Thyroid-Stimulating Antibodies Predict Hyperthyroidisma

Terry F. Davies, MD, FRCP, FACE

Mount Sinai School of Medicine New York, New York


    Introduction
 Top
 Introduction
 Interpreting the medical...
 Potential factors interfering...
 The cost effectiveness factor
 The pregnancy conundrum
 Summary
 References
 
OVER 50% of patients with hyperthyroid Graves’ disease relapse after a 12-month course of antithyroid drugs, the actual percentage varying among populations and with their iodine intake. The measurement of TSH receptor antibodies (TSHR-Abs) by competitive radioreceptor assay or by thyroid cell bioassay (1) in patients with Graves’ disease can be a useful predictor of relapse and remission (2). However, controversy has surrounded the clinical utility of TSHR-Ab measurements ever since my colleagues and I published the first report of their use in the Lancet in 1977 (2). In that study, we analyzed the usefulness of TSHR-Abs to predict relapse and remission after 6 months of antithyroid drug treatment and found that highly positive titers of the autoantibody were indeed useful predictors of recurrence (Fig. 1Go). Similar data had been published previously using indirect measurements of TSHR-Abs—for example, the pioneering work of Alexander, who had showed clearly that the failure of T3 to suppress radioiodine uptake increased the chances of recurrence because a non-TSH thyroid-stimulator of radioiodine uptake was present (3). Similar studies continue to confirm such observations (4) of the failure to suppress T4 levels with T3 therapy. The assay for TSHR-Ab that we had used at that time, the original simple thyroid membrane radioreceptor assay of Smith and Hall (5), was new, but the conclusions were similar to the data of the time. The presence of an abnormal thyroid stimulator indicated a higher chance of recurrence than in patients in whom such a stimulator could not be detected. While a number of the publications that followed our report confirmed this observation (6, 7, 8, 9, 10, 11, 12, 13, 14), others were less enthusiastic (15, 16, 17, 18, 19, 20, 21). The original observation certainly had a number of problems. The most serious was that the follow-up time for detecting recurrence was only 6 months. But I had a fellowship to move to at the NIH, and the study had to end! Furthermore, sensitive TSH assays were not available to help us in the management of the patients. Nevertheless, I believe time has confirmed the original general conclusions. The presence of significant titers of biologically active TSHR-Abs accurately predicts recurrence in up to 90% of such patients.



View larger version (11K):
[in this window]
[in a new window]
 
Figure 1. The original data showing prediction of recurrent Graves’ disease after 6 months treatment with antithyroid drugs (combined carbimazole and thyroxine). Data were expressed as a TSHR-Ab Index, where a low index indicated high titers of TSHR-Abs. On the left are shown those patients who remained in remission, and on the right are those patients who had recurrence defined as increased T3 and T4 levels. (Data redrawn from ref. 2).

 

    Interpreting the medical literature
 Top
 Introduction
 Interpreting the medical...
 Potential factors interfering...
 The cost effectiveness factor
 The pregnancy conundrum
 Summary
 References
 
As discussed above and elsewhere, there have been many publications examining the clinical usefulness of TSHR-Ab measurement. I would like to focus here on a recent meta-analysis published in JCEM that has not, to my knowledge, been discussed at length in the literature (22). The conclusion from this attempted meta-analysis of 18 different reports was that "the available methods for TSHR-Ab do not allow sufficiently high prediction of relapse or remission after antithyroid drugs for the individual patient." I find this conclusion inconsistent with the data presented by the authors and other commentators (23).

The meta-analysis included only certain of the available studies according to a number of clear criteria:

  1. TSHR-Ab measurements were available at the end of antithyroid drug treatment.
  2. The follow up period after stopping therapy was at least 1 yr.
  3. Individual rather than grouped data were available.
  4. No thyroid hormone therapy was given after stopping antithyroid drugs.

Each of these criteria was appropriate and strictly adhered to in the studies selected. However, no criteria at all were applied to the assay of TSHR-Abs except to divide the data into studies with radioreceptor assays and studies with bioassays. The overall assumption of the meta-analysis, therefore, was that all the different assays used in the 18 different studies were of equal sensitivity, equal precision and, most importantly, equal specificity. However, many of the studies used their own "in house" methods for the measurement of TSHR-Abs, sometimes for the first time. We lack, therefore, essential data on the assay performances, which would make a meta-analysis possible to interpret. For example, in one study (24) only 3 out of 184 patients had detectable TSHR-Abs at the end of antithyroid drug treatment. In contrast, in another study 142 out of 391 patients had TSHR-Abs detected (18). In addition, the actual titer of antibody was not taken into consideration in any of the studies analyzed. Borderline positive results were given the same weight as very high titers. Furthermore, what did the different authors mean by remission? Was the TSH normalized in these patients? Relapse and remission were also not defined in this meta-analysis of relapse and remission.

The body of literature on this subject is, therefore, difficult to interpret without careful scrutiny of each study, and inclusion/exclusion criteria would need to be so strict that there would be insufficient data to perform a meta-analysis. Yet we have to conclude that some reports have clearly shown that more than 90% of higher titer TSHR-Ab positive patients had a recurrence of Graves’ disease after withdrawal of antithyroid drugs (13, 25), while this was not seen in other reports (18). How can we explain these disparities? I would like to begin with the only logical conclusion there can be. The presence of TSH receptor antibodies with stimulating activity in the serum of a patient with recent hyperthyroid Graves’ disease must influence thyroid function, unless there is some interference in the transduction of such a signal.


    Potential factors interfering with the action of TSHR-Abs
 Top
 Introduction
 Interpreting the medical...
 Potential factors interfering...
 The cost effectiveness factor
 The pregnancy conundrum
 Summary
 References
 
There are a number of possible reasons for thyrotoxic recurrence after antithyroid drugs not being as predictable as my logic suggests:

1. Antibody quantity and affinity. In theory it is possible that the level of TSHR-Abs could be so insignificant or their affinity so low that their influence on the TSH receptor would be negligible. However, under such circumstances it is unlikely that the TSHR-Abs would be detectable using currently available assays, which remain relatively insensitive to TSH equivalents of less than 10 mU/L. Furthermore, even a slightly increased TSH level is known to be capable of increasing thyroid hormone levels, as seen in patients with thyroid hormone resistance. Hence, the very presence of detectable TSHR-Abs with stimulating activity should lead to hyperthyroidism unless the assay systems employed are providing false positive results or unless there is a target defect (i.e. thyroid dysfunction).

2. Antibody quality. Not all TSHR-Abs are stimulating antibodies. Blocking TSHR-Abs have been well characterized in many laboratories, and some patients with Graves’ disease have been shown to have both types of TSHR-Abs present in their serum (26). Of course, in order for hyperthyroidism to develop there must be a lower concentration or affinity of the blocking variety (27, 28, 29, 30). This has led to discussions on the factors responsible for the balance between stimulating and blocking antibodies and the differing methods for measuring TSHR-Abs, discussions that have occupied clinical thyroidologists for many years. The technique used for the measurement of biologically active TSHR-Abs is indeed critical. Thyroid cell assays have been replaced with more sensitive TSH receptor transfected Chinese Hamster Ovary (CHO) cell stimulation assays (31) and remain the only in vitro way to distinguish stimulating from blocking activity. Many argue that it is never possible to infer the biological activity of TSHR-Abs from competitive binding assays, and in some circumstances this is certainly true. However, if the patient is already thyrotoxic or has recently been thyrotoxic, it does not take a great brain to deduce that the patient’s TSHR-Abs are predominantly of the thyroid-stimulating variety. The real question, therefore, is whether TSHR-Abs change from stimulating to blocking during the course of antithyroid drug treatment. The literature to date is relatively unhelpful in this regard, but this appears to be an unusual event (32, 33). Hence, the measurement of TSHR-Abs by an expensive bioassay is not necessary for everyday clinical practice in patients with recent hyperthyroid Graves’ disease.

3. Problems with thyroid function. There are two major potential disturbances in the thyroid reserve that may prevent an appropriate thyroid hormone response to TSHR-Abs of the stimulating variety—iodine deficiency and autoimmune thyroiditis (34).

Iodine deficiency remains the major cause of global thyroid disease, and moderate iodine deficiency continues to affect thyroid function even in many parts of Western Europe such as Germany, where 15% of the population may have iodine deficiency goiters (35). It is not surprising, therefore, that low titers of TSHR-Abs would have difficulty inducing hyperthyroidism in certain parts of the world because of the lack of intrathyroidal iodine. However, in countries such as the USA, Japan, and Korea, where there are high iodine diets, this restriction should not apply. This certainly complicates the meta-analysis from Feldt-Rasmussen (22), where many of the studies were from countries with mild iodine deficiency.

Similarly, the onset of significant autoimmune thyroiditis would reduce the thyroid reserve considerably and would require enhanced thyroid stimulation to raise thyroid hormone levels above normal. Hence, the presence of TSHR-Abs would not always predict the hyperthyroid state. No data on the onset of hypothyroidism was provided for the 55 patients in remission with detectable TSHR-Abs in large multicenter studies (18) or in any of the other published prediction studies, but the geographic location of the patients would suggest that iodine deficiency was more important that thyroiditis. Furthermore, it appears to take many years for thyroiditis to induce thyroid failure in some Graves’ patients (36). We have all seen examples where this can happen much more suddenly during treatment, but it is unusual.

4. Assay sensitivity. I believe the major problem today, with easily available and relatively precise and specific assays, is not false positive TSHR-Ab results but the large group of patients who develop hyperthyroidism in the absence of detectable TSHR-Abs. Clearly, in these patients a lack of thyroid reserve is not an issue. Accepting the evidence that the presence of TSHR-Abs often predicts recurrence in the presence of iodine sufficiency, does their absence predict remission? Many studies of this problem have shown that a negative result, as so often in medicine, is of little help to the clinician. Of course the loss of previously detectable TSHR-Abs indicates that the immune system may have become more tolerant, but such patients may still have a high recurrence rate after discontinuing antithyroid drugs. Is this due to problems with detecting low levels of TSHR-Ab? Following the earlier discussion, there is much evidence to suggest that only small increases in TSH (and, therefore, TSHR-Ab) lead to thyroid hyperfunction. However, as also mentioned earlier, the bioassays and radioreceptor assays used for the measurement of TSHR-Abs are insensitive and/or imprecise to TSH concentrations of less than 10 mU/L. Low levels of receptor antibody may not be detected. In addition, thyroid autoantibodies in the circulation are merely a reflection of their intrathyroidal production. For example, patients with Hashimoto’s disease and negative anti-Tg and anti-TPO still have intrathyroidal B cells capable of secreting such antibodies (37). No doubt the same applies to patients with Graves’ disease and negative TSHR-Abs. Hence, assay sensitivity is likely to be the major factor in the analysis of recurrent hyperthyroidism. Studies based on in-house assays of dubious validity from inexperienced investigators are likely to accentuate this intrinsic problem. Of course TSHR-Ab negative patients may be patients who do not have Graves’ disease. The diagnostic criteria used in published studies should also be closely scrutinized.


    The cost effectiveness factor
 Top
 Introduction
 Interpreting the medical...
 Potential factors interfering...
 The cost effectiveness factor
 The pregnancy conundrum
 Summary
 References
 
One of the most common arguments against the clinical use of TSHR-Abs has been that the test is not cost-effective. In the United States, laboratory costs are not reflected in the cost of the test performed, but are based more on what the market will pay. In essence, the radioreceptor assay for TSHR-Abs need only cost $10–20, but in practice the patient is more likely to pay $70–80. Under laboratory capitation contracts, this problem is eliminated. Furthermore, both costs are low compared with testing methods in other diseases and specialties. Assuming that all Graves’ patients are tested for TSHR-Abs at diagnosis, then only those patients with higher titers (>20% binding inhibition) need be tested again at the end of their antithyroid drug regimen. Those patients who remain positive should continue on antithyroid drugs for another 6–12 months before retesting, as shown in the algorithm (Fig. 2Go). With a 90% prediction success in positive patients the cost-effectiveness of this approach is self evident. However, if only 5% of patients are TSHR-Abs positive at the end of a course of antithyroid drug treatment, then this approach may be impractical. This is not the clinical experience in most centers. Physicians using this approach with antithyroid drug regimens of 18 months or more may find many of their patients to be TSHR-Ab negative and the test unhelpful as many such patients will remit after such long-term treatment. However, in my experience, a majority of patients can be withdrawn from treatment much earlier, and those in whom such discontinuation should be avoided can be identified by using the TSHR-Ab testing algorithm (Fig. 2Go). Nothing can be more cost effective than treating patients for as short a time as possible while avoiding unnecessary thyrotoxic recurrences.



View larger version (24K):
[in this window]
[in a new window]
 
Figure 2. A suggested clinical pathway for the treatment of Graves’ disease with antithyroid drugs using the TSHR-Ab test as an aid to decision making. In areas of iodine deficiency or where drugs are used for very prolonged periods, the pathway may need to be adjusted.

 

    The pregnancy conundrum
 Top
 Introduction
 Interpreting the medical...
 Potential factors interfering...
 The cost effectiveness factor
 The pregnancy conundrum
 Summary
 References
 
The validity of predicting neonatal Graves’ disease using the measurement of biologically active TSHR-Abs in the mother has been well documented over many years (38, 39). The rarity of this problem demonstrates the success of modern medical care with the early treatment of Graves’ disease. This is, however, the one clinical situation where the bioactivity of the TSHR-Abs may need to be known to predict the neonatal thyroid function (40). Luckily, pregnancy is a time when thyroid autoantibodies generally decrease in titer, due to the secretion of trophoblast factors, which are immunosuppressive, and bioactivity assessment is not necessary in many patients. However, in those Graves’ patients with higher titers of TSHR-Ab by radioassay in the third trimester, particularly those patients who have not been recently hyperthyroid, a subsequent bioassay will allow the prediction of neonatal hyperthyroidism or blocking antibody-induced hypothyroidism (27, 41). This approach applies most importantly to women previously treated for Graves’ disease and who are euthyroid or on thyroxine replacement but have persisting titers of TSHR-Abs.


    Summary
 Top
 Introduction
 Interpreting the medical...
 Potential factors interfering...
 The cost effectiveness factor
 The pregnancy conundrum
 Summary
 References
 
Many physicians would pay dearly to obtain a marker for the disease that interests them most. Those of us interested in caring for patients with Graves’ disease have such a marker. Knowing the titer of TSHR-Abs in patients is useful in the prediction of thyrotoxic recurrence after antithyroid drugs. A careful review of the literature shows strong evidence for the observation that their accurately measured presence predicts hyperthyroidism in more than 90% of cases in iodine-sufficient areas. The assay of TSHR-Abs, therefore, remains a most useful addition to the clinical armamentarium and a low-cost help in treatment planning. Assays to distinguish blocking from stimulating antibodies remain expensive but are only clinically important in patients with pregnancy (40). The major hurdle remains one of increasing the sensitivity of the available assays for TSHR-Ab so that their usefulness can be applied successfully to an even greater proportion of patients with Graves’ disease.


    Footnotes
 
Address all correspondence regarding these controversies and requests for reprints to: Terry F. Davies, M.D., Mount Sinai Medical Center, Box 1055, 1 Gustave L. Levy Place, New York, New York 10128.

a Supported in part by: the National Institutes of Health Grants DK52464, DK35764, and DK45011 from NIDDK. Please note that Dr. Davies is a Board member of the Kronus company, San Clemente, California, which sells assays for the measurement of thyroid autoantibodies including TSH receptor autoantibodies. The views expressed are those of the author only.

Accepted May 20, 1998.


    References
 Top
 Introduction
 Interpreting the medical...
 Potential factors interfering...
 The cost effectiveness factor
 The pregnancy conundrum
 Summary
 References
 

  1. Davies TF. 1996 The pathogenesis of Graves’ disease. In: Braverman LE, Utiger RD, eds. The thyroid: a fundamental text. Philadelphia: Lipincott; 525–536.
  2. Davies TF, Yeo PP, Evered DC, Clark F, Smith BR, Hall R. 1977 Value of thyroid-stimulating-antibody determinations in predicting short-term thyrotoxic relapse in Graves’ disease. Lancet. 1:1181–1182.[Medline]
  3. Alexander W, McLarty DG, Robertson J, et al. 1970 Prediction of the long-term results of antithyroid drug therapy for thyrotoxicosis. J Clin Endocrinol. 30:540–543.[Medline]
  4. Prakash R. 1996 Prediction of remission in Graves’ disease treated with long-term carbimazole therapy: Evaluation of technetium-99m thyroid uptake and TSH concentrations as prognostic indicators. Eur J Nucl Med. 23:118–122.[Medline]
  5. Smith BR, Hall R. 1974 Thyroid-stimulating immunoglobulins in Graves’ disease. Lancet. ii:427–429.
  6. Wilson R, McKillop JH, Pearson DW, Cuthbert GF, Thomson JA. 1985 Relapse of Graves’ disease after medical therapy: predictive value of thyroidal technetium-99m uptake and serum thyroid-stimulating hormone receptor antibody levels. J Nucl Medicine. 26:1024–1028.[Abstract]
  7. Zakarija M, McKenzie JM, Banovac K. 1980 Clinical significance of assay of thyroid-stimulating antibody in Graves’ disease. Ann Intern Med. 93:28–32.[Medline]
  8. Teng CS, Yeung RTT. 1980 Changes in Thyroid-Stimulating Antibody Activity in Graves’ Disease Treated with Antithyroid Drug and Its Relationship to Relapse: A Prospective Study. J Clin Endocrinol Metab. 50:144–147.[Abstract]
  9. Edan G, Massart C, Hody B, Poirier JY, Le Reun M, Hespel JP, Leclech G, Simon M. 1989 Optimum duration of antithyroid drug treatment determined by assay of thyroid stimulating antibody in patients with Graves’ disease. Brit Med J. 298:359–361.[Medline]
  10. Schleusener H, Finke R, Kotulla P, Wenzel KW, Meinhold H, Roedler HD. 1978 Determination of thyroid stimulating immunoglobulins (TSI) during the course of Graves’ disease. A reliable indicator for remission and persistence of this disease? J Endocrinol Invest. 2:155–160.
  11. Karlsson FA, Dahlberg PA. 1980 Thyroid stimulating antibodies (TSAb) in patients with Graves’ disease undergoing antithyroid drug treatment: indicators of activity of disease. Clin Endocrinol. 14:579–585.
  12. McGregor AM, Rees-Smith B, Hall R, Petersen MM, Miller M, Dewar P. 180 Prediction of relapse in hyperthyroid Graves’ disease. The Lancet. 1101–1103.
  13. Wilson R, McKillop JH, Henderson N, Pearson DW, Thomson JA. 1986 The ability of the serum thyrotropin receptor antibody (TRAb) index and HLA status to predict long-term remission of thyrotoxicosis following medical therapy for Graves’ disease. Clin Endocrinol. 25:151–156.[Medline]
  14. Cho B, Shong MH, Yi KH, Lee HK, Koh C-S, Minard K. 1992 Evaluation of serum basal thyrotropin levels and thyrotrophin receptor antibody activities as prognostic markers for discontinuation of antithyroid drug treatment in patients with Graves’ disease. Clin Endocrinol. 36:585–590.[Medline]
  15. Gossage AAR, Crawley JCW, Copping S, Hinge D, Himsworth RL. 1982 Graves’ disease: thyroid function and immunologic activity. J Nucl Med. 23:973–977.[Abstract]
  16. Croxson MS, Lim TMT, Graham FM, Ibbertson HK. 1980 Thyrotrophin displacement activity of serum immunoglobulins in health disease. Aust N Z J Med. 10:151–156.[Medline]
  17. Schleusener H, Schwander J, Holl G, et al. 1987 Do HLA-DR-typing and measurement of TSH-receptor antibodies help in the prediction of the clinical course of Graves’ thyrotoxicosis after antithyroid drug treatment? Acta Endocrinol (Copenh). 281:318–324 [Suppl].
  18. Schleusener H, Schwander J, Fischer C, et al. 1989 Prospective multicentre study on the prediction of relapse after antithyroid drug treatment in patients with Graves’ disease. Acta Endocrinol (Copenh). 120:689–701.[Medline]
  19. Young ET, Steel NR, Taylor JJ, et al. 1988 Prediction of remission after antithyroid drug treatment in Graves’ disease. Q J Med. 250:175–189.
  20. deBruin TW, Bolk J, Bussemaker J, et al. 1988 Graves’ disease: immunological and immunogenetic indicators of relapse. Br Med J. 296:1292–1295.[Medline]
  21. Weetman A, Ratanachaiyavong S, Middleton GW, et al. 1985 Prediction of outcome in Graves’ disease after carbimazole treatment. Quarterly Journal of Medicine 228:409–419.
  22. Feldt-Rasmussen U, Schleusner H, Carayon P. 1994 Meta-analysis evaluation of the impact of thyrotropin receptor antibodies on long term remission after medical therapy of Graves’ disease. J Clin Endocrinol Metab. 78:98–102.[Abstract]
  23. Braverman LE. 1995 Thyroid. In: Bagdade JD, ed. Yearbook of endocrinology. St. Louis, MO: Mosby Yearbook; 266–267.
  24. Murakami M, Koizumi Y, Aizawa T. 1988 Studies of thyroid function and immune parameters in patients with hyperthyroid Graves’ disease in remission. Journal of Clinical Endocrinology and Metabolism. 66:103–108.[Abstract]
  25. Rapoport B, Greenspan F, Filetti S, Pepitone M. 1984 Clinical Experience with a Human Thyroid Cell Bioassay for Thyroid-Stimulating Immunoglobulin. J Clin Endocrinol Metab58 :332–338.
  26. Kraiem Z, Lahat N, Glaser B, Baron E, Sadeh O, Sheinfeld M. 1987 Thyrotropin receptor blocking antibodies: incidence, characterization, and in-vitro synthesis. Clin Endocrinol. 27:409–421.[Medline]
  27. Zakarija M, McKenzie JM, Eidson MS. 1990 Transient neonatal hypothyroidism: characterization of maternal antibodies to the thyrotropin receptor. J Clin Endocrinol Metab. 70:1239–1246.[Abstract]
  28. Davies TF, Platzer M, Schwartz A, Friedman E. 1983 Functionality of thyroid-stimulating antibodies assessed by cryopreserved human thyroid cell bioassay. J Clin Endocrinol Metab. 57:1021–1027.[Abstract]
  29. Creagh F, Teece M, Williams S, Didcote S, Perkins W, Hashim F, Rees-Smith B. 1985 An analysis of thyrotrophin receptor binding and thyroid stimulating activities in a series of Graves sera. Clinical Endocrinology 23:395–404.
  30. Creagh F, Howells RD, Williams S, et al. 1986 IgG TSH receptor antibody activity in Graves’ disease: a study of TSH agonist and antagonist activities by isolectric focusing. Clinical Endocrinology 24:79–88.
  31. Kim MR, Faiman C, Hoogwerf BJ, Gupta MK. 1997 Thyroid-stimulating antibody assay using a human thyrotropin receptor transfected cell line: Relationship to clinical features of Graves’ disease. Endocr Pract. 3:337–343.
  32. Kasagi K, Lida Y, Koshishi J. 1986. Paired determination of thyroid-stimulating and TSH binding inhibitory activities in patients with Graves’ disease during antithyroid drug treatment. Acta Endocrinol (Copenh). 111:474–480.
  33. Tamai H, Hirota Y, Kasagi K. 1987 The mechanism of spontaneous hypothyroidism in patients with Graves’ disease after antithyroid drug treatment. J Clin Endocrinol Metab. 64:718–722.[Abstract]
  34. Rees Smith B, McLachlan SM, Furmaniak J. 1988 Autoantibodies to the thyrotropin receptor. Endocr Rev. 9:106–121.[Abstract]
  35. European Thyroid Association. 1985 Goiter and iodine deficiency. Lancet. i:1289–1293.
  36. Wood LC, Ingbar SH. 1979 Hypothyroidism as a late sequela in patients with Graves’ disease treated with antithyroid agents. J Clin Invest. 64:1429–1436.[Medline]
  37. Baker JR, Saunders NB, Wartofsky L, Tseng Y-C, Burman KD. 1988 Seronegative Hashimoto thyroiditis with thyroid autoantibody production localized to the thyroid. Ann of Intern Med. 108:26–30.[Medline]
  38. Munro DS, Dirmikis SM, Humphries H, Smith T, Broadhead GD. 1978 The role of thyroid-stimulating immunoglobulins of Graves’ disease in neonatal thyrotoxicosis. Br J Obstet Gynaecol85 :837–843.
  39. Zakarija M, McKenzie JM. 1983 Pregnancy-associated changes in thyroid-stimulating antibody of Graves’ disease and the relationship to neonatal hyperthyroidism. J Clin Endocrinol Metab. 57:1036–1040.[Abstract]
  40. Davies TF. 1988 Thyroid-stimulating hormone receptor antibodies—the calm after the storm after the calm? Mayo Clin Proc. 63:736–739.[Medline]
  41. Matsoura N, Yamada Y, Nohara Y, et al. 1980 Familial neonatal transient hypothyroidism due to maternal TSH-binding inhibitor immunoglobulins. N Engl J Med. 303:738–742.[Medline]