is sufficient, euthyroid control subjects with normal thyroid glands [∼10 to 15 g (96)] have serum Tg concentrations that average 10 to 13 ng/mL (μg/L) (Table 13B.1) (61,76). It follows that a gram of normal thyroid tissue gives rise to approximately 1.0 μg/L (ng/mL) Tg in the circulation at euthyroid TSH levels. Thus, 1 g of normal thyroid tissue would be expected to produce 1.0 ng/mL (μg/L) Tg in the circulation when TSH is normal. Given the ten- to twenty-fold stimulation expected during high TSH states, the 1 to 2 g thyroid remnant typically remaining following near-total thyroidectomy (97) would be expected to produce a serum Tg below ∼2.0 μg/L (ng/mL) in the absence of TSH stimulation, and below ∼20 μg/L (ng/mL) 72 hours after rhTSH administration or thyroid hormone withdrawal (Fig. 13B.1) (10).

The use of IMA methodology did not initially improve Tg assay sensitivity, as had been the case for TSH (17,24,75,98). Assay detection limits were difficult to compare before the protocol for determining a realistic estimate of the detection limit (functional sensitivity) was established (76). Functional sensitivity is now defined by guidelines as: “the serum Tg value that can be measured with 20 percent between-run precision over a period of more than 6 months using at least two different reagent lots and two different calibrator lots” (76). Most current assays remain first-generation, defined as having a functional sensitivity between 0.5 and 1.0 μg/L (ng/mL) (Table 13B.1) (17,24,71,75). The sensitivity limit of

such assays is only marginally below the reference range for euthyroid controls with intact thyroid glands and thereby suboptimal for detecting small amounts of persistent/recurrent tumor remaining after thyroidectomy for DTC (Fig. 13B.2). During the last decade, more sensitive second-generation Tg IMA methods [functional sensitivities ≤ 0.1μg/L (ng/mL)] have become available (14,15,16,17,19,24,61,75). Despite some controversy concerning how to interpret low but detectable Tg values in the 0.1 to 1.0 μg/L (ng/mL) range that were hitherto believed to be “undetectable” (10,11,12,13,18), there is growing recognition that the need for rTSH-stimulation is reduced by using a second-generation method to monitor bTg trends (14,15,16,17,19,20,22,23,61,75,99,100).

An unintended consequence of adopting IMA methodology for Tg measurement has been an exacerbation of interference problems, involving both HAMA (31,32,33) and TgAb (24,28,29,30,101,102). Whereas HAMA interference usually results in a falsely high serum Tg that may prompt unnecessary investigations for disease (32,33), TgAb interference causes falsely low or undetectable Tg IMA measurements that have more serious consequences by masking the presence of disease (24,28,101,102,103,104). Whereas blocker reagents are added to IMA methods to minimize HAMA interference (19,31,105), there are still no reliable measures to overcome TgAb interference with Tg measurements.