| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Editorial |
Department of Endocrinology and Diabetes Monash University Department of Medicine Alfred Hospital Melbourne, Victoria 3004, Australia
Address all correspondence and requests for reprints to: Jim R. Stockigt, Department of Endocrinology and Diabetes, Alfred Hospital, Melbourne, Victoria 3004, Australia.
Effective long-term management of differentiated thyroid carcinoma (DCT) requires persistence and attention to detail. After radioiodine ablation of the postsurgical thyroid remnant, serum thyroglobulin (Tg) and nuclear imaging are the standard techniques that distinguish patients who require further treatment from those who have no residual tissue and a low chance of disease recurrence. If the latter group can be identified with certainty, it may be possible to modify T4 dosage to avoid long-term adverse effects of subclinical hyperthyroidism, in particular effects on bone density. Studies done in the presence of high TSH have greater sensitivity in picking up residual tissue and better specificity in indicating freedom from disease (1, 2, 3, 4). There is now a growing consensus that serum Tg has greater sensitivity than diagnostic whole-body scanning in detecting occult residual thyroid tissue or metastases, when patients who show no evidence of residual tumor during suppressive therapy with T4 are studied under comparable high TSH conditions (1, 2, 3, 4). With TSH stimulation, the estimated difference in sensitivity was 91 vs. 19% in favor of Tg measurement (1). Hence, the inference is that a TSH-stimulated serum Tg should become the principal test in the follow-up of low-risk patients with DCT, in preference to diagnostic whole-body scanning (1, 2, 3, 4). If this is to become standard practice, the strengths and limitations of Tg assays will need to be clearly understood. The issues are complex.
Serum Tg needs to be followed up over long periods, and method variations may influence interpretation; for example, a less sensitive Tg assay or a change in the apparent lower limit of deection can give a misleading impression of disease activity. Long-term comparison can be difficult because of uncertainties in establishing a uniform standard for Tg assays (5). Heterophilic antibodies may distort the apparent Tg concentration, predominantly upward (6). Most important, interference from Tg antibody (TgAb), found in about 20% of patients with DCT (7), can invalidate serum Tg assays (7, 8). Antibody interference in assays for human thyroglobulin has been recognized for three decades (9, 10), and the problem remains one of the most difficult challenges in diagnostic laboratory endocrinology. The bias from antibody interference is different for conventional RIA Tg methods and the immunometric assay (IMA) methods that are now more widely used. In IMA methods, antibody interference, if present, always tends to underestimate serum Tg, probably because the complex of Tg with its antibody cannot form the bridge between two antibodies that is required for the assay signal. By contrast, RIA with double-antibody separation can overestimate serum Tg in the presence of antibody (2, 10), although some assays are claimed to be free of interference. A demonstrable difference between IMA and RIA can give a clue to an antibody-dependent artifact, even when antibody is not detected.
Clinicians generally assume that a negative TgAb report, irrespective of method, rules out the possibility that antibody interference may be causing a misleading underestimate of serum Tg by IMA. This assumption is drawn into question by the extensive study by Spencer et al. (11), across multiple laboratories on three continents using 14 Tg assays (10 IMA, four RIA) and 12 different methods to assess TgAb. A key part of their study reports detailed intermethod Tg and TgAb correlations in 42 samples that were positive by at least one of the 12 TgAb methods. It is notable that these test samples were from TgAb-positive euthyroid subjects rather than a DCT cohort. The results show that both antibody detection and antibody-dependent underestimation of serum Tg by IMA are strongly method dependent. Comparison of Tg concentrations from various assays was complex in the absence of a uniform Tg standard. Across the various assays, the authors corrected each Tg result for assay bias by relating it to the all-method mean for that sample. Each Tg IMA value was then compared with the mean of the four Tg RIA results, taken as the reference value for that sample.
Detailed examination of the data (Figs. 1 and 5 in Ref.11) shows numerous discrepancies between the 12 IMA Tg and 10 TgAb methods. Correlations between various assays and data for individual samples did not show any consistent pattern. In 14 of 42 samples, TgAb was positive by only one of 12 methods, whereas only four of 42 samples were positive by all methods. Some samples showed only trivial interference; nine of the 42 samples that were TgAb positive by at least one method showed no significant underestimate of serum Tg by any IMA method. More important, some samples gave potentially troublesome underestimates of serum Tg by IMA in relation to the RIA mean, without consistently showing detectable TgAb. Of particular concern were samples that were TgAb positive in only some of the 12 test methods yet showed major underestimates of serum Tg by numerous IMA methods. For example, samples 32, 33, and 36 showed underestimates of serum Tg by virtually all IMA methods yet were each classified as antibody negative by three or four of the 12 TgAb methods. When methods from the same source [e.g. Access (Beckman, Palo Alto, CA) and Immulite (Diagnostic Product Corp., Los Angeles, CA], were used in tandem, at least four of the 42 test samples still showed the misleading combination of Tg underestimate without demonstrating positive TgAb. No combination of the multiple test methods appeared to be trouble free. Although it was clear that more potent antibody interference was associated with a greater likelihood of underestimation of serum Tg by IMA, the 42 samples that were variously antibody positive gave inconsistent underestimates of serum Tg across different methods.
The fact that the samples were from antibody-positive euthyroid subjects rather than DCT patients may limit the conclusions that can be drawn because the configuration, iodination, and antigenicity of Tg may differ between benign and malignant thyroid tissue (12). It seems implausible that the antigenic structure of Tg in DCT would be less heterogeneous and would cause less discrepancy than Tg from benign tissue, but the question remains to be answered.
A comparable between-method study in patients being followed up for DCT would further define the clinical implications of these methodological issues, but some conclusions can be drawn. The results do not diminish the importance of detectable serum Tg by IMA as a reliable marker of residual thyroid tissue, either benign or malignant. However, an undetectable serum Tg by an IMA method, in conjunction with an apparently negative TgAb, cannot be regarded as a guarantee that all thyroid tissue has been ablated. If we accept that these data from TgAb-positive subjects without thyroid cancer also apply to the DCT population, confidence in the specificity of a negative antibody report with an undetectable serum Tg becomes less secure. Furthermore, the wide variation between TgAb methods diminishes the value of changes in antibody titer as an index of remaining thyroid tissue (13) unless one assay is used consistently. In view of these methodological vagaries, it is important that clinicians and clinical chemists collaborate to maintain uniform methods over long periods to identify significant trends, which may be more informative than absolute values from different methods.
With the combination of negative serum Tg by IMA with positive TgAb, the radioiodine scan, either diagnostic or after therapy in the presence of high TSH, may be definitive in showing residual tissue. However, in view of the demonstrated lack of sensitivity of both TgAb methodology (11) and diagnostic radioiodine scanning (1), a difficult situation arises in patients with high-risk pathology who show no iodine-avid tissue in the presence of high TSH and are apparently negative for serum Tg by IMA under the same conditions. Additional information can be gained by using a RIA Tg method that does not underestimate serum Tg in the presence of TgAb. Such RIAs are no longer widely available, but there is clearly a need to retain them as reference methods in specialized laboratories. Detailed serial ultrasound of the postoperative neck anatomy, noncontrast computed tomography, or positron emission tomography, probably best done in the presence of high TSH (14), can offer additional information. It is clear that effective long-term management of DCT will continue to depend on multidisciplinary collaboration, especially in patients with high-risk pathology.
Footnotes
Abbreviations: DCT, Differentiated thyroid carcinoma; IMA, immunometric assay; Tg, thyroglobulin; TgAb, Tg antibody.
Received July 13, 2005.
Accepted July 19, 2005.
References
This article has been cited by other articles:
 |
|
 |
|
  A. C.M. Persoon, J. M.W. Van Den Ouweland, J. Wilde, I. P. Kema, B. H.R. Wolffenbuttel, and T. P. Links Clinical utility of an automated immunochemiluminometric thyroglobulin assay in differentiated thyroid carcinoma. Clin. Chem., April 1, 2006; 52(4): 686 - 691. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
