November 2009: Volume 35, Number 11
AACC’s Expert Access
Contemporary Issues in Thyroid Disease Measurements
Each month, AACC’s Expert Access Live Online Program features a different hot topic. Visit AACC’s website for more information and an archive of past presentations.
The following is an excerpt from the July 2009 presentation by Carole A. Spencer, MT, PhD, FACB, director of the University of Southern California Endocrine Laboratories.
If TSH value is trending towards the high end of the reference value, would you consider repeat testing to see if the assay is good without any interference from circulating anti-mouse antibodies? After confirming everything is in order as far as the TSH test result is concerned would you further investigate other causes of slight elevation trend in successive TSH assay results?
When the TSH is trending >3.0 mIU/L the first check should be for TPO antibodies (TPOAb) and an autoimmune etiology. Family history is also important. Measuring TSH on a blood specimen drawn 2 months or so later would establish that the TSH abnormality is persistent and not related to TSH lability or nonthyroidal illness. Measuring the TSH using a different manufacturer’s method is a good first step to checking for human anti-mouse antibodies. FT4 is unlikely to be helpful unless the TSH is grossly abnormal, >10 mIU/L or <0.05 mIU/L. Note: not all high TSH levels are abnormal. There can be inactivating polymorphisms of the TSH receptor requiring a higher TSH to maintain euthyroidism. Judgments regarding any treatment or follow-up interval should relate to patient-specific factors.
What should be taken into account in a study designed to assess the TSH reference range in a population 60–80 years old? Does the NACB guideline approach for TSH reference intervals apply? Do we need 120 euthyroid volunteers for each decade if we also want to know the TSH reference intervals for each decade of life?
The NHANES population found a TSH upper limit of 7.5 mIU/L for 80+ year olds (JCEM 2007;92:4575 and JCEM 2009;94:1251). I do not believe we need a TSH population reference range for every decade, but we do need to educate physicians to expect higher TSH in older patients. The question to treat or not to treat ultimately depends on symptoms, TPOAb, medications, and other patient-specific factors. I think we need to abandon the concept of a fixed reference range for TSH or tailoring a reference range for every condition. Furthermore, different assays detect different TSH isoforms. In any situation, not all of the TSH that we are measuring is biologically active, and an increase in bioinactive TSH may be involved with aging.
What is your opinion regarding manufacturers’ addition of albumin to their FT4 reagent?
Manufacturers add a variety of proprietary components to try to engineer a FT4 immunoassay to overcome TBG effects. I believe it is not unusual to add albumin. From my reading, it is apparent that current FT4 immunoassays are all albumin dependant to varying extents. The tests work fairly well with uncomplicated hypo- or hyperthyroidism but are prone to artifacts with many drug therapies or in low albumin states. We need to remember FT4 immunoassays are only estimate tests. Under most circumstances TSH is the more reliable FT4 biosensor provided that thyroid status is stable and hypothalamic-pituitary function is intact.
Is the T3 uptake coming back? How do we prevent its rebirth?
Thyroid Hormone Binding Ratio (THBR) tests used to be called “T3 uptakes”—they estimate TBG concentrations. THBR tests are available on many platforms. Free T4 indexes (FT4I) calculated by dividing total T4 by THBR is particularly useful for evaluating pregnant patients because the non-pregnant FT4I reference range applies to pregnancy whereas the FT4 immunoassay values can fall below the nonpregnant lower limit by the third trimester causing unnecessary anxiety. FT4I is also useful for assessing thyroid status of hospitalized patients for the same reasons. Our hospital allows our endocrinologists and obstetricians to request FT4I as a write in test for these reasons.
Could you please share the most current recommendations for thyroid reflex testing along with the appropriate cut-off values for abnormal TSH? We would like to offer the option for physicians to order TSH with reflex to avoid delays in follow-up testing.
Because of the log/linear TSH/FT4 relationship, you would not expect FT4 to be abnormal unless TSH was >10 mIU/L or <0.05 mIU/L. There isn’t much point to reflex test unless TSH is grossly abnormal.
Did you seen any correlations between thyroid and vitamin D?
I have seen no correlations between thyroid and vitamin D.
What tests are essential for thyroid disorders?
It depends on the disorder. The most sensitive test of thyroid dysfunction is TSH. Because of the log/linear relationship between TSH and FT4 you don’t expect FT4 to be abnormal unless TSH is >10 mIU/L or <0.05 mIU/L. TPO antibodies (TPOAb) is the most sensitive marker for thyroid autoimmunity. T3 is used in some cases when it is necessary to determine the etiology of hyperthyroidism (the TT3/TT4 ratio is high (>20:1) in cases of Graves’ hyperthyroidism). TSH +TPOAb is becoming important for preconception and pregnancy evaluations for thyroid dysfunction. Thyroglobulin is used as a tumor marker for thyroid cancer.
What would be the risks of a rapid home test for elevated TSH? If such a test were available, what would be a good cut off level?
I do not believe that rapid home TSH tests have optimal sensitivity and precision. It is important not only to be able to detect high levels of TSH but also to reliably determine whether TSH is above 2.5 mIU/L for preconception and pregnancy evaluations and be able to reliably detect low TSH <0.1 mIU/L (subclinical hyperthyroidism) in patients taking too much levothyroxine replacement therapy for hypothyroidism
What are the advantages of a third generation TSH method versus a second generation TSH method? Should a third generation method be reported to three decimal places? What makes a method a third generation TSH method?
A third generation TSH method is a method that has a functional sensitivity ≤0.01 mIU/L and a second generation TSH method has a functional sensitivity of 0.1 mIU/L calculated according to NACB guidelines (Thyroid 2003;13:34). Only 2 decimal places are relevant measuring in the 0.01–1.0 range. I suggest one decimal place above 1.0 and whole numbers above 20 mIU/L. In short the functional sensitivity limit is the TSH that can be measured in human sera with a between-run 20% CV over 6–8 weeks using at least two lots of reagents and two instrument calibrations. Most current instruments can achieve this. It is critical to have reliable low range measurement for detecting iatrogenic hyperthyroidism (a problem for 15–20% of LT4 treated hypothyroid patients), thyroid cancer patients (in whom a suppressed TSH is often the goal) and hospitalized patients with nonthyroidal illness (in whom a TSH <0.01 mIU/L likely indicates hyperthyroidism whereas a low but detectable TSH more likely indicates a transient state of illness).
Disclaimer—The opinions and information are the sole responsibility of the presenter. AACC reviews the presentations for oval appropriateness, but this should not be construed as an endorsement by the association or its employees of the opinions and information offered here.
Supported in part by an education grant from Siemens Healthcare Diagnostics.