166 Thyroid Gland Disorders

Thyroid storm and myxedema coma are life-threatening emergencies that represent the extreme ends of the spectrum of thyroid dysfunction in the decompensated patient. Their presentation is usually dramatic and is often precipitated by a nonthyroidal-related illness or event. Recognition of these disorders requires a high degree of clinical suspicion, because thyroid function abnormalities, as well as other biochemical parameters, do not differ significantly from uncomplicated thyrotoxicosis and hypothyroidism. As thyroid storm and myxedema coma are clinical diagnoses, measurement of serum thyroid hormones serve as confirmatory tests in the appropriate setting.

In contrast to these dramatic clinical presentations, critical illness also causes multiple nonspecific alterations in thyroid hormone concentrations in patients without intrinsic thyroid dysfunction that relate to the severity of the illness. Since a wide variety of illnesses tend to result in the same changes in serum thyroid hormones, such alterations in thyroid hormone indexes have been termed the sick euthyroid syndrome. The differentiation between patients with the sick euthyroid syndrome and those with intrinsic thyroid disease is a frequent diagnostic challenge in the intensive care unit (ICU).

This chapter will review normal thyroid physiology, the changes in thyroid hormone metabolism seen with critical illness, and the evaluation of thyroid function in critically ill patients. Finally, diagnosis and management of the sick euthyroid syndrome, thyroid storm, and myxedema coma will be reviewed.

Normal Thyroid Hormone Economy

Regulation

Synthesis and secretion of thyroid hormone is under the control of the anterior pituitary hormone, thyrotropin (or thyroid-stimulating hormone [TSH]). Following a classic negative feedback system, TSH secretion increases when serum thyroid hormone levels fall and decreases when they rise (Figure 166-1). TSH secretion is also under the regulation of the hypothalamic hormone, thyrotropin-releasing hormone (TRH). The negative feedback of thyroid hormone is targeted mainly at the pituitary level but likely affects TRH release from the hypothalamus as well. In addition, input from higher cortical centers can affect hypothalamic TRH secretion.

Figure 166-1 Diagram of the hypothalamic-pituitary-thyroid axis.

Inhibitory effect of T₄ and T₃ on TSH secretion is shown by dashed line and minus sign, and stimulatory effects of TRH on TSH secretion and TSH on thyroid secretion are shown by solid lines and plus signs. T₄ and T₃ may also have an inhibitory effect on TRH secretion.

Under the influence of TSH, the thyroid gland synthesizes and releases thyroid hormone. Thyroxine (T₄, 65% iodine by weight) is the principal secretory product of the thyroid gland, comprising about 90% of secreted thyroid hormone under normal conditions.¹ Whereas T₄ may have direct actions in some tissues, it primarily functions as a hormone precursor that is metabolized in peripheral tissues to the transcriptionally active 3,5,3′-triiodothyronine (T₃, 59% iodine by weight).

The major pathway of metabolism of T₄ is by sequential monodeiodination.² At least three deiodinases, each with its unique expression in different organs, catalyze the deiodination reactions involved in the metabolism of T₄. Removal of the 5′-, or outer ring, iodine by type I iodothyronine 5′-deiodinase (D1) or type II iodothyronine 5′-deiodinase (D2) is the “activating” metabolic pathway leading to formation of T₃. Removal of the inner ring, or 5-, iodine by type III iodothyronine deiodinase D3 is the “inactivating” pathway producing the metabolically inactive hormone, 3,3′,5′-triiodothyronine (reverse T₃, rT₃). D1 is found most abundantly in the liver, kidneys, and thyroid. It is up-regulated in hyperthyroidism and down-regulated in hypothyroidism. D2 is found primarily in the brain, pituitary, and skeletal muscle and is down-regulated in hyperthyroidism and up-regulated in hypothyroidism. D3 is expressed primarily in the brain, in skin, and in placental and chorionic membranes. The actions of D3 also include inactivation of T₃ to form T₂, another inactive metabolite. Under normal conditions, about 41% of T₄ is converted to T₃, about 38% is converted to rT₃, and about 21% is metabolized via other pathways, such as conjugation in the liver and excretion in bile.^4,⁵

T₃ is the metabolically active thyroid hormone and exerts its actions via binding to chromatin-bound nuclear receptors and regulating gene transcription in responsive tissues.³ Important in understanding the alterations in circulating thyroid hormone levels seen in critical illness is the fact that only around 10% of circulating T₃ is secreted directly by the thyroid gland while more than 80% of T₃ is derived from conversion of T₄ in peripheral tissues.^1,² Thus, factors that affect peripheral T₄-to-T₃ conversion will have significant effects on circulating T₃ levels. Serum levels of T₃ are approximately 100-fold less than those of T₄, and like T₄, T₃ is metabolized by deiodination to form diiodothyronine (T₂) and by conjugation in the liver. The half-lives of circulating T₄ and T₃ are 5 to 8 days and 1.3 to 3 days, respectively.⁴

Serum Binding Proteins

Both T₄ and T₃ circulate in the serum as hormones bound to several proteins synthesized by the liver.⁵ Thyroid-binding globulin (TBG) is the predominant transport protein and binds roughly 80% of the circulating serum thyroid hormones. The affinity of T₄ for TBG is about 10-fold greater than that of T₃ and is part of the reason circulating T₄ levels are higher than T₃ levels. Other serum binding proteins include transthyretin,⁶ which binds some 15% of T₄ but little if any T₃, and albumin, which has a low affinity but a very large binding capacity for T₄ and T₃. Overall, 99.97% of circulating T₄ and 99.7% of circulating T₃ is bound to plasma proteins.

Free Hormone Concept

Essential to an understanding of the regulation of thyroid function and the alterations of circulating thyroid hormones seen in critical illness is the “free hormone” concept, which is that only the unbound hormone has any metabolic activity. Under regulation by the pituitary, overall thyroid function is affected when there are any changes in free hormone concentrations. Changes in either the concentrations of binding proteins or the binding affinity of thyroid hormone to the serum binding proteins have significant effects on total serum hormone levels, owing to the high degree of binding of T₄ and T₃ to these proteins. Despite these changes, this does not necessarily translate into thyroid dysfunction.

Thyroid Hormone Economy in Critical Illness

Widespread changes in thyroid hormone economy in the critically ill patient occur as a result of (1) alterations in peripheral metabolism of thyroid hormones, (2) alterations in TSH regulation, and (3) alterations in the binding of thyroid hormone to TBG.

Peripheral Metabolic Pathways

One of the initial alterations in thyroid hormone metabolism in acute illness is the acute inhibition of D1, resulting in the impairment of T₄-to-T₃ conversion in peripheral tissues.⁷ D1 is inhibited by a wide variety of factors, including acute illness (Box 166-1),² resulting in the acute decrease in T₃ production in critically ill patients. In contrast, inner ring deiodination by D3 may be increased by acute illness, resulting in increased levels of rT₃.⁸ Additionally, because rT₃ is subsequently deiodinated by D1, degradation of rT₃ decreases, and levels of this inactive hormone rise in proportion to the fall in T₃ levels. Finally, there is impaired transport of T₄ to peripheral tissues such as the liver and kidney, where much of the circulating T₃ is produced, further contributing to the decrease in production of T₃.⁹

Thyrotropin Regulation

Serum TSH levels are usually normal early in acute illness.¹⁰ Decreased TRH secretion due to inhibitory signals from higher cortical centers, impaired TRH metabolism,¹¹ the alteration of pulsatile TSH,¹² and the decrease or absence of a nocturnal TSH surge^12,¹³ may all further lower TSH levels. Serum levels of leptin, the ob gene product that has been shown to vary directly with thyroid hormone levels,¹⁴ also falls as illness progresses¹⁵ and hypothalamic TRH secretion falls, which in turn leads to lowered TSH levels.¹⁶

The decrease of hypothalamic TRH gene expression in animal models is, however, not associated with increased serum T₄ and T₃ levels.¹⁷ Finally, certain thyroid hormone metabolites that are increased during acute nonthyroidal illness may play a role in the inhibition of TSH and TRH secretion.¹⁸

Common medications used in the treatment of the critically ill patient may also have inhibitory effects on serum TSH levels (Box 166-2). Van den Berghe et al.¹⁹ reported that intravenous (IV) administration of dopamine for as short a time as 15 to 21 hours can acutely decrease TSH levels, and its withdrawal results in a 10-fold increase in serum TSH levels. In one study, children who received dopamine infusions during a pediatric ICU admission for meningococcal sepsis had lower TSH levels than those who did not.^20,²¹ Increased levels of glucocorticoids, whether from endogenous or exogenous sources, also have direct inhibitory effects on TSH secretion.

Serum Binding Proteins

The affinity of thyroid hormones binding to transport proteins and the concentrations of serum binding proteins are altered with acute illness (Table 166-1). Serum levels of transthyretin and albumin decrease, especially during prolonged illness, malnutrition, and in high catabolic states. TBG levels may be increased, as seen with liver dysfunction and human immunodeficiency virus (HIV) infection, or decreased, as seen with severe or prolonged illness.⁵ TBG may also be rapidly degraded by protease cleavage during cardiac bypass, thereby partially explaining the rapid fall of serum T₃ levels in patients undergoing cardiac surgery.²²

TABLE 166-1 Factors That Alter Binding of T₄ to Thyroid-Binding Globulin

	Increase Binding	Decrease Binding
Drugs	Estrogens	Glucocorticoids
Methadone	Androgens
Clofibrate	L-Asparaginase
5-Fluorouracil	Salicylates
Heroin	Mefenamic acid
Tamoxifen	Antiseizure medications (phenytoin, Tegretol)
Raloxifene	Furosemide
	Heparin
	Anabolic steroids
Systemic Factors	Liver disease	Inherited
Porphyria	Acute illness
HIV infection	Nonesterified free fatty acids (NEFAs)
Inherited

An acquired binding defect of T₄ to TBG is commonly seen in patients with critical illness. This is thought to result from the release of some as yet unidentified factor from injured tissues that has the characteristics of unsaturated nonesterified fatty acids (NEFA),²³ which also inhibit T₄-to-T₃ conversion.²⁴ In systemically ill patients, NEFA levels rise in parallel with the severity of the illness,²⁵ and drugs such as heparin stimulate the generation of NEFA.²⁶ Many drugs including high-dose furosemide, antiseizure medications, and salicylates also alter binding of T₄ to TBG. The alterations in serum binding proteins in critical illness make estimating free hormone concentrations difficult (see later).

Evaluation of Thyroid Function in the Critically Ill Patient

Diagnostic Tests

Thyrotropin Assays

Abnormal thyroid function tests have been reported in 20% to 40% of acutely ill patients, more than 80% of whom have no intrinsic thyroid dysfunction after resolution of the illness.²⁷^–²⁹ In a study of 1580 hospitalized patients, only 24% of patients with suppressed TSH values (TSH < assay limit of detection) and 50% of patients with TSH values over 20 mU/L were found to have thyroid disease.^27,²⁸ More importantly, none of the patients with subnormal but detectable TSH values and only 14% of patients with elevated TSH values less than 20 mU/L were subsequently diagnosed with intrinsic thyroid dysfunction. The development of sensitive third-generation TSH assays have led to small improvements in discerning between overt hyperthyroidism and nonthyroidal illness.²⁷ Overall, however, while a normal TSH level has a high predictive value of normal thyroid function, an abnormal TSH value alone is not helpful in evaluating thyroid function in the critically ill patient.

Serum T₄ and T₃ Concentrations

Measurement of free thyroid hormone concentrations in the patient with nonthyroidal illness is fraught with difficulty.³⁰ The gold standard for determination of free hormone levels is equilibrium dialysis. However, this technique is labor intensive and time consuming and thus is rarely used. The most commonly available laboratory tests of thyroid hormone concentrations, the free T₄ index, free T₄, and free T₃, are measured by analog methods which represent estimates of the free hormone concentration and are therefore subject to inaccuracies.^31,³²

The free T₄ index is determined by multiplying the total T₄ concentration by the T₃ or T₄ resin uptake, which is an inverse estimate of serum TBG concentrations.³² Recent developments have allowed the measurement of free T₄ levels by the analog method, a less expensive alternative to the free T₄ index,³³ but the two tests are likely comparably accurate.³⁴ In a healthy population, there is a close correlation between the free T₄ index and free T₄ levels. In the critically ill patient, this association is no longer seen, mainly because of difficulties in estimating TBG binding with resin uptake tests. In spite of this, the sensitivity of the free T₄ index in a large study of hospitalized patients was 92.3%, compared to 90.7% for the sensitive TSH test.²⁷

Serum T₃ concentrations are affected to the greatest degree by alterations in thyroid hormone economy resulting from acute illness. Therefore, there is no indication for routine measurement of serum T₃ levels in the initial evaluation of thyroid function in the critically ill patient. This test should only be obtained if thyrotoxicosis is clinically suspected in the presence of a suppressed sensitive TSH and elevated (or high normal) free T₄ index or free T₄ values. The total T₃ assay is preferable to the free T₃ (analog) assay, owing to the variability between laboratories with the latter test.³²

Although some investigators have reported that serum rT₃ levels are a significant prognostic indicator of mortality in the ICU,³⁵ rT₃ levels are generally unreliable and should not be used to distinguish between intrinsic thyroid dysfunction and nonthyroidal illness.³⁶

Serum Thyroid Autoantibodies

Autoantibodies to thyroglobulin and thyroid peroxidase (TPO), two intrinsic thyroid proteins, are commonly ordered tests.³² Significant titers of either or both of these antibodies indicate the presence of autoimmune thyroid disease, but the presence of thyroid autoantibodies alone does not necessary indicate thyroid dysfunction, as they are present in approximately 12% to 26% of the general population.³⁷ Thyroid autoantibodies do, however, add to the sensitivity of abnormal TSH and FTI values in diagnosing known intrinsic thyroid disease.^27,²⁸

Imaging Studies

Imaging studies are rarely essential to the diagnosis of thyroid disorders in the critically ill patient. Occasionally, functional analysis of the thyroid gland using the radioisotope, iodine-123 (¹²³I), may be useful in the patient with suspected thyrotoxicosis and equivocal laboratory tests. However, these studies are labor intensive, and managing the underlying acute illness often overshadows the benefits of obtaining these studies. Anatomic studies such as ultrasound, isotopic imaging, computed tomography (CT), and magnetic resonance imaging (MRI) are useful in the evaluation of thyroid nodules and goiter, but these conditions rarely are the cause of acute illness; as such, these studies are not usually helpful in the critically ill patient.

Diagnosis

Routine screening of an ICU population for the presence of thyroid dysfunction is not recommended because of the high prevalence of abnormal thyroid function tests and low prevalence of true thyroid dysfunction. When thyroid function tests are ordered in a hospitalized patient, it should only be done if there is a high clinical index of suspicion for thyroid dysfunction. Whenever possible, it is best to defer evaluation of the thyroid-pituitary axis until the patient has recovered from the acute illness. Because every test of thyroid hormone function can be altered in the critically ill patient, no single test can definitively rule in or rule out the presence of intrinsic thyroid dysfunction.

If there is a high clinical suspicion for intrinsic thyroid dysfunction in the critically ill patient, reasonable initial tests would include either free T₄ index or free T₄ and TSH measurements. Assessment of these values in the context of the duration, severity, and stage of illness of the patient will allow the correct diagnosis in most patients. For example, a mildly elevated TSH coupled with a low free T₄ index or free T₄ is more likely to indicate primary hypothyroidism early in an acute illness, as opposed to the same values obtained during the recovery phase of the illness. Similarly, the combination of an elevated TSH and low-normal free T₄ index or free T₄ is more likely to indicate thyroid dysfunction in the hypothermic, bradycardic patient than the tachycardic, normothermic individual. If both the free T₄ index or free T₄ and TSH are normal, thyroid dysfunction is effectively eliminated as a significant contributing factor to the clinical picture. If the diagnosis is still unclear, measurement of thyroid antibodies is helpful as a marker of intrinsic thyroid disease and increases the sensitivity of both the free T₄ index or free T₄ and the TSH. Only in the case of a suppressed TSH and a mid- to high-normal free T₄ index or free T₄ are measurement of serum T₃ levels indicated.

Sick Euthyroid Syndrome

As discussed earlier, critical illness causes multiple nonspecific alterations in thyroid hormone concentrations in patients without intrinsic thyroid dysfunction that relate to the severity of the illness.^18,^38,³⁹ One author has postulated that sick euthyroid syndrome may be a compensatory mechanism in response to the oxidative stress of acute illness.⁴⁰ Whatever the underlying cause, these alterations in thyroid hormone parameters represent a continuum of changes that depends on the severity of the illness and can be categorized into several distinct stages (Figure 166-2).¹⁸ The wide spectrum of changes observed often results from the differing points in the course of the illness when the thyroid function tests were obtained. Importantly, these changes are rarely isolated and often associated with alterations of other endocrine systems, such as decreases in serum gonadotropin and sex hormone concentrations⁴¹ and increases in serum ACTH and cortisol levels.⁴² Thus, the sick euthyroid syndrome should not be viewed as an isolated pathologic event but as part of a coordinated systemic reaction to illness involving both the immune and endocrine systems.

Figure 166-2 Alterations in thyroid hormone concentrations with critical illness.

Schematic representation of the continuum of changes in serum thyroid hormone concentrations in patients with nonthyroidal illness. Alterations become more pronounced with increasing severity of illness, and return to normal range as illness subsides and patient recovers. A rapidly rising mortality accompanies the fall in total and free T₄ concentrations. rT₃, reverse triiodothyronine (3,3′,5′-triiodothyronine); T₃, 3,5,3′-triiodothyronine; TSH, thyroid-stimulating hormone (thyrotropin).

(From Farwell AF. Sick euthyroid syndrome in the intensive care unit. In: Irwin RS, Rippe JM, editors. Intensive care medicine. 5th ed. Philadelphia: Lippincott Williams & Wilkins; 2003.)

Low T₃ State

Common to all of the abnormalities in thyroid hormone concentrations seen in critically ill patients is a substantial depression of serum T₃ levels, which can occur as early as 24 hours after the onset of illness. Over half of patients admitted to the medical service will demonstrate depressed serum T₃ concentrations.^27,²⁸ Development of the low T₃ state arises from impairment of peripheral T₄-to-T₃ conversion through inhibition of type 1 deiodinase (discussed earlier). This results in marked reduction of T₃ production and rT₃ degradation,⁴³ thereby leading to reciprocal changes in serum T₃ and serum rT₃ concentrations. Low T₃ levels are also found in peripheral tissues.³⁵ Thyroid hormone receptor expression is also decreased in acute nonthyroidal illness,⁴⁴ possibly in response to the decrease in tissue T₃ levels.

High T₄ State

Serum T₄ levels may be elevated early in acute illness due to either the acute inhibition of type 1 deiodinase or increased TBG levels. This is seen most often in the elderly and in patients with psychiatric disorders. As the duration of illness increases, non-deiodinative pathways of T₄ degradation increase serum T₄ levels to the normal range.²⁸

Low T₄ State

As the severity and duration of the illness increases, serum total T₄ levels decrease into the subnormal range. Contributors to this decrease in serum T₄ levels are (1) a decrease in the binding of T₄ to serum carrier proteins, (2) a decrease in serum TSH levels, leading to decreased thyroidal production of T₄, and (3) an increase in non-deiodinative pathways of T₄ metabolism. The decline in serum T₄ levels correlates with prognosis in the ICU, with mortality increasing as serum T₄ levels drop below 4 µg/dL and approaching 80% in patients with serum T₄ levels below 2 µg/dL.⁴⁵^–⁴⁷ Despite marked decreases in serum total T₄ and T₃ levels in the critically ill patient, free hormone levels have been reported to be normal or even elevated,^30,³¹ providing a possible explanation for why most patients appear eumetabolic despite thyroid hormone levels in the hypothyroid range. Thus, the low T₄ state is unlikely to be a result of a hormone-deficient state and is probably more of a marker of multisystem failure in these critically ill patients.

Recovery State

As acute illness resolves, so do the alterations in thyroid hormone concentrations. This stage may be prolonged and is characterized by modest increases in serum TSH levels.⁴⁸ Full recovery with restoration of thyroid hormone levels to the normal range may require several weeks⁴⁹ or months after hospital discharge.²⁷ One study reported that 35 of 40 patients with nonthyroidal illness after coronary artery bypass grafting were able to regain normal thyroid function 6 months after surgery.⁵⁰

Treatment Of The Sick Euthyroid Syndrome

The question of whether the sick euthyroid syndrome in critically ill patients represents pathologic alterations in thyroid function that negatively impact these patients or simply reflects the multisystem failure (i.e., respiratory, cardiac, renal, hepatic failure) that occurs in critically ill patients is still debatable.⁵¹^–⁵⁴ What is not debatable is that thyroid hormone replacement therapy has not been shown to be of benefit in the vast majority of these patients in the published studies to date (Box 166-3).⁵⁴ Evidence does suggest a beneficial effect of liothyronine (L-T₃) on increasing organs available for harvest from brain-dead organ donors. While L-T₃ appears to slightly improve hemodynamic and neurohumoral parameters in patients with congestive heart failure, these benefits may represent a pharmacologic effect of T₃ rather than a physiologic replacement hormonal effect. Further, the studies involving patients with congestive heart failure are more remarkable for a lack of deleterious effect of L-T₃ treatment then for any sustained clinical benefit. However, future studies do appear to be warranted in this patient population. At the present time, in the absence of any clinical evidence of hypothyroidism, there does not appear to be any compelling evidence for the use of thyroid hormone therapy in any patient with decreased thyroid hormone parameters due to the sick euthyroid syndrome.