Corticosteroids—glucocorticoids

General information

Nomenclature

The two main classes of adrenal corticosteroids are properly known as glucocorticoids and mineralocorticosteroids. The former are often known by shorter names and are commonly referred to as “glucocorticoids”, “corticosteroids”, “corticoids”, or even simply “steroids”; the latter are often referred to as “mineralocorticoids”. Here we shall use the terms “glucocorticoids” and “mineralocorticoids”. When referring to both we shall use the term “corticosteroids”.

Relative potencies

The main human anti-inflammatory corticosteroid, the glucocorticoid cortisol (hydrocortisone), as secreted by the adrenal gland, has generally been replaced by related glucocorticoids of synthetic origin for therapeutic purposes. These Δ1-dehydrated glucocorticoids are designed to imitate the physiological hormone. They have marked glucocorticoid potency but only minor effects on sodium retention and potassium excretion; the relative glucocorticoid and mineralocorticoid potencies of the best-known compounds, insofar as these potencies are agreed, are compared in Table 1 .

Table 1

Relative potencies of glucocorticoids

Compound	Glucocorticoid potency relative to hydrocortisone	Mineralocorticoid potency	Equivalent doses (mg)
Cortisone	0.8	++	25
Hydrocortisone	1.0	++	20
Prednisone	4	+	5
Prednisolone	4	+	5
Methylprednisolone	5	0	4
Triamcinolone	5	0	4
Paramethasone	10	0	2
Fluprednisolone	10	0	1.5
Dexamethasone	30	0	0.75
Betamethasone	30	0	0.6

Over many years, a great deal of research has been devoted to producing better glucocorticoids for therapeutic use. Those endeavors have succeeded only in part; from the start the mineralocorticoid effects were sufficiently minor to be non-problematic; the fact that successive synthetic glucocorticoids had an increasing potency in terms of weight was not of direct therapeutic significance; and the most hoped-for aim, that of dissociating wanted from unwanted glucocorticoid effects has not been achieved [ ]. Most untoward effects, such as those due to the catabolic and gluconeogenic activities of the glucocorticoid family, probably cannot be dissociated entirely from the anti-inflammatory activity [ ] it is possible that myopathy and muscle wasting are actually more common when triamcinolone or dexamethasone are used, but this may merely reflect overdosage of these potent drugs. However, some progress in achieving a dissociation of effects has been made. Beclomethasone has a relatively greater local than systemic effect. Deflazacort, one of the few new glucocorticoids to have been developed in recent years, originally promised reduced intensity of adverse effects, for example on bone mineral density, but the early promise has not held up [ ]. Cloprednol seems to affect the hypothalamic–pituitary–adrenal axis much less than other glucocorticoids, and to cause less excretion of nitrogen and calcium [ ].

Uses

Most patients who are treated therapeutically with glucocorticoids do not have glucocorticoid deficiency. Adverse reactions to glucocorticoids depend very largely on the ways in which, and the purposes for which, they are used. There are four groups of uses.

1.
Substitution therapy is used in cases of primary and secondary adrenocortical insufficiency; the aim is to provide glucocorticoids and mineralocorticoids in physiological amounts, and the better the dosage regimen is adapted to the individual’s needs, the less the chance of adverse effects [ ].
2.
Anti-inflammatory and immunosuppressive therapy exploits the immunosuppressive, anti-allergic, anti-inflammatory, anti-exudative, and anti-proliferative effects of the glucocorticoids [ ]. The desired pharmacodynamic effects reflect a general influence of these substances on the mesenchyme, where they suppress reactions that result in the symptoms of inflammation, exudation, and proliferation; the nonspecific effects of glucocorticoids on the mesenchyme are part of their physiological actions, but they can only be obtained to a clinically useful extent by using dosages at which the more specific (and unwanted) physiological effects also occur. High doses sufficient to suppress immune reactions are used in patients who have undergone organ transplantation.
3.
Hormone suppression therapy can be used, for example, to inhibit the adrenogenital syndrome [ ]. Higher doses are used. The treatment of the adrenogenital syndrome is only partly substitutive and has to be adapted to the individual case, but doses are needed at which various hormonal effects of the glucocorticoids and mineralocorticoids are likely to become troublesome.
4.
Massive doses of glucocorticoids, far exceeding physiological amounts, are given in the immediate management of anaphylaxis, although their beneficial effects are delayed for several hours. This is because, in severely ill patients, early administration of hydrocortisone 100–300 mg as the sodium succinate salt can gradually enhance the actions of adrenaline [ ]. Glucocorticoids have been used as an adjunct to the use of inotropic and vasopressor drugs for septic shock. Their efficacy, as well as their proposed mechanisms of action, is controversial; inhibition of complement-mediated aggregation and resultant endothelial injury, and inhibition of the release of beta-endorphin are current theories of their mechanism of action. However, controlled studies have not indicated a beneficial effect of high-dose glucocorticoid therapy in treating septic shock [ , ]. Hence, there is no established role for glucocorticoids in the treatment of shock, except shock caused by adrenal insufficiency.

Routes of administration

Glucocorticoids can be given by many different routes. The following routes are dealt with in this monograph:

oral;
rectal;
intravenous;
intramuscular;
intradermal injection;
intracapsular (breast) injection.

Separate monographs deal with the following routes:

epidural and intrathecal injection;
inhalation;
intra-articular and periarticular injection;
intralesional injection;
intranasal administration;
topical administration (skin, eyes).

General adverse effects and adverse reactions

The incidence and severity of adverse reactions to glucocorticoids depend on the dose and duration of treatment. Even the very high single doses of glucocorticoids, such as methylprednisolone, which are sometimes used, do not cause serious adverse effects, whereas an equivalent dose given over a long period of time can cause many long-term effects.

The two major risks of long-term glucocorticoid therapy are adrenal suppression and Cushingoid changes. During prolonged treatment with anti-inflammatory doses, glucose intolerance, osteoporosis, acne vulgaris, and a greater or lesser degree of mineralocorticoid-induced changes can occur. In children, growth can be retarded, and adults who take high doses can have mental changes. There may be a risk of gastroduodenal ulceration, although this is much less certain than was once thought. Infections and abdominal crises can be masked. Some of these effects reflect the catabolic properties of the glucocorticoids, that is their ability to accelerate tissue breakdown and impair healing. Allergic reactions can occur.

Anyone who prescribes long-term glucocorticoids should have a checklist in mind of the undesired effects that they can exert, both during treatment and on withdrawal, so that any harm that occurs can be promptly detected and countered. The main groups of risks arising from long-term treatment with glucocorticoids are summarized in Table 2 .

Table 2

Risks of long-term glucocorticoid therapy

1. Cushing’s syndrome

Moon face (facial rounding)

Central obesity

Striae

Hirsutism

Acne vulgaris

Ecchymoses

Hypertension

Osteoporosis

Proximal myopathy

Disorders of sexual function

Diabetes mellitus

Hyperlipidemia

Disorders of mineral and fluid balance (depending on the type of glucocorticoid)

2. Adrenal insufficiency

Insufficient or absent stress reaction

Withdrawal effects

3. Unwanted results accompanying desired effects

Increased risk of infection

Impaired wound healing

Peptic ulceration, bleeding, and perforation

Growth retardation

4. Other adverse reactions

Mental disturbances

Encephalopathy

Increased risk of thrombosis

Posterior cataracts

Increased intraocular pressure and glaucoma

Aseptic necrosis of bone

The adverse reactions that were reported in a study of 213 children are listed in Table 3 [ ].

Table 3

Adverse reactions in 213 children given intravenous methylprednisolone

Adverse reaction	Number
Behavioral changes	21
Abdominal disorders	11
Pruritus	9
Urticaria	5
Hypertension	5
Bone pain	3
Dizziness	3
Fatigue	2
Fractures	2
Hypotension	2
Lethargy	2
Tachycardia	2
Anaphylactoid reaction	1
“Grey appearance”	1

Drug studies

Observational studies

A study has been undertaken to clarify whether glucocorticoid excess affects endothelium-dependent vascular relaxation in glucocorticoid treated patients and whether dexamethasone alters the production of hydrogen peroxide and the formation of peroxynitrite, a reactive molecule between nitric oxide and superoxide, in cultured human umbilical endothelial cells [ ]. Glucocorticoid excess impaired endothelium-dependent vascular relaxation in vivo and enhanced the production of reactive oxygen species to cause increased production of peroxynitrite in vitro. Glucocorticoid-induced reduction in nitric oxide availability may cause vascular endothelial dysfunction, leading to hypertension and atherosclerosis.

All consecutive patients (n = 88) starting long-term (≥ 3 months), high-dosage prednisone therapy (≥ 20 mg/day) were enrolled in a study of the frequency of adverse events, susceptibility factors, and patients’ opinions in two French centers [ ]. Lipodystrophy was the most frequent adverse event (63%, range 51–73); it was considered the most distressing by the patients and was most frequent in women and young patients. Neuropsychiatric disorders occurred in 42 patients (53%, 41–64), and needed hospitalization in five cases. There were skin disorders in 37 patients (46%, (35–58), more often in women. Muscle cramps and proximal muscle weakness were reported by 32% (22–44) and 15% (8–25) respectively. New hypertension occurred in 8.7% (2.9–20). Lastly, 39% (20–61) of the premenopausal women reported menstrual disorders.

Long-term adverse effects of glucocorticoids are well characterized, but there are few data on the incidence and likelihood of short-term adverse effects. The records of all eligible kidney or pancreas–kidney transplant recipients have been examined, to determine the incidence of adverse effects potentially related to early administration of glucocorticoids and to determine the probability that the adverse effect was due to the glucocorticoid [ ]. Adverse reactions were identified in all 103 patients by an average of 8 days. The mean number of adverse reactions per patient was 3.26. There was weight gain in 80%, hypertension in 72%, diabetes mellitus in 52%, hyperglycemia in 48, leukocytosis in 31%, insomnia in 27%, anxiety in 11%, and psychosis in 1.9%. According to the Naranjo algorithm, leukocytosis was judged as “probable” and weight gain and psychosis were “possible to probable”. Diabetes, hyperglycemia, hypertension, and insomnia were “possible”, and anxiety was “possible to doubtful”.

Comparative studies

In another randomized trial, the effects and adverse effects of early dexamethasone on the incidence of chronic lung disease have been evaluated in 50 high-risk preterm infants [ ]. The treated infants received dexamethasone intravenously from the fourth day of life for 7 days (0.5 mg/kg/day for the first 3 days, 0.25 mg/kg/day for the next 3 days, and 0.125 mg/kg/day on the seventh day). The incidence of chronic lung disease at 28 days of life and at 36 weeks of postconceptional age was significantly lower in the infants who were given dexamethasone, who also remained intubated and required oxygen therapy for a shorter period. Hyperglycemia, hypertension, growth failure, and left ventricular hypertrophy were the transient adverse effects associated with early glucocorticoid administration. Early dexamethasone administration may be useful in preventing chronic lung disease, but its use should be restricted to preterm high-risk infants.

Placebo-controlled studies

Patients taking glucocorticoids have an increased risk of infections, including those produced by opportunistic and rare pathogens. However, it has been suggested that glucocorticoid administration in severe community-acquired pneumonia could attenuate systemic inflammation and lead to earlier resolution of pneumonia and a reduction in sepsis-related complications. In a placebo-controlled study in 46 patients with severe community-acquired pneumonia who received protocol-guided antibiotic treatment hydrocortisone (intravenous 200 mg bolus followed by infusion at a rate of 10 mg/hour) for 7 days produced significant clinical improvement [ ]. Adverse effects were not described.

Although there have been several trials of early dexamethasone to determine whether it would reduce mortality and chronic lung disease in infants with respiratory distress, the optimal duration and adverse effects of such therapy are unknown. The purpose of one study was: (a) to determine if a 3-day course of early dexamethasone therapy would reduce chronic lung disease and increase survival without chronic lung disease in neonates who received surfactant therapy for respiratory distress syndrome and (b) to determine the associated adverse effects [ ]. This was a prospective, placebo-controlled, multicenter, randomized study of a 3-day course of early dexamethasone therapy, beginning at 24–48 hours of life in 241 neonates, who weighed 500–1500 g, had received surfactant therapy, and were at significant risk of chronic lung disease or death. Infants randomized to dexamethasone received a 3-day tapering course (total dose 1.35 mg/kg) given in six doses at 12-hour intervals. Chronic lung disease was defined by the need for supplementary oxygen at a gestational age of 36 weeks. Neonates randomized to early dexamethasone were more likely to survive without chronic lung disease (RR = 1.3; CI = 1.0, 1.7) and were less likely to develop chronic lung disease (RR = 0.6; CI = 0.3, 0.98). Mortality rates were not significantly different. Subsequent dexamethasone therapy was less in early dexamethasone-treated neonates (RR = 0.8; CI = 0.70, 0.96). Very early (before 7 days of life) intestinal perforations were more common among dexamethasone-treated neonates (8 versus 1%). The authors concluded that an early 3-day course of dexamethasone increases survival without chronic lung disease, reduces chronic lung disease, and reduces late dexamethasone therapy in high-risk, low birthweight infants who receive surfactant therapy for respiratory distress syndrome. The potential benefits of early dexamethasone therapy in the regimen used in this trial need to be weighed against the risk of early intestinal perforation.

Although dexamethasone is commonly associated with transient adverse effects, several randomized trials have shown that it rapidly reduces oxygen requirements and shortens the duration of ventilation. A randomized study was designed to evaluate the effects of two different dexamethasone courses on growth in preterm infants [ ]. The first phase included 30 preterm infants at high risk of chronic lung disease, of whom 15 (8 boys) were given dexamethasone for 14 days, from the tenth day of life; they received a total dose of 4.75 mg/kg; 15 babies were assigned to the control group (8 boys). The second phase included 30 preterm infants at high risk of chronic lung disease, of whom 15 babies (7 boys) were treated with dexamethasone for 7 days, from the fourth day of life; they received a total dose of 2.38 mg/kg; 15 babies were assigned to the control group (9 boys). Infants given dexamethasone had significantly less weight gain than controls, but they caught up soon after the end of treatment. At 30 days of life, the gains in weight and length in each group were similar to those in control infants, but those given dexamethasone had significantly less head growth. There were no differences between the groups at discharge. The longer-term impact of postnatal dexamethasone on mortality and morbidity is less clear. Better data, from larger clinical trials with longer follow-up, will determine whether this kind of treatment enhances lives, makes little difference, causes significant harm, or does several of these things [ ].

Systematic reviews

A systematic review of glucocorticoid adjunctive therapy in adults with acute bacterial meningitis has been published [ ]. Five trials involving 623 patients were included (pneumococcal meningitis = 234, meningococcal meningitis = 232, others = 127, unknown = 30). Treatment with glucocorticoids was associated with a significant reduction in mortality (RR = 0.6; 95% CI = 0.4, 0.8) and in neurological sequelae (RR = 0.6; 95% CI = 0.4, 1), and with a reduction in case-fatality in pneumococcal meningitis of 21% (RR = 0.5; 95% CI = 0.3, 0.8). In meningococcal meningitis, mortality (RR = 0.9; 95% CI = 0.3, 2.1) and neurological sequelae (RR = 0.5; 95% CI = 0.1, 1.7) were both reduced, but not significantly. Adverse events were similar in the treatment and placebo groups (RR = 1; CI = 0.5, 2), with gastrointestinal bleeding in 1% of glucocorticoid-treated patients and 4% of the rest. The authors recommended the early use of glucocorticoid therapy in adults in whom acute community-acquired bacterial meningitis is suspected.

A systematic review of randomized controlled trials has been performed to determine whether dexamethasone therapy in the first 15 days of life prevents chronic lung disease in premature infants [ ]. Studies were identified by a literature search using Medline (1970–97) supplemented by a search of the Cochrane Library (1998, Issue 4). Inclusion criteria were: (a) prospective randomized design with initiation of dexamethasone therapy within the first 15 days of life; (b) report of the outcome of interest; and (c) less than 20% crossover between the treatment and control groups during the study period. The primary outcomes were mortality at hospital discharge and the development of chronic lung disease at 28 days of life and 36 weeks postconceptional age. The secondary outcomes were the presence of a patent ductus arteriosus and treatment adverse effects. Dexamethasone reduced the incidence of chronic lung disease by 26% at 28 days (RR = 0.74; CI = 0.57, 0.96) and 48% at 36 weeks postconceptional age (RR = 0.52; CI = 0.33, 0.81). These reductions were more significant when dexamethasone was started in the first 72 hours of life. The 24% relative risk reduction of deaths was marginally significant (RR = 0.76; CI = 0.56, 1.04). The 27% reduction in patent ductus arteriosus and the 11% increase in infections were not statistically significant, nor were any other changes. The conclusion from this meta-analysis was that systemic dexamethasone given to at-risk infants soon after birth may reduce the incidence of chronic lung disease. There was no evidence of significant short-term adverse effects.

The effects of systemic glucocorticoids in chronic obstructive pulmonary disease (COPD) have been summarized in two Cochrane systematic reviews assessing chronic stable and acute exacerbations of the disease [ ]. In stable COPD, compared with placebo, oral glucocorticoid treatment increased mean FEV ₁ by 53 ml and mean 12-minute walking distance by 29 m, but at an increased risk of any drug-related adverse event (OR = 7.7; 95% CI = 2.3, 25.7). In acute exacerbations, oral glucocorticoids reduced the risk of treatment failure (OR = 0.48; 95% CI = 0.34, 0.68), improved mean FEV ₁ at 72 hours by 140 ml, and improved arterial blood gases, but increased the risk of drug-related adverse events (OR = 2.3; 95% CI = 1.6, 3.3). The authors commented that treatment of stable and acute exacerbations of COPD with systemic glucocorticoids results in statistically significant average benefits, but at an increased risk of adverse effects. In stable COPD, there is little support for the use of systemic glucocorticoid treatment, as data on long-term outcomes are lacking. For acute exacerbations, the evidence in support of the use of systemic glucocorticoids is stronger, but further research is required to define the optimum dose, route, and duration.

Organs and systems

Cardiovascular

The considerable body of evidence that glucocorticoids can cause increased rates of vascular mortality and the underlying mechanisms (increased blood pressure, impaired glucose tolerance, dyslipidemia, hypercoagulability, and increased fibrinogen production) have been reviewed [ ]. In view of their adverse cardiovascular effects, the therapeutic options should be carefully considered before long-term glucocorticoids are begun; although they can be life-saving, dosages should be regularly reviewed during long-term therapy, in order to minimize complications.

The benefit of glucocorticoid therapy is often limited by several adverse reactions, including cardiovascular disorders such as hypertension and atherosclerosis. Plasma volume expansion due to sodium retention plays a minor role, but increased peripheral vascular resistance, due in part to an increased pressor response to catecholamines and angiotensin II, plays a major role in the pathogenesis of hypertension induced by glucocorticoid excess. However, the molecular mechanism remains unclear.

Long-term systemic administration of glucocorticoids might be expected, because of their effects on vascular fragility and wound healing, to increase the risk of vascular complications during percutaneous coronary intervention. To assess the potential risk of long-term glucocorticoid use in the setting of coronary angioplasty, 114 of 12 883 consecutively treated patients who were taking long-term glucocorticoids were compared with those who were not. Glucocorticoid use was not associated with an increased risk of composite events of major ischemia but was associated with a threefold risk of major vascular complications and a three-to fourfold risk of coronary perforation [ ].

Hypertension

The secondary mineralocorticoid activity of glucocorticoids can lead to salt and water retention, which can cause hypertension. Although the detailed mechanisms are as yet uncertain, glucocorticoid-induced hypertension often occurs in elderly patients and is more common in patients with total serum calcium concentrations below the reference range and/or in those with a family history of essential hypertension [ ].

Hemangioma is the most common tumor of infancy, with a natural history of spontaneous involution. Some hemangiomas, however, as a result of their proximity to vital structures, destruction of facial anatomy, or excessive bleeding, can be successfully treated with systemic glucocorticoids between other therapies. The risk of hypertension is poorly documented in this setting. In one prospective study of 37 infants (7 boys, 17 girls; mean age 3.5 months, range 1.5–10) with rapidly growing complicated hemangiomas treated with oral prednisone 1–5 mg/kg/day, blood pressure increased in seven cases [ ]. Cardiac ultrasound examination in five showed two cases of myocardial hypertrophy, which was unrelated to the hypertension and which regressed after withdrawal of the prednisone.

Myocardial ischemia

Cortisone-induced cardiac lesions are sometimes reported and electrocardiographic changes have been seen in patients taking glucocorticoids [ ]. Whereas abnormal myocardial hypertrophy in children has perhaps been associated more readily with corticotropin, it has been seen on occasion during treatment with high dosages of glucocorticoids, with normalization after dosage reduction and withdrawal.

Fatal myocardial infarction occurred after intravenous methylprednisolone for an episode of ulcerative colitis [ ].

A day after a dose of intravenous methylprednisolone 60 mg a 79-year-old woman developed acute thoracic pain and collapsed. An electrocardiogram showed signs of a myocardial infarction and her cardiac enzyme activities were raised. She died within several hours. Autopsy showed an anterior transmural myocardial infarction and mild atheromatous lesions in the coronary arteries.

This report highlights the risk of cardiovascular adverse effects with short courses of glucocorticoid therapy in elderly patients with inflammatory bowel disease, even with rather low-dosage regimens. Acute myocardial infarction occurred in an old man with coronary insufficiency and giant cell arteritis after treatment with prednisolone [ ] but could well have been coincidental.

Myocardial ischemia has been reportedly precipitated by intramuscular administration of betamethasone [ ]. It has been suggested that long-term glucocorticoid therapy accelerates atherosclerosis and the formation of aortic aneurysms, with a high risk of rupture [ ].

Patients with seropositive rheumatoid arthritis taking long-term systemic glucocorticoids are at risk of accelerated cardiac rupture in the setting of transmural acute myocardial infarction treated with thrombolytic drugs [ ].

Two women and one man, aged 53–74 years, died after they received thrombolytic therapy for acute myocardial infarction. All three had a long history of seropositive rheumatoid arthritis treated with prednisone 5–20 mg/day for many years.

In a nested case-control study of the risk of acute myocardial infarction associated with oral glucocorticoids in 404 183 patients, aged 50–84 years, without cancer from the general UK population 4795 deaths from acute myocardial infarction or coronary heart disease were included [ ]. A sample of 20 000 controls was randomly selected, frequency matched by age, sex, and calendar year. Relative risks were estimated using unconditional logistic regression. The adjusted odds ratio for acute myocardial infarction in current users of oral glucocorticoids compared with non-users was 1.42 (95% CI = 1.17, 1.72). The risk during the first 30 days of use (OR = 2.24; 95% CI = 1.56, 3.20) was greater than with longer duration (OR = 1.22; 95% CI = 0.98, 1.52). The risk was more pronounced (OR = 2.15; 95% CI = 1.45, 3.14) among users of prednisolone equivalent doses over 10 mg/day. The dose effect was observed among patients with and without coronary heart disease or chronic obstructive pulmonary disease/asthma. These results suggest a small increased risk of acute myocardial infarction with oral glucocorticoids, with a greater risk in users of high doses.

Oral glucocorticoids reduce the risk of re-stenosis after percutaneous coronary intervention. Of 220 patients, 28 had adverse effects that were probably related to prednisone: gastric pain (4%), increased arterial pressure needing increased antihypertensive treatment (4%), edema (1.8%), and concomitant infections (1.4%) [ ]. In three asymptomatic patients (1.4%), who were also taking a thienopyridine, there was neutropenia, detected during routine blood cell count 4 weeks after the intervention. It resolved completely after withdrawal of prednisone and thienopyridine in all cases. Neutropenia after prednisone had never been reported before. However, a direct cause-and-effect relation between thienopyridines or prednisone (alone or in combination) and neutropenia cannot be proved because of lack of rechallenge. The authors hypothesized that prednisone, by increasing CYP3A4 activity, may enhance the formation of an active thienopyridine metabolite, possibly amplifying the risk of adverse effects.

Cardiomyopathy

Postnatal exposure to glucocorticoids has been associated with hypertrophic cardiomyopathy in neonates. Such an effect has not previously been described in infants born to mothers who received antenatal glucocorticoids. Three neonates (gestational ages 36, 29, and 34 weeks), whose mothers had been treated with betamethasone prenatally in doses of 12 mg twice weekly for 16 doses, 8 doses, and 5 doses respectively, developed various degrees of hypertrophic cardiomyopathy diagnosed by echocardiography [ ]. There was no maternal evidence of diabetes, except for one infant whose mother had a normal fasting and postprandial blood glucose before glucocorticoid therapy, but an abnormal 1-hour postprandial glucose after 8 weeks of betamethasone therapy, with a normal HbA1C concentration. There was no family history of hypertrophic cardiomyopathy, no history of maternal intake of other relevant medications, no hypertension, and none of the infants received glucocorticoids postnatally. Follow-up echocardiography showed complete resolution in all infants. The authors suggested that repeated antenatal maternal glucocorticoids might cause hypertrophic cardiomyopathy in neonates. These changes appear to be dose-and duration-related and are mostly reversible.

Transient hypertrophic cardiomyopathy is a rare sequel of the concurrent administration of glucocorticoid and insulin excess [ ]. The heart is also almost certainly a site for myopathic changes analogous to those that affect other muscles.

Transient hypertrophic cardiomyopathy has been attributed to systemic glucocorticoid administration for a craniofacial hemangioma [ ].

A 69-day-old white child presented with a rapidly growing 2.5 × 1.5 cm hemangioma of the external left nasal side wall. He was normotensive and there was no family history of cardiomyopathy or maternal gestational diabetes. Because of nasal obstruction and possible visual obstruction, he was given prednisolone 3 mg/kg/day. After 10 weeks his weight had fallen from 7.6 to 7.1 kg and 2 weeks later he became tachypneic with a respiratory rate of 40/minutes. A chest X-ray showed cardiomegaly and pulmonary venous congestion. An echocardiogram showed hypertrophic cardiomyopathy. The left ventricular posterior wall thickness was 10 mm (normal under 4 mm), and the peak left ventricular outflow gradient was 64 mmHg. He was given a β-blocker and a diuretic and the glucocorticoid dose was tapered. The cardiomyopathy eventually resolved.

Dilated cardiomyopathy caused by occult pheochromocytoma has been described infrequently.

A 34-year-old woman had acute congestive heart failure 12 hours after administration of dexamethasone 16 mg for an atypical migraine [ ]. The authors postulated that the acute episode had been induced by the dexamethasone, which increased the production of adrenaline, causing β ₂ -adrenoceptor stimulation, peripheral vasodilatation, and congestive heart failure.

In an addendum the authors reported another similar case.

Obstructive cardiomyopathy has been attributed to a glucocorticoid in a child with subglottal stenosis [ ].

A 4-month-old boy (weight 4 kg) developed fever, nasal secretions, and stridor due to a subglottal granuloma. Dexamethasone 1 mg/kg/day was started and tapered over 1 week. The mass shrank to 25% of its original size but the symptoms recurred 2 weeks later. The granuloma was excised and dexamethasone 1 mg/kg/day was restarted. After 5 days he developed a tachycardia (140/minute) and a new systolic murmur. Echocardiography showed severe ventricular hypertrophy with dynamic left ventricular outflow tract obstruction. The dexamethasone was weaned over several days. Over the next 3 weeks several echocardiograms showed rapid resolution of the outflow tract obstruction and gradual improvement of the cardiac hypertrophy. After 8 months there was no further problem.

Cardiac dysrhythmias

Serious cardiac dysrhythmias and sudden death have been reported with pulsed methylprednisolone. Oral methylprednisolone has been implicated in a case of sinus bradycardia [ ].

A 14-year-old boy received an intravenous dose of methylprednisolone 30 mg/kg for progressive glomerulonephritis. After 5 hours, his heart rate had fallen to 50/minute and an electrocardiogram showed sinus bradycardia. His heart rate then fell to 40/minutes and a temporary transvenous pacemaker was inserted and methylprednisolone was withdrawn. His heart rate increased to 80/minutes over 3 days. After a further 3 days, he was treated with oral methylprednisolone 60 mg/m ² /day and his heart rate fell to 40/minutes in 5 days. Oral methylprednisolone was stopped on day 8 of treatment and his heart rate normalized.

Hypokalemia, secondary to mineralocorticoid effects, can cause cardiac dysrhythmias and cardiac arrest.

Recurrent cardiocirculatory arrest has been reported [ ].

A 60-year-old white man was admitted for kidney transplantation. Immediately after reperfusion and intravenous methylprednisolone 500 mg, he developed severe bradycardia with hypotension and then cardiac arrest. After resuscitation, his clinical state improved quickly, but on the morning of the first postoperative day directly after the intravenous administration of methylprednisolone 250 mg, he had another episode of severe bradycardia, hypotension, and successful cardiopulmonary resuscitation. A third episode occurred 24 hours later after intravenous methylprednisolone 100 mg, again followed by rapid recovery after resuscitation. Two weeks later, during a bout of acute rejection, he was given intravenous methylprednisolone 500 mg, after which he collapsed and no heartbeat or breathing was detectable; after cardiopulmonary resuscitation he was transferred to the intensive care unit, where he died a few hours later.

If patients at risk are identified, glucocorticoid bolus therapy should be avoided or, if that is not possible, should only be done under close monitoring.

Andersen’s syndrome is a form of long QT syndrome. It is a rare genetic disorder, inherited in an autosomal dominant pattern, and caused by mutations in the KCNJ2 gene that encodes for Kir2.1, an inward rectifier potassium channel. Periodic paralysis, cardiac arrhythmias, and bone features are the hallmarks of this syndrome. Rest following strenuous physical activity, carbohydrate ingestion, emotional stress, and exposure to cold are precipitating triggers. Two families with this disorder have been identified [ ]. They had periodic paralysis and cardiac abnormalities, but only discrete developmental features. One patient reported having had symptoms twice during the day after glucocorticoids treatment and alleviated after withdrawal.

Cardiac rhythm disturbances, both tachydysrhythmias and bradydysrhythmias, have been reported in adults after high-dose intravenous pulse methylprednisolone, but much less often in children. Five children with rheumatic diseases developed sinus bradycardia during daily therapy with intravenous pulse methylprednisolone [ ]. There were reductions in resting heart rate of 35–50% of baseline in each case. All the patients were asymptomatic and all recovered spontaneously over a variable period of time after the end of pulse therapy.

Pericarditis

Disseminated Varicella and staphylococcal pericarditis developed in a previously healthy girl after a single application of triamcinolone cream 0.1% to relieve pruritus associated with Varicella skin lesions [ ].

Vasculitis

Long-term treatment with glucocorticoids can cause arteritis, but patients with rheumatoid arthritis have a special susceptibility to vascular reactions, and cases of periarteritis nodosa after withdrawal of long-term glucocorticoids have been reported [ ].

Respiratory

Local adverse effects are common in patients with asthma who use inhaled glucocorticoids, as suggested by a survey of the prevalence of throat and voice symptoms in patients with asthma using glucocorticoids by metered-dose pressurized aerosol [ ].

There have been no reports of an increased frequency of lower respiratory tract infections. However, patients with aspiration of gastric material who were treated with glucocorticoids did not have improved survival but had a higher incidence of pneumonia [ ].

In cases of pneumothorax with closed thoracotomy tube drainage, chronic glucocorticoid treatment has been reported to delay and impede re-expansion of the lung [ ].

Hiccup is a rare complication of glucocorticoid therapy; five cases have been published at various times [ ].

A 59-year-old man had intractable hiccups during treatment with dexamethasone for multiple myeloma [ ].
Persistent hiccupping has been described in a 30-year-old man after the administration of a single intravenous dose of dexamethasone (16 mg) [ ]. The symptom was resistant to metoclopramide and resolved spontaneously after 4 days. On rechallenge, the hiccups recurred within 2 hours and disappeared after 36 hours.

Low-dose metoclopramide can be effective and may allow a patient to continue beneficial therapy without the discomfort and exhaustion that can accompany intractable hiccups.

Ear, nose, throat

Atrophic changes and fungal and other infections can alter the nasal mucosa after aerosol treatment [ ], and since most systematic published documentation on these intranasal products is limited to 1–2 years of experience (although they have been in use for a far longer period), some reserve is warranted with respect to their long-term safety and the wisdom of continual use.

Nervous system

Cerebral venous thrombosis associated with glucocorticoid treatment has rarely been reported. A relation between glucocorticoids and venous thrombosis has already been suggested but has never been clearly understood. Three young patients, two women (aged 28 and 45) and one man (aged 38 years), developed cerebral venous thrombosis after intravenous high-dose glucocorticoids [ ]. All presented with probable multiple sclerosis according to clinical, CSF, and MRI criteria. All had a lumbar puncture and were then treated with methylprednisolone 1 g/day for 5 days. All the usual causes of cerebral venous thrombosis were systematically excluded. The authors proposed that glucocorticoids interfere with blood coagulation and suggested that the administration of glucocorticoids after a lumbar puncture carries a particular risk of complications.

Dexamethasone is widely used for the prevention and treatment of chronic lung disease in premature infants, in whom follow-up studies have raised the possibility of an association with alterations in neuromotor function and somatic growth. In 159 survivors (mean age 53 months) of a previous placebo-controlled study, the children who had received dexamethasone had a significantly higher incidence of cerebral palsy (39/80 versus 12/79; OR = 4.62; 95% CI = 2.38, 8.98) [ ]. The most common form of cerebral palsy was spastic diplegia. Developmental delay was more frequent in the dexamethasone group (44/80 versus 23/79; OR = 2.9; CI = 1.5, 5.4). In a systematic review the authors concluded that postnatal dexamethasone at currently recommended doses should be avoided because of long-term neurological adverse effects [ ]. Lower doses of dexamethasone or inhaled glucocorticoids might be indicated for ill ventilator-dependent infants with chronic lung disease after the age of 2 weeks.

In 146 children who participated in a placebo-controlled trial of early postnatal dexamethasone therapy for the prevention of the chronic lung disease of prematurity, follow-up at school age (mean age 8 years old) showed that the children who had received dexamethasone were significantly shorter than the controls (mean height 122.8 cm versus 126.4 cm for boys and 121.3 cm versus 124.7 cm for girls) and had a significantly smaller head circumference (49.8 cm versus 50.6 cm) [ ]. They also had significantly poorer motor skills, motor coordination, and visuomotor integration. Compared with the controls, the children who had received dexamethasone had significantly lower IQ scores, including full scores (mean 78.2 versus 84.4), verbal scores (84.1 versus 88.4), and performance scores (76.5 versus 84.5). The frequency of clinically significant disabilities was higher among the children who had received dexamethasone than among the controls (39% versus 22%). The authors did not recommend the routine use of dexamethasone therapy for the prevention or treatment of chronic lung disease.

Long-term treatment with glucocorticoids can cause cerebral atrophy [ ].

Severe organic brain syndrome has been seen in six patients taking long-term glucocorticoids [ ]. The manifestations included confusion, disorientation, apathy, confabulation, irrelevant speech, and slow thinking; the symptoms occurred abruptly.

Latent epilepsy can be made manifest by glucocorticoid treatment. Seizures in patients with lung transplants were related to glucocorticoids, which had been used in high dosages to prevent organ rejection. There was an increased risk of seizures in younger patients (under 25 years) and with intravenous methylprednisolone [ ].

Long-term glucocorticoid treatment can result in papilledema and increased intracranial pressure (the syndrome of pseudotumor cerebri or so-called “benign intracranial hypertension”), particularly in children.

Benign intracranial hypertension occurred in a 7-month-old child after withdrawal of topical betamethasone ointment and in a 7-year-old boy treated with a 1% cortisol ointment in large amounts.
A 6-year-old girl, who had taken prednisone for 2.5 years for nephrotic syndrome with seven relapses in 3 years, developed symptoms of benign intracranial hypertension after oral glucocorticoid dosage reduction over 10 months from 30 mg/day to 2.5 mg/every other day [ ]. Laboratory studies and head CT scan were normal, but there was bilateral papilledema and the cerebrospinal fluid pressure was increased. She was given prednisone 1 mg/kg/day initially, with acetazolamide, and 25 ml of cerebrospinal fluid was removed. All her symptoms resolved and treatment was gradually withdrawn. She developed no further visual failure.

The symptoms can simulate those of an intracranial tumor. All patients taking large doses of glucocorticoids who complain of headache or blurred vision, particularly after a reduction in dosage, should have an ophthalmoscopic examination to exclude this complication. Paradoxically, cerebral edema occurring during a surgical procedure can be partly prevented by glucocorticoids [ ].

An encephalopathy can occur at any age [ ], not necessarily in association with intracranial hypertension.

There have been repeated reports of epidural lipomatosis, which can lead to spinal cord compression [ , ] or spinal fracture [ ]; in one instance, the excised lipomata contained brown fat, a phenomenon that may prove to be not unusual in glucocorticoid-induced lipomata [ ].

A 40-year-old woman with ulcerative colitis took cortisone 20 mg/day and developed progressive paraplegia [ ]. There was kyphosis of the thoracic spine from T7 to T9, with pathological fractures. An MRI scan showed massive epidural fat extending from T1 to T9. She recovered 3 months after surgical removal of the epidural fat.
A 78-year-old man was given methylprednisolone (60 mg/day reducing to 8 mg/day) for temporal arteritis [ ]. After 4 months, he developed numbness and paresis of the legs and hyperalgesia at dermatomes T3 and T4. After 10 months he had marked disturbance of proprioception combined with spinal ataxia and an increasing loss of motor bladder control. There was an intraspinal epidural lipoma in the dorsal part of the spine from T1 to T10. The fat was removed surgically and within 4 weeks his gait disturbance and proprioception improved, the sensory deficit abated, and the bladder disorder disappeared completely.
A 57-year-old man took prednisone 20–30 mg/day for 13 years for rheumatoid arthritis [ ]. He had been treated unsuccessfully with gold, azathioprine, hydroxychloroquine, and sulfasalazine; tapering his glucocorticoid dosage had been unsuccessful. He developed worsening back pain in his thoracic spine and lateral leg weakness. He was unable to walk. He was Cushingoid and had marked thoracic kyphosis associated with multiple vertebral body fractures in T5–T8. An MRI scan at T5–T6 showed displacement and compression of the spinal cord by high-signal epidural fat, which had caused anterior thecal displacement and total effacement of cerebrospinal fluid.

The authors of the last report commented on the high dose of prednisone used.

Glucocorticoid-induced spinal epidural lipomatosis is not very common in children. Spinal magnetic resonance imaging was performed in 125 children with renal diseases (68 boys); they had either back pain or numbness, were obese, or had taken a cumulative dose of prednisone of more than 500 mg/kg; there was lipomatosis in five patients [ ].

In the past there was reason to think that glucocorticoids might precipitate multiple sclerosis. However, this has not been confirmed, and there is evidence that a special glucocorticoid regimen can actually be capable of retarding deterioration in multiple sclerosis [ ].

A Guillain–Barré-like syndrome occurred in a patient receiving high-dose intravenous glucocorticoid therapy [ ]. Although glucocorticoids have been used successfully to treat weakness due to chronic inflammatory demyelinating sensorimotor neuropathy, other types of acquired chronic demyelinating neuropathies can be impaired by these drugs.

In four patients with a pure motor demyelinating neuropathy treated with oral prednisolone (60 mg/day) motor function rapidly deteriorated within 4 weeks of starting prednisolone [ ]. Intravenous immunoglobulin some months later in two of them produced clear improvement in strength and motor nerve conduction.

A rare but reversible complication of glucocorticoids therapy, posterior reversible encephalopathy syndrome, has been reported in patients with renal diseases or hematological malignancies and in patients with bone marrow and solid organ transplants. However, the patients were either acutely ill or had been exposed to other agents implicated in causing the syndrome (i.e., immunosuppressive agents or cytotoxic drugs). It has now been reported with high-dose dexamethasone in a patient who was not taking concomitant medications [ ].

Sensory systems

The eye can be involved in generalized adverse reactions to systemically administered glucocorticoids. For example, conjunctivitis can occur as part of an allergic reaction and infections of the eye can be masked as a result of anti-inflammatory and analgesic effects. Ophthalmoplegia can occur as one of the consequences of glucocorticoid myopathy [ ]. Two complications that require special discussion are cataract and glaucoma.

Cataract

Oral glucocorticoid treatment is a risk factor for the development of posterior subcapsular cataract. A review of nine studies including 343 asthmatics treated with oral glucocorticoids showed a prevalence of posterior subcapsular cataracts of 0–54% with a mean value of 9% [ ]. In a 1993 study in children taking low-dose prednisone there were cataracts in seven of 23 cases [ ]. Some studies have shown a clear correlation with the duration of treatment and total dosage, others have not [ ]. The use of inhaled glucocorticoids was associated with a dose-dependent increased risk of posterior subcapsular and nuclear cataracts in 3654 patients aged 49–97 years [ ]. Data on glucocorticoid use were available for 3313 of these patients; glucocorticoid use was classified as none in 2784 patients, inhaled only in 241, systemic only in 177, and both inhaled and systemic in 111. Compared with non-use, current or prior use of inhaled glucocorticoids was associated with a significant increase in the prevalence of nuclear cataracts (adjusted relative prevalence = 1.5; 95% CI = 1.2, 1.9) and posterior subcapsular cataracts (1.9; 1.3, 2.8), but not cortical cataracts. The increased prevalence of posterior subcapsular cataracts was significantly associated with current use of inhaled glucocorticoids (2.6; 1.7, 4.0); there was no association with past use. Current use of inhaled glucocorticoids was also associated with an increased prevalence of cortical cataracts (1.4; 1.1, 1.7). The highest prevalences of posterior subcapsular and grade 4 or 5 nuclear cataracts were found in patients who had taken a cumulative dose of beclomethasone over 2000 mg.

It has been suggested that the risk of cataract is higher in patients with rheumatoid arthritis than in patients with bronchial asthma, and it is also higher in children. The reversibility of the lenticular changes has often been discussed [ , ], but even without glucocorticoid withdrawal regression has been found in children taking long-term treatment [ ]. Nevertheless, some 7% of the patients who develop cataract caused by glucocorticoid treatment have to be operated on. A change in permeability of the lens capsule, followed by altered electrolyte concentrations in the lens and a change in the mucopolysaccharides in the lens have been advanced as reasons for the development of cataract.

Increased intraocular pressure and glaucoma

Ocular hypertension and open-angle glaucoma are well-known adverse reactions to ophthalmic administration of glucocorticoids [ ].Glucocorticoids can increase intraocular pressure, which can cause interlamellar stromal edema after laser in situ keratomileusis (LASIK) [ ].

Frequency . A total of 113 patients with angiographically proven subretinal neovascularization were enrolled into a prospective study of the effects of intravitreal triamcinolone [ ]. About 30% developed a significant rise in intraocular pressure (at least 5 mmHg) above baseline during the first 3 months.

In a large case-control study, in which 9793 elderly patients with ocular hypertension or open-angle glaucoma were compared with 38 325 controls, there was an increased risk of these complications with oral glucocorticoids [ ]. The risk of ocular hypertension or open-angle glaucoma increased with increasing dose and duration of use of the oral glucocorticoid. There was no significant increase in the risk of ocular hypertension or open-angle glaucoma in patients who had stopped taking oral glucocorticoids 15–45 days before. The authors estimated that the excess risk of ocular hypertension or open-angle glaucoma with current oral glucocorticoid use is 43 additional cases per 10 000 patients per year. However, in patients taking over 80 mg/day of hydrocortisone equivalents, the excess risk is 93 additional cases per 10 000 patients per year. Monitoring of intraocular pressure may be justified in long-term users of oral glucocorticoids, as it is in long-term users of topical glucocorticoids.

Prolonged use of high doses of inhaled glucocorticoids also increases the risk of ocular hypertension and open-angle glaucoma [ ]. In a case-control study of the records of 9793 elderly patients with ocular hypertension or open-angle glaucoma over a 6-year period, there was a significantly increased risk of ocular hypertension and open-angle glaucoma in patients who had taken high doses of inhaled glucocorticoids (1500–1600 micrograms) for 3 months or longer (OR = 1.44; 95% CI = 1.01, 2.06). Both a high dosage of inhaled glucocorticoid and prolonged continuous duration of therapy had to be present to increase the risk.

Pathogenesis . The pathogenesis of glucocorticoid-induced glaucoma is still unknown, but there is reduced outflow, and excessive accumulation of mucopolysaccharides may be a major factor. An association with cataract and papilledema has often been observed. The rise in intraocular pressure is variable: in the pediatric study of low dose cited above there was a reversible effect in only two of 23 subjects compared with controls, but in other studies serious increases in pressure have occurred, with a risk of blindness.

There is almost certainly a genetic predisposition to glucocorticoid-induced glaucoma, as there is to glaucoma in general.

Susceptibility factors . Children have more frequent, more severe, and more rapid ocular hypertensive responses to topical dexamethasone than adults. In one case a systemic glucocorticoid caused significant but asymptomatic ocular hypertension in a child [ ].

A 9-year-old girl with acute lymphoblastic leukemia received a 5-week course of oral prednisolone 60 mg/day (2.3 mg/kg/day). She did not receive any other systemic medications that have a known effect on intraocular pressure. Her baseline pressures in the right and left eyes were 16 and 17 mmHg with visual acuities of 20/20 and 20/15 respectively. The cup-to-disk ratio was 0.5 in both eyes, with normal visual fields. She was not myopic and had no family history of glaucoma or glucocorticoid responsiveness. After 8 days of systemic glucocorticoid therapy, her intraocular pressures increased to 39 mmHg and 38 mmHg in the right and left eyes respectively. Gonioscopy confirmed an open drainage angle in both eyes. She was given topical betaxolol 0.25% and dorzolamide 2% bd. However, her intraocular pressure continued to increase to 52 mmHg in the right eye and 47 mm Hg in the left eye on day 10. Topical latanoprost 0.001% od and brimonidine 0.2% bd were added, and the intraocular pressures fell to 38 mmHg and 36 mmHg. Two days after withdrawal of the prednisolone, the intraocular pressure returned rapidly to 17 mm Hg in both eyes. Over the next 6 weeks, this was maintained despite stepwise withdrawal of all glaucoma medications. Four months later, she was given a 4-week course of oral dexamethasone 10 mg/day and had similar patterns of changes in intraocular pressure. Oral acetazolamide was prescribed. She remained largely asymptomatic throughout, except for one episode of reduced visual acuity from 20/20 to 20/40 in the right eye when the intraocular pressure reached 52 mmHg.

Chorioretinopathy

Systemic glucocorticoid treatment can cause severe exacerbation of bullous exudative retinal detachment and lasting visual loss in some patients with idiopathic central serous chorioretinopathy [ ]. The atypical presentation of this condition can include peripheral retinal capillary non-perfusion and retinal neovascularization. The treatment of choice in patients with idiopathic central serous chorioretinopathy is laser photocoagulation.

In a prospective, case-control study 38 consecutive patients (28 men and 10 women), aged 28–63 years with central serous chorioretinopathy, were compared with 38 age-and sex-matched controls (28 men and 10 women) aged 27–65 years [ ]. Eleven patients (29%; eight men and three women) with central serous chorioretinopathy were taking glucocorticoids, compared with two patients (5.2%; one man and one woman) in the control group (OR = 7.33, 95% CI = 1.49, 36).

Endophthalmitis . Intravitreal triamcinolone injection is safe and effective for cystoid macular edema caused by uveitis, diabetic maculopathy, and central retinal vein occlusion, and for pseudophakic cystoid macular edema. Potential risks include glaucoma, cataract, retinal detachment, and endophthalmitis. Infectious endophthalmitis is extremely rare when appropriate sterile technique is practised. Seven patients developed a clinical picture simulating endophthalmitis after intravitreal injection of triamcinolone [ ]. The authors believed that this effect was a toxic reaction to the injected material and explained that the differential diagnosis of infectious endophthalmitis in eyes that have been injected with triamcinolone under sterile conditions includes a sterile toxic endophthalmitis that requires careful monitoring, perhaps every 8–12 hours, in order to determine whether the inflammation is worsening or improving. Resolution occurs spontaneously, and in the absence of eye pain unnecessary intervention can be avoided.

Pseudohypopyon and sterile endophthalmitis after intravitreal injection of triamcinolone for pseudophakic cystoid macular edema has been reported [ ].

An 88-year-old woman underwent phacoemulsification surgery, which was complicated by posterior capsule rupture. Anterior vitrectomy was performed, with implantation of a silicone intraocular lens into the sulcus. Postoperatively, she developed cystoid macular edema, which failed to respond to topical dexamethasone, topical ketorolac, and posterior sub-Tenon injection of triamcinolone, limiting visual acuity to 6/24 at 7 months after the surgery. An intravitreal injection of triamcinolone acetonide (4 mg in 0.1 ml) (Kenalog®, Bristol-Myers Squibb, Middlesex, UK) was administered through the pars plana with a 30-gauge needle using a sterile technique. Three days later she reported painless loss of vision, which had developed immediately after the injection. Visual acuity was reduced to perception of hand movements. There was minimal conjunctival injection and the cornea was clear. A 3 mm pseudohypopyon, consisting of refractile crystalline particles, was visible in the anterior chamber, associated with 3 + anterior chamber cells (or particles). Severe vitreous haze prevented visualization of the retina. Because infectious endophthalmitis could not be excluded, she was treated with intravitreal injections of ceftazidime and vancomycin. Vitreous and aqueous taps were performed and the pseudohypopyon was completely aspirated from the anterior chamber. The next day a 2 mm pseudohypopyon had reformed. The position of the pseudohypopyon depended on gravity and shifted with changes in head position. Aqueous and vitreous cultures were negative. Microscopy of the aspirated pseudohypopyon showed triamcinolone particles with no cells. The pseudohypopyon, vitreous haze, and cystoid macular edema (as demonstrated on optical coherence tomography) resolved spontaneously over 6 weeks and visual acuity recovered to 6/12.

The pseudohypopyon was a unique feature of this case and was due to the presence of a posterior capsule defect enabling the passage of triamcinolone from the vitreous cavity into the anterior chamber. The authors commented that presumably the triamcinolone crystals had been carried into the anterior chamber by currents generated by saccadic eye movements in the partially vitrectomized vitreous cavity. In this case the pseudohypopyon was distinguishable from an infective or inflammatory hypopyon by its ground glass appearance, the presence of refractile particles, and its shifting position, which depended on the patient’s head position. The absence of ocular pain, photophobia, ciliary injection, or iris vessel dilatation suggested a non-inflammatory response and perhaps it would be appropriate to monitor such patients closely rather than administering intravitreal antibiotics.

Retinal damage

An apparent association between severe retinopathy of prematurity and dexamethasone therapy has been shown in a retrospective study [ ]. Infants treated with dexamethasone required longer periods of mechanical ventilation (44 versus 26 days), had a longer duration of supplemental oxygen (57 versus 29 days), had a higher incidence of patent ductus arteriosus (28/38 versus 18/52), and required surfactant therapy more often for respiratory distress syndrome (17/38 versus 11/52). Prospective, randomized, controlled studies are needed to correct for differences in severity of cardiorespiratory disease. Until such studies are available, careful consideration must be given to indications, dosage, time of initiation, and duration of treatment with dexamethasone in infants of extremely low birthweight.

Toxic optic neuropathy

Toxic optic neuropathy can occur and may underlie various reports of sudden blindness in patients taking glucocorticoids. In one case, transient visual loss occurred on several occasions, each time after administration of a glucocorticoid [ ]. In another case, blindness occurred suddenly and paradoxically after glucocorticoid injections into the nasal turbinates [ ]. Although glucocorticoids are sometimes used successfully to relieve pre-existing optic neuritis, a number of such patients react adversely with increased episodes of visual loss.

Exophthalmos

Exophthalmos has been described incidentally as a complication of long-term glucocorticoid therapy and there has been a series of 21 cases [ ].

Psychological

The psychostimulant effects of the glucocorticoids are well known [ ], and their dose dependency is recognized [ ]; they may amount to little more than euphoria or comprise severe mental derangement, for example mania in an adult with no previous psychiatric history [ ] or catatonic stupor demanding electroconvulsive therapy [ ]. In their mildest form, and especially in children, the mental changes may be detectable only by specific tests of mental function [ ]. Mental effects can occur in patients treated with fairly low doses; they can also occur after withdrawal or omission of treatment, apparently because of adrenal suppression [ , ].

A 32-year-old woman developed irritability, anger, and insomnia after taking oral prednisone (60 mg/day) for a relapse of ileal Crohn’s disease [ ]. The prednisone was withdrawn and replaced by budesonide (9 mg/day), and the psychiatric adverse effects were relieved after 3 days. A good clinical response was maintained, with no relapse after 2 months of budesonide therapy.

Seventeen patients taking long-term glucocorticoid therapy (16 women, mean age 47 years, mean prednisone dose 16 mg, mean length of current treatment 92 months) and 15 matched controls were assessed with magnetic resonance imaging and proton magnetic resonance spectroscopy, neurocognitive tests (including the Rey Auditory Verbal Learning Test, Stroop Colour Word Test, Trail Making Test, and estimated overall intelligent quotient), and psychiatric scales (including the Hamilton Rating Scale for Depression, Young Mania Rating Scale, and Brief Psychiatric Rating Scale) [ ]. Glucocorticoid-treated patients had smaller hippocampal volumes and lower N -acetylaspartate ratios than controls. They had lower scores on the Rey Auditory Verbal Learning Test and Stroop Colour Word Test (declarative memory deficit) and higher scores on the Hamilton Rating Scale for Depression and the Brief Psychiatric Rating Scale (depression). These findings support the idea that chronic glucocorticoid exposure is associated with changes in hippocampal structure and function.

Mental and psychomotor development

Dexamethasone has been used in ventilator-dependent preterm infants to reduce the risk and severity of chronic lung disease. Usually it is given in a tapering course over a long period (42 days). The effects of dexamethasone on developmental outcome at 1 year of age has been evaluated in 118 infants of very low birth weights (47 boys and 71 girls, aged 15–25 days), who were not weaning from assisted ventilation [ ]. They were randomly assigned double-blind to receive placebo or dexamethasone (initial dose 0.25 mg/kg) tapered over 42 days. A neurological examination, including ultrasonography, was done at 1 year of age. Survival was 88% with dexamethasone and 74% with placebo. Both groups obtained similar scores in mental and psychomotor developmental indexes. More dexamethasone-treated infants had major intracranial abnormalities (21 versus 11%), cerebral palsy (25 versus 7%; OR = 5.3; CI = 1.3, 21), and unspecified neurological abnormalities (45 versus 16%; OR = 3.6; CI = 1.2, 11). Although the authors suggested an adverse effect, they added other possible explanations for these increased risks (improved survival in those with neurological injuries or at increased risk of such injuries).

Behavioral disorders

Children have marked increases in behavioral problems during treatment with high-dose prednisone for relapse of nephrotic syndrome, according to the results of a study conducted in the USA [ ]. Ten children aged 2.9–15 years (mean 8.2 years) received prednisone 2 mg/kg/day, tapering at the time of remission, which was at week 2 in seven patients. At baseline, eight children had normal behavioral patterns and two had anxious/depressed and aggressive behavior using the Child Behaviour Checklist (CBCL). During high-dose prednisone therapy, five of the eight children with normal baseline scores had CBCL scores for anxiety, depression, and aggressive behavior above the ninety-fifth percentile for age. The two children with high baseline CBCL scores had worsening behavioral problems during high-dose prednisone. Behavioral problems occurred almost exclusively in the children who received over 1 mg/kg every 48 hours. Regression analysis showed that prednisone dosage was a strong predictor of increased aggressive behavior.

Intravenous methylprednisolone was associated with a spectrum of adverse reactions, most frequently behavioral disorders, in 213 children with rheumatic disease, according to the results of a US study [ ]. However, intravenous methylprednisolone was generally well tolerated. The children received their first dose of intravenous methylprednisolone 30 mg/kg over at least 60 minutes, and if the first dose was well tolerated they were given further infusions at home under the supervision of a nurse. There was at least one adverse reaction in 46 children (22%) of whom 18 had an adverse reaction within the first three doses. The most commonly reported adverse reactions were behavioral disorders (21 children), including mood changes, hyperactivity, hallucinations, disorientation, and sleep disorders. Several children had serious acute reactions, which were readily controlled. Most of them were able to continue methylprednisolone therapy with premedication or were given an alternative glucocorticoid. The researchers emphasized the need to monitor treatment closely and to have appropriate drugs readily available to treat adverse reactions.

Large doses are most likely to cause the more serious behavioral and personality changes, ranging from extreme nervousness, severe insomnia, or mood swings to psychotic episodes, which can include both manic and depressive states, paranoid states, and acute toxic psychoses. A history of emotional disorders does not necessarily preclude glucocorticoid treatment, but existing emotional instability or psychotic tendencies can be aggravated by glucocorticoids. Such patients as these should be carefully and continuously observed for signs of mental changes, including alterations in the sleep pattern. Aggravation of psychiatric symptoms can occur not only during high-dose oral treatment, but also after any increase in dosage during long-term maintenance therapy; it can also occur with inhalation therapy [ ]. The psychomotor stimulant effect is said to be most pronounced with dexamethasone and to be much less with methylprednisolone, but this concept of a differential psychotropic effect still has to be confirmed.

Memory

Glucocorticoids can regulate hippocampal metabolism, physiological functions, and memory and there is evidence of memory loss during glucocorticoid treatment and correlations between memory and cortisol concentrations in certain diseases, although it is unclear whether exposure to the endogenous glucocorticoid cortisol in amounts seen during physical and psychological stress in humans can inhibit memory performance in otherwise healthy individuals.

The effects of prednisone on memory have been assessed [ ]. Glucocorticoid-treated patients performed worse than controls in tests of explicit memory. Pulsed intravenous methylprednisolone (2.5 g over 5 days, 5 g over 7 days, or 10 g over 5 days) caused impaired memory in patients with relapsing-remitting multiple sclerosis, but this effect is reversible, according to the results of an Italian study [ ]. Compared with ten control patients, there was marked selective impairment of explicit memory in 14 patients with relapsing-remitting multiple sclerosis treated with pulsed intravenous methylprednisolone. However, this memory impairment completely resolved 60 days after methylprednisolone treatment.

In an elegant experiment on the effect of cortisol on memory, 51 young healthy volunteers (24 men and 27 women) participated in a double-blind, randomized, crossover, placebo-controlled trial of cortisol 40 mg/day or 160 mg/day for 4 days [ ]. The lower dose of cortisol was equivalent to the cortisol delivered during a mild stress and the higher dose to major stress. Cognitive performance and plasma cortisol were evaluated before and until 10 days after drug administration. Cortisol produced a dose-related reversible reduction in verbal declarative memory without effects on nonverbal memory, sustained or selective attention, or executive function. Exposure to cortisol at doses and plasma concentrations associated with physical and psychological stress in humans can reversibly reduce some elements of memory performance.

Prednisone, 10 mg/day for 1 year, has been evaluated in 136 patients with probable Alzheimer’s disease in a double-blind, randomized, placebo-controlled trial [ ]. There were no differences in the primary measures of efficacy (cognitive subscale of the Alzheimer Disease Assessment Scale), but those treated with prednisone had significantly greater memory impairment (Clinical Dementia sum of boxes), and agitation and hostility/suspicion (Brief Psychiatric Rating Scale). Other adverse effects in those who took prednisone were reduced bone density and a small rise in intraocular pressure.

In healthy individuals undergoing acute stress, there was specifically impaired retrieval of declarative long-term memory for a word list, suggesting that cortisol-induced impairment of retrieval may add significantly to the memory deficits caused by prolonged treatment [ ].

In 52 renal transplant recipients (mean age 45 years, 34 men and 18 women) taking prednisone (100 mg/day for 3 days followed by 10 mg/day for as long as needed; mean dose 11 mg/day) there was a major reduction in immediate recall but not delayed recall [ ]. However, there was a significant correlation between mean prednisone dose and delayed recall. In animals, phenytoin pretreatment blocks the effects of stress on memory and hippocampal histology.

In a double blind, randomized, placebo-controlled trial 39 patients (mean age 44 years, 8 men) with allergies or pulmonary or rheumatological illnesses who were taking prednisone (mean dose 40 mg/day) were randomized to either phenytoin (300 mg/day) or placebo for 7 days [ ]. Those who took phenytoin had significantly smaller increases in a mania self-report scale. There was no effect on memory. Thus, phenytoin blocked the hypomanic effects of prednisone, but not the effects on declarative memory.

Psychiatric

Use of glucocorticoids is associated with adverse psychiatric effects, including mild euphoria, emotional lability, panic attacks, psychosis, and delirium. Although high doses increase the risks, psychiatric effects can occur after low doses and different routes of administration. Of 92 patients with systemic lupus erythematosus (78 women, mean age 34 years) followed between 1999 and 2000, psychiatric events occurred in six of those who were treated with glucocorticoids for the first time or who received an augmented dose, an overall 4.8% incidence [ ]. The psychiatric events were mood disorders with manic features (delusions of grandiosity) (n = 3) and psychosis (auditory hallucinations, paranoid delusions, and persecutory ideas) (n = 3). Three patients were first time users (daily prednisone dose 30–45 mg/day) and three had mean increases in daily prednisone dose from baseline of 26 (range 15–33) mg. All were hypoalbuminemic and none had neuropsychiatric symptoms before glucocorticoid treatment. All the events occurred within 3 weeks of glucocorticoid administration. In five of the six episodes, the symptoms resolved completely after dosage reduction (from 40 to 18 mg) but in one patient an additional 8-week course of a phenothiazine was given. In a multivariate regression analysis, only hypoalbuminemia was an independent predictor of psychiatric events (HR = 0.8, 95% CI = 0.60, 0.97).

Although mood changes are common during short-term, high-dose, glucocorticoid therapy, there are virtually no data on the mood effects of long-term glucocorticoid therapy. Mood has been evaluated in 20 out-patients (2 men, 18 women), aged 18–65 years taking at least 7.5 mg/day of prednisone for 6 months (mean current dose 19 mg/day; mean duration of current prednisone treatment 129 months) and 14 age-matched controls (1 man, 13 women), using standard clinician-rated measures of mania (Young Mania Rating Scale, YMRS), depression (Hamilton Rating Scale for Depression, HRSD), and global psychiatric symptoms (Brief Psychiatric Rating Scale, BPRS, and the patient-rated Internal State Scale, ISS) [ ]. Syndromal diagnoses were evaluated using a structured clinical interview. The results showed that symptoms and disorders are common in glucocorticoid-dependent patients. Unlike short-term prednisone therapy, long-term therapy is more associated with depressive than manic symptoms, based on the clinician-rated assessments. The Internal State Scale may be more sensitive to mood symptoms than clinician-rated scales.

Psychoses

Mania has been attributed to glucocorticoids [ ].

A 46-year-old man, with an 8-year history of cluster headaches and some episodes of endogenous depression, took glucocorticoids 120 mg/day for a week and then a tapering dosage at the start of his latest cluster episode. His headaches stopped but then recurred after 10 days. He was treated prophylactically with verapamil, but a few days later, while the dose of glucocorticoid was being tapered, he developed symptoms of mania. The glucocorticoids were withdrawn, he was given valproic acid, and his mania resolved after 10 days. Verapamil prophylaxis was restarted and he had no more cluster headaches.

The authors commented that the manic symptoms had probably been caused by glucocorticoids or glucocorticoid withdrawal. They concluded that patients with cluster headache and a history of affective disorder should not be treated with glucocorticoids, but with valproate or lithium, which are effective in both conditions. They suggested that lamotrigine, an anticonvulsive drug with mood-stabilizing effects, may prevent glucocorticoid-induced mania in patients for whom valproate or lithium are not possible.

Glucocorticoids can cause neuropsychiatric adverse effects that dictate a reduction in dose and sometimes withdrawal of treatment. Of 32 patients with asthma (mean age 47 years) who took prednisone in a mean dosage of 42 mg/day for a mean duration of 5 days, those with past or current symptoms of depression had a significant reduction in depressive symptoms during prednisone therapy compared with those without depression [ ]. After 3–7 days of therapy there was a significant increase in the risk of mania, with return to baseline after withdrawal.

The management of a psychotic reaction in an Addisonian patient taking a glucocorticoid needs special care [ ]. Psychotic reactions that do not abate promptly when the glucocorticoid dosage is reduced to the lowest effective value (or withdrawn) may need to be treated with neuroleptic drugs; occasionally these fail and antidepressants are needed [ ].

However, in other cases, antidepressants appear to aggravate the symptoms.

Two patients with prednisolone-induced psychosis improved on giving the drug in three divided daily doses [ ]. Recurrence was avoided by switching to enteric-coated tablets.

This suggests that in susceptible patients the margin of safety may be quite narrow. It is possible that reduced absorption accounted for the improvement in this case, but attention should perhaps be focused on peak plasma concentrations rather than average steady-state concentrations.

Two women developed secondary bipolar disorder associated with glucocorticoid treatment and deteriorated to depressive–catatonic states without overt hallucinations and delusions [ ].

A 21-year-old woman, who had taken prednisolone 60 mg/day for dermatomyositis for 1 year developed a depressed mood, pessimistic thought, irritability, poor concentration, diminished interest, and insomnia. Although the dose of prednisolone was tapered and she was treated with sulpiride, a benzamide with mild antidepressant action, she never completely recovered. After 5 months she had an exacerbation of her dermatomyositis and received two courses of methylprednisolone pulse therapy. Two weeks after the second course, while taking prednisolone 50 mg/day, she became hypomanic and euphoric. She improved substantially with neuroleptic medication and continued to take prednisolone 5 mg/day. About 9 months later she developed depressive stupor without any significant psychological stressor or changes in prednisolone dosage. She had mutism, reduction in contact and reactivity, immobility, and depressed mood. Manic or mixed state and psychotic symptoms were not observed. She was initially treated with intravenous clomipramine 25 mg/day followed by oral clomipramine and lithium carbonate. She improved markedly within 2 weeks with a combination of clomipramine 100 mg/day and lithium carbonate 300 mg/day. Prednisolone was maintained at 5 mg/day.
A 23-year-old woman with ulcerative colitis and no previous psychiatric disorders developed emotional lability, euphoria, persecutory delusions, irritability, and increased motor and verbal activity 3 weeks after starting to take betamethasone 4 mg/day. She improved within a few weeks with bromperidol 3 mg/day. After 10 months she became unable to speak and eat, was mute, depressive, and sorrowful, and responded poorly to questions. There were no neurological signs and betamethasone had been withdrawn 10 months before. She was treated with intravenous clomipramine 25 mg/day and became able to speak. Intravenous clomipramine caused dizziness due to hypotension, and amoxapine 150 mg/day was substituted after 6 days. All of her symptoms improved within 10 days. Risperidone was added for mood lability and mild persecutory ideation.

In one case, glucocorticoid-induced catatonic psychosis unexpectedly responded to etomidate [ ].

A 27-year-old woman with myasthenia gravis taking prednisolone 100 mg/day became unresponsive and had respiratory difficulties. She was given etomidate 20 mg intravenously to facilitate endotracheal intubation. One minute later she became alert and oriented, with normal muscle strength, and became very emotional. Eight hours later she again became catatonic and had a similar response to etomidate 10 mg. Glucocorticoid-induced catatonia was diagnosed, her glucocorticoid dosage was reduced, and she left hospital uneventfully 4 days later.

The effect of etomidate on catatonia, similar to that of amobarbital, was thought to be due to enhanced GABA receptor function in patients with an overactive reticular system.

A case report has suggested that risperidone, an atypical neuroleptic drug, can be useful in treating adolescents with glucocorticoid-induced psychosis and may hasten its resolution [ ].

A 14-year-old African-American girl with acute lymphocytic leukemia was treated with dexamethasone 24 mg/day for 25 days. Four days after starting to taper the dose she had a psychotic reaction with visual hallucinations, disorientation, agitation, and attempts to leave the floor. Her mother refused treatment with haloperidol. Steroids were withdrawn and lorazepam was given as needed. Nine days later the symptoms had not improved. She was given risperidone 1 mg/day; within 3 days the psychotic reaction began to improve and by 3 weeks the symptoms had completely resolved.

Obsessive-compulsive disorder

Obsessive-compulsive behavior after oral cortisone has been described [ ].

A 75-year-old white man, without a history of psychiatric disorders, took cortisone 50 mg/day for 6 weeks for pulmonary fibrosis and developed severe obsessive-compulsive behavior without affective or psychotic symptoms. He was given risperidone without any beneficial effect. The dose of cortisone was tapered over 18 days. An MRI scan showed no signs of organic brain disease and an electroencephalogram was normal. His symptoms improved 16 days after withdrawal and resolved completely after 24 days. Risperidone was withdrawn without recurrence.

Endocrine

The endocrine effects of the glucocorticoids variously involve the pituitary–adrenal axis, the ovaries and testes, the parathyroid glands, and the thyroid gland.

Pituitary gland

Empty sella syndrome occurred in a boy who developed hypopituitarism after long-term pulse therapy with prednisone for nephrotic syndrome [ ].

A 16-year-old Japanese boy’s growth and development was normal until the age of 2 years. He then developed nephrotic syndrome and was treated with pulsed glucocorticoid therapy nine times over the next 14 years. After the age of 3 years, his rate of growth had fallen. At 16 years, when he was taking prednisone 60 mg/m2/day he was given prednisone on alternate days and the dose was gradually tapered. The secretion of pituitary hormones, except antidiuretic hormone, was impaired and an MRI scan of his brain showed an empty sella and atrophy of the pituitary gland.
Two patients developed hypopituitarism and empty sella syndrome during glucocorticoid pulse therapy for nephrotic syndrome [ ].

When markedly impaired growth is noted in patients treated with glucocorticoids long-term or in pulses, it is necessary to assess pituitary function and the anatomy of the pituitary gland. Children who receive glucocorticoid pulse therapy may develop an empty sella more frequently than is usually recognized.

Pituitary–adrenal axis

Raised glucocorticoid plasma concentrations usually result, after 2 weeks, in the first signs of iatrogenic Cushing’s syndrome. The characteristic symptoms can occur individually or in combination. Whereas in Cushing’s disease or corticotropin–induced Cushing’s syndrome, the predominant symptoms are in part determined by hyperandrogenicity and tend to comprise hypertension, acne, impaired sight, disorders of sexual function, hirsutism or virilism, striae of the skin, and plethora, Cushing’s syndrome due to glucocorticoid therapy is likely to cause benign intracranial hypertension, glaucoma, subcapsular cataract, pancreatitis, aseptic necrosis of the bones, and panniculitis. Obesity, facial rounding, psychiatric symptoms, edema, and delayed wound healing are common to these different forms of Cushing’s syndrome.

It has been said that Cushing-like effects are to be expected if the function of the adrenal cortex is suppressed by daily doses of more than 50 mg hydrocortisone or its equivalent. However, pituitary–adrenal suppression has been described at lower dosage equivalents, for example during prolonged intermittent therapy with dexamethasone [ ]. The secondary adrenal insufficiency caused by therapeutically effective doses can be observed even after giving prednisone 5 mg tds for only 1 week; after withdrawal, adrenal suppression lasts for some days. If one continues this treatment for about 20 weeks, maximal atrophy of the adrenal cortex results, and lasts for some months. This effect begins with inhibition of the hypothalamus, and culminates in true atrophy of the adrenal cortex. It can occur even with glucocorticoids given by inhalation [ ]. Inhaled fluticasone is associated with at least a twofold greater suppression of adrenal function than inhaled budesonide microgram for microgram, according to the results of a crossover study [ ]. Patients with liver disease may experience adrenal suppression with lower doses of glucocorticoids [ ]. It is advisable to use alternate-day therapy to avoid suppression of corticotropin secretion in patients who will need long-term therapy; it will produce the same therapeutic effect as daily dosage. It can be helpful to measure the degree of suppression of corticotropin secretion during long-term glucocorticoid treatment of asthmatic children, as a means of optimizing therapy and avoiding excessive dosage [ ]. The period of time during which the patient should be considered at risk of adrenal insufficiency after withdrawal of oral prednisolone treatment in childhood nephrotic syndrome is still controversial. A study in such patients has suggested that adrenal insufficiency may occur up to 9 months after treatment has ended [ ].

Many protocols for treating children with early B cell acute lymphoblastic leukemia involve 28 consecutive days of high-dose glucocorticoids during induction. The effect of this therapy on adrenal function has been prospectively evaluated [ ] in 10 children by tetracosactide stimulation before the start of dexamethasone therapy and every 4 weeks thereafter until adrenal function returned to normal. All had normal adrenal function before dexamethasone treatment and impaired adrenal responses 24 hours after completing therapy. Each child felt ill for 2–4 weeks after completing therapy. Seven patients recovered normal adrenal function after 4 weeks, but three did not have normal adrenal function until 8 weeks after withdrawal. Thus, high-dose dexamethasone therapy can cause adrenal insufficiency lasting more than 4 weeks after the end of treatment. This problem might be avoided by tapering doses of glucocorticoids and providing supplementary glucocorticoids during periods of increased stress.

Tolerance to glucocorticoids in this, as in some other respects, varies from individual to individual; some patients tolerate 30 mg of prednisone for a long time without developing Cushing’s syndrome, while others develop symptoms at 7.5 mg; the doses recommended today to avoid Cushing’s syndrome in most patients are usually equivalent to hydrocortisone 20 mg. Cushing’s syndrome and other systemic adverse effects can occur not only from oral and injected glucocorticoids, but also from topical and intranasal treatment [ ] and intrapulmonary or epidural administration [ , ].

Glucocorticoid-treated patients with inadequate adrenal function who have an intercurrent illness or are due to undergo surgery will have an inadequate reaction to the resulting stress and need to be temporarily protected by additional glucocorticoid [ ].

Iatrogenic Cushing’s syndrome after a single low dose is exceptional [ ].

A 45-year-old woman was given a single-dose of intramuscular triamcinolone acetonide 40 mg for acute laryngitis and 1 month later was noted to have a cushingoid appearance. Endocrinological tests confirmed hypothalamic–pituitary–adrenal (HPA) axis suppression. Eight months later, the cushingoid appearance had completely disappeared and HPA function had spontaneously recovered.

Pseudohyperaldosteronism has been reported even after intranasal application of 9-alpha-fluoroprednisolone [ ].

Parathyroid function

There is antagonism between the parathyroid hormone and glucocorticoids [ ]. Latent hyperparathyroidism can be unmasked by glucocorticoids [ ].

Thyroid function

Even a single dose of corticotropin briefly inhibits the secretion of thyrotrophic hormone. The uptake of radioactive iodine is also suppressed by corticotropin and by glucocorticoids, but this has no clinical relevance. Pathological changes in thyroid function induced by glucocorticoid treatment are reportedly rare.

Metabolism

Glucose metabolism

All glucocorticoids increase gluconeogenesis. The turnover of glucose is increased, more being metabolized to fat, and blood glucose concentration is increased by 10–20%. Glucose tolerance and sensitivity to insulin are reduced, but provided pancreatic islet function is normal, carbohydrate metabolism will not be noticeably altered. So-called “steroid diabetes,” a benign diabetes without a tendency to ketosis, but with a low sensitivity to insulin and a low renal threshold to glucose, only develops in one-fifth of patients treated with high glucocorticoid dosages. Even in patients with diabetes, ketosis is not to be expected, since glucocorticoids have antiketotic activity, presumably through suppression of growth hormone secretion.

Glucocorticoid treatment of known diabetics normally leads to deregulation, but this can be compensated for by adjusting the dose of insulin. The increased gluconeogenesis induced by glucocorticoids mainly takes place in the liver, but glucocorticoid treatment is especially likely to disturb carbohydrate metabolism in liver disease.

When hyperglycemic coma occurs it is almost always of the hyperosmolar non-ketotic type. After termination of glucocorticoid treatment, steroid diabetes normally disappears. An apparent exception to these findings is provided by the case of a patient in whom glucocorticoid treatment was followed by severe diabetes with diabetic nephropathy, but this was a seriously ill individual who had already undergone renal transplantation [ ]. Gestational diabetes mellitus was more common in women who had received glucocorticoids with or without β-adrenoceptor agonists for threatened preterm delivery compared with controls [ ].

Glucocorticoids probably have more than one effect on carbohydrate metabolism. An increase in fasting glucagon concentration has been observed in volunteers given prednisolone 40 mg/day for 4 days, and this effect may be involved, alongside gluconeogenesis, in glucocorticoid-induced hyperglycemia.

Deflazacort, an oxazoline derivative of prednisolone, was introduced as a potential substitute for conventional glucocorticoids in order to ameliorate glucose intolerance. In a randomized study in kidney transplant recipients with pre-or post-transplantation diabetes mellitus, 42 patients who switched from prednisone to deflazacort (in the ratio 5:6 mg) were prospectively compared with 40 patients who continued to take prednisone [ ]. During the mean follow-up period of 13 months, neither graft dysfunction nor acute rejection developed in the conversion group, and there was improvement in blood glucose control. When the conversion group was stratified into those with pre-or post-transplantation diabetes, there were promising effects in the patients with post-transplantation diabetes. More than a 50% dosage reduction of hypoglycemic drugs was possible in 42% of those with post-transplantation diabetes.

The risk of hyperglycemia requiring treatment in patients receiving oral glucocorticoids has been quantified in a case-control study of 11 855 patients, 35 years of age or older, with newly initiated treatment with a hypoglycemic drug [ ]. The risk for initiating hypoglycemic therapy increased with the recent use of a glucocorticoid. The risk grew with increasing average daily glucocorticoid dosage (in mg of hydrocortisone equivalents): 1.77 for 1–39 mg/day, 3.02 for 40–79 mg/day, 5.82 for 80–119 mg/day, and 10.34 for 120 mg/day or more.

The prevalence of glucocorticoid-induced diabetes mellitus and susceptibility factors have been determined in patients with primary renal diseases [ ]. During glucocorticoid therapy (initial dose of prednisolone 0.75 mg/kg/day), diabetes mellitus was newly diagnosed in 17 of 42 patients by hyperglycemia 2 hours after lunch, although they had normal fasting blood glucose concentrations. Age (OR = 1.40, 95% CI = 1.06, 1.84) and body mass index (OR = 1.87, 95% CI = 1.03, 3.38) were independent susceptibility factors for glucocorticoid-induced diabetes mellitus.

Lipid metabolism

High-dose glucocorticoid therapy can cause marked hypertriglyceridemia, with milky plasma [ ]. It has been suggested that this is caused by abnormal accumulation of dietary fat, reduced postheparin lipolytic activity, and glucose intolerance [ ]. An association between glucocorticoid exposure and hypercholesterolemia has been found in several studies [ ] and can contribute to an increased risk of atherosclerotic vascular disease.

Most premature neonates need intravenous lipids during the first few weeks of life to acquire adequate energy intake and prevent essential fatty acid deficiency before they can tolerate all nutrition via enteral feeds. Dexamethasone is associated with multiple adverse effects in neonates, including poor weight gain and impairment of glucose and protein metabolism. In ten neonates (four boys, mean age 17.3 days) taking dexamethasone for bronchopulmonary dysplasia, intravenous lipids (3 g/kg/day) caused hypertriglyceridemia in the presence of hyperinsulinemia and increased free fatty acid concentrations [ ]. Because of concomitant hyperinsulinemia, the authors speculated that dexamethasone reduced fatty acid oxidation, explaining poor weight gain.

Altered fat deposition has been repeatedly reported. Fat can be deposited epidurally and at other sites. Adiposis dolora, which involves the symmetrical appearance of multiple painful fat deposits in the subcutaneous tissues, has on one occasion been attributed to glucocorticoids [ ].

The authors of a prospective study in two French tertiary centers have described corticosteroid-induced lipodystrophy [ ]. They enrolled 88 consecutive patients who started long-term (≥ 3 months), high dosage (≥ 20 mg/day) systemic glucocorticoid therapy and assessed the development of lipodystrophy from standardized head and neck photographs. Arterial blood pressure and fasting blood glucose concentrations were assessed at baseline and then every 3 months until month 12. Total cholesterol, HDL-cholesterol, LDL-cholesterol, and triglycerides were recorded at baseline, month 3, and month 12.The mean age of the patients was 57 years and the mean baseline dosage of prednisolone was 56 mg/day; 64 patients were still taking glucocorticoids at month 12 (mean dosage 11 mg/day). The cumulative incidences of lipodystrophy at months 3 and 12 were 61% and 69% respectively. Baseline characteristics were similar in those who developed lipodystrophy and those who did not, except with regard to baseline body mass index, which was higher in the former (24 versus 21 kg/m ² ). Blood pressure was significantly higher in patients with lipodystrophy at month 9 (135/78 versus 127/73 mmHg) and month 12 (141/81 versus 128/72 mmHg). Those with lipodystrophy had significantly higher plasma concentrations of fasting blood glucose, triglycerides, and total cholesterol and lower HDL-cholesterol concentrations during follow-up.

Tumor lysis syndrome

Acute tumor lysis syndrome is a life-threatening metabolic emergency that results from rapid massive necrosis of tumor cells. There have been repeated reports of an acute tumor lysis syndrome when glucocorticoids are administered in patients with pre-existing lymphoid tumors [ ].

A 60-year-old woman took dexamethasone 4 mg 8-hourly for dyspnea due to a precursor T lymphoblastic lymphoma-leukemia with bilateral pleural effusions and a large mass in the anterior mediastinum [ ]. She developed acute renal insufficiency and laboratory evidence of the metabolic effects of massive cytolysis. She received vigorous hydration, a diuretic, allopurinol, and hemodialysis. She recovered within 2 weeks and then underwent six courses of CHOP chemotherapy. The mediastinal mass regressed completely. She remained asymptomatic until she developed full-blown acute lymphoblastic leukemia, which was resistant to treatment.

Electrolyte balance

The severity of potassium loss due to glucocorticoids depends partly on the amount of sodium in the diet; the most widely used synthetic glucocorticoids cause less potassium excretion than natural hydrocortisone does. Prednisone and prednisolone have a glucocorticoid activity 4–5 times that of hydrocortisone, but their mineralocorticoid activity is less (see Table 1 ); even at high dosages they do not cause noteworthy sodium and water retention. Of the major synthetic glucocorticoids, dexamethasone has the strongest anti-inflammatory, hyperglycemic, and corticotropin-inhibitory activity; sodium retention is completely absent; the degree of glucocorticoid-induced metabolic alkalosis may also be less with dexamethasone than with hydrocortisone or methylprednisolone [ ].

Mineral balance

There can be increases in calcium and phosphorus loss because of effects on both the kidney and the bowel, with increased excretion and reduced resorption [ ]. Tetany, which has been seen in patients receiving high-dose long-term intravenous glucocorticoids, has been explained as being due to hypocalcemia, and there are also effects on bone. Tetany has also been reported in a patient with latent hyperparathyroidism after the administration of a glucocorticoid [ ].

Hypocalcemic encephalopathy occurred in a 35-year-old woman with hypoparathyroidism. It was believed that the administration of methylprednisolone intramuscularly had precipitated severe hypocalcemia, which had led to a metabolic encephalopathy [ ].

The administration of large doses of glucocorticoids to patients with major burns presenting with low cardiac output has been reported to produce a reversible drop in serum zinc, which might lead to impaired tissue repair [ ], but it is not clear whether this has clinical effects.

You're Reading a Preview

Become a Clinical Tree membership for Full access and enjoy Unlimited articles

Become membership

If you are a member. Log in here

General information

Nomenclature

Relative potencies

Uses

Routes of administration

General adverse effects and adverse reactions

Drug studies

Observational studies

Comparative studies

Placebo-controlled studies

Systematic reviews

Organs and systems

Cardiovascular

Hypertension

Myocardial ischemia

Cardiomyopathy

Cardiac dysrhythmias

Pericarditis

Vasculitis

Respiratory

Ear, nose, throat

Nervous system

Sensory systems

Cataract

Increased intraocular pressure and glaucoma

Chorioretinopathy

Retinal damage

Toxic optic neuropathy

Exophthalmos

Psychological

Mental and psychomotor development

Behavioral disorders

Memory

Psychiatric

Psychoses

Obsessive-compulsive disorder

Endocrine

Pituitary gland

Pituitary–adrenal axis

Parathyroid function

Thyroid function

Metabolism

Glucose metabolism

Lipid metabolism

Tumor lysis syndrome

Electrolyte balance

Mineral balance

You're Reading a Preview

Related Posts

Repirinast

Renin inhibitors, direct

Remifentanil

Remacemide