College of American Pathologists
Printable Version

  Untangling thyroid problems, test by test


cap today



February 2007
Cover Story

William Check, PhD

In a patient sample in which thyroid-stimulating hormone is significantly elevated and free tetraiodothyronine (FT4, or free thyroxine) is decreased, the pathologist would diagnose primary hypothyroidism. Conversely, a sample with a depressed TSH and an abnormally high FT4 concentration would make the pathologist consider the diagnosis of primary hyperthyroidism. But what are the most important and common causes of each condition? And which tests should be performed next? Whereas most thyroid abnormalities can be diagnosed by measuring TSH and FT4, when should triiodothyronine, T3, be determined?

Solutions to these and other conundrums in thyroid testing were revealed at CAP ’06 by William E. Winter, MD, professor of pathology, immunology, and laboratory medicine and pediatrics, University of Florida, Gainesville. Insights from Dr. Winter’s workshop, along with interviews with other experts in thyroid testing, provide the basis for explaining the workings of the thyroid and its many abnormal states—basic conditions like hypo- and hyperthyroidism; how to test for autoimmune thyroid disease, which encompasses Hashimoto’s thyroiditis and Graves’ disease; subclinical thyroid disease; sick euthyroid syndrome; postpartum thyroiditis; factitious thyroid disorders; and monitoring thyroid cancer after treatment. Newer developments in thyroid testing include continuing reduction of the upper limit of the reference range for TSH, greater attention to thyroid dysfunction in pregnancy, and a trend to screen older persons for thyroid disease.

Asked to highlight one pivotal issue from this menu, Dr. Winter goes straight to basics. “From the viewpoint of the clinical pathologist,” he says, “one of the most pertinent questions is, what is appropriate laboratory testing when thyroid dysfunction is a consideration? If the clinician believes that the hypothalamus and pituitary are functioning normally, you can screen for both hypo- and hyperthyroidism by measuring TSH, which is the single most important test of thyroid function. If TSH is abnormal, you can go on to measure free T4.” A T3 level is not routinely needed. However, measuring T3, preferably total, would be the next option in a potentially hyperthyroid patient when the TSH is low but the FT4 is normal. A normal T3 in this setting suggests subclinical hyperthyroidism. Alternatively, if the T3 is elevated with a depressed TSH and normal FT4, T3 toxicosis is recognized. In a severely ill hospitalized patient, a reduced T3 in the presence of a normal TSH and normal FT4 suggests sick euthyroid syndrome, also known as nonthyroidal illness.

Dr. Winter’s second message is that autoimmunity directed against the thyroid gland is the most common cause of thyroid disease—both hypo- and hyperthyroidism—in adults and children. “To look for the etiology of thyroid disease, you should order thyroid autoantibody studies,” he says. Thyroid microsomal autoantibodies, or TMAb, and thyroperoxidase autoantibodies, or TPOAb, provide similar information, he advises, so you don’t need to order both tests. If either of these is negative, you could then measure thyroglobulin autoantibodies, TgAb. If this “waterfall” or sequenced approach is not available, you can do the initial evaluation with a two-test panel—either TMAb or TPOAb combined with TgAb. “Thyroid autoantibody panels adequate for diagnostic purposes are available,” he says.

Another basic consideration—reference ranges—tops the list of important issues for Carole A. Spencer, PhD, professor of medicine and technical director of the endocrine laboratory, University of Southern California. “In suspected thyroid disease, the test of choice is clearly thyrotropin [TSH] —there is no argument about that,” Dr. Spencer says. “But there is considerable and growing debate about what is an appropriate TSH reference range. The TSH reference range has contracted after adoption of the newer immunometric assays in the 1980s.” In particular, the upper limit of the reference range has undergone a progressive lowering. Current guidelines of the National Academy of Clinical Biochemistry and American Association of Clinical Endocrinologists recommend a value around 2.5 mIU/L, says Dr. Spencer. “Most laboratories still quote an upper limit for TSH of 4, 4.5, and sometimes even 5 or 5.5,” Dr. Spencer says. “This can cause a problem for physicians to diagnose the early stages of hypothyroidism, so-called subclinical hypothyroidism.” Having a more appropriate upper reference limit of 2.5 to 3 (depending on the assay) is especially important for detecting mild thyroid deficiency in early pregnancy, when TSH levels are lowered by hCG stimulation of the thyroid gland. “There is growing concern that it is necessary to identify women with mild thyroid deficiency in the first trimester,” Dr. Spencer says, “since it can be detrimental to the mother and the fetus. This is adding to the debate about where to set the TSH upper limit.”

Dr. Spencer’s second major thought concerns her particular area of expertise, monitoring differentiated thyroid carcinoma with thyroglobulin, Tg. “Thyroglobulin doesn’t really play any role in the diagnosis of thyroid carcinoma,” she says, “because virtually everyone with some thyroid tissue will produce it. On the other hand, it is clear that because thyroglobulin is produced only by thyroid tissue, it is a good post-op tumor marker.” The value of using a sensitive Tg assay is becoming increasingly apparent. In fact, Dr. Spencer draws an analogy between the early TSH assays in which TRH stimulation was needed to overcome assay insensitivity and current Tg assays, in which expensive recombinant human TSH (rhTSH) stimulation is used to overcome the insensitivity of the typical Tg assay. As with TRH stimulation of TSH, the rhTSH-stimulated Tg response is merely proportional to the basal Tg level. This means that with the use of one of the newer, more sensitive Tg assays (functional sensitivity <0.1 ng/mL), the rhTSH-stimulated Tg can be predicted and becomes unnecessary.

Like Dr. Winter, Laurence M. Demers, PhD, highlights the issue of appropriate testing. “If you look at the community hospital level, you see that many physicians are still ordering T3 uptake and calculating free hormone binding and measuring total T4,” says Dr. Demers, distinguished professor of pathology and medicine and director of the core endocrine laboratory, Penn State Milton S. Hershey Medical Center. “Most of these assays have gone by the boards with the greater ability of today’s assays. With a good test of TSH and a good test of FT4, you can make the diagnosis of virtually all forms of thyroid disease,” he emphasizes. “Unfortunately, many older physicians are still relying on T3 uptake tests that we abandoned some time ago. And reference laboratories continue to do them to appease these clinicians and to satisfy their needs.” In fact, Dr. Demers notes, calculating the free thyroxine index can be misleading and can cause false interpretation. “Pathologists at the community hospital level need to understand better the tests being offered,” Dr. Demers says. He and Dr. Spencer recently led a panel that put together a guideline for appropriate use of thyroid function testing (Baloch Z, et al. Thyroid. 2003;13:3–126).

Unraveling thyroid disease proceeds directly from an understanding of the physiology of the hypothalamic-pituitary-thyroid axis. The hypothalamus makes TRH (thyrotropin-releasing hormone, a tripeptide), which stimulates the pituitary to make TSH, which acts on the thyroid gland to make T4 and T3. T4 is enzymatically converted to T3 in the peripheral tissues. T3 affects the ability of virtually all tissues to metabolize and use oxygen. Both of these thyroid hormones feed back on the pituitary and hypothalamus to shut off production of TRH and TSH, although T4 may be more important in this regard. If T4 is low, the pituitary becomes more sensitive to TRH and makes more TSH. In the circulation, >99 percent of all thyroid hormone is bound to thyroxine-binding globulin, TBG, so measuring FT4 requires high analytical sensitivity.

Very good assays for TSH and FT4 are available. Third-generation assays for TSH are sensitive to 0.01 mIU/L, 100 times more sensitive than first-generation assays. “I’m not sure that getting lower than 0.01 significantly benefits patient care,” Dr. Winter said. Assays for FT4 come in three types: dialysis equilibrium, ultracentrifugation, and second-generation immunoassays. Dr. Winter doesn’t use dialysis equilibrium assays, since they are labor-intensive, expensive, and time-consuming. Ultracentrifugation assays are also labor-intensive. His choice is an immunoassay. “There are lots of good FT4 immunoassays on the market,” he says.

As Dr. Spencer noted, the challenge with TSH assays is selecting an appropriate upper reference limit (d’Herbomez, M, et al. Clin Chem Lab Med. 2005;43:102–105; Volzke H, et al. Thyroid. 2005;15:279–285). Population surveys have addressed this issue (Hollowell JG, et al. J Clin Endocrinol Metab. 2002;87:489–499). However, Dr. Spencer writes in an upcoming article, “[C]urrent analysis shows that an accurate TSH reference range cannot be determined from population study data because occult thyroid dysfunction skews the TSH upper reference limit.” Using ultrasound to detect occult thyroid disease would reduce this problem, but not completely. Furthermore, it is not feasible to do ultrasound on all participants in a population study. Thus, the recommendation to adopt an empiric upper limit of ~3 mIU/L appears justified, especially considering the new recommended upper limit for first trimester of pregnancy (2.5 mIU/L).

The recommendation to reduce the upper TSH reference limit has met with mixed reaction. “One group says we don’t want to label so many asymptomatic persons with a TSH between 3.5 and 4 as having thyroid problems,” Dr. Spencer relates. “Others say that symptoms are an insensitive endpoint for assessing thyroid status and we should use the TSH level more as a risk factor. We must interpret TSH not in terms of a number but relative to patient-specific risk factors.” In her view, it is important to take a nuanced view of TSH results. “I think it is important to set an upper limit somewhere between 2.5 and 3,” she says, “but physicians need to be educated that just because a patient has a TSH of 3.5 doesn’t necessarily mean that patient is abnormal.” Dr. Spencer emphasizes that the TSH reference range doesn’t define a “normal range.” “The clinician must look at the TSH number in the context of whether the patient has thyroid autoantibodies or other contributing risk factors, such as diabetes, obesity, or advanced age. The treatment threshold should be patient-specific,” she advises.

Dr. Spencer says laboratories can’t lower the TSH reference range without educating clinicians. “When we recently lowered the upper limit to 3, we issued a one-page justification for our medical staff quoting the supporting literature,” she says. “We have had no resistance and no problems with doing that. It’s not the big deal that many labs think it is.”

According to Dr. Demers, “There is a feeling that most patients who have subclinical hypothyroidism are being missed with today’s reference range set so high. This is a controversy that will rage on until well-defined population-based studies are done.” Closer to home, Dr. Demers says that every laboratory needs to establish its own reference range for thyroid tests. “There are geographical differences as we go around the country,” he notes, “with different iodine levels in the soil.” A reference range can be determined by assaying 100 or 200 individuals who have had a physical examination and a clinical history suggesting that they are free of thyroid disease. “You should also run all the tests that would be used to make a diagnosis of thyroid disease to make sure these individuals are free of any thyroid abnormality,” he adds. He includes an assay for thyroid autoantibodies.

Pregnancy is a special case of needing a reduced upper limit for the TSH reference range. “In the first 13 weeks of pregnancy, the fetus is totally dependent on maternal thyroxine, especially for development of the brain and neuronal connections,” Dr. Spencer says. “So it is critical that the developing fetus have sufficient thyroxine. Even mild thyroxine deficiency, so-called subclinical hypothyroidism, is associated with preterm delivery and can have a deleterious effect on the IQ of a child.” According to Dr. Spencer, loss of IQ points has been demonstrated in children whose mothers had subclinical hypothyroidism in pregnancy, perhaps secondary to the high prevalence of preterm delivery. “Some are resisting this additional testing,” Dr. Spencer says. “But if you think of the high cost of preterm delivery and neonatal ICU care, clearly it would be beneficial to identify mothers who have mild thyroid deficiency in early pregnancy. Of course, you have to use TSH to detect thyroid hormone deficiency and use a lower TSH upper limit, around 2.5 mIU/L.”

Dr. Demers says the American College of Obstetricians and Gynecologists endorses measuring TSH in early pregnancy (Obstet Gynecol. 2002; 100:387–396). However, he adds, “I don’t think the ACOG people understand that whole reference range issue, so they are leaving it up to laboratorians to define the upper limit.”

Some endocrinologists have suggested also measuring autoantibodies in pregnant women, since 80 to 85 percent of persons with subclinical hypothyroidism are positive for thyroid autoantibodies. This idea, too, is controversial. In addition to confirming autoimmune thyroid disease, the presence of TPOAb in the first trimester of pregnancy is a strong predictor of postpartum thyroiditis, PPT. “Both postpartum thyroiditis and postpartum depression are common,” Dr. Winter says. Add to that the fact that many women are tired after giving birth from lack of sleep and taking care of a newborn, and you can see the need for accurate diagnostic skills. “It could be difficult to diagnose unless you are thinking of postpartum thyroid dysfunction,” Dr. Spencer says. Postpartum thyroiditis is typically associated with thyroid autoantibodies. Hashimoto’s disease can show up as PPT and can be associated with pernicious anemia and type 1 diabetes. “As pregnancy progresses, most autoimmune conditions get better due to dampening of the immune system,” Dr. Spencer says. But about six weeks after delivery, there is a flare-up of autoimmune thyroiditis. The presence of thyroid autoantibodies during pregnancy is a risk factor for PPT: About 30 percent of women who have TPOAb in the first trimester will develop PPT. If TPOAb is still detected in the third trimester, the risk of PPT is 50 percent.

Usually, PPT resolves within a year. But within that time frame the course can be highly variable, covering practically every permutation: transient hyperthyroidism sinking into transient hypothyroidism before recovery (the most common pattern); hyperthyroidism with direct restoration of euthyroidism; or hypothyroidism followed by recovery. In some patients hypothyroidism becomes permanent.

Hypothyroidism, and to a lesser extent hyperthyroidism, constitute the vast majority of indications for testing for thyroid disease. Since thyroid hormone affects practically every tissue, manifestations of these conditions are typically described as protean. For hypothyroidism, clinical findings can include a puffy face, coarse hair, mental dullness, tiredness, weight gain, menstrual irregularities, increased sleeping, cold intolerance, and constipation. Hyperthyroidism is the opposite, reflecting a revved-up metabolism—weight loss; nervousness; tachycardia; sleeplessness; smooth and shiny hair and skin; and diarrhea. A common clinical feature of Graves’ disease is exophthalmos, which is due to increased retro-orbital fat and muscle swelling.

Hashimoto’s thyroiditis, HT, is the most common cause of primary hypothyroidism, while Graves’ disease, GD, is the most common cause of primary hyperthyroidism. What is most striking is that both conditions are associated with the same thyroid autoantibodies that serve as markers of thyroid autoimmunity: TMAb, TPOAb, and TgAb. Both are autoimmune conditions with a familial pattern. Both conditions can occur in the same family, Dr. Winter noted, giving rise to the concept of AITD (autoimmune thyroid disease), where there’s a spectrum of forms of autoimmunity directed against the thyroid gland that lead to a variety of clinical disorders. There can certainly be a genetic tendency to have an autoimmune response to thyroid antigens that plays out differently in different individuals, perhaps due to different HLA gene alleles. “Thyroid autoantibodies identify AITD,” Dr. Winter says.

In HT, a cell-mediated autoimmune response against follicular cells destroys the cells. Ultimately the production of thyroid hormone declines. (Autoantibodies are not themselves pathogenic in this condition, but serve as markers of the autoimmune response.) In GD, autoantibodies with agonistic activity against the TSH receptor produce the characteristic triad of hyperthyroidism, exophthalmia, and infiltrative dermopathy (pretibial myxedema). The diagnosis of GD is based on the presence of diffuse goiter, hyperthyroid symptoms, and a positive thyroid autoantibody test.

End-stage HT results in frank hypothyroidism due to destruction of the thyroid gland. In some cases of HT, an accelerated rate of thyroid gland destruction can cause a transient state of hyperthyroidism from release of thyroid hormone—so-called Hashitoxicosis. This condition is transient and must be distinguished from GD because Hashitoxicosis is self-limited and only symptomatic treatment is required. Highly sensitive third-generation TSH assays differentiate GD from the destructive phase of HT: in GD, TSH levels are typically <0.01 mIU/L, while in destructive HT they are >0.01–0.34 mIU/L (based on a lower limit of the TSH reference interval of 0.35 mIU/L).

If a diagnosis of GD is equivocal, a functional autoantibody assay can resolve this question. A cellular assay for the TSH-receptor agonistic activity of the patient’s serum, called the thyroid-stimulating immunoglobulin test, is preferred, Dr. Winter says. A membrane-based functional assay for blocking autoantibodies to the TSH receptor, the thyrotropin-binding inhibitory immunoglobulins, is not as helpful. “TBII does not distinguish agonist from antagonist autoantibodies,” Dr. Winter points out. Autoantibodies that block binding of TSH to its receptor produce atrophic thyroiditis and are a not-uncommon cause of primary hypothyroidism.

Rarely, one may see a sample from a clinically euthyroid (or only mildly hypothyroid) person in which TSH is high normal or mildly elevated even though both the FT4 and T3 are above normal. One explanation for such an occurrence is resistance to thyroid hormone due to a mutant form of the T3 receptor. Dr. Winter has written about this disorder and other heritable molecular causes of thyroid disease (Winter WE, et al. Ann Clin Lab Sci. 2001;31:221–244). “While molecular diagnosis of such conditions is rarely indicated for clinical management,” he wrote, “knowledge of the molecular mechanisms of these diseases can greatly enhance the clinical laboratory scientist’s ability to advise clinicians about appropriate thyroid testing.”

A puzzling but not-rare finding would be a euthyroid person with persistently elevated TSH but normal FT4 (and normal T3 if it is measured). These results might also occur in a person with minimal symptoms of hypothyroidism, perhaps lipid abnormalities and neurological effects, such as depression or mental slowing. The reverse situation—a person with very mild (if any) hyperthyroid symptoms, depressed TSH but normal FT4 and T3—can also occur. The diagnoses are subclinical hypo- and hyperthyroidism, respectively. Over the longer term, subclinical hypothyroidism presages increased cardiovascular morbidity, while the hyperthyroid version confers an increased risk of atrial fibrillation and osteoporosis. About five percent of persons with euthyroid Hashimoto’s thyroiditis convert to full-blown hypothyroidism each year.

“Our primary care community is a believer in subclinical hypothyroidism,” says Cynthia Lais, MD, pathologist with Hospital Pathology Associates, PA and former medical director, clinical chemistry division, Allina Medical Laboratories, Minneapolis. “And our endocrinologists are certainly believers.” In the context of subclinical hypo- or hyperthyroidism, if TSH is abnormal and FT4 is normal, Dr. Lais recommends repeating the TSH measurement. “If the TSH remains abnormal,” she says, “that patient can be classified as subclinical hypothyroid or hyperthyroid.” Early detection of subclinical hypothyroidism is enhanced by a lower upper limit of the reference range for TSH, Dr. Lais notes. A few years ago, in consultation with area endocrinologists, AML lowered the upper limit for TSH to 2.5, consonant with the NACB recommendation.

“In the last 10 years, even the last five, we have learned more about the impact of hyperthyroidism on the risk of fractures from osteoporosis and the risk of atrial fibrillation. Also, there is more evidence showing linkage between hypothyroidism and cardiovascular risk and diabetes, even in the range of a TSH 3 to 10. These are huge medical problems in this country,” she continues. “If someone with subclinical thyroid disease could be treated earlier and as a result be at less risk of major medical problems, this to me justifies screening.”

Dr. Lais finds that some clinicians are screening annually for thyroid disease in women over 50 and men over 60. “Ten percent of women will have an abnormal TSH by age 50,” she says, “and 10 percent of men by age 60.” If there is any indication of increased risk, such as autoimmune disease, diabetes, or a family history of thyroid disease, many physicians are screening people younger than 50 or 60.

“Thyroid autoimmunity, especially Hashimoto’s thyroiditis, increases in prevalence with age,” says Dr. Spencer. “Increasingly HMOs are recognizing the benefits of being proactive with screening. Kaiser adds a routine TSH test much sooner than age 50. And now,” she continues, “with the issue of pregnancy, it is not a bad idea to check thyroid status in young women in their reproductive years. Usually the women who get into problems with thyroid hormone insufficiency in early pregnancy are those who have autoantibodies before pregnancy.”

Ideally, women should be screened with both TSH and TPOAb tests, Dr. Spencer says. “There are a number of individuals with slightly elevated TSH and ultrasound evidence of autoimmune thyroid disease who don’t have antibodies detected (Vejbjerg P, et al. Eur J Endocrinol. 2006;155:547–552). Thyroid antibody assays are still fairly crude,” she says. “You can detect autoantibodies with one manufacturer’s test but not another.”

On the other end of the spectrum is thyroid testing of critically ill patients in the hospital. “Don’t screen for thyroid dysfunction in hospitalized patients,” Dr. Winter cautions. Such patients often have a condition called sick euthyroid syndrome: normal TSH and FT4 levels with subnormal T3. This is because severe or chronic illness can produce a decline in the enzyme that converts T4 to T3; TSH doesn’t rise because FT4 is normal. This decline is interpreted as an adaptive mechanism that turns down metabolism. “Measure thyroid function in hospitalized patients only when thyroid disease is a serious consideration in explaining the patient’s acute problem,” Dr. Winter says. Unexplained coma or unexplained congestive heart failure might qualify.

One case presented by Dr. Winter involved an obese health care professional who had a low TSH and FT4 but an elevated T3. The most likely diagnosis in this case is exogenous T3 ingestion (presumably for weight loss), consistent with Munchausen syndrome. A low thyroglobulin level and a reduced radioactive iodine unit would support the diagnosis of exogenous T3 ingestion. “A low FT4 and TSH with an increased T3 cannot usually occur physiologically and must represent T3 ingestion,” Dr. Winter says. “I am not aware of a toxic nodule secreting T3 alone,” he adds.

The differential diagnosis of a thyroid mass is a large topic, and one that Dr. Winter covered thoroughly. A malignant thyroid mass can be papillary or follicular carcinoma, both of which are differentiated; anaplastic carcinoma, which is undifferentiated; medullary carcinoma; or lymphoma. All are histologic diagnoses. Thyroid testing plays no role in initial diagnosis. “Thyroid function studies are almost always normal in patients with a thyroid mass,” Dr. Winter says, “but we do them anyway.” However, thyroglobulin, a protein secreted by the thyroid gland, is a useful postsurgical tumor marker in papillary and follicular thyroid cancers, he notes.

A newer concept is to measure baseline Tg prior to surgery, Dr. Spencer says. “Use of ultrasound is increasingly picking up small nodules that lead to biopsy, which yields a diagnosis of thyroid carcinoma. In these cases, it is useful to get a pre-op thyroglobulin level, which gives you some sort of idea of the tumor’s ability to produce the thyroglobulin molecule.” Dr. Spencer’s group looks at the relationship between the size of the nodule, the amount of tumor that patient will be found to have at surgery, and the preoperative Tg concentration. “If a tiny, 1-cm nodule turns out to be papillary cancer and the pre-op thyroglobulin is, say, 300 [the upper limit of the reference range is 40], one can assume that tumor is very efficient at producing thyroglobulin,” she explains. “So after surgery, we can consider that thyroglobulin will be a very sensitive marker of recurrence for that particular tumor.” Another patient may have a large differentiated tumor with preoperative Tg in the reference range. If Tg is undetectable after thyroidectomy in that patient, it doesn’t necessarily mean all tumor was removed.

“Perhaps the biggest influence on thyroglobulin secretion is TSH,” Dr. Spencer continues. A helpful development in monitoring thyroid carcinoma was the introduction of a test that uses recombinant human TSH to stimulate the thyroid to take up radioactive iodine for imaging, which is one way to detect residual thyroid tissue or recurrence. Highly sensitive Tg assays are providing an alternative to rhTSH stimulation testing, Dr. Spencer says. She draws an analogy between improvements in Tg assays and the TSH story of the 1980s and 1990s. “Early TSH assays were very insensitive,” she recalls. “As immunometric assays were brought in, TSH sensitivity improved 100-fold. This is also happening with thyroglobulin.” She further sees an analogy between the rhTSH stimulation test and the TRH stimulation test that was used to overcome TSH assay insensitivity. “As TSH assays became more sensitive, there was no need to use TRH stimulation testing,” she says. “The situation with rhTSH stimulation of thyroglobulin is very similar.” As highly sensitive Tg assays are brought in, there will be no need to use rhTSH stimulation, which will save considerable expense. “This is a major advance that we will see in the future,” Dr. Spencer predicts.

“It is becoming evident if you have a sensitive thyroglobulin assay, one that measures 0.1 ng/mL or lower, monitoring the basal thyroglobulin level when TSH is suppressed will become recognized as a better indicator of whether the tumor is growing than measuring rhTSH-stimulated thyroglobulin,” Dr. Spencer says. Following thyroidectomy for a thyroid tumor, most patients are placed on high doses of levothyroxine to keep TSH suppressed or in the low range because it is known that TSH is a trophic factor for these tumors. “Studies show that keeping TSH suppressed with a high level of thyroid hormone reduces the recurrence rate,” Dr. Spencer says. She foresees that the hallmark for disease will become a rising basal thyroglobulin level in the face of chronic TSH suppression.

Dr. Spencer cites one further wrinkle with Tg assays. About 20 percent of patients with thyroid carcinomas have antibodies to Tg. Unfortunately, immunometric Tg assays, which can be automated, suffer from interference by the antibodies and give false results. So when Tg antibodies are detected, radioimmunoassays, which give valid Tg measurements in the presence of Tg antibodies, must be used. “Only a few labs offer these RIAs,” Dr. Spencer says, “but they are being increasingly requested for monitoring thyroid carcinoma patients with antibodies to thyroglobulin.” Two laboratories that do offer the RIA for Tg are Esoterix and Dr. Spencer’s laboratory.

Dr. Winter notes that traditionally it has been advised that Tg not be measured as a tumor marker in patients with differentiated thyroid cancer who are TgAb positive because TgAb interferes in unpredictable ways in Tg immunoassays. In some case where TgAb is positive, the Tg may be depressed, whereas in other cases the Tg may rise, possibly leading to erroneous clinical interpretations of the patient's thyroid cancer status.

Dr. Lais has her own fairly radical vision of the future of thyroid testing. “So far most TSH screening has been age-related,” she says. “I think that in a matter of years we will be doing TSH screening just as we do glucose screening. The number of patients with thyroid disease eclipses those with diabetes—the numbers are astounding.”

William Check is a medical writer in Wilmette, Ill..