The adrenal gland was first described in 1552 by Bartolomeu Estaquio as the “glandulae renis incumbents” in Opuscula Anatômica [42], although its function remained a mystery for centuries. The mystery began to be solved in 1885, however, when Thomas Addison described the clinical features of 11 patients with primary adrenal insufficiency [57]. In 1949, the synthesis of cortisone facilitated the treatment of this condition [50]. The adrenal gland is made of two tissue types, namely, the adrenal medulla and the adrenal cortex, which have different embryonic origins. By 4–5 weeks of gestation, cells from the mesoderm aggregate to form a primitive cortex between the posterior part of the dorsal mesentery and the gonadal ridge [7]. Shortly thereafter, this primitive cortex becomes surrounded by a narrow band of cells termed the permanent cortex. By 7–8 weeks of fetal life, the primitive cortex is invaded by chromaffin cells that develop rapidly and eventually replace most of the primitive cortex, forming the medulla. At this time the adrenal gland is close to the cranial part of the primitive kidney and not far from the genital ridge. The adrenal medulla, which originates from ectodermal cells, has an entirely different function from the mesodermal adrenal cortex.

In mammals, the adrenal cortex is made of three zones. The first region is the outer zone, the zona glomerulosa, which is responsible for the production and secretion of the mineralocorticoid aldosterone. The inner region is divided into the zona fasciculata and the zona reticularis and is responsible for synthesis and production of glucocorticoids (cortisol, corticosterone, and adrenal androgens). We first focus our discussion on the physiological function and regulation of cortisol production. Then we consider the more common disorders related to primary or secondary hypocortisolism, their potential therapeutic approach, and therapy-related complications in childhood.

Cortisol is the principal glucocorticoid hormone produced by the adrenal cortex in humans. The production of cortisol is the result of a series of reactions that involve the concerted action of several enzymes within the adrenals. In this complex process of steroidogenesis, the uptake of cholesterol to the mitochondria represents the first and critical step that is facilitated by the action of a regulatory protein called the steroidogenic acute regulatory protein [75]. The biosynthetic pathway of the adrenal steroids is shown in Fig. 1. Thus, during steroidogenesis, cholesterol is the precursor of a number of steroid hormones of both gonadal and adrenocortical origin. Although they share similar chemical formulae, small differences in their molecular structure characterize each steroid hormone and give them specific functions [86]. The pathway from cholesterol to the end steroid products requires five cytochrome P450 enzymes [cholesterol side-chain cleavage enzyme (20-hydroxylase, 22-hydroxylase, 20,22-lyase, CYP11A), 3β-hydroxysteroid dehydrogenase (3β-HSD), 17α-hydroxylase and 17,20-lyase (CYP17), 21-hydroxylase (CYP21), 11β-hydroxylase (CYP11B1), aldosterone synthetase (11β-hydroxylation, 18-hydroxylation, 18-oxidation, CYP11B)]. Cholesterol is stored in the adrenal cell as cholesterol esters [75, 86]. Under the influence of an esterase, cholesterol becomes available and is transported to the mitochondria, where it is converted into pregnenolone [75, 86]. This steroid then moves into the endoplasmic reticulum, where 3β-hydroxysteroid dehydrogenase, 21-hydroxylase, and 17-3β-hydroxylase enzymes are located. The resulting steroids include 11-deoxycorticosterone (DOC) and 11-deoxycortisol as well as two C-19 carbon steroids, androstenedione and dehydroepiandrosterone. At this point, DOC and 11-deoxycortisol return to the mitochondria, where they are converted into corticosterone and cortisol, respectively. This is the end of the biosynthetic process in the cells of the fasciculata. In the cells of the zona glomerulosa in the mitochondria, DOC is transformed into corticosterone, 18-hydroxycorticosterone, and aldosterone [75, 86].

Synthesis and secretion of cortisol are regulated by the pituitary hormone adrenocorticotropin (ACTH), which in turn is regulated by hypothalamic corticotropin-releasing hormone (CRH) with the synergistic action of arginine vasopressin (AVP). These hormones comprise the hypothalamic–pituitary–adrenal axis that is directly related to a complex closed-loop system. Indeed, CRH is synthesized in the hypothalamus and carried to the anterior pituitary, where it stimulates ACTH release. Finally, ACTH stimulates the adrenal cortex to secrete cortisol. Cortisol inhibits the synthesis and secretion of both CRH and ACTH in a negative feedback regulation system [64, 67, 85, 109].

CRH is a 41-amino acid straight-chain peptide secreted mainly by the median eminence into the portal vessels. Via specific receptors (CRH-R1), CRH activates the formation of cyclic adenosine monophosphate, which then activates a series of protein kinases, resulting in increased transcription of the pro-opiomelanocortin gene and in ACTH formation [12, 64, 67, 85, 109]. ACTH has a half-life in blood of a few minutes and like other hormones binds specific receptors on the adrenal cortex, type 2 melanocortin receptors (MC2-R), and increases cyclic adenosine monophosphate formation to initiate the synthesis of cortisol, which is released immediately into the systemic circulation by diffusion [12, 64, 67, 85, 109]. ACTH stimulation of cortisol on the adrenal includes both an immediate and a chronic phase. Acutely, over a few minutes, steroidogenesis is stimulated through a steroidogenic acute regulatory protein (STAR-)-mediated increase in cholesterol delivery to the CYP11A1 enzyme in the inner mitochondrial membrane [14]. In the more chronic phase, over 24–26 h of exposure, ACTH leads to an increase in the synthesis of all steroidogenic CYP enzymes (CYP11A1, CYP17, CYP21A2, CYP11B1) in addition to adrenodoxin, and these effects are mediated at the transcriptional level. Additional effects of ACTH include: (a) increased synthesis of the low-density lipoprotein and high-density lipoprotein receptors, and possibly also HMG-CoA (3-hydroxy-3-methyl-glutaryl-CoA) reductase, the rate-limiting step in cholesterol biosynthesis; (b) increased adrenal weight by inducing both hyperplasia and hypertrophy [12, 14, 64, 67, 85, 109].

Glucocorticoid synthesis is mostly affected by two variables: the secretion patterns and the secretion rate. The former is related to three main physiological mechanisms affecting the secretion of cortisol: pulsative secretion and diurnal variation, stress, and negative feedback. The normal pattern of glucocorticoid secretion includes both a diurnal rhythm and a pulsatile ultradian rhythm. In fact, the natural cortisol peak in humans occurs early, before awakening, and falls progressively during the day, reaching low levels in the evening [9, 112]. The circadian rhythm of glucocorticoid secretion is accompanied by a pulsatile ultradian rhythm throughout the 24-h cycle [20]. As documented by automated frequent blood-sampling techniques, the pulses vary in amplitude throughout the day, with the amplitude generally decreasing during the diurnal trough. Of note, the two components are separable secretory modes. Thus the “pulsatile” and “circadian” rhythms are independently regulated [118]. Ultradian rhythmicity has been shown in rats [116], monkeys [94], and humans [9, 38, 46, 112]. Among the most relevant practical consequences [118] of the pulsatility is that the underlying pattern of spontaneous pulses might not be detected if sampling is infrequent and/or conducted over a short period. This might also have an additional effect on the tissue specificity. In fact, the two glucocorticoid-related tissue receptors have different affinities. Therefore, according to the circulating level of ligands, the receptors will be differentially occupied and activated [118], especially affecting the occupancy of the lower-affinity glucocorticoids receptors [118]. Prolonged versus intermittent exposure seems to also affect steroid-responsive hepatic enzymes. Studies have documented that short exposure to glucocorticoids may have different effects on tyrosine aminotransferase, an enzyme involved in the catalysis of the first step in tyrosine catabolism [88, 107]. Finally, prolonged exposure to glucocorticoids has been shown to downregulate glucocorticoids receptors [88].

This complex regulation system is further characterized by the ability of the adrenal glands to secrete steroids in a stress-related way [37]. Surgical stress such as trauma and tissue destruction, medical stress such as acute illness, fever, and hypoglycemia, and emotional stress related to psychological upset result in a significant increase in cortisol secretion in most cases. The hypothalamic–pituitary–adrenal axis in conjunction with the sympathetic system connects the brain with the periphery of the body. Of note, the body responses to a stressor – physical or emotional – that disrupts the homeostatic balance of the organism are mainly related to the hypothalamic–pituitary–adrenal axis activity [37, 93]. All the complex activities characterizing the individual’s adaptive response to excessive stress are stereotypical and usually defined as the “general adaptation syndrome” [37, 93]. This physiological response involves interactions between hormones and the central nervous system. Glucocorticoids along with catecholamines (the end product of sympathetic nervous system activation) secreted by the adrenal medulla and sympathetic nerves orchestrate the “fight or flight” response, which is the first stage of the general adaptation syndrome [69]. The fight or flight response refers to different factors including: a quick mobilization of energy from storage to different systems, such as the heart, muscles, and the brain; a prompt transport of nutrients and oxygen to relevant tissues facilitated by accelerated cardiac output and breathing rate; and increased blood pressure [23]. According to the theory of Munck and colleagues [77], the physiological function of stress-induced increase in glucocorticoid levels is to defend the body against the normal defense reactions that are activated by stress and not against the stress itself. According to this theory, glucocorticoids accomplish this function by turning off these defense reactions, thus preventing them from overshooting and threatening homeostasis. Therefore, it is now commonly accepted that glucocorticoid secretion in a stress situation plays a double and complementary function: a permissive and suppressive effect, the former preparing or priming defense mechanisms for action and the latter limiting these actions [78]. CRH and AVP neurons of the hypothalamic paraventricular nuclei and the noradrenergic neurons of the locus coeruleus/norepinephrine–central sympathetic systems in the brain stem represent the main apparatus of the stress system. In addition, the peripheral branches of this system consist of the hypothalamic–pituitary–adrenal axis and the systemic sympathetic and adrenomedullary nervous system [21]. Both central components of the stress system are stimulated by cholinergic and serotonergic neurotransmitters and inhibited by γ-aminobutyric acid, benzodiazepine, and arcuate nucleus pro-opiomelanocortin peptides [19, 31]. Activation of the central stress system results in the secretion of CRH and AVP into the hypophyseal portal circulation, thus inducing glucocorticoid secretion by the hypothalamic–pituitary–adrenal axis. In this complex event the systemic sympathetic and adrenomedullary nervous systems are also activated as a direct consequence of central stress system stimulation, which in turn results in a peripheral secretion of catecholamines and several neuropeptides. At rest the stress system is still active, assisting the body in responding to various distinct signals, for example, circadian, neurosensory, blood-borne, and limbic [22]. The activation of the stress system has thus several effects: it increases arousal, accelerates motor reflexes, improves attention and cognitive function, decreases appetite and sexual arousal, and also increases the tolerance of pain [23]. Although these types of stress are well known to affect cortisol production, research is still ongoing to define all the regulatory mechanisms involved. Among the reported results, studies of the immune system have shown that leukocytes may play a relevant regulatory action by secreting a series of interleukins able to significantly affect the adrenal axis [110].

The negative feedback represents a relevant feedback control able to constantly equilibrate the secretion rate of both ACTH and cortisol. When plasma concentrations of cortisol increase markedly, a negative feedback effect on the secretion of CRH and ACTH is induced [64, 67, 85, 109].

The secretion rate and cortisol metabolism represent an additional variable that needs to be considered in defining glucocorticoid synthesis in the young. The daily cortisol production rate ranges between 5 and 10 mg/m² body surface area [16, 28, 51, 55]. Circulating cortisol in humans is about 90 % plasma protein bound, mostly to cortisol-binding globulin and less to albumin, while only 5–10 % circulates unbound as a free active hormone [64, 67, 85, 109]. The free cortisol concentration ranges from approximately 1 nmol/l at the diurnal trough to approximately 100 nmol/l at the diurnal peak [96]. Estimations of the circulating half-life of cortisol vary between 70 and 120 min. Cortisol is cleared through several distinct pathways, including A-ring reduction to form tetrahydrocortisol and its 5α-isomer, allotetrahydrocortisol, hydroxylation to yield 6-β-hydroxycortisol, and the reduction of the 20-oxo group to produce cortisol [34]. Cortisone is an inactive steroid that circulates at concentrations of around 60 nmol/l, largely unbound to plasma proteins and without marked diurnal variation. The main source of cortisone is 11-β-hydroxysteroid dehydrogenase-type 2 (11-β-HSD-2) in the kidney [96, 108], which gates glucocorticoid access to nuclear receptors by a prereceptor mechanism. 11-β-HSD-1 converts cortisone to cortisol, amplifying the steroid signal in target cells [97]. Additionally, cortisol derives from circulating cortisone via conversion in peripheral tissues expressing the enzyme 11-β-HSD-1. The cortisol secretion rate in children also shows some peculiarities. Several studies in children have shown that in normal children and adolescents, the cortisol secretion rate is directly related to body size [28, 51, 74]. Migeon et al. showed that when the values are corrected for body surface area, the rates are similar at various ages; in fact, the average ± standard deviation was 12 ± 2 with a range of 8–16 mg/m²/24 h [74]. Using stable isotope dilution/mass spectroscopy, Esteban et al. showed that the cortisol secretion rate for 12 normal subjects was lower, accounting for 5.7 ± 1.5 mg/m²/24 h [74]. Kerrigan et al. also investigated the daily cortisol production and clearance rates in a group of 18 normal unstressed pubertal male subjects by applying deconvolution analysis to serum cortisol concentrations obtained every 20 min for 24 h [51] and found similar results to Esteban’s data. In addition, they showed that the estimated cortisol production rate for the early puberty group was indistinguishable from that of the late puberty subjects [51]. No difference was observed between the two pubertal groups in the secretory burst frequency and half-duration, mass of cortisol released per secretory episode, average maximal rate of hormone secretion, and serum cortisol half-life [51]. A significant diurnal pattern of cortisol secretion was observed for all subjects, manifested by nyctohemeral variations in the frequency of adrenocortical secretory bursts, the amplitude (maximal rate of cortisol secretion) and the mass of cortisol released per secretory episode. In this age group, maximum serum hormone concentrations occurred between 07:06 and 11:14 h [51]. Similar results were also reported by Linder et al., who evaluated the cortisol production rate in 33 normal children and adolescents, using a stable isotope-dilution technique with high-performance liquid chromatography-mass spectrometry [61].

Glucocorticoids are essential for the maintenance of homeostasis and enable the organism to prepare for, respond to, and manage physical or emotional stress. These hormones affect nearly every organ and tissue in the body and have diverse life-sustaining effects throughout the life span. Glucocorticoid access to nuclear receptors is gated by the 11-β-HSD enzymes. Corticosteroids are highly lipophilic and are thought to diffuse readily across biological membranes to access their intracellular receptors [41, 81]. At the cellular level, the myriad effects of corticosteroids are largely a consequence of transcriptional actions mediated via binding to two types of intracellular receptors: the high-affinity mineralocorticoid receptor and the lower-affinity glucocorticoid receptors [32, 70]. On binding ligand, glucocorticoid receptors and mineralocorticoid receptors dissociate from complexes with chaperone proteins, translocate to the nucleus, and bind directly or indirectly to the regulatory regions of target genes: ≈2 % of the human genome is regulated by glucocorticoids [89], although few, any genes are exclusively controlled by corticosteroids. Rapid glucocorticoid signaling via membrane binding has also been postulated [18].

Cortisol is involved in peripheral glucose uptake and utilization (gluconeogenesis and glycogenolysis). Cortisol also affects the maintenance of proper cardiovascular tone, endothelial integrity, and the distribution of fluids within the vascular compartment. Moreover, cortisol potentiates the vasoconstrictor action of catecholamines and decreases the production of nitric oxide [25, 36]. Therefore, cortisol deficiency results in hypoglycemia, hypotension, lethargy, decreased appetite, absolute leukocytosis, eosinophilia, and anemia. Cortisol influences the activity and direction of the reactions underlying intermediary metabolism and many functions of the central nervous system, including arousal, cognition, mood, and sleep. Physiological amounts of glucocorticoids are also essential for normal renal tubular function and thus for water and electrolyte homeostasis. Studies have shown that 15–20 % of the human leukocyte transcriptome is influenced by glucocorticoids [24, 33], and almost two thirds of them are induced, whereas the rest are suppressed. Through their genomic actions, glucocorticoids regulate cellular metabolism primarily through catabolic actions in the liver, muscle, and adipose tissue [24, 33]. Finally, multiple components regulating the quantity and quality of immune/inflammatory responses are well-recognized glucocorticoid targets, providing the basis for the wide use of glucocorticoids as potent anti-inflammatory/immunosuppressive drugs in the treatment of inflammatory diseases and cancer [90].

The first treatment for adrenal insufficiency was introduced in the 1930s when lipid extracts from adrenal glands were tested, leading to a drastic and rapid drop of the mortality rate from 100 % to a seemingly normal life expectancy. In 1937 and 1949, the synthesis of 11-deoxycortisone (11-DOC) and cortisone, respectively, represented major improvements in therapy. Since the first published report of the efficacy of cortisone in the treatment of rheumatoid arthritis in 1949 [92], patients with adrenal insufficiency have been treated with glucocorticoid replacement, and after the introduction of fludrocortisone in the 1950s replacement therapy has remained virtually unchanged [66]. Hydrocortisone is now used in many centers around the world.

As mentioned, following the first report of the efficacy of cortisone in treating rheumatoid arthritis, glucocorticoids have been used widely in several autoimmune diseases and in the treatment of a spectrum of disorders in childhood. In particular, glucocorticoid replacement remains the cornerstone of treatment for certain life-threatening endocrinopathies in childhood, such as congenital adrenal hyperplasia, Addison disease, and as replacement therapy for those subjects with secondary hypothalamic–pituitary–adrenal axis deficit. In the next sections, a short description of these main disorders in childhood is provided, evidencing the role of glucocorticoids in their treatment.

Adrenal insufficiency is a clinical condition characterized by a state of failure of the adrenal cortex to provide sufficient amounts of steroid hormones, in particular glucocorticoids. Several causes might be responsible for the development of adrenal insufficiency in childhood. According to the localization of its underlying cause, adrenal insufficiency in childhood can be essentially categorized into two major groups: primary and secondary. The most frequent causes of primary and secondary adrenal insufficiency are summarized in Table 1.

The group of primary adrenal insufficiency includes: autoimmune adrenalitis (Addison disease, which can arise in isolation or as part of an autoimmune polyglandular syndrome), infections (tuberculosis, cryptococcosis, mycosis, AIDS), congenital conditions (adrenoleukodystrophy, adrenomyeloneuropathy, congenital adrenal hyperplasia), bilateral adrenalectomy, bilateral adrenal hemorrhage, metastases and surgery, and drug-induced adrenal insufficiency (treatment with mitotane, etomidate, ketoconazole, aminoglutethimide). Secondary adrenal insufficiency results from hypothalamic–pituitary impairment, with consecutive lack of CRH and/or ACTH. Thus, this group mainly includes: pituitary tumors or other tumors of the hypothalamic–pituitary region often associated with panhypopituitarism (caused by tumor growth or treatment with surgery or irradiation), exogenous glucocorticoids leading to suppression of CRH/ACTH release, head trauma, and pituitary infiltration.

Although most of these conditions rarely occur in childhood, adrenal insufficiency related to congenital adrenal hyperplasia, Addison disease, and Cushing’s syndrome are not uncommon, thus requiring clinicians, health-care planners, and patients to understand these life-threatening disorders and the proper management of adrenal insufficiency in the various clinical settings. While etiological aspects characterize the different causes of adrenal insufficiency, glucocorticoids traditionally represent the main therapeutic option in all forms of adrenal insufficiency, including acute and chronic states.

Acute adrenal insufficiency is a life-threatening disease that involves severe hypotension or hypovolemia, acute abdominal pain, nausea and vomiting, lack of stamina, and weight loss [4]. Anorexia, fever, weakness, fatigue, lethargy, and confusion may also be associated with this condition. Dizziness, irritability, and postural hypotension are frequent complaints; these symptoms can be triggered by several predisposing factors such as trauma, surgery, and infections, which suddenly increase the need for corticosteroids. Acute adrenal insufficiency-related shock is often unresponsive to volume replacement and vasoconstrictor agents [54, 84, 111]. Hyperpigmentation and salt-craving are also often detected. According to the underlying cause, the onset of the disease can be insidious, taking years to diagnose, or can lead to the development of an acute crisis following an intercurrent illness [54, 84].

Congenital adrenal hyperplasia (CAH) is an autosomal recessive disorder caused by the deficiency of an enzyme involved in steroidogenesis within the adrenal cortex [72, 115]. Although several enzymatic defects have been described, the most common is cytochrome P450 21-hydroxylase (CYP21) deficiency. The defect accounts for approximately 95 % of cases and results from mutations [5, 44, 56, 103] of the CYP21A2 gene located on chromosome 6p21.3. The enzyme adrenal insufficiency converts 17-hydroxyprogesterone into 11-deoxycortisol and progesterone into 11-deoxycortisone, which are precursors for cortisol and aldosterone, respectively (Fig. 1). Therefore, defects of the enzymatic activity result in an impaired adrenal synthesis of cortisol often associated with aldosterone deficiency, which in turn leads to increased ACTH secretion by the pituitary gland. The impaired cortisol/aldosterone synthesis and the increased ACTH production directly induce: severe salt wasting (SW) and Addisonian crisis, related to cortisol and aldosterone deficiency; adrenal gland hyperplasia, related to ACTH oversecretion; and accumulation of steroid precursors, inducing a variable degree of virilization as a direct consequence of adrenal androgen overproduction [115].

According to the degree of the enzyme deficiency, different clinical phenotypes can be defined including: classic SW, classic simple virilizing (SV), and nonclassic (NC) CAH. The classic SW form is the most severe form of enzymatic activity deficiency, resulting from a residual activity of less than 1 %. In this form, severe cortisol deficiency and decreased aldosterone synthesis are detected. Female patients are virilized prenatally owing to adrenal androgen excess. Neonates (boys and girls) also suffer from life-threatening Addisonian crisis. In those forms characterized by a residual enzyme activity of 1–2 % (simple virilizing) CAH, the residual activity is enough for sufficient aldosterone production, thus preventing SW. By contrast, cortisol synthesis is impaired and this results in the development of genital ambiguity in affected female patients due to prenatal virilization. In those forms with a residual enzymatic activity around 20–50 % (nonclassic), cortisol and aldosterone production are normal. In these subjects a mild androgen excess may be detected and may induce premature pubarche, cystic acne, hirsutism, and menstrual disorders in some subjects in childhood/adolescence, or may even be asymptomatic. Some patients present first in adulthood with fertility problems.

Autoimmune adrenalitis, or autoimmune Addison disease (AAD), is a rare condition in childhood. Both humoral and cellular immunity play a role in AAD pathogenesis, with presence of adrenal cortex autoantibodies in the serum of patients [3]. These adrenal cortex autoantibodies are of the immunoglobulin subclasses IgG1, IgG2, and IgG4 and are directed against the steroidogenic enzymes, with steroidogenic 21-hydroxylase being the most prevalent [100, 117]. Although the presence of adrenal cortex autoantibodies is a main feature of the disease, their role in the pathogenesis of autoimmune Addison disease is still debated. Studies have shown that the destruction of adrenocortical cells is mainly mediated by T-lymphocytes. Thus the secondary release of peptides may result in the production of antibodies [13]. Autoimmune adrenalitis may present in 60 % of cases as part of an autoimmune polyendocrine syndrome, while in the remaining 40 % it is isolated [4]. During the first two decades of life, isolated AAD is predominantly observed in male subjects (70 %); however, after the third decade of life, there is a substantial female preponderance (81 %) [102]. Spontaneous recovery of adrenal function has been described but is rare. Addison disease is the final result of AAD; the initial phase is subclinical, and after at least 90 % of the adrenal gland has been destroyed, symptoms of adrenal failure occur [11]. This condition can easily be misdiagnosed in childhood, thus negatively affecting data on its true prevalence; autoimmune adrenalitis is the main cause of adrenal insufficiency after the introduction of antituberculosis therapy, and is responsible for 68–94 % of the cases in European and North American reports [10, 29, 79]. Determination of inappropriately low cortisol production associated with the presence of high titers of adrenal cortex autoantibodies is strongly suggestive of autoimmune adrenalitis. The diagnosis is confirmed by excluding other causes of adrenal failure, using other tests as necessary. Treatment is based on corticosteroid replacement, and the prognosis following treatment is the same as for the normal population. Thus the standard initial therapy is corticosteroid replacement.

Cushing’s syndrome refers to a large group of clinical conditions characterized by the presence of signs and symptoms associated with prolonged exposure to inappropriate levels of the hormone cortisol [67]. In children with Cushing’s syndrome, the hypothalamic–pituitary–adrenal axis has lost its ability for self-regulation. Thus, the impaired hypothalamic–pituitary–adrenal axis function may result from an excessive secretion of either ACTH or cortisol and from the loss of the negative feedback function [67].

Cushing’s syndrome is a rare entity; its overall incidence is approximately two to five new cases per million people per year. Characteristically, in older children, a female predominance has been described that decreases with younger age and seems to switch to a male predominance in infants and young toddlers [67, 85, 109].

Although both exogenous and endogenous causes can induce Cushing’s syndrome, the former are certainly more common in children. In particular, exogenous or iatrogenic causes might result from chronic administration of glucocorticoids or ACTH (such as in the treatment of many nonendocrine diseases including neoplastic, hematologic, pulmonary, autoimmune, epileptic, and dermatologic disorders). Among the endogenous causes of Cushing’s syndrome in children, ACTH overproduction from the pituitary (called Cushing’s disease) is the most common, and results from an ACTH-secreting pituitary microadenoma or, rarely, a macroadenoma. Cushing’s disease is more common in children older than 7 years of age, accounting for approximately 75 % of all cases of Cushing’s syndrome in this age group. By contrast, in children younger than 7 years adrenal causes of Cushing’s syndrome (adenoma, carcinoma, or bilateral hyperplasia) are the most frequent. Ectopic ACTH/CRH production occurs rarely in young children and adolescents, and for some forms they have never been described in young children [85, 104, 105, 109]. A few additional rare diseases, such as primary pigmented adrenocortical nodular disease (PPNAD), massive macronodular adrenal hyperplasia (MMAD), McCune–Albright syndrome, might be related to Cushing’s syndrome in childhood. PPNAD is a genetic disorder and the majority of cases are associated with Carney complex, a syndrome of multiple endocrine abnormalities in addition to lentigo and myxomas. Periodic, cyclical, or otherwise atypical Cushing’s syndrome is often documented in children and adolescents with PPNAD. MMAD is another rare bilateral disease that leads to Cushing’s syndrome [105]. In children with MMAD, the adrenal glands are massively enlarged, with multiple huge nodules that are typical yellow-to-brown cortisol-producing adenomas. Data have shown that in some patients with MMAD, cortisol levels seem to increase with food ingestion (food-dependent Cushing’s syndrome), which might result from an aberrant expression of the gastric inhibitory polypeptide receptor in the adrenal glands. In the majority of patients with MMAD, however, the disease does not appear to be gastric inhibitory polypeptide receptor dependent.

In children with McCune–Albright syndrome, adrenal adenomas or, more frequently, bilateral macronodular adrenal hyperplasia can also be seen [30, 53]. In this syndrome, there is a somatic mutation of the GNAS1 gene leading to constitutive activation of the Gsα protein and continuous, non-ACTH-dependent activation of steroidogenesis by the adrenal cortex.

The treatment of choice varies according to the underlying cause [49, 85, 104, 105, 109]. Transsphenoidal surgery (TSS) with or without irradiation of the pituitary gland represents the treatment of choice for almost all patients with ACTH-secreting pituitary adenomas (Cushing’s disease). Surgical resection with or without radiotherapy is also the treatment of choice for benign adrenal tumors. The treatment of choice in bilateral micronodular or macronodular adrenal disease, such as PPNAD and MMAD, is usually bilateral total adrenalectomy. In addition, in subjects with Cushing’s disease or ectopic ACTH-dependent Cushing’s syndrome in whom surgery or radiotherapy has failed, or in whom the tumor has not been localized, adrenalectomy is a potential treatment. Finally, pharmacotherapy is also an option if surgery fails for Cushing’s disease or in ectopic ACTH secretion where the source cannot be identified. Several molecules can be used, such as mitotane, aminoglutethimide, metyrapone, trilostane, and ketoconazole, which may act by: inhibiting the biosynthesis of corticosteroids by blocking the action of 11-β-hydroxylase and cholesterol side chain cleavage enzymes; destroying adrenocortical cells that secrete cortisol; blocking the conversion of cholesterol to pregnenolone in the adrenal cortex; inhibiting the synthesis of cortisol, aldosterone, and androgens; preventing the conversion of 11-deoxycortisol to cortisol; inhibiting the conversion of pregnenolone to progesterone; or blocking adrenal steroidogenesis.

Although the treatment of choice varies according to the underlying cause of Cushing’s syndrome or disease, the hypothalamic–pituitary–adrenal axis is often negatively affected [49, 85, 104, 105, 109]. Hypopituitarism is the most common adverse effect, and it is more frequent when surgery precedes radiotherapy. In addition, after the completion of successful TSS in Cushing’s disease or excision of an autonomously functioning adrenal adenoma, there will be a period of adrenal insufficiency while the hypothalamic–pituitary–adrenal axis recovers. Therefore, in this situation glucocorticoids might be replaced. Treatment is aimed at restoring physiological changes, with a usual replacement dose of 12–15 mg/m²/day two or three times daily [63]. In addition, in the immediate postoperative period, cortisol treatment should be started initially at stress doses of glucocorticoids and then weaning relatively rapidly to a physiological replacement dose.

According to the underlying alteration, glucocorticoid replacement might be temporarily adopted only for a short period [49, 85, 104, 105, 109]. Thereafter, patients should be closely followed up with a systematic assessment of the adrenocortical function. Clinicians might consider discontinuing glucocorticoid treatment if normal responses to a 1-h ACTH test are documented (cortisol level over 18 μg/dl at 30 or 60 min after ACTH stimulation) [49, 85, 104, 105, 109].

In children with unilateral adrenalectomy as in patients with Cushing’s disease post-TSS, a similar replacement regimen is needed for a single adrenocortical tumor. By contrast, for those who have undergone bilateral adrenalectomy, lifetime replacement with both glucocorticoids (as described previously) and mineralocorticoids (fludrocortisone 0.1–0.3 mg daily) is needed. In these patients, too, glucocorticoids at stress doses are needed immediately postoperatively, with a relatively quick weaning to physiological replacement doses. In addition, for temporary and permanent adrenal insufficiency, acute illness, trauma, or surgical procedures, stress doses must be adopted in all patients [49, 85, 104, 105, 109].

The main aim of treatment of adrenal insufficiency in childhood is to restore the impaired hypothalamic–pituitary–adrenal axis, without impairing growth while allowing for normal pubertal development and fertility. In addition, in subjects with congenital adrenal hyperplasia a proper suppression of androgen production is needed to minimize the peripheral effects of hyperandrogenism secretion.

The available evidence suggests that conventional treatment of patients with hypoadrenalism may result in adverse effects on some surrogate markers of disease risk, such as a lower bone mineral density, than in age- and sex-matched controls, and in increased postprandial glucose and insulin concentrations. Although the quality of life of patients with hypoadrenalism may be impaired, there is no evidence of an improvement with higher doses of steroids, although quality of life is better if the hydrocortisone dose is split up, with the highest dose taken in the morning. Thus the evidence suggests that most patients may safely be treated with a low dose of glucocorticoids in two or three divided doses, along with education about the appropriate course of action in the event of intercurrent illnesses.

The glucocorticoid of choice in childhood is hydrocortisone, which is short acting and hence has the lowest growth-suppressing effect (Table 2) [48, 101]. During infancy, especially in subjects needing an initial reduction of markedly elevated adrenal sex hormones, up to 25 mg of hydrocortisone/m² may be required. This is more than the daily physiological secretion of 7–9 mg/m² in newborns and 6–8 mg/m² in older infants and children [48, 101]. Hydrocortisone oral suspension is not recommended [73]; divided or crushed tablets of hydrocortisone should be used in growing children. Cortisone acetate requires conversion to cortisol for bioactivity [82]; thus hydrocortisone is considered the drug of first choice. To mimic the circadian cortisol secretion, the daily hydrocortisone dose is divided into two or three doses, with administration of one half to two thirds of the total daily dose in the morning. The short elimination half-life of hydrocortisone (approximately 1.5 h) when given in traditional immediate-release preparations, however, leads to high peaks with low values in between. A twice-daily regimen with administration of the second dose 6–8 h after the morning dose is recommended. The timing of the second dose may be changed slightly according to the patient’s activities. Some authors postulate that a thrice-daily administration is more beneficial [2, 6, 43, 59, 87], although there is no hard evidence available yet to support this. Whereas hydrocortisone is preferred during infancy and childhood, longer-acting glucocorticoids may be recommended at or near the completion of linear growth, such as in older adolescents or young adults (Table 3). Prednisone and prednisolone should be given twice daily [48, 101]. Prednisolone may be preferable since it is the active drug. The dose (2–4 mg/m²/day) should be one-fifth that of hydrocortisone. The dosage of dexamethasone is 0.25–0.5 mg/m²/day given once daily. These steroids have minimal mineralocorticoid effects compared with hydrocortisone. In children with advanced bone age, such as in boys with non-salt-losing CAH, initiation of therapy may precipitate central precocious puberty, requiring additional treatments, such as with a GnRH agonist. In some children with treatment refractory to hydrocortisone, long-acting glucocorticoids may be effective [91]. In symptomatic patients with non-classic-CAH, treatment with glucocorticoids is recommended. In these patients, chronic steroid treatment may suppress the hypothalamic–pituitary–adrenal axis, so they require stress dosing during surgery or severe illness. For asymptomatic patients with non-classic-CAH, hydrocortisone treatment is not required during stress [101]. All patients with classic CAH require mineralocorticoid replacement with fludrocortisone at a dose of 0.05–0.2 mg/day. The dose is slightly higher (up to 0.3 mg/day) in newborns and small infants because of their increased metabolism and end-organ resistance to mineralocorticoids. Such therapy will reduce vasopressin and ACTH levels and lower the dosage of glucocorticoid required. The need for continuing mineralocorticoids should be assessed based on plasma renin activity (PRA) and blood pressure [47]. Although aldosterone levels are normal in patients with NSW CAH, these patients also benefit from mineralocorticoid replacement as it helps to decrease the dose of glucocorticoid required to suppress androgens. Hence, published guidelines recommend that all children with classic CAH be treated with fludrocortisone [48, 101]. Owing to the obligatory urinary sodium loss, sodium chloride supplementation should be provided to infants. Sodium chloride supplements are often needed in infancy at 1–3 g/day (17–51 mEq/day; 1 g = 17 mEq of sodium), divided with each feed [48, 76, 101]. Older infants and children generally do not require salt supplementation.