Approach to a low TSH level: Patience is a virtue
ABSTRACTConfronted with a low serum level of thyrotropin (thyroid-stimulating hormone, TSH), physicians should not jump to the conclusion that it is due to a hyperthyroid state, as other conditions and some drugs can be associated with a TSH level that is slightly low (0.1–0.4 μIU/mL) or frankly suppressed (< 0.1 μIU/mL). This review discusses how to approach a low TSH, stressing the frequent need to reassess thyroid function before making a diagnosis, the underlying processes and the drugs that can be responsible, and the degree of TSH suppression and its role in the evaluation.
KEY POINTS
- A low TSH value should always be followed up by measuring the thyroid hormones, ie, thyroxine (T4) and triiodothyronine (T3).
- Serum levels of free thyroid hormones should be used when interpreting an abnormal TSH level, especially in the acute and inpatient settings.
- A low TSH level is not always the result of suppression by elevations in circulating thyroid hormones.
- A low TSH level in the setting of normal levels of free thyroid hormones should always be reassessed in 4 to 6 weeks before making a diagnosis.
- Overt hyperthyroidism is usually associated with a frankly suppressed TSH (< 0.1 μIU/mL).
LOW TSH, NORMAL FREE T4, NORMAL FREE T3
Subclinical hyperthyroidism
Subclinical hyperthyroidism is defined as low TSH, normal free T4, and normal free T3 levels. Symptoms of hyperthyroidism such as fatigue, insomnia, weight loss, palpitations, tremor, or heat intolerance generally play a role in whether therapy is considered, but not in making the diagnosis of subclinical hyperthyroidism. To make the correct diagnosis, it is crucial to confirm that this pattern of test results persists by repeating these tests over the next few months.
Exogenous thyrotoxicosis, by far the most common form of subclinical thyrotoxicosis, results from taking levothyroxine (T4) or liothyronine (T3), or both, either in unintentional supratherapeutic doses in patients with hypothyroidism or in intentionally high doses to suppress TSH in patients with a history of differentiated thyroid cancer.
Endogenous thyrotoxicosis. Subclinical hyperthyroidism from an endogenous cause is the result of an underlying pathophysiologic process, the same processes responsible for overt states of hyperthyroidism (eg, Graves disease, toxic nodular thyroid disease) (see the discussion of overt hyperthyroidism in a later section).
The course of endogenous subclinical hyperthyroidism depends on the underlying cause and on the level of TSH suppression.3–5 Subclinical hyperthyroidism secondary to a multinodular goiter is estimated to progress to overt hyperthyroidism in about 5% of patients per year,6 but in patients with nodular thyroid disease and TSH levels of 0.1 μIU/mL or lower, one study reported progression to overt hyperthyroidism in approximately 10% of patients per year.3
The risk of subclinical Graves disease progressing to overt hyperthyroidism has been difficult to estimate, given the relapsing and remitting nature of the disease. Rosario3,4 reported that subclinical Graves disease progressed to overt hyperthyroidism in 2 years in 6 (40%) of 15 patients who had TSH levels lower than 0.1 μIU/mL, but in no patients who had TSH levels of 0.1 to 0.4 μIU/mL. These patients were younger than 65 years. In a group age 60 and older with endogenous subclinical hyperthyroidism and a TSH level between 0.1 and 0.4 μIU/mL, Rosario4 reported that progression to overt hyperthyroidism was uncommon, occurring in about 1% of patients per year.
Thus, periodic reassessment of thyroid function tests in patients with subclinical hyperthyroidism is crucial in monitoring for disease progression, especially in those with frankly suppressed TSH values (< 0.1 μIU/mL).
Adverse outcomes associated with subclinical hyperthyroidism are mainly cardiac arrhythmias (atrial fibrillation) and accelerated loss of bone mineral density.
Cooper7 notes that definitive treatment (radioactive iodine ablation, antithyroid drugs, or surgery) “seems reasonable” for older patients (age > 60 years) with a TSH level lower than 0.1 μIU/mL and for certain patients with TSH levels of 0.1 to 0.4 who are at high risk, eg, those with a history of heart disease, osteoporosis, or symptoms of hyperthyroidism.
Normal variant
The normal range for TSH, as for other substances, is defined as the mean value in the general population plus or minus 2 standard deviations. This range includes 95% of the population, so that 2.5% of people have a level higher than this range, and 2.5% have a level lower than this range.
But some people with lower levels of TSH, especially in the range of 0.1 to 0.4 μIU/mL (3 standard deviations below the mean) are actually euthyroid. These people have historically been classified as having subclinical hyperthyroidism, as there is no means of differentiating these “normal” euthyroid people from people with asymptomatic subclinical hyperthyroidism. They need to be followed, since they may have true subclinical hyperthyroidism that may manifest symptomatically in the future, possibly warranting treatment.
Euthyroid sick syndrome
Euthyroid sick syndrome is common during critical illness. However, thyroid disease is common in the general population, and often no test results from before the onset of a critical illness are available to help the clinician separate overt thyroid disease from euthyroid sick syndrome. Furthermore, patients are often unable to provide a history (or to relate their symptoms) of overt thyroid disease, making abnormal thyroid function tests difficult to interpret in the hospital. When previous values are available, they can be invaluable in correctly interpreting new abnormal results.
Thyroid function test values in euthyroid sick syndrome can vary depending on the severity of illness. A low free T3, a normal free T4, and a low-normal TSH are the most common abnormalities seen in euthyroid sick syndrome. The free T3 level is low because of decreased peripheral conversion of T4 to T3 during critical illness. However, euthyroid sick syndrome can present with a spectrum of abnormal thyroid function tests, further complicating interpretation and diagnosis. Serum TSH levels have been reported to be normal in about 50%, low in 30%, and high in 12% of patients with nonthyroidal illness.8 However, marked suppression of serum TSH (< 0.1 μIU/mL) was observed only in about 7% of patients, mainly in those whose clinical picture was confounded by medications (dopamine or corticosteroids, or both) that have independent TSH-lowering effects (see below).8
Drugs that suppress TSH
Many drugs used in the hospital and intensive care unit can alter thyroid function tests independently of systemic illness, further complicating the clinical picture.
Glucocorticoids, in high doses, have been shown to transiently suppress serum TSH.9,10
Octreotide (Sandostatin) and other somatostatin analogues also transiently suppress TSH.11–14 However, these drugs (and glucocorticoids) do not appear to result in central hypothyroidism.10,15–17
Dopamine, given in pharmacologic doses for a prolonged time, has been shown to reduce the serum TSH level in both critically ill and normal healthy people.18
Dobutamine (Dobutrex) in pharmacologic doses has been likewise shown to lower TSH levels, although the serum TSH level was noted to remain within the normal range in those who had a normal TSH value at baseline.19
Amiodarone. Although most patients who take amiodarone (Cordarone, Pacerone) remain euthyroid, the drug can cause hypothyroidism or hyperthyroidism. Initially, amiodarone usually causes a decrease in T3 via inhibition of 5′-deiodinase, with a transient reciprocal increase in TSH.20
When amiodarone induces thyrotoxicosis, the condition can be subclinical, manifested by a low TSH in the setting of normal levels of thyroid hormones, or as overt thyrotoxicosis with a low TSH and elevated levels of thyroid hormones. See further discussion below on amiodarone’s effects on thyroid function.
Patients taking drugs that lower TSH are often critically ill and may also have a component of euthyroid sick syndrome, resulting in a mixed picture.
