| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Special Articles |
Centro per lo Studio, Prevenzione, Diagnosi e Cura delle Tireopatie Universita degli Studi di Parma 43100 Parma, Italy
Lewis E. Braverman
Genetics Division, Brigham & Women’s Hospital Harvard Medical School Boston, Massachusetts 02115
Leslie J. DeGroot, M.D.
Thyroid Study Unit The University of Chicago Chicago, Illinois 60637
 |
Introduction |
|---|
 Top Introduction ConclusionReferences |
|---|
All of the above listed criteria for the diagnosis of Graves’ hyperthyroidism are rarely present in all patients with classic Graves’ disease. In a European multicenter study, the "full picture" of Graves’ disease was present in only 10.3% of 271 patients (1). The prevalence of positive TBII, a marker of immunogenic hyperthyroidism, was found in 89% of the patients with Graves’ disease, a value not much different from that of a homogenous thyroid scan (78.1%) and a diffusely enlarged gland by palpation (74.8%). In the same study, 31.2% of hyperthyroid patients could not be classified as having Graves’ hyperthyroidism by the above criteria. However, half of these patients, despite the absence of positive TBII, were finally classified as having Graves’ disease. Other studies in Graves’ hyperthyroid patients have reported positive and negative TBII tests in 76–95% and 10%, respectively (2, 3, 4, 5, 6, 7, 8, 9, 10). Usually patients with negative TBII had mild hyperthyroidism and small goiters (10). Recently, 95.7% of 277 untreated Graves’ patients had positive TBII tests (11), and TBII and/or TSab were positive in 91.7% of the patients. The receptor assay was positive in 100% of untreated hyperthyroid Graves’ patients (12). In another review, the TSab test was positive in 71–100% of the cases (13). In juvenile untreated Graves’ patients, TRab were positive in 93% (14), whereas in another study TSAb were detected in only 73% of the patients (15). TSab were present in all patients with subclinical Graves’ hyperthyroidism, i.e. patients with normal serum thyroid hormone concentrations, suppressed serum TSH concentrations, and absent ophthalmopathy (16). Other recent studies have reported a far lower positivity of the TBII test, ranging from 25–75% of untreated Graves’ patients (17, 18). A striking variation in positive TRab results has been observed in Graves’ patients residing in the same country, England, ranging from 35% in Preston to 92% in Southampton (13).
To determine the real value of TRab testing in the diagnosis of Graves’ disease, it is worthwhile to compare the predictivity of this test to other tests in revealing the presence of thyroid autoimmunity. Positive TRab were detected in 80% of newly diagnosed Graves’ patients, and antimicrosomal antibodies (Mab) were present in 76.5% (19). In another study TSAb were present in 84–100% of Graves’ patients, whereas Mab were positive less frequently (31–77%) (20). As previously mentioned, all patients with subclinical Graves’ hyperthyroidism had positive TSab. However, the Mab test was also positive in all patients (16). TRab were present in 88.2% of untreated Graves’ hyperthyroid patients, whereas Mab were positive in 64.7% and thyroglobulin antibodies (Tgab) in 53% of the cases (21). Thus, in our opinion palpation, clinical findings, measurement of TPOab, and less commonly, a 123I uptake and scan can give sufficient information to establish a diagnosis of Graves’ disease. Whether the TRab test should be included in a panel for the diagnosis of Graves’ disease is best revealed by surveys of the European Thyroid Association (ETA), the Japan Thyroid Association (JTA), and the American Thyroid Association (ATA) (22). TSRab were requested by the responders of the above societies with the following frequency: 38% for ETA members, 32% for JTA members, and 9% for ATA members. Finally, in treatment guidelines published by the American Thyroid Association and the Association of Clinical Endocrinologists, the use of TRab as a first line test in the diagnosis of Graves’ disease is not recommended (23, 24).
TRab have also been suggested as a useful test to predict the recurrence of Graves’ hyperthyroidism after drug withdrawal. In one case study, 83% of patients with positive TRab relapsed, whereas only 5% remained euthyroid (25). Furthermore, 89% and 81% of patients with negative TRab remained euthyroid for 1 yr and 3 yr, respectively, after antithyroid treatment was discontinued. In another study, positive TRab were present in 74% of relapsed patients and in 17% of those in remission (19%). However, other studies found that TRab measurement had a low predictive value in assessing the outcome of Graves’ disease after a course of antithyroid drugs (26, 27). More recently, it has been reported that, when TSab and TBII decline and disappear during methimazole treatment, remission occurs (11). In contrast, patients who have positive TSab and TBII values at the end of treatment remained hyperthyroid (11). However, positive TSab and TBII values at the onset of Graves’ hyperthyroidism had a low predictive value on the course of the disease. Similarly, Zingrillo et al. (21) reported that TRab values before antithyroid treatment had no value in predicting the remission of Graves’ disease after 1 and 2 yr of antithyroid drug therapy. They also observed that 1 yr after methimazole (MMI) withdrawal, 70% of relapsed patients had positive TRab, and 21% of patients in remission had positive values. Furthermore, TRab values were of even lower value in predicting the outcome of the disease at 2 yr after antithyroid drug withdrawal. Others have reported that, in patients with negative TBII at the onset of Graves’ disease, the appearance of positive TBII, and its concentrations during MMI administration did not predict the outcome of the disease (20). More recently, positive TBII values were observed in 16% of Graves’ patients who relapsed and in 5% of those who remained in remission (17). This study also demonstrated that TBII determinations at the beginning of antithyroid treatment did not identify patients who relapsed or remained euthyroid after antithyroid drugs were discontinued. It is of interest that patients with euthyroid Graves’ disease who developed hyperthyroidism following the administration of pharmacological quantities of iodine did not have positive TRab (28). However, others have reported that TRab levels increased slightly in Graves’ hyperthyroid patients treated with excess iodine from iopanoic acid (29). In 67 patients with Graves’ disease treated with antithyroid drugs for at least 1 yr, TBII results at the time of drug withdrawal had specificity and sensitivity values in predicting relapse of 94.5% and 45%, respectively (30). Furthermore, TBII levels at the beginning of antithyroid treatment had a low value in predicting the outcome of the disease. In a multicenter survey of 451 patients, the relapse rate of Graves’ hyperthyroidism within 1 yr, after 1 yr therapy with antithyroid drugs, varied from 30–67% of the patients (27). In this large cohort of patients, TBII tests had sensitivity and specificity values in predicting relapse of 49% and 73%, respectively. These values were too low to suggest that TBII assays are reliable in predicting outcome of the disease in individual patients. In patients who underwent subtotal thyroidectomy for Graves’ disease, TSab and TBII values either before or after thyroidectomy did not predict the recurrence of hyperthyroidism (31). Weetman et al. (32) found no value in the use of TSab in predicting remission or relapse of Graves’ hyperthyroidism after 6 months of treatment with carbimazole and triiodothyronine. Similar conclusions were reached in Graves’ hyperthyroid patients treated with carbimazole and thyroxine (33).
In a prospective study, the remission rate of patients with Graves’ disease was 51.9% in patients whose antithyroid treatment was discontinued when TBII and TSH levels were normal and 63.1% in patients treated for 24 months, irrespective of their TBII and TSH levels (34). A meta-analysis evaluation of the value of TRAb results in predicting long-term remission in patients with Graves’ hyperthyroidism following antithyroid drug therapy concluded that the absence of TRab are suggestive that the disease has remitted but that the TRab test was not sufficiently reliable to predict relapse or remission of Graves’ hyperthyroidism in the individual patient (35). In a more recent study of 306 patients, relapse of Graves’ hyperthyroidism was observed in 97.5% of patients with elevated TRab levels but also in 41.4% of those with low TRab levels (36). Positive TRab before medical treatment had sensitivity and specificity values in predicting the recurrence of Graves’ disease of 42.5% and 87%, whereas clinical parameters such as age older than 40 yr, goiter size less than 40 mL, and absent ophthalmopathy had sensitivity values of 20, 59.7, and 42.4% and a specificity of 89.2, 62.5, and 71.6%, respectively, in predicting the relapse of hyperthyroidism.
The TRab test has also been evaluated in special clinical conditions such as pregnant and postpartum women and children. During pregnancy, it has been reported that TSab values decline toward term (37). This finding was associated with reduced doses of antithyroid drug to control the hyperthyroidism. We believe that adjustment of the antithyroid drug dose can be obtained by serum TSH and free thyroxine (FT4) concentrations rather than measuring TRab. Of far greater importance, elevated TSab levels at 28–30 weeks of pregnancy predict an increased frequency of neonatal Graves’ hyperthyroidism (37). Thus, measurement of serum TSab in women with active or previous Graves’ disease during the last trimester of pregnancy is useful in predicting the occurrence of neonatal Graves’ disease. Fetal ultrasound to detect a goiter and fetal heart rate are helpful in detecting fetal hyperthyroidism. TSH blocking antibodies in pregnant women with autoimmune thyroid disease, especially patients with Hashimoto’s thyroiditis and hypothyroidism, may also induce transient neonatal hypothyroidism (38). In contrast, the occurrence of Graves’ disease in postpartum women was not accompanied by an increment in TSRab (39). Furthermore, relapse of Graves’ disease after parturition was accompanied by an increase in FT4I and FT3I values followed by an increase in TSab and TBII in 60% of the patients. In the other patients, TSab and TBII levels increased concomitantly or before the rise in serum thyroid hormone concentrations. In juvenile Graves’ disease, the TRab test correctly predicted the outcome of the disease after antithyroid drugs were discontinued in 72% of the patients (14). Early remission of Graves’ disease in children within 2 yr of beginning medical treatment was accompanied by positive TSI in 50% of patients, whereas 93% of children who required antithyroid treatment for more than 2 yr had positive TSI (40). Children with lower serum T4 and T3 concentrations had an early remission. Furthermore, children with an early remission were older, had higher body mass index (BMI), lower heart rate, and a smaller goiter. Clinical parameters such as BMI and goiter size were independent predictors of early remission.
We believe that the cumulative evidence suggests that the routine use of serum TRab in the diagnosis and in predicting the course of Graves’ hyperthyroidism following antithyroid drug treatment is not indicated for the individual patient. The presence of a diffusely enlarged thyroid with suppressed serum TSH and elevated free T4 and T3 (or the T3 only in T3 toxicosis) values, positive TPO antibodies, and when indicated, an elevated thyroid 123I uptake, and ophthalmopathy when present, is sufficient to diagnose hyperthyroid Graves’ disease. Measurement of TRab adds little to the initial diagnosis and subsequent management of the patient unless the etiology of the thyrotoxicosis is truly in question. Measurement of serum TRab at the conclusion of antithyroid drug therapy for Graves’ disease is of marginal predictive value of remission in a given patient (41, 42). Clinical parameters such as a small goiter, low doses of antithyroid drug to control the hyperthyroidism, normal serum TSH values, and a longer course of therapy are more often associated with remission (43, 44). We certainly agree that measurement of TRab during the last trimester of pregnancy is helpful in predicting the possibility of neonatal Graves’ disease (37, 45). Finally, under many managed care plans, the cost of obtaining a serum TRab is sufficiently high so that payment is refused.
Note Added in Proof
Since submission and acceptance of this manuscript, two papers have been published that suggest a possible role for measuring TRab. Yoshida et al. (Clin Endocrinol. 48: 17–22, 1998) report that the development of TSH blocking antibodies shortly after 131I therapy for Graves’ disease may cause hyperthyroidism and that their disappearance results in recovery of thyroid function. However, Chiovato et al. (J Clin Endocrinol Metab. 83: 40–46, 1998) reported that, although quite common, the appearance of TSH blocking antibodies and the disappearance of thyroid-stimulating antibody result in early post 131I hypothyroidism in a minority of patients. Although they suggest that high TSHab levels before 131I therapy are associated with relative resistance to 131I, and a rise in TSab after treatment is associated with the development of hypothyroidism, significantly larger thyroids were observed in those patients who remained hyperthyroid or were euthyroid 1 yr after 131I therapy than in patients who became hypothyroid. Again, we doubt that these two observations warrant the measurement of thyroid-stimulating or TSH-blocking antibodies in the routine management of patients with Graves’ disease before or after 131I therapy.
|
Conclusion |
|---|
|
TopIntroductionConclusionReferences |
|---|
Regarding the first question, in most cases the combination of elevated FTI, suppressed TSH, goiter, and often some eye signs, allows definitive diagnosis, and the relatively cheap and readily available TPO antibody assay adds some certainty and reassurance that autoimmunity is the cause. Many clinicians do not routinely order TPO antibodies. TRab assay is commonly done for special problems including those patients who present with initial unilateral exophthalmos, or sometimes subclinical hyperthyroidism, or for example, hyperemesis with thyrotoxicosis, and sometimes in differential diagnosis of postpartum painless thyroiditis. Often in such cases, the values are indeterminate when the clinical signs are equivocable.
Actually, the TRAb assay is currently not much used in the United States. The reason for this is not certain. It was introduced much later than the TPO and Tg antibody assays. There is also a question that, in some areas of the world where it is widely employed, there may be financial incentive to its use, which is not the case in the United States. There is not much evidence that lack of use alters our approach to diagnosis or treatment. Perhaps, if the assay was as generally available and comparably costly to the TPO assay, it would serve the same purpose (recognizing anti-thyroid immunity) and might be substituted. At the present time, the TPO assay appears to be equally sensitive, more available, and less costly than any of the TRAb assays.
Regarding the second question, it is useful to remember that only a minority of patients with Graves’ disease are treated with antithyroid drugs to induce cure. This group represents less than 20% or so of the total population and is mainly comprised of younger patients. Measurement of an "initial value" would only apply to this subgroup. The data provided by Roti and Braverman point out the generally low predictive value of TRAb measurement at initiation of antithyroid drug therapy. A value at the beginning and end of antithyroid therapy does not provide any more information than a single value at the end of the contemplated period of treatment.
Certainly TRAb assay, at the end of a planned period of antithyroid drug therapy, is a useful predictor of outcome for a group of patients. For the individual patient, it does not appear to be much superior to an observed reduction in the size of the goiter or to ease of control on low doses of antithyroid drugs. The basic problem that has limited its widespread use is lack of precision. When we contemplate either giving up antithyroid drug therapy for another treatment such as radioactive iodide, or continuing drugs for another year, we want to have, effectively, 100%—not 90%—precision in our interpretation of the clinical situation. The degree of imprecision is seen in Fig. 1 in Davies’ article. Many more reports show a much greater lack of accuracy in predicting outcome of a course of therapy, and these are fully reported by Roti and Braverman. Roughly 20% of patients enter apparent clinical remission with detectable TRab in the serum, and about 10–20% relapse even though TRab are not measurable by available assays. While it is probably true, as Davies points out, that high or significant levels would be more predictive of relapse, high levels occur in only a small proportion of patients.
General practice is to gradually reduce antithyroid drugs over a period of 2 or 3 months, intermittently checking TSH or FTI to see if they have altered. This allows certain knowledge whether another course of therapy or different therapy is required, or if the patient has at least temporarily entered remission.
One problem pointed out by both discussants is that there may be imprecise correlation between TSab antibody levels and function of the thyroid. As assays improve, TSab levels that do not increase function above normal will become measurable, and some patients will have thyroiditis that limits the response to even elevated levels. Thus it is quite possible that improved sensitivity and precision in the assay may accentuate rather than diminish the overlap and thereby reduce diagnostic certainty.
Cost effectiveness is, much to our distaste, an important consideration, and it is not easy to prove that TRab assay is going to provide information that saves money, when the same information can be provided very easily by reducing antithyroid drug therapy and observing the patient.
For the last of the four questions posed above, it is considered prudent by most clinicians to measure a TRAb assay in pregnant women who, because of past or present Graves’ disease, could pass thyroid stimulatory antibodies to their infant. However, in this situation and despite the fair correlation between binding and stimulating antibody levels, it is the TSAb assay that is really required and not a binding assay. We do not want to guess, as suggested by one discussant, but rather we want to know for certain whether stimulating antibodies are present in high concentration.
We can all agree that, in a situation in which cost was no object, the addition of TRab assays would provide marginal benefit. But experience over the past two decades has shown that the assays are rarely required and have a specific effect on therapy in only a very small fraction of cases of Graves’ disease.
|
Footnotes |
|---|
|
References |
|---|
|
TopIntroductionConclusionReferences |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 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 |
