Evaluation of Endocrine Function

Growth Hormone Deficiency

Idiopathic growth hormone deficiency is the most common cause of GH deficiency (GHD) in children, whereas a pituitary adenoma is the most common etiology in adult-onset GHD. No simple reproducible method for determining abnormal GH secretory patterns exists. In healthy normal individuals, 70% to 80% of GH results are below 1 ng/mL (<1 μg/L), and secretory peaks typically reach 20 to 40 ng/mL (20–40 g/L) ( ). Thus, in a child with decreased growth velocity, a low or nondetectable GH does not necessarily indicate GHD. Similar to GH, IGF-1 declines with age. IGF-1 is more diagnostically useful in patients younger than 40 years of age; however, it is still not sensitive enough to be used as a standalone test to make the diagnosis of GHD ( ). Manipulation of the endocrine system through stimulation and suppression of the various axes is often required for diagnosing conditions of hormonal deficiency and surfeit. In this vein, GHD is diagnosed by showing failure of GH to increase adequately in response to pharmacologic stimulation. Several endocrine organizations have published guidelines on GH deficiency, detailing which patients to test, which test to utilize, and what cutoffs to apply ( ; ). The insulin tolerance test (ITT) has long been considered the gold standard for diagnosing GHD. However, it is most unpleasant for the patient, requires the attendance of a physician throughout the testing period, and is contraindicated in those with a history of seizures or cardiac or cerebrovascular disease. Failure of GH to rise above 3 to 5 ng/mL in adults when glucose drops to <40 mg/dL ( , ) and above 10 ng/mL in children is considered to be abnormal. In patients unable to undergo an ITT and in those needing a second confirmatory test, arginine stimulation alone or in combination with GHRH is usually the next step. Combination testing with GHRH plus arginine was preferred, as many normal adults failed stimulation with arginine alone ( ), however since GHRH is no longer commercially available in the United States, the glucagon stimulation test (GST) has become the accepted alternative. The mechanism by which glucagon stimulates GH release is not well understood. In this test, glucagon is administered IM to the patient, and GH serum levels are determined every 30 minutes over a 4-hour period. GH usually peaks after 120 to 180 minutes. The initial dose given depends on the BMI. For BMI ≤25 mg/M ₂ , a dose of 1 mg is given; for BMI values >25 mg/M ₂ , a dose of 1.5 mg is given. GHD is diagnosed by failure of GH to rise above 3 μg/L (a lower cut-point of 1 μg/L in overweight/obese patients improves sensitivity and specificity) ( ). Two caveats about this test should be noted: (1) glucagon is contraindicated in patients with pheochromocytoma, and (2) no studies have been performed in patients with diabetes or pre-diabetes (see Chapter 17 ), so the diagnostic accuracy of this test in these patients has not been established ( ). Common side effects of GST included nausea, vomiting, and delayed hypoglycemia. Glucagon is contraindicated in persons with pheochromocytoma. Other pharmacologic stimuli such as L -dopa, clonidine, and arginine are weak GH secretagogues requiring even lower cutoff points for diagnosis and are not widely recommended for use in the United States ( ). No studies have been performed in persons with diabetes or prediabetes, and therefore the diagnostic accuracy in these populations remains unclear (Yuen, 2011). Other methods to assess for GHD include 24-hour or nighttime monitoring of GH, and provocative tests using clonidine or L -dopa. All of these tests carry significant individual variability. They have not been systematically studied and are used mainly in the pediatric population ( ).

Because body mass index (BMI) affects the GH response to stimulation with GHRH plus arginine, newer guidelines recommend the use of GH cutoffs based on BMI for interpretation of results ( ). There are two settings in which an argument can be made against the need to test for GHD. The first is in patients with known pituitary disease who have a low IGF-1 level. The other scenario is in patients with evidence of three or more pituitary hormone deficiencies; studies have shown these patients to have a greater than 96% chance of having GHD ( ).

Because of perceived and actual deficiencies in these tests, the ghrelin-receptor agonist test is a current avenue of investigation. Ghrelin, sometimes called the “hunger hormone,” is produced by the stomach. Release of ghrelin increases appetite, stimulates the release of GH, and regulates glucose metabolism. Ghrelin has been shown to stimulate growth hormone release through the growth hormone secretagogue receptor (GHSR)1a in the anterior pituitary and hypothalamus (Muller et al., 2015). Macimorelin acetate is an oral ghrelin receptor agonist. When administered in an oral form at 0.5 mg/kg, macimorelin is readily absorbed and stimulates GH secretion with good tolerability ( ). It has been shown to be a reproducible safe diagnostic test for adult growth hormone deficiency, with accuracy comparable to that of the ITT ( ). In a post hoc analysis, using a GH cutoff of 5.1 ng/mL, it showed 92% sensitivity and 96% specificity ( ). A GH cutoff of ≤2.8 μg/L is suggested by the AACE 2019 guidelines ( ).

For children, it is common to use exercise testing in screening for GHD. Further testing is not indicated if the results are normal. However, if GH fails to increase adequately, pharmacologic testing is warranted ( ). Healthy children may not respond to any single GH provocative test; therefore, GHD in childhood is defined by failure of serum GH to reach defined levels when at least two different pharmacologic stimuli are used. IGF-1 measurements have been used to screen for GHD in children; however, since a variety of factors can decrease IGF-1, it is not used to confirm the diagnosis of GHD. IGF-1 levels decline in malnutrition, hypothyroidism, hepatic disease, uncontrolled diabetes mellitus, renal disease, and with age. Levels of IGF-BP3 usually parallel those of IGF-1 and are often low in GHD. However, results falling within the reference range do not exclude the diagnosis of GHD. Up to 18% of subjects with GHD may have normal IGF-1 levels. Care should be taken when interpreting values for IGF-1, IGF-BP3, and GH because the assays themselves and the cutoffs used to define normal differ greatly between laboratories. To address these issues, there is a move to have laboratories utilize a World Health Organization international standard for GH and IGF-1 and to create a normative range from a statistically valid control group ( ).

It is important to understand the difference in the physiology behind testing with GHRH as compared to the other secretagogues. ITT and arginine stimulate the production of GHRH; clonidine increases GH secretion by inhibiting the release of somatostatin. For these stimuli to elicit a rise in GH, the hypothalamus must be intact. When GHRH is used for testing, the hypothalamus is bypassed and the pituitary gland is directly stimulated; therefore, it is possible to miss up to 50% of patients who have tertiary (hypothalamic) GHD.

Individuals with childhood-onset idiopathic GHD are less likely to have permanent GHD. It is recommended that unless they have a known genetic mutation, embryopathic lesion, or irreversible lesion/damage to the pituitary gland/hypothalamus, they should be taken off GH replacement and retested when they reach early adulthood ( ).

Growth Hormone Excess

Growth hormone overproduction can result in the condition called acromegaly . If the condition develops before closure of the epiphyses, these individuals may be exceedingly tall ( gigantism ). More commonly, it presents during adulthood, causing diffuse enlargement of soft tissues and organs throughout the body. Characteristic features include prognathism, frontal bossing, and spade-like hands. The screening test for clinically suspected acromegaly is a randomly collected IGF-1. A normal IGF-1 with respect to the appropriate age- and gender-matched reference ranges rules out the diagnosis of acromegaly ( ). However, it should be noted that a physiologic increase in IGF-1 may be seen during pregnancy and late-stage adolescence ( ). If the level of IGF-1 is elevated or borderline with respect to the appropriate age- and gender-related reference range, it becomes necessary to confirm the diagnosis using an oral glucose tolerance test (OGTT). The OGTT is performed by obtaining baseline blood samples for glucose and GH, administering 75 g of glucose orally, and then obtaining blood for glucose and GH every 30 minutes over the next 2 hours. A normal response is suppression of GH to <1 ng/mL (1 μg/L) at any time during the test. If GH fails to drop to below 1 ng/mL (1 μg/L), the patient is diagnosed as having acromegaly ( ; , ). GH may fail to adequately suppress in the setting of puberty, uncontrolled diabetes mellitus, malnutrition, hepatic disease, and renal disease. However, all but puberty are associated with a normal serum IGF-1 ( ).

The difficulty comes in diagnosing mild disease. Freda had studied 60 postoperative patients with acromegaly, 22 patients with active disease (elevated IGF-1), 38 patients in remission (normal IGF-1), and 25 healthy controls. The highest nadir GH was 0.13 μg/L in the controls and 0.3 μg/L in those with active disease. Of those with active disease, 50% had GH values <1 μg/L, leading to misclassification of these individuals as being normal when the criterion of a GH value <1 μg/L was applied ( ; ). In a more recent study of 16 patients with mild disease (the majority having microadenomas), Dimaraki and colleagues (2002) found that 50% of patients with acromegaly were able to suppress their GH levels to <1 μg/L following an OGTT. A proposed reason for these patients with active disease showing normal suppression is that they may have subtle changes in GH secretion that are not elicitable using this short suppression test. To address this issue, the following suggestion has been proffered ( ). If the initial random IGF-1 is normal and GH is <0.3 μg/L, then acromegaly is excluded. If either test is abnormal, repeat the IGF-1 and proceed to an OGTT. Failure of GH to be suppressed below 0.3 μg/L, accompanied by an elevated IGF-1, is diagnostic of acromegaly. Suppression of GH below 0.3 μg/L with normal IGF-1 excludes acromegaly. For those who adequately suppress their GH but have an elevated IGF-1, close follow-up is recommended ( ). Oral, but not transdermal, estrogen can reduce IGF-1 concentrations ( ); therefore, temporarily withholding oral estrogen therapy during testing may be prudent.

Acromegalic patients may show a paradoxical rise, an absent response, or a partial decline in GH in response to an OGTT ( ). Random GH sampling is not adequate to establish the diagnosis of acromegaly because considerable overlap in GH exists between healthy persons and patients with the disease. In addition, GH can be increased in patients with a number of disorders, including renal disease, cirrhosis, malnutrition, and uncontrolled diabetes mellitus, as well as in patients under physical or emotional stress. Although larger adenomas are generally associated with higher GH levels, the correlation between GH levels and acromegalic manifestations is poor. In most patients with untreated acromegaly, GH levels remain above basal levels throughout the day, ranging from 10 to 100 ng/mL (10–100 μg/L). Serum GH levels in excess of 50 to 100 ng/mL (50–100 μg/L) are virtually never seen in nonacromegalic persons ( ). Other diagnostic maneuvers using TRH, GnRH, and GHRH yield discordant results in about 50% of patients with acromegaly. Therefore, their use for diagnostic purposes or for monitoring disease activity is not recommended ( ). In patients with documented acromegaly but no evidence of a pituitary tumor, GHRH should be measured to detect a hypothalamic or ectopic source for acromegaly.

Patients treated for acromegaly require lifelong monitoring for disease activity. Patients treated with surgery or radiation can be monitored by the GH response to OGTT and a random IGF-1 measurement. The goal of therapy is to normalize IGF-1 and suppress the GH response to OGTT to <1 ng/mL (<1 μg/L). Several studies have shown that when these goals are attained, mortality rates become comparable with those of the normal population ( ; ; ; ). If an ultrasensitive assay is available, an OGTT GH nadir of <0.4 μg/L can be used. However, this lower cutoff has not been demonstrated to improve metabolic outcome or better diagnose remission (Melmed, 2018; Ku, 2016). As mentioned previously, caution is required in interpreting IGF-1 levels in women on oral but not topical estrogen treatment, as oral estrogens lower IGF-1 concentrations ( ). For patients on somatostatin receptor analogs, monitoring is performed by measuring GF-1 and GH just prior to the next treatment dose. Patients on GH receptor blockers are monitored by a random IGF-1 alone. In rare instances, acromegaly may be due to ectopic secretion of GHRH (e.g., bronchial carcinoid, pancreatic islet cell tumors, small cell lung cancer), in which case the measurement of GHRH may be useful.

Oxytocin

Oxytocin secretion is stimulated by stretching of the cervix and vagina during parturition; this is known as the Fergusson reflex. Oxytocin contributes to uterine contractions late in labor both by direct action on the myometrium and by stimulation of prostaglandin secretion by the decidua ( ). It also plays a role in hemostasis at the placental site following delivery. Throughout pregnancy, estrogen upregulates the numbers of oxytocin receptors on the uterus and decidua, resulting in increased sensitivity to oxytocin toward the end of the term.

Oxytocin stimulates the myoepithelial cells surrounding the mammary glands and lactiferous ducts to contract, resulting in milk ejection. In response to suckling, neurogenic stimuli are transmitted from the nipple to the third, fourth, and fifth thoracic nerves, to the spinal cord, to the midbrain, and on to the hypothalamus, where oxytocin release is triggered. Psychological stimuli can bypass this pathway, with bursts of oxytocin secretion occurring with anticipation of nursing or on hearing a baby cry ( ; ). Stressful situations can inhibit its secretion, resulting in decreased milk ejection.

Pathologic conditions associated with oxytocin excess or deficiency are rare and are limited to case reports. Its function in males remains unknown. Clinical demand for the measurement of oxytocin levels is extremely rare. Because it has a half-life of 3 to 5 minutes and is subject to rapid degradation by oxytocinase, individual reference laboratory procedures for its collection must be strictly adhered to if meaningful results are to be obtained.

Arginine Vasopressin/Antidiuretic Hormone

AVP (ADH) is synthesized within the paraventricular and supraoptic nuclei of the hypothalamus. The major function of ADH is to maintain osmotic homeostasis by regulating water balance. It does this by stimulating the V2 receptors on principal epithelial cells that line the cortical collecting ducts of the kidney. As discussed in Chapter 15 , ADH induces an increase in the production of cyclic adenosine monophosphate (cAMP), which, in turn, causes the water channels (aquaporins) to fuse with the apical membrane of the principal cells, increasing water permeability and reabsorption. At much higher blood concentrations, AVP also acts as a potent pressor by causing vasoconstriction (V1 receptor), stimulates ACTH secretion from the anterior pituitary gland, and stimulates the production of clotting factor VIII.

ADH secretion is modulated by changes in serum osmolality and by alterations in intravascular volume. Changes in plasma osmolality are detected by osmoreceptors located within the anterior hypothalamus. Plasma ADH rises and falls in direct proportion to the serum osmolality. A concordantly linear relationship exists between ADH and the urine osmolality (Uosm). The Uosm is capped at about 1200 mOsm/kg, the maximum Uosm allowed by the renal medullary concentration gradient.

ADH secretion is maximally stimulated at a serum osmolality greater than 295 mOsm/kg and is suppressed when the osmolality falls below 284 mOsm/kg. Thirst can be considered the fail-safe mechanism for fluid and osmotic homeostasis, as it is initiated at a greater plasma osmolality (Posm) than is needed to stimulate ADH secretion. The thirst receptors are located within the hypothalamus and are distinct from the osmoreceptors that stimulate ADH secretion.

Changes in intravascular volume are detected by baroreceptors. There are two sets of baroreceptors: low-pressure volume receptors located in the right atrium and pulmonary venous system, and high-pressure arterial baroreceptors located in the carotid sinus and aortic arch. These receptors are normally under tonic inhibition; a drop in intravascular volume removes the inhibition and results in a rise in ADH, leading to increased water reabsorption. ADH secretion is much more sensitive to changes in osmolality than to changes in intravascular volume. A 1% to 2% increase in osmolality will cause a rise in ADH secretion. A much greater stimulus, such as a 5% to 10% drop in blood volume or blood pressure, is needed before the baroreceptors trigger the release of ADH.

In addition to changes in osmolality and intravascular volume, other physiologic factors that can stimulate ADH secretion include nausea, cytokines, interleukin (IL)-6, hypoglycemia, hypercarbia, and nicotine ( ; ).

Basal plasma vasopressin normally ranges from 0.5 to 2 pg/μL. In concert with atrial natriuretic hormone, thirst, and the renin-angiotensin-aldosterone axis, ADH works to maintain blood pressure, volume, and tonicity. Accordingly, when a patient is evaluated for a disorder of water homeostasis, these multiple interactive factors must be taken into consideration. Normal plasma osmolality ranges from 280 to 295 mOsm/kg and plasma sodium from 135 to 145 mmol/L. It can be determined directly by measuring the plasma freezing-point depression or the plasma vapor pressure. It can also be determined indirectly, by measuring the plasma sodium, blood urea nitrogen (BUN), and glucose, and applying the following formula (also, see Chapter 9 ):

This equation is discussed at length in Chapter 15, Chapter 9 .

Because sodium is the major plasma solute, hyponatremia is tantamount to hypoosmolality. There is usually excellent agreement between the directly measured and calculated osmolality. However, there are two situations when this does not hold true: in pseudohyponatremia (also termed factitious hyponatremia ), and in the presence of high concentrations of other effective solutes. In pseudohyponatremia, the plasma volume is displaced by the presence of high concentrations of lipids or proteins (e.g., hyperglobulinemia), resulting in a low plasma concentration of Na+, although the activity of Na+ in water is normal. In these instances, if Na+ is directly measured with an ion-specific electrode (ISE), the activity reading is normal. In order to extend electrode life, laboratories using high-throughput analyzers generally perform ISE measurements on diluted plasma (so-called indirect ISE measurement), making the resultant plasma concentration measurement vulnerable to pseudohyponatremia. The measured Posm is unaffected by the presence of hyperlipidemia or paraproteinemia, as the concentration of solute particles per volume of fluid is unchanged.

The presence of high concentrations of osmotically active solutes—such as glucose, radiographic contrast agents, mannitol, and ethylene glycol—causes a shift of water from the intracellular to the extracellular compartment, leading to dilutional hyponatremia (this situation, though quite different from that described earlier, is also sometimes referred to as pseudohyponatremia ). Similar translocational hyponatremia occurs with the absorption of glycine during transurethral prostate resection as well as in gynecologic and orthopedic procedures ( ; ). The calculated Posm is unreliable in these situations; Posm must be measured directly instead. A difference between the calculated and measured Posm of greater than 10 mOsm/kg supports the presence of pseudohyponatremia or other osmolar solutes ( ). The most common situation causing a large “osmolal gap” is ethanol intoxication.

In hyperglycemia, the measured Na+ should be corrected by the addition of 1.6 mmol/L for each 100-mg/dL increase in serum glucose above normal (100 mg/dL). A more recent analysis suggested that this formula may underestimate the decrease in Na+ caused by hyperglycemia and supported the use of a 2.4-mmol/L correction factor ( ). True hyponatremia is present if, after the correction formula is applied, the serum Na+ is still <135 mmol/L.

During pregnancy, there is a resetting of the osmostat, with a drop in osmolality of approximately 10 mOsm/kg ( ). The precise mechanism for this is not known; however, studies have suggested a role for the hormone relaxin and for the enzyme oxytocinase. During pregnancy, there is a progressive increase in placental production of oxytocinase, which, in addition to breaking down oxytocin, breaks down vasopressin, possibly resulting in a change in the hypothalamic AVP osmolality feedback loop.

Diabetes Insipidus

Diabetes insipidus (DI) is classified as being either central (due to absent or decreased ADH secretion from the hypothalamus or neurohypophysis) or nephrogenic (due to renal resistance to the actions of ADH). Each is further subcategorized as being complete or partial. All variants are characterized by the passage of large volumes of dilute urine (>2.5 L/day) in the face of an inappropriately elevated Posm. Provided that the thirst mechanism is intact and there is free access to water, most individuals can avoid dehydration and maintain a normal Posm and serum sodium levels.

The water deprivation test is the preferred diagnostic test for identifying the presence of DI ( Table 25.2 ). Central and nephrogenic DI can be distinguished by the response to either endogenous or exogenous vasopressin ( Table 25.3 ). In neurogenic DI, ADH levels are low and the kidney rapidly acts to conserve water in response to exogenous ADH administration. In contrast, nephrogenic DI is associated with normal or increased levels of ADH, and administration of additional ADH has little or no effect on renal water reabsorption. Differentiating partial nephrogenic DI from primary polydipsia can be more difficult. Unlike DI, in which the underlying problem is difficulty in holding on to free water, in primary polydipsia the pathophysiology arises from increased intake of fluids. Primary polydipsia (PP) can be due to either dipsogenic DI or psychogenic polydipsia. In dipsogenic DI, the setpoint for ADH secretion is normal; however, a resetting of the thirst threshold occurs so that it is now below the threshold for ADH secretion. In psychogenic polydipsia, the osmostat for ADH secretion is normal. However, because of underlying psychiatric illness, the subject has a compulsion to drink excessive volumes of liquid. This intake of copious volumes of fluid as seen in PP leads to a washout in the medullary concentrating gradient within the kidney, resulting in a decreased ability to concentrate the urine in response to dehydration. Differentiating between nephrogenic DI and PP, as well as among those who have an equivocal response to ADH, is best achieved by plotting the basal and postdehydration Uosm and plasma ADH on nomograms created by Zerbe and Robertson ( Figs. 25.3 and 25.4 ) ( , ). Direct measurement of plasma ADH may be helpful in confirming the diagnosis of a pathologic ADH deficiency or excess in the presence of a known or suspected disorder of renal function or fluid and electrolyte homeostasis, or when a patient is unable to tolerate physiologic testing. In cases of congenital DI, family members who are heterozygote carriers of ADH gene mutations may be identified by demonstrating decreased ADH levels, which may be the only subclinical manifestation of ADH deficiency.

TABLE 25.2

Water Deprivation Test for Diagnosing and Classifying Diabetes Insipidus

Patients with mild polyuria may be instructed to withhold all fluid intake from 10 pm onward. For those with more severe polyuria (8 to 10 L/day), water deprivation should be started early in the morning under close observation.
1. During testing, the patient is forbidden to take in anything by mouth. 2. Obtain the following baseline parameters: urine volume (Uvol) and urinary osmolality (Uosm), plasma osmolality (Posm), and plasma sodium (PNa). Also, record the weight and blood pressure (BP)/pulse (P) seated and standing. 3. Urine and plasma are collected hourly for Uvol, Uosm, and Posm. Weight, BP, and P are also recorded. Record requests for fluid. 4. When the Uosm has plateaued (e.g., hourly increase <30 mOsm/kg for 3 consecutive hours), or when body weight has decreased by 3% to 5%, or the patient develops a >20–mmHg drop in systolic BP, obtain samples for Uvol, Uosm, Posm, PNa, and AVP (plasma). 5. Administer 1 μg of desmopressin IV or IM, or 5 μg of AVP SC Uosm; urine output and Posm are recorded at 30, 60, and 120 minutes after the injection. The highest Uosm value is used to evaluate the patient’s response to AVP.
Precautions
When possible, discontinue any medications that can influence ADH secretion. Observe for hypotension and nausea, which may stimulate ADH secretion. The patient should not be permitted to smoke during the test.
Interpretation
Normal	Final Uosm before AVP challenge is higher than Posm. Following AVP challenge, the Uosm is less than 10% higher than the maximal Uosm achieved with water restriction alone.
Neurogenic DI	Final Uosm before AVP challenge is less than Posm. Following AVP administration, there is a greater than 50% increase in Uosm.
Nephrogenic DI	Final Uosm before AVP challenge is less than Posm. Following AVP administration, there is a less than 10% increase in Uosm.
Partial central DI	Uosm may be higher than Posm following dehydration; however, there is only a 10% to 50% increase in Uosm following administration of AVP.
Partial nephrogenic DI	Uosm may be higher than Posm following dehydration; however, there is a greater than 10% increase in Uosm following administration of AVP.
Plotting basal and postdehydration Uosm and plasma ADH on the nomograms from Zerbe and Robertson will permit further distinction between partial nephrogenic DI, partial central DI, and primary polydipsia (see Figs. 25.3 and 25.4 ).

ADH, Antidiuretic hormone; AVP, arginine vasopressin; DI, diabetes insipidus; IM, intramuscular; IV, intravenous; SC, subcutaneous.

TABLE 25.3

Tests in the Differential Diagnosis of Disorders of Water Homeostasis

Disorder	BASELINE			AFTER 12-HOUR FLUID RESTRICTION			Urine Osmolality Post-AVP Challenge
	Serum Na+ and Osmolality	Urine Na+ and Osmolality	Serum ADH	Serum Na+ and Osmolality	Urine Na+ and Osmolality	Serum ADH
Normal control	N	N	N	N	High	High	Same
SIADH	Low	N–High	High	Low–N	High	High	—
Neurogenic DI	N-High	Low	Low	High	Low–N	Low	Increased
Nephrogenic DI	N–High	Low	N–High	High	Low–N	High	Same
Psychogenic polydipsia	Low–N	Low	Low	N	N–High	N–High	Same

ADH, Antidiuretic hormone; AVP, arginine vasopressin; DI, diabetes insipidus; N, normal; SIADH, syndrome of inappropriate secretion of ADH.

Due to instability of ADH (Morgenthaler, 2006), low accuracy (38%) in differentiation between DI and PP (Fenske, 2012), and the lack of a clinically useful description of the relationship between plasma AVP levels and osmolality ( ), other avenues for diagnosing DI vs. PP have been sought. Among ADH biomarkers is copeptin, the 39 amino-acid C-terminal segment of pre-pro-arginine vaspression. It is secreted in response to the same stimuli as AVP and in equimolar amounts to AVP. Its high ex-vivo stability and ease of measurement make it an excellent surrogate for AVP. The normal range for copeptin is 1.0-13.8 pmol/L, with a higher median value seen in men ( ). A baseline copeptin level of >21.4 pmol/L (without pre-thirsting) is considered diagnostic of nephrogenic DI ( ). Measurement of copeptin in conjunction with the hypertonic infusion test has been shown to be superior to the standard water deprivation test in distinguishing PP from central DI, with a diagnostic accuracy of 97% when using a cutoff of >4.9pmol/L. In complete and partial DI the copeptin is ≤4.9 pmol/L and >4.9 pmol/L in PP with a sensitivity of 93.2% and specificity of 100% (Fenske, 2018). In CKD, P(cop) and P(vp) both increased with decreasing estimated glomerular filtration rate (eGFR), but P(cop) increased much faster than P(vp) ( ). Copeptin also rises in acute illness, myocardial infarction, and stroke. Another proposed approach is copeptin measurement following stimulation by arginine infusion, which has the advantage of already being a standard clinical test used to evaluate GHD and has been shown to differentiate between DI and PP with high accuracy (95%) (Winzeler 2019).

Nephrogenic DI is either congenital or acquired. An X-linked receptor defect is the most common etiology for congenital nephrogenic DI; medications are the most common cause of acquired nephrogenic DI ( ). Medications known to be associated with acquired nephrogenic DI include lithium, demeclocycline, and methoxyflurane anesthetics. Other causes of acquired nephrogenic DI include hypercalcemia, hypokalemia, sickle cell disease, and postobstructive uropathy ( Box 25.4 ).

BOX 25.4

Causes of Diabetes Insipidus

Central

• Primary

• Familial

• Idiopathic

• Acquired

• Tumors—craniopharyngioma, pituitary tumors, metastases (i.e., lung, breast), Rathke cleft cyst, nonlymphocytic leukemia

• Granulomatous disorders—sarcoidosis, Langerhans cell histiocytosis, Wegener granulomatosis

• Traumatic—head trauma, surgery

• Infectious—tuberculosis, meningitis, encephalitis

• Vascular—cerebral aneurysm, sickle cell anemia, Sheehan syndrome

• Drugs—alcohol, diphenylhydantoin, chlorpromazine, and β-adrenergic agonists

• Other—lymphocytic hypophysitis, hypoxic encephalopathy

Nephrogenic

• Hereditary

• Mutation in the V ₂ receptor

• Mutation in the aquaporin gene

• Acquired

• Drugs—lithium, phenytoin, demeclocycline, vinblastine, cisplatin, propoxyphene, colchicine, gentamicin, amphotericin, ethanol, atrial natriuretic hormone, norepinephrine, methoxyflurane anesthetics, furosemide

• Electrolyte disorders—hypercalcemia, hypokalemia

• Systemic illnesses—sickle cell, multiple myeloma, amyloidosis, sarcoidosis, Sjögren syndrome, polycystic kidney disease

• Other—low-protein diet, postobstructive uropathy

Neurogenic DI may be congenital, the result of an autosomal-dominant mutation in the ADH signal peptide or in the exons that code for neurophysins ( ; ). More commonly, neurogenic DI is induced by lesions affecting the hypothalamus (metastatic tumors, trauma, granulomatous disorders) or medications. Drugs that may cause neurogenic DI by suppressing ADH release include phenytoin (Dilantin), chlorpromazine, and α-adrenergic agonists. It has been suspected that autoantibodies directed against the ADH-producing cells of the hypothalamus may be responsible for some cases of idiopathic neurogenic DI ( , ; ). These antibodies are detectable by indirect immunofluorescence antibody.

Another common condition in which decreased ADH levels have been found is primary nocturnal enuresis in children. Although these patients have ADH levels only slightly lower than controls (2.9 ng/L vs. 3.6 ng/L), the clinical symptoms are reliably ameliorated by vasopressin therapy ( ). The determination of ADH levels in diagnosing this condition is not always helpful because of moderate overlap between enuretic and normal children. Measuring the specific gravity in the first morning urine is more reliable for identifying children who may benefit from ADH supplementation; they will have a urine specific gravity below 1.015 ( ).

Syndrome of Inappropriate Secretion of ADH

The syndrome of inappropriate secretion of ADH (SIADH) is characterized by a euvolemic hypoosmolar hyponatremia associated with hyperosmolar urine (the result of continued inappropriate natriuresis). By definition, SIADH cannot be diagnosed until nonosmotic stimuli for ADH secretion and other pathologies that interfere with free water clearance have been excluded ( ). Physiologic triggers of ADH secretion include nausea, pregnancy, hypoglycemia, intracranial hypertension, mechanical ventilation, and hypoxia. Hypothyroidism and glucocorticoid deficiency cause a decrease in free water clearance, leading to a dilutional hyponatremia. Mineralocorticoid deficiency can lead to hyponatremia caused by increased renal sodium loss. It is therefore extremely important to test for deficiencies of these axes and to institute appropriate hormonal replacement therapy before making a diagnosis of SIADH. Other confounding factors that can hamper the diagnosis of SIADH include renal disease, cardiac disease, and medications such as diuretics. In these situations, one can cautiously perform a water load test ( Table 25.4 ). In the absence of medications or other conditions that may impair diuresis, failure to excrete 80% to 90% of the administered water load within 4 hours and to suppress the Uosm to <100 mOsm/kg is consistent with the diagnosis of SIADH.

TABLE 25.4

Water Load Test

1. Baseline: The patient must be euvolemic and have a serum Na + between 125 and 150 mmol/L and Posm >275 mOsm/kg. 2. The patient is given 20 mL of water per kilogram of body weight (max = 1500 mL) to drink over 30 minutes. 3. With the patient maintained in the recumbent position, urine output is collected every hour for the next 5 hours. Record the volume and osmolality of each specimen.
Interpretation
Normals	Excrete 80% to 90% of the administered water load within 4 hours and suppress the Uosm to <100 mOsm/kg.
SIADH	Failure to meet these criteria in the absence of medications or other conditions that may impair diuresis is consistent with the diagnosis of SIADH.

SIADH, Syndrome of inappropriate secretion of antidiuretic hormone.

In the right clinical setting, the diagnosis of SIADH often can be confidently made on the basis of serum and urine electrolyte determinations alone. A spot urine Na+ less than 30 mmol/L can generally distinguish those with hyponatremia due to volume depletion from those with SIADH, in whom urine Na+ is greater than 30 mmol/L ( ). In contrast, differentiating the patient with euvolemia and SIADH from the one with hypovolemia and concomitant renal salt wasting (e.g., salt-losing nephropathy, diuretic use) can be much more difficult. In addition to hyponatremia, both have a spot UNa+ greater than 30 mmol/L and a fractional excretion of Na+ greater than 1. The fractional excretion of sodium (FENa [%]) is calculated as follows:

$\frac{U_{\left[N{a}^{+}\right]}\times P_{\left[Cr\right]}}{P_{\left[N{a}^{+}\right]}\times U_{\left[Cr\right]}}\times 100$

where U [Na ⁺ ] = urine sodium concentration, P [Na ⁺ ] = plasma sodium concentration, U [Cr] = urine creatinine concentration, and P [Cr] = plasma creatinine concentration.

One way to distinguish between the two is to administer 1 L of 0.9% NaCl intravenously over 24 hours for 2 days. A rise in serum Na+ by more than 5 mmol/L is suggestive of hypovolemia. In SIADH, no change or an increase of less than 5 mmol/L occurs. Alternatively, fluid restriction to 600 to 800 mL/day for 2 to 3 days will lead to improvement in hyponatremia in SIADH but not in renal salt wasting (see Table 25.3 ).

The cause of renal salt wasting in SIADH is twofold. At the onset of the disease, there is volume expansion, which inhibits the renin-aldosterone axis and leads to greater improved delivery of sodium to the distal tubules. The continued sodium loss is due to secretion of atrial natriuretic protein ( ), which also is stimulated by increased intravascular volume.

Because of hyperfiltration and the local actions of ADH on the renal V2 receptor, SIADH is one of the two states in which some of the lowest uric acid levels occur; the other is pregnancy. In SIADH, the serum levels of ADH may be variably increased, usually out of proportion to Posm. However, approximately 20% of cases meeting the physiologic diagnosis of SIADH will not have detectable levels of ADH ( ). This may be due to insensitivity of the assay to low levels of ADH, increased renal sensitivity to ADH, or perhaps the presence of another hormone with antidiuretic activity ( ).

The diagnosis of SIADH is largely a diagnosis of exclusion, with the clinician having ruled out other causes of hyponatremia ( Fig. 25.5 ). It most often occurs as a manifestation of the paraneoplastic syndrome. However, it may also occur in central nervous system trauma or infection, in lung disease, or as the result of medications (vinca alkaloids, tricyclic antidepressants, selective serotonin reuptake inhibitors; Box 25.5 ).

BOX 25.5

Causes of SIADH

• CNS disease

• Neoplasm, infection, trauma, cerebrovascular accident

• Oat cell carcinoma and adenocarcinoma of the lung, pancreatic cancer, lymphoma

• Pulmonary infection

• TB, pneumonia, positive-pressure ventilation

• Idiopathic

Medications

• Oral hypoglycemics—chlorpropamide, tolbutamide

• Antineoplastics—vincristine, cyclophosphamide

• Diuretics—hydrochlorothiazide

• Psychotropics—amitriptyline, phenothiazines, SSRIs, monoamine oxidase inhibitors

• Other—morphine, barbiturates, clofibrate, nicotine, acetylcholine, anesthetic agents, β-adrenergic stimulants, metoclopramide, desmopressin

CNS, Central nervous system; SIADH, syndrome of inappropriate secretion of antidiuretic hormone; SSRIs, selective serotonin reuptake inhibitors.

Thyrotropin-Releasing Hormone

TRH is a modified tripeptide (pyroglutamyl-histidyl-proline-amide), released from the prepro molecule by a peptidase ( ). TRH is found in the hypothalamus, brain, C cells of the thyroid gland, δ cells of the pancreas, myocardium, prostate, testis, and spinal cord. The neuron bodies producing TRH are innervated by catecholamines, leptin, and somatostatin-containing axons, which ultimately influence the rate of TRH synthesis.

Thyroid hormone production is self-regulated through feedback inhibition at the level of the hypothalamus (TRH) and pituitary (TSH). TRH also affects the production of other pituitary hormones, for example, it stimulates the release of prolactin from the lactotrophs. Leptin plays a significant role in the regulation of the TRH gene ( ), affecting the individual’s appetite for food intake. The precise interaction of leptin and the thyroid axis remains to be elucidated.

It is difficult to develop a specific antibody for TRH, and assays are not clinically useful.

Thyroid-Stimulating Hormone

TSH is a glycoprotein consisting of two mono-covalently linked α and β subunits. The α subunit shares the same amino acid sequences as LH, FSH, and human chorionic gonadotropin (hCG). It is the β subunit that carries specific information to the binding receptors for expression of hormonal activities.

The radioimmunoassay for measuring TSH was first developed by Odell and colleagues in 1965. This first-generation assay did not have sufficient sensitivity to distinguish normal TSH levels from the suppressed levels seen in primary hyperthyroidism. By the mid-1980s, a “sensitive” or second-generation immunometric TSH method using monoclonal or polyclonal antibodies was developed; sensitivity was improved to 0.1 to 0.2 mIU/L. Refinements in the immunometric method led to the third-generation assays now in common use, which allow measurement down to about 0.005 mIU/L. Fourth-generation TSH assays have been developed displaying sensitivities as low as 0.0004 mIU/L. However, they are not widely available and they fail to provide clinically useful information above and beyond that offered by third-generation assays. The American Thyroid Association (ATA) recommendations state that third-generation assays should be able to quantitate TSH in the 0.010- to 0.020-mIU/L range on an interassay basis with a coefficient of variation of 20% or less. In reporting assay results, the ATA recommends using functional sensitivity, defined as the point at which the interassay precision has a coefficient of variation equal to or less than 20% ( ).

Owing to improvements in the sensitivity of the TSH assay and the fact that TSH has an inverse log/linear relationship with FT ₄ (e.g., small changes in FT ₄ lead to large changes in TSH), this test alone can identify virtually all instances of hyperthyroidism and hypothyroidism. Exceptions exist in individuals with structural hypothalamic or pituitary disease, thyroid hormone resistance, disruption in HPT axis function due to medications, or when results are factitious as a result of substances causing assay interference (e.g., biotin, heterophile antibodies). Improved sensitivity of the third-generation assays facilitates distinguishing profound TSH suppression typical of frank Graves thyrotoxicosis (TSH <0.01 mIU/L) from modest degrees of TSH suppression (0.01–0.1 mIU/L) observed with mild (subclinical) primary hyperthyroidism and some cases of nonthyroidal illness. Although assay sensitivity has improved, the normal range has remained unchanged at approximately 0.35 and 5.0 mIU/L for most laboratories.

Serum TSH levels can be assayed using immunoassays and by mass spectroscopy, the latter being considered to be the gold standard with which all assays are compared. Reference intervals (RI) are specific for each methodology and manufacturer; this can prove problematic as patients move between doctors and different labs. To that end, the International Federation of Clinical Chemistry (IFCC) has formed a Committee for Standardization of Thyroid Function Tests. They are in the later stages of creating a process that harmonizes and standardizes TSH and FT ₄ measurements. Until this is accomplished, it is recommended, when possible, to longitudinally adhere to using one particular lab for monitoring patients with thyroid disease ( ; ).

In hyperthyroidism, TSH is suppressed while FT ₄ and/or FT ₃ are elevated, while in hyperthyroidism caused by a TSH-producing tumor or pituitary resistance, the elevated FT ₄ and/or FT ₃ are accompanied by an inappropriately normal to slightly elevated TSH. In subclinical hyperthyroidism, the TSH is low, while both the FT ₄ and FT ₃ are in the normal range ( Table 25.6 ).

In primary hypothyroidism, in which the problem lies with the thyroid gland itself, the TSH elevation is accompanied by low FT ₄ . In the setting of secondary (pituitary) and tertiary (hypothalamic) hypothyroidism the TSH is inappropriately normal for the low levels of T ₄ and T ₃ . As the thyroid gland fails, the hypothalamus and pituitary compensate by increasing the synthesis and secretion of their respective hormones. This results in a state termed subclinical hypothyroidism in which the TSH is elevated while the T ₄ , T ₃ , and FT ₄ remain normal.

An important cause of both increased and decreased TSH is nonthyroidal illness (NTI). Patients with NTI tend to have low TSH levels during the acute phase of their illness, rising to within or above the reference range with resolution of the underlying process and returning to normal once the acute illness has resolved. The interpretation of thyroid function tests (TFTs) in this setting is often confounded by the use of medications such as opioids, glucocorticoids, and dopamine, which suppress TSH secretion. New to this list of medications are checkpoint and tyrosine kinase inhibitors, which can cause secondary hypothyroidism (hypophysitis) as well as primary thyroid dysfunction.

The TSH-centered strategy has two primary limitations. First, it assumes that hypothalamic-pituitary function is intact and normal. Second, it assumes that the patient is stable (i.e., the patient has not had recent adjustments in therapy for hyperthyroidism or hypothyroidism) ( ). If either of these criteria is not met, the serum TSH result can be misleading (see Table 25.6 ). In these instances, measurement of FT ₄ can often help clarify the diagnosis. In situations in which the TSH is low and the FT ₄ or TT ₄ is normal, measurement of the total T ₃ (TT ₃ ) or FT ₃ is fitting, as the patient may have T ₃ thyrotoxicosis.

Thyroxine

After release from the follicles, T ₄ binds to various proteins in the blood (thyroid-binding globulin, albumin, thyretin). T ₄ can be measured by immunoassay after the hormone is separated from the carrier protein. The reference range is 5 to 12.5 μg/dL in adults, with slightly lower results for certain pediatric age groups. Although TSH is the principal test for assessing thyroid function, T ₄ measurements are often used along with TSH and can be of importance in interpreting TSH results. The combination of a low T ₄ and increased TSH is indicative of primary hypothyroidism, whereas elevated T ₄ and/or T ₃ combined with a decreased TSH is characteristic of primary hyperthyroidism. Hyperthyroidism, in which the T ₄ is elevated while the serum T ₃ is at or below the reference range, is termed T ₄ thyrotoxicosis . This can occur in patients with iodine-induced thyrotoxicosis (e.g., patients on amiodarone or following iodine contrast media). A suppressed TSH in association with a normal to low-normal T ₄ and a high T ₃ characterizes T ₃ thyrotoxicosis . This is more common in a toxic nodule and the elderly (see Table 25.6 ).

Severe NTI is associated with low T ₄ and T ₃ levels. This so-called low T ₄ and T ₃ syndrome, or euthyroid sick syndrome, is felt to be an adaptive response to reduce metabolic demands and conserve protein stores ( ) and is associated with a poor prognosis ( ). It is thought to arise from a maladjusted central inhibition of TRH ( ; ) ( Fig. 25.7 ).

Euthyroid hyperthyroxinemia is distinguished by the presence of an increased serum T ₄ in association with a normal TSH in an otherwise euthyroid individual. This entity has a variety of causes, including increased synthesis of binding proteins, as seen with certain drugs (e.g., estrogen) or medical conditions (e.g., liver disease), and decreased conversion of T ₄ and T ₃ (e.g., high doses of β blockers or steroids). It is also seen in patients who are acutely hospitalized for psychiatric illness, in those with familial dysalbuminemia, and in thyrotoxic patients with concomitant NTI. When medications or other factors cause increased protein binding, there is a consequent increase in serum total T ₄ ; conversely, a decrease in binding capacity leads to a decrease in serum total T4. These perturbations do not have any physiologic effect in persons with a normal thyroid axis, as the bioactive, free hormone fraction remains unaffected. In these situations, free T ₄ correlates better with thyroid functional status than does serum total T ₄ .

The T ₃ uptake (T ₃ -UP) test is a classical method of adjusting the total T ₄ measurement for alterations in binding protein. The T ₃ -UP test yields a relative measurement of unoccupied T ₄ binding sites. It was performed by saturating available binding sites with radiolabeled T ₃ , then measuring the unbound T ₃ by adsorbing it onto a resin (hence the term T ₃ resin uptake). Procedures giving analogous results can be performed via nonisotopic methods on automated analyzers. The free thyroxine index (FTI) is then calculated as (T ₃ -UP of the patient)/(mean T ₃ -UP of the reference population) multiplied by the patient’s total T ₄ . The FTI (reference range, 5.4–9.7) is only a relative estimate of the free T ₄ concentration. In general, during thyrotoxicosis, both total T ₄ and T ₃ -UP are elevated. Conversely, in hypothyroidism, both total T ₄ and T ₃ -UP are decreased and, when there is a binding globulin abnormality, these two values diverge (e.g., total T ₄ is elevated and the T ₃ -UP is decreased or the total T ₄ is decreased and the T ₃ -UP is increased). The FTI has been largely replaced by the direct measurement of free T ₄ by immunoassay, dialysis, and ultrafiltration.

Free Thyroxine

FT ₄ is the biologically active fraction of T ₄ . Equilibrium dialysis is the traditional reference technique for measuring FT ₄ . It is unaffected by changes in binding protein concentration. It is carried out by dialyzing serum against a nonprotein buffer using a membrane impermeable to proteins so that FT ₄ will achieve equal concentrations on both sides of the membrane. The protein-free solution is then collected and analyzed for FT ₄ . Even this laborious method does not give a true measure of FT ₄ in all cases because small molecules that may affect T ₄ binding are diluted in the procedure. Refinement of this technique to include tandem mass spectroscopy (ED/MS-MS), has made this the new gold standard for measuring several hormones, including FT ₄ .

ED/MS-MS is costly, time-consuming, and ill-suited to meet the high-volume demands of clinical laboratories. Immunoassays performed on automated clinical chemistry platforms can rapidly provide appropriate results in most clinical circumstances and have become the mainstay for measuring all thyroid hormones. There are various circumstances in which immunoassays can be unreliable. Spurious results can occur with abnormal variants of the binding proteins (e.g., familial dysalbuminemic hyperthyroxinemia, a condition with high T ₄ but normal FT ₄ , caused by an albumin variant that binds T ₄ abnormally) or in the presence of T ₄ or T ₃ autoantibodies ( ). Increases or decreases in serum FT ₄ without concomitant changes in metabolic state have been reported in pregnancy, NTI, and in patients being treated with certain medications. For example, prolonged administration of phenytoin or carbamazepine results in a 15% to 30% decrease in both serum T ₄ and FT ₄ . In these circumstances, dialysis or ultrafiltration methods are indicated.

Biotin, a commonly used supplement, has recently been documented to interfere with immunoassays that use streptavidin and biotinylated antibodies. Depending on whether a competitive or sandwich technique is used, the measured FT ₄ or TSH will be falsely elevated or falsely low, respectively. Options to remedy this include having the patient stop biotin-containing supplements 3 to 5 days prior to testing or referring to a laboratory that does not use a streptavidin-biotin methodology ( ).

The RI for FT ₄ is highly assay specific. In addition, RIs vary throughout pregnancy, infancy and childhood. It is therefore important to use the proper RI corresponding to each trimester of pregnancy and age range for infants and children. As with TSH, a process to harmonize FT ₄ assays using the FT ₄ obtained by ED/MS-MS as the universal standard, is currently under development ( ). Until this standardization is finalized, it is recommended that one maintain consistency in the commercial lab and methodology utilized for measuring FT ₄ and TSH.

Triiodothyronine

T ₃ , generally measured by immunoassay, has a reference interval typically in the range of 60 to 160 μg/dL (0.9–2.46 nmol/L). T ₃ is much less bound to protein than is T ₄ ; thus, a relatively greater proportion of T ₃ than T ₄ exists in a free, diffusible state. Since assays for FT ₃ are less widely available and less widely standardized, the current practice is to measure total T ₃ as opposed to FT ₃ . Total T ₃ concentrations are measured by competitive immunoassay methods that are now mostly nonisotopic and use enzymes, fluorescence, or chemiluminescence molecules as signals ( ). The measurement of T ₃ is useful in confirming the diagnosis of hyperthyroidism when TSH is suppressed but there is only minimal to no elevation in T ₄ .

In 90% of patients with hyperthyroidism, both T ₄ and T ₃ are increased, with the increase in T ₃ usually greater than that of T ₄ . Serum T ₃ can be normal or low in patients with hyperthyroidism in the setting of NTI or if they are taking medications that decrease the conversion of T ₄ to T ₃ (e.g., high-dose propranolol, amiodarone). Hyperthyroidism in association with an elevated T ₃ but normal T ₄ and FT ₄ is termed T ₃ thyrotoxicosis ( ). Patients with this syndrome are a heterogeneous group with no distinctive signs or symptoms. Although most patients with T ₃ thyrotoxicosis have Graves disease, it may also be seen with toxic multinodular goiter or toxic adenoma. T ₃ thyrotoxicosis has been reported to occur in ≈1% to 4% of all thyrotoxic patients, with a much higher incidence in regions with iodine deficiency.

The measurement of T ₃ is generally not useful in evaluating patients suspected of having hypothyroidism as levels fall within the RI in 15% to 30% of patients ( ). However, decreased T ₃ levels are seen in patients with severe hypothyroidism, whose T ₄ < 2 μg/dL (32 nmol/L) ( ). In acute NTI, such as myocardial infarction, T ₃ declines rapidly to about 50% of the reference value within a matter of days ( ). T ₃ concentrations are also low in cord blood but increase rapidly during the first few hours of life.

Blood concentrations of T ₃ rise rapidly following the ingestion of T ₃ -containing preparations (e.g. Cytomel, Amour thyroid), peaking within 2 to 4 hours. Clearance is also rapid, with a half-life of 0.68 to 1.38 days ( ). As a result, knowing both the time of hormone administration and consistency of blood test timing in relation to hormone ingestion is essential for proper interpretation of serum T ₃ results. In contrast, the T ₃ level is stable in patients treated with T ₄ (levothyroxine). As in this situation, T ₃ is derived from the peripheral conversion of T ₄ ; stable levels are reached after only a few weeks of therapy.

A number of pharmacologic agents interfere with the conversion of T ₄ to T ₃ (i.e., glucocorticoids, amiodarone, large dosages of propranolol), causing low levels of T ₃ , low FT ₃ , and normal or even high levels of serum T ₄ and FT ₄ , with normal TSH.

Reverse Triiodothyronine (rT ₃ )

rT ₃ is a biologically inactive metabolite of thyroxine that is produced through 5-deiodination of T ₄ . T ₃ and rT ₃ tend to vary reciprocally. In NTI, rT ₃ is elevated while T ₃ is diminished. rT ₃ is increased in healthy newborns, in patients with hyperthyroidism (including factitious hyperthyroidism), and in patients taking certain drugs, such as amiodarone and propranolol. Measurements of rT ₃ are of little clinical utility and rarely if ever indicated.

Thyroglobulin

Tg is synthesized and secreted by thyroid follicular cells. It is present in the serum of normal individuals in the range of up to about 30 ng/mL (45 pmol/L). The serum Tg concentration reflects thyroid mass, thyroid injury, and TSH receptor stimulation ( ). Tg is increased in a variety of disorders, including Graves disease, thyroiditis, and nodular goiter. The routine measurement of Tg is never indicated. Its utility resides in monitoring for recurrence of certain variants of thyroid cancer, in the diagnosis of thyroid dysgenesis in congenital hypothyroidism, and in helping to distinguish thyrotoxicosis factitia from other forms of thyrotoxicosis.

While Tg is helpful in monitoring for recurrence/persistence of well-differentiated thyroid carcinoma, it is of no utility in patients with undifferentiated thyroid malignancies or medullary thyroid cancer. The clinical value of Tg measurement is limited in the presence of Tg antibodies (Tg Abs), found in 20% to 25% of patients with thyroid carcinoma. Tg Abs can interfere with Tg measurements by immunoassay, resulting in falsely low serum Tg concentrations ( ). It is therefore extremely important to always test for Tg Abs whenever measuring Tg. LC/MS-MS may eventually allow for accurate measurement of Tg in the presence of Tg Abs ( ). Even though the presence of Tg Abs may invalidate the value of Tg in monitoring recurrence and response to treatment in differentiated thyroid cancer, because Tg Ab reflects thyroid tissue mass, Tg Ab itself may be used as a surrogate marker ( ; ; ).

Recombinant human TSH (rhTSH) is commonly used to test patients with thyroid cancer for the presence of residual/recurrent disease. This is more practical, more convenient, and markedly less unpleasant for the patient than stopping thyroid hormone suppression over the span of several weeks to increase the TSH level; an elevated TSH is required for radioactive iodine (RAI) uptake to occur in remaining remnant or recurrent tumor. Measurement of Tg 72 hours following the last dose of rhTSH correlates highly with the presence of recurrence; some authorities support the use of this test alone without a concomitant whole-body scan (WBS) for iodine for those with “low-risk” tumors (<40 years of age, no distant metastases, size <1 cm, well differentiated). All other patients with thyroid cancer need to undergo WBS and measurement of Tg, which can either be accomplished using rhTSH or thyroid hormone withdrawal. Using a highly sensitive assay for Tg in patients who have undergone remnant ablation, a post-rhTSH Tg > 0.5 to 1.0 ng/mL, or Tg > 0.1 ng/mL while on suppression is suspicious for recurrence/persistence of disease and needs to be worked up accordingly (i.e., full restaging while hypothyroid off of suppressive therapy) ( ). Poorly differentiated tumors can cease production of Tg, in which case other modalities for monitoring—such as positron emission tomography scan and thyroid ultrasound—may be utilized.

In the early weeks post-thyroidectomy, serum Tg levels typically fall, with a half-life of 2 to 4 days. The pattern of change in serum Tg while on T ₄ suppression is a better indicator of a change in tumor burden than any single serum Tg value ( ). Because of inter-assay variability, it is extremely important to use the same laboratory for monitoring TSH, Tg, and Tg Ab.