Adrenocortical Function

Björn P. Meij

Department of Clinical Sciences of Companion Animals
Faculty of Veterinary Medicine
Utrecht University
Utrecht, The Netherlands

Jan A. Mol

Department of Clinical Sciences of Companion Animals
Faculty of Veterinary Medicine
Utrecht University
Utrecht, The Netherlands

FIGURE 19-1 Numbers of the carbon atoms and letters designating the rings of the pregnenolone molecule (IUPAC-IUB, 1989).

In birds aldosterone and corticosterone are the main corticosteroids secreted. Zonation of the avian adrenal is less clear than in the mammalian adrenal. However, the outer subcapsular cells, looping in a manner similar to the zona glomerulosa, appear to be the predominant aldosterone secretors. The cells reaching toward the central part of the gland form corticosterone (Kime, 1987).

FIGURE 19-2 Biosynthetic pathways for adrenal steroid production. Numbers by arrows denote specific enzymes: 1 = CYP11A1; 2 = CYP17; 3 = 3β-HSD; 4 = CYP21; 5 = CYP11B1; 6 = CYP11B2.

In birds, corticosterone, a glucocorticoid with minor mineralocorticoid properties, is by far the main adrenal secretory product. In the peripheral blood of adult birds, no cortisol could be demonstrated with radioimmunoassay (Walsh et al., 1985; Zenoble et al., 1985) or HPLC analysis (Lumeij et al., 1987). In chickens, the 17α-hydrolyase activity is present in embryonic life but decreases after hatching (Carsia et al., 1987; Nakamura et al., 1978).

FIGURE 19-3 Physiological control of aldosterone secretion (Müller, 1986).

Apart from these factors, other intra-adrenal factors (Ehrhart-Bornstein et al., 1998) and environmental factors may regulate aldosterone production not only at the level of conversion of cholesterol to pregnenolone but also at the level of aldosterone synthase (CYP11B2) (Williams, 2005).

At normal concentrations, only about 10% of the total blood cortisol and corticosterone is in the free form (i.e., susceptible to ultrafiltration). At body temperature, 70% of the plasma cortisol is bound to a globulin called transcortin or corticosteroid-binding globulin (CBG). Transcortin has a high affinity for cortisol and corticosterone, but its binding capacity is limited. Another 20% of plasma cortisol is bound to albumin although its affinity for cortisol is much less than that of transcortin. In line with these percentages, in the dog the free fraction has been estimated to range from 5% to 12% (Kemppainen et al., 1991; Meyer and Rothuizen, 1993).

Transcortin is ubiquitous in mammals, but plasma concentrations vary considerably, resulting in species differences in total cortisol concentration. Most domestic animals have little corticosteroid-binding activity compared to humans (Rosner, 1969). As the free rather than the protein-bound steroid is biologically active, methods have been developed to measure free cortisol. By employing the combination of ultrafiltration and equilibrium dialysis, it was demonstrated that in dogs with portosystemic encephalopathy the associated hyperadrenocorticism is not only characterized by an increased total cortisol concentration in plasma but also by an increase in the free fraction of plasma cortisol (Meyer and Rothuizen, 1994). Between various pig breeds, a two-fold difference in plasma CBG binding capacity was found and related to increased drip loss in the Meishan breed (Geverink et al., 2006).

The physiological significance of protein binding probably lies in a buffering effect, which prevents rapid variations of the plasma cortisol level. Transcortin restrains the active cortisol from reaching the target organ and also protects it from rapid inactivation by the liver and excretion through the kidneys.

Plasma aldosterone is predominantly bound to albumin, which has a low affinity. The relatively low degree of protein binding of plasma aldosterone partially explains the very low plasma concentration and the short biological half-life of this hormone.

Only unbound cortisol and its metabolites are filterable at the glomerulus. Most of this filtered cortisol is reabsorbed, whereby a tubular maximum is only achieved at very high filtered loads of free cortisol (Boonayathap and Marotta, 1974). Less than 20% of the filtered cortisol is excreted unchanged in the urine. Nevertheless, in most mammals the kidneys account for 50% to 80% of the excretion of the metabolized steroids. The remainder is lost via the gut. To render them suitable for renal elimination, the steroids are inactivated and made more water soluble through enzymatic modifications. The liver is the major organ responsible for steroid inactivation and conjugation to form water-soluble compounds, although in the dog—contrary to humans—the kidney and the gastrointestinal tract also contribute to the metabolic clearance of cortisol (McCormick et al., 1974). In the canine, kidney cortisol glucuronide is both secreted and reabsorbed, without a tubular maximum or a plasma threshold (Boonayathap and Marotta, 1974).

FIGURE 19-4 Metabolism of cortisol in the dog (simplified).

Apart from this 11β-hydroxylation, cortisol metabolism in the dog involves the following: (1) reduction of ring A to tetrahydro derivatives, (2) reduction of the 20-keto group to a hydroxyl, and (3) conjugation with glucuronic acid to form glucuronides (Gold, 1961). In addition, unconjugated 20-hydroxycortisol/cortisone has been found in canine urine.

In total glucuronide fraction of urinary corticoids in dogs, at least steroids reduced at C-20 represent 60%. This is of prime importance when it comes to assessing adrenocortical function by measuring urinary cortisol metabolites. Measurements directed at steroids containing a 17α,21-dihydroxy, 20-keto arrangement (i.e., the Porter-Silber reaction) will detect only a small part of the cortisol metabolites and therefore have limited value (Siegel, 1965). Instead, preferably, measurements are preformed involving reduction of the urine metabolites at C-20, followed by oxidation to 17-ketosteroids, which are then quantitated. Further details on this measurement of total 17-hydroxycorticosteroids are given in Section IV.B.3.

Aldosterone is converted not only to tetrahydroaldosterone-3-glucuronide but also to aldosterone-18-glucuronide. Most of aldosterone metabolism takes place in liver and kidney, but the intestine and spleen might also contribute to a minor degree (Balikian, 1971).

In domestic animals, the catabolism of androgens has not been studied in any detail. For the dog, it is known that little of the secreted androgens can be measured as 17-ketosteroids in urine (Rijnberk et al., 1968b; Siegel, 1967). Probably most of the excretion occurs via the bile.

In the cat, glucocorticoid excretion is also largely biliary (Rivas and Borrell, 1971; Taylor, 1971). Feline liver function has several specific features, including a relatively low glucuronyl-transferase activity. Hence, formation of glucuronide conjugates is low, and the conjugates are mainly sulphates, which are mostly excreted via the biliary route. Despite this low urinary excretion of metabolites and conjugates, it was shown that renal excretion of free cortisol is sufficient for diagnostic purposes (Goossens et al., 1995) (see also Section IV.B).

Apart from the use of urinary excretion to examine the glucocorticoid status, fecal glucocorticoid metabolite measures are increasingly being used, especially for zoo and wildlife animals. Interpretation of fecal glucocorticoids may be confounded by a large variety of factors that should be taken into account (Millspaugh and Washburn, 2004).

1. Glucocorticoids

FIGURE 19-5 Structure-function relationships of corticosteroids. The bold lines and letters indicate the structure common to all glucocorticoids. Substituents that may enhance glucocorticoid activity are presented by light lines and letters.

In their target organ cells, glucocorticoids act in the same manner as other steroid hormones: diffusion through the cell membrane; binding to the cytoplasmic, glucocorticoid (GR), and mineralocorticoid (MR) receptors; translocation of the hormone receptor complex to the nucleus; and subsequent stimulation of the synthesis of specific RNAs leading to synthesis of specific enzymes. The latter include key enzymes in gluconeogenesis such as fructose-1,6-disphosphatase, glucose-6-phosphatase, and pyruvate carboxylase. In the dog, corticosteroid excess also results in induction of an isoenzyme of alkaline phosphatase. This corticosteroid-induced form of alkaline phosphatase has not yet been found in other species. The marked stability at 65°C allows easy quantitation in plasma as a diagnostic screening tool for cases of suspected hyperadrenocorticism (Teske et al., 1986, 1989).

The overall effect of glucocorticoids on metabolism is to supply glucose to the organism by the transformation of proteins. This occurs via the above-mentioned induction of gluconeogenic enzymes in the liver. Thus, glucocorticoids divert metabolism from a phase of growth and storage toward increased physical activity and energy consumption, whereas chronic excess leads to catabolic effects such as muscle wasting, skin atrophy, and osteoporosis. The tendency to hyperglycemia is opposed by increased secretion of insulin, which in turn tends to enhance fat synthesis. This along with the increased food intake owing to central appetite stimulation (Debons et al., 1986) explains the (centripetal) fat deposition, manifested by abdominal enlargement.

In situations of glucocorticoid deficiency, water excretion is impaired, whereas glucocorticoid excess may result in polyuria, being most pronounced in the dog. This is in part due to antagonism of cortisol to the action of vasopressin. In addition, glucocorticoid excess causes loss of the sensitivity of the osmoregulation of vasopressin release (Biewenga et al., 1991). Even physiological increases in cortisol may inhibit basal vasopressin release in dogs (Papanek and Raff, 1994).

Glucocorticoids have long been known to have effects on blood cells, including a reduction in the numbers of eosinophils and lymphocytes and an increase in the number of neutrophils and hence in the total number of leukocytes. Glucocorticoid deficiency leads to normochromic, normocytic anemia.

As far as effects on other endocrine glands are concerned, it is shown that canine hyperadrenocorticism results in reversible suppression of growth hormone secretion (Peterson and Altszuler, 1981). The frequently observed lowering of circulating thyroxine concentrations has been ascribed to changes in the thyroid hormone binding capacity of the plasma and to inhibition of lysosomal hydrolysis of colloid in the thyroid follicular cell (Kemppainen et al., 1983; Woltz et al., 1983). Glucocorticoids have multiple effects on peripheral transfer, distribution, and metabolism (Kaptein et al., 1992), but thyroid function does not seem to be affected (Rijnberk, 1996). The glucocorticoid prednisone was found to inhibit LH secretion in male dogs, leading to reduced testosterone concentrations (Kemppainen et al., 1983).

2. Adrenal Androgens

In health, adrenocortical production of androgens is trivial in comparison with the production of these hormones by the gonads. However, a pathological excess of adrenal androgens might induce virilization (i.e., the development of masculine secondary sex characteristics), which would be most noticeable in the female or immature male. So far, no virilization as a consequence of enzyme deficiency has been reported in domestic animals, but in equine hyper-adrenocorticism, hirsutism is a common feature (Pauli et al., 1974; Van der Kolk et al., 1993).

3. Mineralocorticoids

The widespread MR has equal affinity for aldosterone and the glucocorticoids cortisol and corticosterone, whereas the latter two hormones circulate at much higher concentrations than aldosterone. However, in the classical aldosterone targets (kidney, colon, and salivary gland), the enzyme 11β-hydroxysteroid dehydrogenase (11βHSD) converts cortisol and corticosterone (but not aldosterone!) to the 11-keto analogues. These analogues cannot bind to MR, thereby enabling aldosterone to occupy this receptor (Funder, 2005). Thus, not the receptor but a prereceptor modification provides the tissue specificity.

In domestic animals and certainly in the dog, initially there has been some controversy as to the occurrence of circadian variation in cortisol concentrations. Later definite evidence for episodic but not circadian fluctuations in plasma cortisol concentrations in dogs was presented (Kemppainen and Sartin, 1984a). In the horse (Dybdal et al., 1994) and the pig (Benson et al., 1986; Bottoms et al., 1972; Janssens et al., 1995b; Larsson et al., 1979) as well as in sheep (Fulkerson and Tang, 1979), bulls (Thun et al., 1981), pigeons (Joseph and Meier, 1973; Westerhof et al., 1994), and chickens (Lauber et al., 1987), the occurrence of circadian variations is generally acknowledged. Also in the cat, there have been reports on the occurrence of an (opposite) diurnal rhythm. However, in later reports this was not confirmed (Johnston and Mather, 1979; Leyva et al., 1984). As in the dog, the secretion is episodic without evidence for a diurnal rhythm (Peterson and Randolph, 1989). Apart from relatively short-term changes such as episodic secretion and diurnal variation, plasma cortisol concentration may change with age and during the estrus cycle, during pregnancy, and around the time of parturition. For example, basal cortisol concentrations in puppies 8 weeks of age or younger are lower than in mature dogs, most probably because of reduced binding to plasma proteins rather than to altered secretion patterns (Randolph et al., 1995). In gilts a discrete cortisol peak occurs during the early follicular phase of the sexual cycle, coinciding with the decline in plasma progesterone levels (i.e., at 4 days before the plasma LH surge) (Janssens et al., 1995c). Plasma cortisol concentrations of pregnant cows decrease significantly during the fourth month of pregnancy. Thereafter they remain fairly constant until the fifth and sixth day before parturition, when a sharp rise is seen, and also at around 24 h before parturition plasma corticoids increase again (Eissa and El-Belely, 1990).

A highly significant increase in basal plasma cortisol concentrations occurs with aging in dogs (Goy-Thollot et al., 2006) with no differences in ACTH-stimulated plasma cortisol concentrations. This confirms earlier studies (Rothuizen et al., 1993) where the increased basal activity of the hypothalamo-pituitary-adrenal axis is attributed to a decrease in limbic type-I MR concentrations, whereby no differences were found for the type II GR in canine brain structures. The increase in pituitary type II GR with aging explains the intact feedback in the aging dog (Rothuizen et al., 1993). The ACTH-stimulated increase in plasma aldosterone concentrations was found to be significantly reduced with aging (Goy-Thollot et al., 2006).

Independent of these physiological variations, it is clear that stress may activate the pituitary-adrenocortical system. Factors such as housing (Carlstead et al., 1993; Cockram et al., 1994; Koelkebeck and Cain, 1984), lactation (Gwazdauskas et al., 1986), exercise (Dybdal et al., 1980; Foss et al., 1971; Rossdale et al., 1982; Sloet van Oldruitenborgh-Oosterbaan et al., 2006), surgery (Robertson et al., 1994), anaesthesia, heat (Gould and Siegel, 1985), emotional strain (Bobek et al., 1986; James et al., 1970; Kirkpatrick et al., 1977), food deprivation (Messer et al., 1995), and anticipation of feeding (Murayama et al., 1986) all lead to increased corticosteroid secretion. In dogs, a visit to a veterinary practice, orthopedic examination, and hospitalization increased the urinary corticoid:creatinine ratio, which is a measure for adrenocortical function (Van Vonderen et al., 1998). In the horse, during exercise increased plasma renin activity results in a considerable rise in aldosterone secretion (Guthrie et al., 1980, 1982). From the lack of any increase in plasma corticosterone in homing pigeons during flight, it was concluded that the animals were not under serious stress (Viswanathan et al., 1987).

With chronic stress, such as during prolonged tethered housing of pigs, an increased responsiveness of the adrenal cortex to ACTH is induced, whereas the sensitivity of the pituitary to stimulation with corticotrophin releasing hormone (CRH) or vasopressin remains unaltered, and the challenge (nose sling)-induced ACTH response is lower than in loosely housed pigs. This indicates that during chronic stress, mitigating mechanisms become operational, most probably mediated by endogenous opioids (Janssens et al., 1995a).

According to Selye’s theory of general adaptation, corticosteroids are needed for the (metabolic and circulatory) defense reaction. Administration of cortisol to improve nonspecific resistance is indicated in established pituitary or adrenal insufficiency. From an immunological point of view, stress suppresses defense reactions. Stress-induced glucocorticoid secretion prevents the overshooting of the animal’s reactions to stress. The glucocorticoids modulate the mediators of the immune system such as the different lymphokines and mediators of the inflammatory reactions: prostaglandins, leukotrienes, kinins, serotonin, and histamine (Munck et al., 1984).

Adrenocortical hypofunction includes all conditions in which the secretion of adrenal steroid hormones falls below the requirements of the animal. It may be divided into two categories:

1. Primary hypoadrenocorticism or Addison’s disease is the result of deficiency of both glucocorticoid and mineralocorticoid secretion from the adrenal cortices (primary adrenocortical failure). “Idiopathic” or immune-mediated atrophy (Schaer et al., 1986) of all zones is the most frequently observed histopathological lesion, although the lymphocytic adrenalitis may be confined to the zona fasciculata and zona reticularis (Kooistra et al., 1995). It is well known in the dog and rare in the cat (Peterson and Randolph, 1989).

2. Secondary hypoadrenocorticism is the result of pituitary ACTH deficiency, causing decreased glucocorticoid secretion. Because of the almost unaltered aldosterone secretion, it will usually remain unnoticed. It is observed occasionally in dogs with pituitary insufficiency resulting from pituitary cystic lesions leading to an empty sella or as a result of a compressive nonfunctioning (endocrine inactive) pituitary macroadenoma.

Animals receiving long-term corticosteroid treatment, despite physical and biochemical hyperadrenocorticoid changes, develop secondary adrenocortical insufficiency because of prolonged hypothalamo-pituitary suppression. Atrophy of the two inner zones of the adrenal cortex results from the loss of endogenous ACTH stimulation. This is observed in the species in which corticosteroid therapy is common practice such as the dog (Greco and Behrend, 1995; Rijnberk, 1996) and the horse (Hoffsis and Murdock, 1970; Toutain and Brandon, 1983). It has also been observed in dogs and cats treated with progestagens, owing to the glucocorticoid activity that is intrinsic to these drugs (Chastain et al., 1981; Mansfield et al., 1986; Middleton et al., 1987; Selman et al., 1994).

In principle, the adrenal cortex of animals might, as in humans, give rise to three distinct clinical syndromes of hyperfunction: mineralocorticoid excess or hyperaldosteronism (Conn’s syndrome), glucocorticoid excess or hypercortisolism (Cushing’s syndrome), and androgen excess (adrenogenital syndrome).

Hypercortisolism is common in the dog though rare in the cat and the horse. In about 15% of hyperadrenocorticoid dogs, the disease is due to a primary adrenocortical tumor (ACTH independent). The excessive secretion of cortisol by these tumors may either be autonomous or associated with aberrant expression of ectopic or overexpression of eutopic adrenocortical receptors (Lacroix et al., 2001). For example, the expression of ectopic receptors for gastric inhibitory polypeptide may result in “food-dependent hypercortisolism.” In the other 85% of cases, the (ACTH-dependent) hypercortisolism is associated with a pituitary lesion, producing excess ACTH. The pituitary corticotroph adenomas may reside in either the anterior lobe or the intermediate lobe (Kemppainen and Peterson, 1994; Peterson et al., 1982). In rare cases, Cushing’s syndrome may also be due to ectopic secretion of ACTH by a neuroendocrine tumor outside the pituitary gland (Galac et al., 2005). In cats, hypercortisolism (Cushing’s syndrome) is usually concurrent with diabetes mellitus and may be either primary adrenocortical or pituitary dependent (Meij et al., 2001, 2004; Meijer et al., 1978b; Peterson and Randolph, 1989).

Primary hyperaldosteronism in cats may be due to an adrenocortical tumor (Ash et al., 2005; Flood et al., 1999; Mackay et al., 1999; Moore et al., 2000; Rijnberk et al., 2001) or to bilateral hyperplasia of zona glomerulosa tissue (Javadi et al., 2005). Prominent clinical features in cats with hyperaldosteronism are hypokalemia, episodic weakness, cervical ventroflexion, arterial hypertension, and vision loss resulting from retinal detachment. The nontumorous form of primary hyperaldosteronism in cats is associated with progressive renal disease (Javadi et al., 2005).

In the horse, hypercortisolism (Cushing’s syndrome) originates almost invariably in the intermediate lobe of the pituitary (Orth et al., 1982). In domestic ferrets, hyper-adrenocorticism is quite common and it is primarily associated with adrenocortical tumor or nodular adrenocortical hyperplasia (Rosenthal et al., 1993). Domestic ferrets with hyperadrenocorticism have no detectable abnormalities in plasma concentrations of ACTH or α-MSH (Schoemaker et al., 2002a) and increased urinary corticoid/creatinine ratios resistant to dexamethasone suppression (Schoemaker et al., 2004), both indicating ACTH-independent hyperadrenocorticism. Positive staining was found with the LH receptor antibody in hyperplastic or neoplastic adrenal glands of hyperadrenocorticoid ferrets supporting the hypothesis that gonadotropic hormones play a role in the pathogenesis of hyperadrenocorticism in ferrets (Schoemaker et al., 2002b).

In both hypoadrenocorticism and hyperadrenocorticism, there are a number of abnormal laboratory findings that are more common than in other diseases. Many of these abnormalities can be derived from the above-described action of glucocorticoids and mineralocorticoids (Section II.F). The essentials are summarized here.

In primary hypoadrenocorticism, the decreased aldosterone production results in hyponatremia and hyperkalemia. The associated hypovolemia leads to prerenal uremia. In hyperadrenocorticism, the classic hematological abnormality is eosinopenia, which may be associated with lymphopenia and occasionally with leukocytosis and erythrocytosis. A common biochemical abnormality is the elevated plasma concentration of alkaline phosphatase (Eckersall and Nash, 1983) that may occur in conjunction with mild elevation of alanine aminotransferase (ALT). In addition, hyperglycemia, hyperlipidemia, and low thyroxine concentrations may be observed. Urine-specific gravity is usually low, and in 5% to 10% of cases glucosuria is found.

Of these abnormalities, the elevated alkaline phosphatase (AP) concentration is the most common laboratory abnormality in dogs with corticosteroid excess (either exogenous or endogenous). This increase is due to the induction of a specific isoenzyme, which has greater heat stability at 65°C than other AP-isoenzymes (Teske et al., 1986) and is therefore easily measured by a routine laboratory procedure. An abnormally elevated AP-65°C value may point to hyperadrenocorticism, but it is unsuitable as a diagnostic test because of its low specificity (Teske et al., 1989). The low thyroxine (T4) concentrations in plasma, which may be observed in hyperadrenocorticism, seem to be a consequence of altered transport, distribution, and metabolism of T₄ rather than the result of hyposecretion (Rijnberk, 1996).

There may be other conditions causing these abnormalities, however, and some cases do not present with typical routine laboratory characteristics. Therefore, a detailed understanding of the specific laboratory tests commonly used to investigate adrenocortical function is essential. In addition, the spontaneous forms of hyperadrenocorticism may arise from pituitary ACTH overproduction or from autonomously hypersecreting adrenocortical tumors. Each of these requires a different mode of treatment, and therefore insight into the tests used in the differential diagnosis is of prime importance.

The availability of highly specific radioimmunoassays has greatly enhanced and simplified assessment of adrenocortical function. Some of the older (chemical) methods, which are now seldom used, are mentioned briefly. Methods in common use are described in more detail, which applies especially to the dynamic tests, part of which is used in the differential diagnosis (Section IV.C).

1. Cortisol Production Rate

In the radionuclide dilution method, the rate of cortisol production is determined through administration of radiolabeled cortisol and isolation of cortisol metabolites from urine that is collected over at least 24h. The specific activities of labeled cortisol and tetrahydrocortisol (THF) or tetrahydrocortisone (THE) are compared. Some preliminary work (Rijnberk et al., 1968b) indicated that this test worked for the dog and might potentially be very useful. However, the laborious character of the procedure rules out general availability.

2. Plasma Corticoids