Treatment of Addison’s disease
The mainstay of treatment for Addison’s disease is to replenish near physiological doses of mineralocorticoids and glucocorticoids both under basal, i.e. daily conditions and during stress. The therapy has been virtually unchanged over the last 50 years, but recently modified-release preparations of hydrocortisone and continuous subcutaneous infusion systems have been tested with promising results. In addition, some women receive dehydroepiandrosterone (DHEA) treatment for androgen deficiency, but such treatment is still controversial. The current understanding of how replacement therapy in Addison’s disease should be given is detailed below.
Aldosterone plays a key role in the water and electrolyte homeostasis. The principal stimulators of aldosterone synthesis and secretion are angiotensin II and potassium. Aldosterone shows diurnal variation with an early night time decline and late night rise that is likely to result to a large extent from diurnal renin activity. Aldosterone binds to the mineralocorticoid receptor (MR), a nuclear receptor which is expressed mainly in the kidneys, colon and salivary glands, but also to some extent by liver, brain, pituitary and peripheral blood mononuclear cells. Activation of the MR results in gene regulation which promotes sodium reabsorption and potassium excretion in these organs. Cortisol and aldosterone binds to the MR with similar affinity, but cortisol is inactivated (i.e. oxidised) locally in mineralocorticoid target tissues by the enzyme 11 b -hydroxysteroid dehydrogenase type 2 (11 b -HSD2). Similarly, prednisolone has mineralocorticoid potential, but it is inactivated by the renal 11 b -HSD2, whereas dexamethasone is a pure glucocorticoid with no intrinsic mineralocorticoid effect.
Most patients with Addison’s disease require mineralocorticoid replacement, whereas mineralocorticoids do not usually have to be replaced in secondary adrenal failure. The natural mineralocorticoids aldosterone is difficult to synthesise. Therefore, in clinical practice, the synthetic mineralocorticoid fludrocortisone is used. Most studies conclude that a daily dose of 0.05 – 0.20 mg fludrocortisone is appropriate in Addison’s disease, in a situation of normal intake and loss of electrolytes. Furthermore, the glucocorticoid replacement contributes to the mineralocorticoid effect, depending on glucocorticoid type and dose and possibly variable activity of the 11 b -HSD2 enzyme
Mineralocorticoid action is evaluated by clinical examination, i.e. symptoms of orthostatic hypotension, and measurement of blood pressure in the lying and supine position. The dosage is further guided by the measurement of sodium and potassium homeostasis. Plasma renin activity (PRA) or concentration has become widely used as effect parameters of mineralocorticoid action; recommended target level is at or slightly above the upper normal reference limit. Normalisation of PRA confers risk of hypokalemia, oedema and hypertension. PRA measurement does not distinguish between adequate and excessive mineralocorticoid replacement, and is therefore most valuable to prevent under-replacement of mineralocorticoids. Furthermore, the renin substrate angiotensinogen is increased by glucocorticoids, and PRA may thus be falsely elevated in glucocorticoid over-replacement.
Patients on mineralcocorticoid replacement therapy who develop hypertension pose a particular challenge, and there are no evidence-based recommendations. A reasonable approach is to make sure that the dose of fludrocortisone is not inappropriately high when assessed as described below. If there is no sign of over-replacement the patient should receive conventional antihypertensive therapy, The aldosterone antagonists spironolactone and eplerenone are contraindicated.
The glucocorticoids have numerous physiological effects, of which the most important are regulation of energy metabolism, bone metabolism, and modulation of the immune system, as well as considerable neuropsychological effects. Pituitary ACTH is the principal regulator of cortisol production and release from the zona fasiculata cells of the adrenal cortex. This causes a diurnal and pulsatile cortisol release with high serum levels in the late night, and a gradual decline during the day. Current estimates of the endogenous cortisol production are 5.4-11 mg/m 2/day. Circulating cortisol is largely bound to cortisol binding globulin (CBG) and albumin, with only 5-10% as unbound and putatively active hormone. Furthermore, unbound cortisone circulates at concentrations similar to peak free cortisol, which provides a substrate for cortisol regeneration in peripheral tissues that express the enzyme 11 b -hydroxysteroid dehydrogenase type 1 (11 b -HSD1).
The conventional glucocorticoid replacement regimens have been hydrocortisone (i.e. cortisol) 20 mg + 10 mg and cortisone a cetate 25 mg + 12,5 mg. Such dosage is higher than the normal endogenous cortisol production and does not restore the normal cortisol biorhythm. A routine starting dose of hydrocortisone should be 10 mg on awakening, 5 mg at lunchtime and 5 mg in the early evening. However, the peak levels of cortisol after a dose vary enormously between individuals; this variation may be reduced by weight-adjusted dosing.
Some oppose the use of cortisone acetate, since this is a pro-drug that has to be activated (i.e. reduced) to cortisol by hepatic 11 b -HSD1. Cortisone reductase (11 b -HSD1) deficiency has been described, a condition that renders cortisone inactive as replacement therapy. Variable action of this enzyme could result in variable and unpredictable cortisol effects after cortisone acetate, but this has not been demonstrated as a clinical problem. On the contrary, the absorption curve of cortisone acetate is blunted and delayed compared to that of hydrocortisone, which could possibly be favourable. The difference in circulating levels of cortisone after hydrocortisone and cortisone acetate replacement and its clinical relevance are not known. Furthermore, there is a wide variety of drugs (rifampicin, several antiepileptics, ketokonazole, hyperforin (St. John’s Wart)), food constituents and xenobiotics that can cause problematic interactions with the glucocorticoid replacement therapy, but the size of the problem is not known.
Some prescribe synthetic glucocorticoids such as prednisolone or dexamethasone in adrenal failure, which give a more stable glucocorticoid effect throughout the day and night. In our experience, these steroids frequently give cushingoid side effects, and the surveillance of the therapy is even more difficult than for cortisone acetate and hydrocortisone. Furthermore, there is increasing evidence that such treatment confers higher risk of long-term adverse metabolic effects than the natural glucocorticoids.
Recently, new strategies for physiological glucocorticoid replacement have been suggested, such as timed-release hydrocortisone tablets and continuous subcutaneous hydrocortisone infusion (CSHI).
Proper tools are lacking for assessing the glucocorticoid replacement therapy and evaluating treatment effects in clinical trials. The therapy is mainly assessed by the evaluation of clinical signs of hypocortisolism (fatigue, abdominal or muscular pain, weight loss, hyperpigmentation) or hypercortisolism (weight gain, muscle weakness, psychiatric symptoms).
The only reason to measure cortisol levels is to determine the absorption and elimination of the replacement dose. Serum cortisol day curves have been advocated as the best objective measures of the glucocorticoid replacement; this is cumbersome in clinical practice, and of no value when synthetic glucocorticoids are used. Salivary cortisol reflects the bioavailable and active cortisol fraction and such measurement is a practical and reliable alternative to serum cortisol day curves.
The human adrenals secrete high amounts of DHEA and DHEAS, and small amounts of androstenedione. The mean daily production has been estimated to 4-14 mg DHEA and 20-25 mg DHEAS. ACTH stimulates adrenal androgen secretion, but unknown factors account for the lifetime changes and variations during intercurrent illnesses. A continuous interconversion between DHEA and DHEAS is taking place, catalysed by widely distributed sulphotransferases. Maximum serum concentrations are reached in the third decade of life, with 20% or less of peak values after 70 years of age.
DHEA(S) are inactive precursor steroids, which are converted to testosterone and estrogens in peripheral tissues. It has been estimated that the adrenals contribute 30-50 % of androgens in men and a very large proportion of androgens in women, particularly after menopause. Interesting immunomodulatory properties of DHEA have been suggested, but the exact mechanism of action is not known. Various neuronal effects have been described in animal models, but specific DHEA receptors have not been identified.
A daily dose of 25-50 mg DHEA is sufficient to restore normal levels of DHEA, DHEA(S), androstenedione and testosterone in adrenal failure, but controlled trails have failed to document significant effects on well being. Patients with secondary adrenal insufficiency usually have more severe androgen deficiency than patients with Addison’s disease, giving a stronger argument for replacement therapy in such cases.
Measurements of androgen concentration are rough estimates, since the normal interindividual variations are large. Moreover, the metabolism of DHEA may not be reflected by serum levels of active androgens, but rather of their metabolites . Such measurements, however, are not widely available and validated, and the practical approach would be to aim for serum levels of DHEAS within the age-matched reference interval, and reduce doses in case of side effects. Serum androstenedione and testosterone levels might also guide the replacement, although that requires assays with acceptable quality in the female range.
Replacement therapy in acute adrenal failure
A common advice is to increase the dose two to three-fold during minor illnesses, tapering down to normal dose over two to three days when the illness subside. A thumb rule is to double the dose if the patient has fever above 38.5 ° C, and to triple the dose when the temperature exceeds 39.5 ° C. In severe illness and in gastrointestinal disease with impaired absorption of tablets, hydrocortisone should be supplied intravenously, either in frequent boluses or as continuous infusion of 100 to 300 mg/day, depending on the severity of the intercurrent illness. It is of utmost importance that these individuals be equipped with a medical alert card that give instruction about treatment of adrenal crises and an emergency kit including hydrocortisone for intramuscular administration. Furthermore, it is advisable that they are offered vaccination against influenza and pneumococcus infection, as well as travel vaccinations. However, neither of these recommendations are based on strict evidence, and the requirements are likely to vary considerably between individuals and type of infection.
Possibly, slightly increased dosage may be beneficial during psychological strain and strenuous physical activity, since cortisol normally increases in these situations. Patients commonly self-medicate in stressful situations, indicating subjective benefit of glucocorticoids. However, short-term glucocorticoid excess is well known to boost well-being, and such use may lead to over-dosage. On the contrary, more frequent extra doses may be required since the recommended basal doses are lower than before. Patient education for optimisation of the therapy is certainly an aim, and clinical research to provide sound evidence for recommendations in various situations is welcomed.
Pregnancy and labour
The serum cortisol level increases up to 8-fold during pregnancy. A maximum plateau is reached at about 24 th gestational week, with a new albeit lower peak at delivery. The elevation of cortisol can be attributed to a 3-4-fold oestrogen stimulated increase in CBG levels, but there is also increased level of unbound cortisol due to reduced clearance of cortisol. Adrenal production of cortisol is not increased during pregnancy, and only a negligible production takes place in the placenta. ACTH levels are lower than normal during pregnancy, but show normal diurnal rhythm, and a rise towards delivery. Aldosterone levels increase 10-40-fold as a consequence of increased PRA and angiotensin levels. Adrenal androgen levels gradually decline in pregnancy due to conversion to estrogens.
A number of single case reports and some patient samples have been published. These reports conclude that both mother and child do well if the mother is given standard replacement therapy. It is generally recommended to continue with unchanged doses of glucocorticoids throughout the pregnancy, and give 100-300 mg hydrocortisone intravenously during labour. Since lower glucocorticoid replacement dose is now recommended in adrenal insufficiency, the risk of under-replacement during pregnancy require special attention. Symptoms of adrenal insufficiency such as nausea, vomiting, dizziness and hyperpigmentation may easily be misinterpreted as common symptoms of pregnancy, and render the assessment of the replacement therapy difficult. Synthetic glucocorticoids as replacement therapy in pregnancy has not been studied and they should be used with caution. Dexamethasone traverses the placenta and suppresses the ACTH and endogenous cortisol production in the foetus. Prednisolone is effectively inactivated by the placental 11 b -HSD, and is probably of less concern during pregnancy.
The fludrocortisone doses often need to be increased in the course of the pregnancy. The fludrocortisone dosage can be guided by evaluation of blood pressure, oedema and serum electrolytes. Renin activity measurements are difficult to interpret, since this normally increases during pregnancy. Androgen replacement in pregnancy is not likely to be hazardous, but has not been studied and cannot be recommended.