Chapter 8 The adrenal glands
The adrenal glands have two functionally distinct parts: the cortex and the medulla. The adrenal cortex is essential to life; it produces three classes of steroid hormone: glucocorticoids, mineralocorticoids and androgens. The medulla, which is functionally part of the sympathetic nervous system, is not essential to life, and its pathological importance is related mainly to the occurrence of rare catecholamine-secreting tumours.
Glucocorticoids, of which the most important is cortisol, are secreted in response to adrenocorticotrophic hormone (ACTH), which is itself secreted by the pituitary in response to the hypothalamic corticotrophin releasing hormone (CRH). Cortisol exerts negative feedback control on ACTH release through inhibiting the action of CRH; it also inhibits CRH secretion. Glucocorticoids have many actions (Fig. 8.1) and are particularly important in mediating the body’s response to stress. Corticosterone, a precursor of aldosterone, is a weak glucocorticoid (30% of the activity of cortisol).
The most important mineralocorticoid is aldosterone. This is secreted in response to angiotensin II, produced as a result of the activation of the renin–angiotensin system by a decrease in renal blood flow and other indicators of decreased extracellular fluid (ECF) volume (Fig. 8.2). Secretion of aldosterone is also directly stimulated by hyperkalaemia. The main action of aldosterone is to stimulate the reabsorption of sodium and the excretion of potassium and hydrogen ions in the distal convoluted tubules of the kidneys; its effect on sodium results in its having a central role in the determination of the ECF volume. ACTH does not have a major physiological role in aldosterone secretion, although it has a role in its synthesis through stimulating cholesterol desmolase, the first step in the biosynthetic pathway of the adrenal steroids. Curiously, the secretion of aldosterone by adrenal tumours is affected by ACTH (see p. 149). 11-Deoxycorticosterone and corticosterone also have mineralocorticoid activity. Cortisol has as high affinity for mineralocorticoid receptors, as does aldosterone, and its concentration in the blood is considerably higher, but renal tubular cells contain 11β-hydroxysteroid dehydrogenase, which converts cortisol to cortisone. The latter has low affinity for mineralocorticoid receptors, thus allowing these to respond primarily to aldosterone and not be overwhelmed by cortisol.
Figure 8.2 Stimulation of aldosterone secretion through activation of the renin–angiotensin system. Renin, released into the plasma from the juxtaglomerular cells of the kidney in response to various stimuli, catalyses the formation of angiotensin I from angiotensinogen, an α2-globulin. Angiotensin I is metabolized to an octapeptide, angiotensin II, by angiotensin-converting enzyme during its passage through the lungs. Angiotensin II stimulates the release of aldosterone from the adrenal cortex; it is a powerful pressor agent and also stimulates thirst and the secretion of vasopressin.
The adrenal cortex is also a source of androgens, including dehydroepiandrosterone (DHEA), DHEA sulphate (DHEAS) and androstenedione. These stimulate libido and the development of pubic and axillary hair in females, but are weak androgens in comparison with testosterone, and have only a minor physiological function in males. The clinical effects of excessive adrenal androgens can be a prominent feature of adrenal disorders in females.
The hormones secreted by the adrenal cortex are synthesized from cholesterol by a sequence of enzyme-catalysed reactions (Fig. 8.3). An awareness of these pathways is important for the understanding of congenital adrenal hyperplasia, a group of conditions each caused by a lack of one of these enzymes.
Figure 8.3 Biosynthesis of adrenal steroid hormones. Cortisol and the androgens are synthesized in the zona reticularis and zona fasciculata of the adrenal glands. Corticosterone methyloxidase II, required for the synthesis of aldosterone, is present only in the zona glomerulosa. Androstenedione can be converted to testosterone in peripheral tissues, but in adult males, adrenal androgens and the testosterone derived from them make only a minor contribution to total androgenic activity.
Adrenal steroid hormones can all be measured by immunoassays, although some cross-reactivity occurs between some steroids in some assays (e.g. between 11-deoxycortisol and cortisol). The plasma concentrations of steroid hormones can fluctuate for various reasons and the results of single estimations must be interpreted with caution.
The measurement of urinary cortisol excretion is valuable in investigations of Cushing’s syndrome. Urinary ‘steroid profiling’, in which steroids are separated and quantified by gas–liquid chromatography, often combined with mass spectrometry, is particularly valuable in the investigation of suspected congenital adrenal hyperplasia; it may also be helpful in the investigation of suspected adrenal carcinoma.
Some 95% of cortisol in the blood is bound to protein, principally to the cortisol-binding globulin, transcortin. Free cortisol concentration, and thus the amount of cortisol that can be excreted unchanged in the urine, is very low. Transcortin is almost fully saturated at normal cortisol concentrations. Because of this, if cortisol production increases, the concentration present in the plasma in the free form, and thus the amount that is excreted, increases to a disproportionately greater extent than the total. For this reason, measurement of the 24-h urinary excretion of cortisol, provided that an accurate urine collection can be made, is a sensitive way of detecting increased, but not decreased, secretion of the hormone.
Plasma cortisol concentrations show a diurnal variation, being highest in the morning and lowest at night (Fig. 8.4). Blood for cortisol measurement should usually be drawn between 08:00 h and 09:00 h; however, samples can be taken at 23:00 h to detect loss of the diurnal variation, an early feature of adrenal hyperfunction (Cushing’s syndrome). Random measurements are rarely of any value in the diagnosis of adrenal disease, except that a high concentration in a sick patient may reasonably be taken to exclude adrenal failure.
Figure 8.4 Diurnal variation in plasma cortisol concentration. Plasma cortisol concentrations are at their highest shortly after waking and then decline throughout the day to reach a nadir in the late evening. Because of this variation, it is important that blood samples are taken at times that coincide either with the peak or the trough, random samples being of little value. The graph shows mean values and the range in a sample of healthy people.
Cortisol is secreted in response to stress, mediated through ACTH, and thus stress must be kept to a minimum if results are to be interpreted correctly. Investigations of adrenal hypo- or hyperfunction often involve measurement of cortisol after attempting to stimulate or suppress its secretion.
When interpreting plasma cortisol results, it should be remembered that the synthetic glucocorticoid prednisolone may cross-react with cortisol in immunoassays for the hormone. Cross-reaction does not occur with dexamethasone, nor with spironolactone, an aldosterone antagonist used as a diuretic.
Aldosterone secretion is stimulated through the action of renin; therefore, it is often helpful to measure the plasma renin activity at the same time as the concentration of aldosterone, to establish whether aldosterone secretion is autonomous or under normal control. Calculation of the plasma aldosterone/renin ratio in a random blood sample is a useful screening test for excessive aldosterone secretion: this is excluded by a low value (see p. 148). Plasma aldosterone concentration varies with posture: the use of samples taken from patients while they are recumbent or ambulant is discussed further in connection with the investigation of excessive secretion of aldosterone (see p. 148).
The common causes and clinical features of this uncommon but life-threatening condition are listed in Figure 8.5. The cases originally described by Addison were caused by tuberculosis, but autoimmune disease is now the major cause in the UK. In such cases, adrenal autoantibodies are usually present, and there may be associated autoimmune disease affecting other tissues (e.g. pernicious anaemia).
The commonest cause of adrenal hypofunction is suppression of the pituitary–adrenal axis by glucocorticoids used therapeutically. Although, during treatment, patients may develop features of Cushing’s syndrome, a sudden withdrawal of steroids or failure to increase the dose during stress (e.g. surgery) may precipitate acute adrenal failure. Normal pituitary–adrenal function is regained only slowly when steroids are withdrawn and it is essential that the dosage is reduced gradually when steroid treatment is to be discontinued.
The majority of the clinical features of adrenal failure are due to the lack of glucocorticoids and mineralocorticoids. Increased pigmentation is a result of the high concentrations of ACTH, which occur because of the loss of negative feedback by cortisol: ACTH has some melanocyte-stimulating activity.
Adrenal failure usually has an insidious onset, with non-specific symptoms, but can develop acutely. Adrenal crisis is a medical emergency. The clinical features include severe hypovolaemia, shock and hypoglycaemia. It can be precipitated by stress (e.g. due to infection, trauma or surgery) in patients with incipient adrenal failure. Patients being treated with glucocorticoids, whether in physiological doses (replacement therapy) or pharmacological doses (e.g. in severe inflammatory conditions) are also susceptible to adrenal failure in these circumstances if the dosage is not increased. Haemorrhage into the adrenal glands may occur as a complication of anticoagulant treatment and in meningococcal septicaemia, and can result in acute adrenal failure. Although the adrenals are a relatively frequent site of tumour metastasis, this only occasionally results in adrenal insufficiency.
Adrenal failure can occur secondarily to pituitary failure as a result of decreased stimulation by ACTH. Other features of hypopituitarism may be present (see p. 127); in contrast to patients with primary adrenal failure, abnormal pigmentation does not occur. In secondary adrenal failure, hypotension can occur because the sensitivity of arteriolar smooth muscle to catecholamines is reduced by a lack of cortisol. Hyponatraemia can occur, as the lack of cortisol reduces the ability of the kidneys to excrete a water load, but there is no renal salt wasting because aldosterone secretion is not dependent on ACTH.
Unless a patient is being treated with synthetic corticosteroids, a plasma cortisol concentration of <50 nmol/L in a blood sample drawn at 09:00 h is effectively diagnostic of adrenal failure, while a concentration of >550 nmol/L excludes the diagnosis. However, in the majority of patients with adrenal failure, whether primary or secondary, the plasma cortisol concentration lies between these extremes, and an ACTH stimulation test must be performed to establish the diagnosis. The normal response to a single dose of soluble ACTH (tetracosactide or Synacthen) (’short Synacthen test’) is shown in Figure 8.6. If the response is in any way abnormal, the patient should be assumed to have adrenal failure. In both primary and secondary adrenal failure, the response in the short ACTH stimulation test is absent or blunted (Case history 8.1). This should be regarded as a screening test for adrenal failure. The distinction between primary and secondary adrenal failure can usually be made on the basis of measurement of the plasma ACTH concentration at 09:00 h: high values (a result of decreased negative feedback by cortisol) are typical of primary adrenal failure; low, or low–normal values, are typical of secondary adrenal failure. Alternatively, a long ACTH stimulation test can be performed (see Fig. 8.6). There are various protocols for this investigation. Typically, a single dose of depot ACTH (1 mg i.m.), which has a longer duration of action, is given and plasma cortisol is measured after 6 and 24 h. A failure to increase is typical of primary adrenal failure, whereas in secondary adrenal failure there is usually an increase at 6 h and a further increase after 24 h. If no increase occurs, but secondary adrenal failure remains a possibility, depot ACTH can be given over three days: a failure of cortisol to increase over this time excludes the diagnosis.
Figure 8.6 ACTH stimulation tests (also known as tetracosactide or Synacthen tests: Synacthen is synthetic ACTH) for the diagnosis of adrenal failure. It is important to note that blood should be taken for ACTH assay before giving ACTH. It is not necessary to withhold any treatment until after the tests have been completed, provided that the drug being used does not cross-react with cortisol, as exogenous steroids do not affect the response of the adrenal gland to ACTH in the short term.
Although, ideally, these tests should be done before starting treatment, when a severely ill patient is judged clinically to have adrenal failure treatment should not be delayed. A blood sample can be taken immediately for later cortisol measurement. Treatment can then be commenced with a synthetic glucocorticoid that does not cross-react with cortisol in the laboratory assay (e.g. dexamethasone) and an ACTH stimulation test performed as soon as is convenient. The results will not be vitiated by the treatment if only a short time elapses before the test is done. Patients presenting acutely with adrenal failure require intravenous hydrocortisone and fluid replacement with 0.9% sodium chloride. Plasma potassium and glucose concentrations should be monitored and intravenous glucose provided if necessary.
Once primary adrenal failure has been diagnosed, the cause should be sought, for example by measuring anti-adrenal antibodies (present in 90% of patients with autoimmune disease) and looking for evidence of tuberculosis.
All patients with adrenal failure require life-long replacement therapy, usually with both hydrocortisone and 9α-fludrocortisone, a synthetic mineralocorticoid. Hydrocortisone replacement is usually given in three unequal doses (e.g. 10 mg in the morning and 5 mg at midday and in the early evening), although a once-daily modified-release formulation is being developed. The adequacy of replacement can be assessed clinically and by measuring plasma cortisol concentration at intervals throughout the day (cortisol ‘day curve’): this allows detection of a concentration that is too high shortly after a dose or too low shortly before the next dose is due. Mineralocorticoid treatment can be assessed by measuring plasma renin activity: elevated activity implies inadequate replacement, and complete suppression implies excessive replacement (which can cause hypertension). Hydrocortisone has some intrinsic mineralocorticoid activity, and occasionally patients may be free of symptoms on hydrocortisone alone, particularly if they maintain a high salt intake. There is currently some interest in the potential value of providing androgen replacement (e.g. with DHEA), particularly in females; short-term benefits (e.g. in mood and libido) have been reported, but such replacement is not yet widely practised.
Long-term follow-up is essential to ensure the continuing adequacy of replacement treatment and to check for the development of other autoimmune endocrine disease. The dose of hydrocortisone should be increased during intercurrent illness, trauma, surgery, etc.
In Cushing’s syndrome, there is overproduction primarily of glucocorticoids, although mineralocorticoid and androgen production may also be excessive. In Conn’s syndrome, mineralocorticoids alone are produced in excess.
The causes and clinical features of Cushing’s syndrome are listed in Figure 8.7. Cushing’s disease, that is adrenal hyperfunction secondary to a pituitary corticotroph adenoma, accounts for 60–70% of cases of spontaneously arising Cushing’s syndrome (i.e. not caused by treatment with steroids). The clinical features are due primarily to the glucocorticoid effects of excessive cortisol, but cortisol precursors and indeed cortisol itself have some mineralocorticoid activity. Thus sodium retention, leading to hypertension, and potassium wasting, causing a hypokalaemic alkalosis, are common findings, except in iatrogenic disease (synthetic glucocorticoids have no mineralocorticoid activity). Increased production of adrenal androgens may also contribute to the clinical presentation.
Case history 8.1
A 17-year-old woman presented with a two-month history of tiredness and lethargy. She had noticed that she became dizzy when she stood up. On examination, she had pigmentation of the buccal mucosa and palmar creases and in an old appendicectomy scar. Her blood pressure was 120/80 mmHg lying down, but fell to 90/50 mmHg when she stood up.
|Serum: sodium||128 mmol/L|
|Blood glucose (fasting)||2.5 mmol/L|
|Plasma cortisol: 09:00 h||150 nmol/L|
|30 min after ACTH||160 nmol/L|
|60 min after ACTH||160 nmol/L|
|Plasma ACTH: (09:00 h) (normal <50 ng/L)||500 ng/L|
On the basis of these results, a diagnosis of primary adrenal failure was made. The patient’s symptoms resolved rapidly after starting glucocorticoid and mineralocorticoid replacement, and she remained well thereafter. Postural hypotension is a common finding in adrenal failure: it is due to a decrease in ECF volume caused by a lack of aldosterone, leading to sodium loss, together with a decrease in arteriolar tone owing to loss of the permissive effect of cortisol on the action of catecholamines. This decrease in ECF volume may also cause a degree of pre-renal uraemia, as demonstrated in this case. Hyponatraemia is not always present in adrenal failure, particularly in the early stages. Sodium is lost isotonically from the kidneys, but the lack of cortisol may cause water retention and, with severe hypovolaemia, vasopressin (antidiuretic hormone, ADH) secretion is stimulated. Deficiency of aldosterone is also responsible for potassium retention and thus hyperkalaemia.
The 09:00 h cortisol is at the lower limit of the reference range and there is virtually no response to ACTH. Except in very severe cases, cortisol is measurable in the plasma, even though the concentration is low–normal or frankly low. However, this represents the maximal output of the adrenal glands, as they are already stimulated by the high concentration of endogenous ACTH.
Some ten years later, the patient’s periods ceased. Her premature menopause was due to autoimmune ovarian failure. There is a recognized association between autoimmune adrenal failure and other organ-specific autoimmune diseases.