Glucocorticoid Therapy for Immune-Mediated Diseases: Basic and Clinical Correlates

Mechanisms of Action

Dr. George P. Chrousos (Chief, Pediatric Endocrinology Section, Developmental Endocrinology Branch, National Institute of Child Health and Human Development, NIH, Bethesda, Maryland): Glucocorticoids exert most of their effects through specific, ubiquitously distributed intracellular receptors [1]. The classic model of glucocorticoid action was described more than two decades ago and is briefly updated here (Figure 1, panel A). Glucocorticoids circulate in blood, is either in the free form or in association with cortisol-binding globulin. The free form of the steroid can readily diffuse through the plasma membrane and can bind with high affinity to cytoplasmic glucocorticoid receptors (the role of receptors primarily residing in the nucleus is controversial). The formation of the ligand-receptor complex is followed by its activation (that is, translocation into the nucleus and binding to what are called acceptor sites). The bound complex modulates transcription of specific genes that encode proteins responsible for the action of glucocorticoids.

Glucocorticoid Receptors

In 1985, the complementary DNA of the human glucocorticoid receptor was cloned [2]; it contains three main functional domains Figure 1, panel B): first, the DNA-binding domain in the center of the molecule that recognizes specific sequences of the DNA called hormone-responsive elements; second, the ligand-binding domain in the carboxyl terminal region that interacts with the specific steroid; and third, the immunogenic domain in the amino terminal region.

The nonactivated glucocorticoid receptor resides in the cytosol in the form of a hetero-oligomer with other highly conserved proteins [3]. This molecular complex comprises receptor, heat-shock proteins, and immunophilin (Appendix Table 1) [4]. The binding of the receptor to the heat-shock protein 90 facilitates its interaction with the ligand [5]. When the ligand binds, the receptor dissociates from the rest of the hetero-oligomer and translocates into the nucleus. Before or after the translocation, the receptor forms homodimers through sequences present in the DNA and ligand-binding domains [6].

Gene Regulation

After specific interaction with pore-associated proteins, the hormone-receptor complexes enter the nucleus through the nuclear pores [7]. The interaction is facilitated by two nuclear localization sequences in the receptor, both in the ligand-binding domain. Inside the nucleus, the hormone-receptor complexes bind to specific glucocorticoid responsive elements within DNA [8]. The complexes modulate the transcription rates of the corresponding glucocorticoid-responsive genes [9], apparently by stabilizing the initiation complex, composed of RNA polymerase II and its ancillary factors A through F. The hormone-receptor complex may interact directly with factor IIB [10], but it also interacts with other nuclear proteins to produce the conditions necessary for effective transcription [11]. These proteins may be able to relax the DNA away from the nucleosome and thus make it easier for the polymerase to exert its effects. In addition, glucocorticoid receptors may interact with DNA-binding proteins that are associated with different regulatory elements of the DNA [12, 13]. At least two such proteins have been described: One is the glucocorticoid modulatory element-binding protein and the other is the CACCC-box-binding protein. Both of these transcription factors potentiate the modulatory effects of glucocorticoids after transcription of specific genes.

Transcription appears to be important in the regulation of genes involved in growth and inflammation. Glucocorticoid response elements can act both positively and negatively on transcription, depending on the gene on which the complex acts [14, 15]. One major way by which glucocorticoids exert down-modulatory effects on transcription is through noncovalent interaction of the activated hormone-receptor complex with the c-Jun/c-Fos heterodimer [16-18], which binds to the activator protein (AP)-1 site of genes of several growth factors and cytokines. The glucocorticoid-receptor complex prevents the c-Jun/c-Fos heterodimer from stimulating the transcription of these genes. Another mechanism by which glucocorticoids may suppress gene transcription is by an interaction between the hormone-receptor complex and glucocorticoid response elements that are in close proximity to responsive elements for other transcription factors [19]. Thus, the promoter region of the glycoprotein hormone- subunit, which is stimulated by cyclic AMP through the cyclic AMP-responsive element, contains a glucocorticoid response element in close proximity, so that when the receptor dimer binds to its own element, it hinders the cyclic AMP-binding protein from exerting its stimulatory effect on that gene.

Post-Transcriptional Effects

In addition to modulating transcription, glucocorticoids also have effects on later cellular events, including RNA translation, protein synthesis, and secretion. They can alter the stability of specific messenger RNAs of several cytokines and other proteins, thereby altering the intracellular steady-state levels of these molecules [20, 21]. This may occur through modulation of transcription of still unknown proteins that bind RNA and alter its translation and degradation rates. Also, glucocorticoids influence the secretion rates of specific proteins through mechanisms that have not yet been defined. Finally, the receptor itself has guanylate cyclase activity, and glucocorticoids can rapidly alter the electrical potential of some cells [22, 23].

Anti-inflammatory and Immunosuppressive Effects

Dr. Dimitrios T. Boumpas: Although the cause and pathogenesis of many immunologically mediated diseases are not completely understood, it is known that the localization of leukocytes at sites of inflammation, their subsequent activation, and the generation of secretory products contribute to tissue damage, as shown in Figures 2 and 3 [24-26]. Glucocorticoids inhibit the access of leukocytes to inflammatory sites, interfere with their function and the function of fibroblasts and endothelial cells at those sites, and suppress the production and the effects of humoral factors. In general, leukocyte traffic is more susceptible to alteration by glucocorticoids than is cellular function; in turn, cellular immunity is more susceptible than humoral immunity to these agents.

Even though the effects of glucocorticoids on the different types of inflammatory cells will be discussed separately, each cell type is actually involved in complex interactions with other cells. Glucocorticoids affect many, if not all, the cells and tissues of the body, thus provoking a wide range of changes that involve several cell types concurrently.

Effects on Nonlymphoid Inflammatory Cells

Dr. Ronald L. Wilder (Chief, Inflammatory Joint Diseases Section, Arthritis and Rheumatism Branch, National Institute of Arthritis and Musculoskeletal and Skin Diseases, NIH, Bethesda, Maryland): Glucocorticoids are among the most potent anti-inflammatory agents available in clinical medicine. Pharmacologic doses of glucocorticoids dramatically inhibit exudation of plasma and accumulation of leukocytes at sites of inflammation. Several factors influence the magnitude of these effects, including the dose and route of administration of the glucocorticoids used, as well as the type and differentiation state of the target cell population [27]. Several host variables also modify the anti-inflammatory response to glucocorticoids. For example, some persons (those with active systemic lupus erythematosus) appear to have an accelerated rate of glucocorticoid catabolism [28]. Various levels of target tissue resistance may exist in some patients with systemic lupus erythematosus and rheumatoid arthritis [29]. These factors, alone or in combination, may explain the observation that different patients and diseases have variable therapeutic responses to glucocorticoids [30, 31].

Macrophages

Glucocorticoids antagonize macrophage differentiation and inhibit many of their functions [27]. These agents 1) depress myelopoiesis and inhibit expression of class II major histocompatibility complex antigens induced by interferon-; 2) block the release of numerous cytokines, such as interleukin-1, interleukin-6, and tumor necrosis factor-; 3) depress production and release of proinflammatory prostaglandins and leukotrienes; and 4) depress tumoricidal and microbicidal activities of activated macrophages.

Neutrophils

The major effect of glucocorticoids on neutrophils appears to be the inhibition of neutrophil adhesion to endothelial cells. This effect diminishes trapping of neutrophils in the inflamed site and probably is responsible for the characteristic neutrophilia. At pharmacologic doses, glucocorticoids only modestly impair important neutrophil functions, such as lysosomal enzyme release, the respiratory burst, and chemotaxis to the inflamed site. Lower doses do not affect these neutrophil functions [27, 30].

Eosinophils, Basophils, and Mast Cells

Just as they affect macrophages, glucocorticoids decrease circulating eosinophil and basophil counts. They also decrease the accumulation of eosinophils and mast cells at sites of allergic reactions. Functionally, glucocorticoids inhibit IgE-dependent release of histamine and leukotriene C₄ from basophils, and they also inhibit degranulation of mast cells [27].

Endothelial Cells

These cells form the barrier between the blood and the tissues and are critical regulators of the inflammatory cascade. They affect hemostasis, vascular permeability, trapping, and exudation of leukocytes into inflammatory sites. Glucocorticoids have profound effects on the activation and subsequent function of these cells [27] and clearly inhibit vascular permeability. Moreover, they inhibit numerous molecular events associated with activation. For example, they inhibit up-regulation of the expression of class II major histocompatibility complex antigens, as well as endotoxin-induced up-regulated expression of the adhesion molecules (endothelial leukocyte adhesion molecule-1 [ELAM-1] and intercellular cell adhesion molecule-1 [ICAM-1]), which are cell surface molecules critical to leukocyte localization [24-26, 32] (see Figure 3). Further, glucocorticoids inhibit the secretion of the complement pathway proteins C3 and factor B; they also inhibit the formation of interleukin-1 and arachidonic acid metabolites and are potent inhibitors of the expression of cyclooxygenase type 2.

Fibroblasts

These cells contribute to the inflammatory process and are major targets of glucocorticoids [27]. At supraphysiologic concentrations, glucocorticoids suppress proliferation and suppress growth factor-induced DNA synthesis and protein synthesis, including synthesis of collagens. They induce fibronectin messenger RNA transcription, inhibit interleukin-1, inhibit tumor necrosis factor--induced metalloproteinase synthesis, and inhibit arachidonic acid metabolite synthesis. They are also potent inhibitors of the up-regulated expression of cyclooxygenase type 2 (Crofford LJ, Wilder RW, Hla T. Unpublished observations). These observations suggest that glucocorticoids might retard bone and cartilage destruction in diseases such as rheumatoid arthritis.

Prostaglandins

The cellular actions of glucocorticoids may be affected by prostaglandins. Prostaglandins are metabolites of arachidonic acid, which is released from phospholipids by the action of phospholipase A₂. Previously, it was assumed that glucocorticoids directly inhibited the enzymatic activity of this enzyme. New evidence, however, indicates that suppression of phospholipase A₂ activity is mediated by the activation of inhibitors of the enzyme itself or by inhibition of enzyme synthesis. The primary mediators of the inhibition of enzyme activity appear to be members of the annexin family of proteins, which includes proteins such as lipocortin-1 [33]. This topic, however, remains controversial [34-36].

A second step in prostaglandin synthesis is the formation of prostaglandin H₂ from arachidonic acid by enzymes called cyclooxygenases. Recent data have shown the existence of at least two cyclooxygenase (COX) genes, COX-1 and COX-2; the proteins have 61% amino acid identity. The COX-2 gene and protein, but not COX-1, are strongly up-regulated in macrophages, endothelial cells, and fibroblasts by mediators such as endotoxin and interleukin-1, although they are inhibited by glucocorticoids. In contrast, COX-1 is constitutively expressed and relatively unaffected by glucocorticoids.

The role of the differential expression of COX-1 and COX-2 in inflammation and their dramatically different responses to glucocorticoids is unclear, but the available data suggest that glucocorticoids affect the production of proinflammatory arachidonic metabolites (37-40; Crofford LJ, Wilder RW, Hla T. Unpublished observations).

Effects on Lymphocytes

Dr. Dimitrios T. Boumpas: The immunosuppressive effects of glucocorticoids are also directed at the traffic and function of lymphocytes.

T Lymphocytes

In humans, a single dose of glucocorticoids produces a marked but transient lymphocytopenia involving all lymphocyte subpopulations [41]. The mechanism of the lymphopenia involves the redistribution of circulating lymphocytes to other lymphoid compartments, particularly the bone marrow [42]. Changes in the expression of adhesion molecules (see Figure 3) may be responsible for that redistribution of lymphocytes [32]. In contrast to other species (such as the mouse, rat, or rabbit), glucocorticoid-induced lymphopenia in humans is not due primarily to cell death. Mature, resting human lymphocytes are not lysed even by suprapharmacologic doses of glucocorticoids. However, immature human T cells (thymocytes and transformed lymphocytes) and, in some instances, activated T cells may be susceptible to lysis [43, 44] by programmed cell death (apoptosis).

In addition to their effects on cell distribution, glucocorticoids also affect the initiation and the progression of the T-cell cycle. During the initiation phase (activation), antigen binds to the T-cell receptor and initiates a cascade of events that leads to production of interleukin-2 (and other cytokines) and induction of high-affinity interleukin-2 receptors. Binding of interleukin-2 to its receptor promotes T-cell proliferation and generation of effector, suppressor, and cytotoxic functions. As shown in Figure 4, glucocorticoids inhibit, through various mechanisms, several events associated with T-cell activation [45-48]. In addition to depressing interleukin-2 production, they interfere with the action of interleukin-2 on activated T cells by inhibiting the interleukin-2-dependent phosphorylation of several intracellular proteins [48]. Several other cytokines secreted by activated T cells (interleukins-3, -4, and -6 and interferon-) are also inhibited by glucocorticoids [49, 50].

Cytokines are essential for T-cell functions. Because glucocorticoids inhibit the production (and in some cases the action) of several cytokines, they also inhibit the function of T cells. Thus, the generation, proliferation, and function of helper and suppressor T cells are inhibited by these agents [51]. Cytotoxic T-cell responses (both auto- and alloreactive) are also inhibited by moderate-to-high doses of glucocorticoids, primarily because of the blockade of cytokine expression and to a lesser extent because of the lysis of reactive T-cell clones [52].

B Lymphocytes

In contrast to T cells, B cells are relatively resistant to the immunosuppressive effects of glucocorticoids. These agents inhibit the proliferation of B cells, provided that they are present in culture during the first 24 hours after stimulation. Once B cells are activated, they differentiate into immunoglobulin-secreting plasma cells; glucocorticoids have only minimal effects on this process [53].

From a clinical standpoint, the most important effect of glucocorticoids on B cells relates to antibody production. Low doses of glucocorticoids do not affect serum immunoglobulin levels or antibody synthesis in vivo after inoculation with various antigens, but a brief course of daily high-dose prednisone will decrease serum immunoglobulin levels with maximal suppression observed 2 to 4 weeks after treatment [54]. This suppression is the result of an initial increase in immunoglobulin catabolism, which is followed by decreased synthesis. The mechanisms for this decrease have not been defined. Because glucocorticoids inhibit the production of several cytokines involved in immunoglobulin synthesis (interleukins-1 through 6 and interferon-), decreased immunoglobulin production could result from decreased accessory or helper T-cell activities. Whether glucocorticoids inhibit immunoglobulin gene expression is not known.

In summary, whereas inhibition of leukocyte traffic and cellular immune responses require lower doses of glucocorticoids, higher doses of these agents are needed to suppress the functions of leukocytes and the humoral immune response. This heterogeneity of response is also apparent among different persons and diseases (as described above).

Therapeutic Use

Dr. Thomas R. Cupps (Department of Medicine, Georgetown University Medical Center, Washington, D.C.): Glucocorticoid use should be based on an evaluation of therapeutic goals and benefits compared with potential drug-associated morbidity. Knowledge of the clinical pharmacology and treatment protocols of glucocorticoids is critical; several excellent reviews have been published [55-60].

In general, topical or compartmental (for example, intra-articular) use of glucocorticoids results in less toxicity than does systemic administration. Among systemic treatments involving the same total dose of glucocorticoids, split-dose regimens are more toxic than single daily-dose protocols, which are themselves more toxic than alternate-day treatment schedules [42, 59]. When administered on a daily basis, glucocorticoid analogs with prolonged biologic half-lives (for example, dexamethasone) have a greater potential for side effects than do analogs with intermediate biologic half-lives (for example, prednisone). Higher doses of systemic glucocorticoids can be given for less than a week with relative safety, although the same dose of drug administered for a more extended period will result in predictable, clinically significant morbidity.

Patient education is an important factor in optimal glucocorticoid therapy. The patient should be made aware of the potential clinical consequences from hypothalamic-pituitary-adrenal (HPA) axis insufficiency that may result from systemic glucocorticoid therapy. Thus, the patient must be cautioned never to stop or rapidly taper glucocorticoids without medical advice. The HPA-axis response to stress may be decreased even after a week of daily glucocorticoid therapy [61]. If regular oral glucocorticoid treatment is interrupted for a period of more than 24 hours, patients may be susceptible to circulatory collapse in response to physiologic stress, trauma, infection, or surgery, and parenteral administration of glucocorticoids may be required. Even though HPA-axis insufficiency is more common with higher cumulative doses of glucocorticoids, it cannot be reliably predicted from the dose of glucocorticoids, the duration of therapy, or the basal plasma cortisol levels. Testing with corticotrophin-releasing hormone may be useful in assessing pituitary-adrenal functions in patients who may have HPA-axis insufficiency [62].

The patient should be warned that glucocorticoid therapy usually stimulates the appetite and causes weight gain; thus, the importance of diet should be emphasized when therapy is begun. The symptoms and signs of diabetes and steroid myopathy, as well as neuropsychiatric and infectious complications associated with glucocorticoid therapy, should also be described to the patient [63].

The choice of a particular glucocorticoid analog for systemic therapy depends, in part, on clinical variables. Hydrocortisone is generally used for physiologic replacement and stress coverage if patients have HPA suppression [64]. Because of its short biologic half-life and sodium- and potassium-retaining effects, this agent is not commonly used for systemic immunosuppressive or anti-inflammatory treatment. Fluorinated analogs, such as dexamethasone, have a long biologic half-life and little sodium-retaining potency [65, 66]. This long biologic half-life, however, may be associated with a greater potential for morbidity, so this group of glucocorticoid analogs is not commonly used in prolonged daily treatment regimens.

The appropriate starting dose of glucocorticoids should be adequate to rapidly suppress disease activity and tissue damage. High daily or split-dose therapy (for example, 0.6 to 1.0 mg/kg per day of prednisone) is indicated in the early phases of particularly aggressive diseases. Within 1 to 2 weeks, therapy should be consolidated into a single morning dose. Subsequent tapering to the minimal effective maintenance dose (preferably alternate-day) is dictated by clinical variables. Tapering too gradually may be associated with greater glucocorticoid-induced morbidity, although tapering too rapidly predisposes patients to disease exacerbations.

The use of adjunctive anti-inflammatory and immunosuppressive drug therapy (so called steroid-sparing agents) as a means to decrease glucocorticoid dose requirements should be considered. The ideal would be to achieve a similar level of disease suppression with lower doses of glucocorticoids and an overall decrease in glucocorticoid-associated morbidity. Several agents, including azathioprine, methotrexate, antimalarial agents, cyclosporin A, cyclophosphamide, and others, have been used. Alternate-day therapy is another option that may help to decrease glucocorticoid-related morbidity [67]. Tapering to an alternate-day protocol is also useful to allow HPA-axis recovery before stopping glucocorticoid therapy. If a dose level, for instance, of 30 mg of prednisone per day effectively suppresses disease activity, switching to an alternate-day regimen should be considered [42]. The dose can be simultaneously increased on the high day as the dose is decreased on the low day, or simply gradually decreased on the low day until a strictly alternate-day regimen is achieved.

The choice of a specific glucocorticoid treatment plan is frequently influenced by the physician's experience in treating the underlying disease. In some instances, alternate-day glucocorticoids can be used from the beginning of treatment, as in membranous glomerulonephritis [68]. More often, tapering to alternate-day therapy is planned as a way to minimize the potential risks for complications from intense glucocorticoid induction therapy, as in systemic vasculitis [69].

Conventional high-dose glucocorticoid therapy may be disappointing because of incomplete responses or severe complications or both. Accordingly, regimens have been modified to improve the therapeutic index of glucocorticoids. One approach is the use of intravenous pulse therapy. Infusions of large doses of glucocorticoids (for example, methylprednisolone in doses up to 1.0 g/m² body surface area for 1 to 5 days) have become widely used in selected clinical situations [70-74]. Pulse therapy is commonly used to initiate therapy in patients with rapidly progressive, immunologically mediated diseases, such as acute transplant rejection, Goodpasture disease, necrotizing glomerulonephritis, and severe lupus nephritis. Unfortunately, no formal studies have compared conventional high-dose and pulse glucocorticoid therapies. The utility of pulse therapy for extended maintenance treatment appears to be limited [72, 74].

Complications of Therapy

Dr. James E. Balow (Chief, Kidney Disease Section, NIDDK, NIH, Bethesda, Maryland): Given the diversity of mechanisms and sites of action, it is not surprising that glucocorticoids cause different types of toxicity. In the early years of aggressive glucocorticoid therapy, iatrogenic Cushing syndrome was common. The consequences of extended high doses of glucocorticoid therapy were catastrophic, including death from infection and cardiovascular complications. Although such extreme complications are less frequent today, side effects are still commonplace and usually cannot be completely avoided [75]. Indeed, some risk for complications is inherent in the attempt to obtain the therapeutic benefits of glucocorticoids.

The medical complications of glucocorticoid therapy are listed in Table 1. The risks for all of these complications are theoretically dose- and time-dependent. These complications range from minor to severe, early to late, flagrant to insidious, expected to rare, and tolerable to intolerable. An attempt has been made to group them in order of their practical importance. Of necessity, this review of glucocorticoid side effects is not comprehensive; it focuses on changing concepts and on recent information about the pathogenesis of certain complications.

Infections

The propensity of glucocorticoids to cause an infection diathesis is controversial. Clearly, the risk for infection is dependent on the dose and duration of glucocorticoid therapy [76]. The relative risk for infection across a number of clinical settings was approximately two times that of controls in a meta-analysis of 71 trials involving more than 2000 glucocorticoid-treated patients [77]. The risk varied according to the type of disease being treated with glucocorticoids: Although relative risk did not increase in renal-limited diseases, it increased almost three times for patients with neurologic diseases (presumably representing attendant deficits in local host-defense mechanisms).

Peptic Ulcer Disease

Ulcers have been long considered a complication of glucocorticoid therapy; however, this side effect has recently been reexamined by also accounting for concurrent drug therapy. A prospective study [78] of 1400 patients receiving steroid therapy showed an overall twofold increased risk for peptic ulcer disease. However, multivariate analysis showed that concomitant use of nonsteroidal anti-inflammatory drugs was the variable that explained the increased risk for peptic ulcer disease in patients taking glucocorticoids.

Atherosclerosis

Accelerated atherosclerosis seems to be associated with glucocorticoid treatment [79]. However, other processes operating in the systemic diseases for which glucocorticoids are used are also atherogenic. For example, in systemic lupus erythematosus, the long-term use of prednisone was found to be an independent contributor to the risk for coronary disease [80], but glucocorticoids may simply be a surrogate for the true risk factor.

Evidence of atherosclerotic risk in steroid-treated patients without complicated systemic diseases is not definitive. For example, in one study of normal volunteers, glucocorticoids did not induce lipoprotein abnormalities associated with cardiovascular risk [81]; 2 weeks of glucocorticoid therapy did not substantively affect total cholesterol or low-density lipoproteins. Very low-density lipoproteins were the most adversely affected, but glucocorticoids increased high-density lipoproteins as well. If glucocorticoids themselves are indeed atherogenic, the process probably involves mechanisms other than a change in lipoprotein profiles.

Skeletal Complications

The bony complications of glucocorticoids are insidious and perplexing. Osteoporosis is predominantly associated with loss of trabecular bone, and its pathogenesis is complex [82]. Steroids cause bone loss by decreasing intestinal calcium absorption and by increasing urinary calcium excretion by the kidney. Decreased blood calcium results in mild hyperparathyroidism, which in turn accelerates bone resorption. There are other disturbances in mineral metabolism, but the key factor in osteoporosis is an imbalance between bone growth and resorption [83].

In a study of asthmatic patients, the effects of intermittent and long-term steroid treatment on mineral density of trabecular and cortical bone were compared [84]. Neither intermittent nor long-term therapy had a substantial effect on cortical bone, and only long-term therapy had a deleterious effect on trabecular bone, the type of bone affected primarily in osteoporosis.

The issues in steroid-induced osteoporosis center on prevention and treatment. As for idiopathic osteoporosis [85], the guidelines for steroid-induced osteoporosis include exercise and adequate diet [85-87]. Dietary calcium should be 1200 to 1500 mg/d [88], a level that usually requires supplementation in the typical U.S. diet.

Sex hormones should be replaced if gonadal failure is present. Although deficiencies of male and female sex hormones foster osteoporosis, pharmacologic doses of these hormones are not indicated because of potential adverse consequences on prostate and breast tumors. Likewise, vitamin D should be replaced if it is deficient, but high doses are not recommended for routine use because of the propensity for calcium urolithiasis.

Calcitonin and bisphosphonates are the recommended therapeutic agents for increasing bone formation and decreasing resorption. Fluoride, although it increases bone density, does not produce structurally strong bone and is generally not recommended. Calcitonin (Cibacalcin; CIBA Pharmaceutical, Company, Summit, New Jersey) may be the safest and most effective agent available to inhibit osteoclastic activity. One controlled trial [89] found that calcitonin effectively increases bone mineral content in patients with steroid-induced osteoporosis. Calcitonin also decreases bone pain in patients with symptomatic osteoporosis. A nasal spray preparation of calcitonin is under investigation and would be easier for patients to use than the current injectable preparation.

Bisphosphonates, which are resistant to hydrolysis and bind to hydroxyapatite crystals in bone, also reverse the unfavorable balance of bone resorption and mineralization. Ingestion of the bisphosphonate-crystal complexes inhibits osteoclast function. Two preparations of bisphosphonates are available in the United States: etidronate (Didronel; Procter & Gamble Pharmaceuticals, Inc., Norwich, New York) and pamidronate (Aredia; Ciba Pharmaceutical Co., Summit, New Jersey). These agents have been approved primarily for use in management of malignancy-associated hypercalcemia. An oral preparation of pamidronate had a salutary effect on bone mineral density in steroid-induced osteoporosis [90], but more data on the effects of these drugs in steroid-induced osteoporosis are clearly needed. Because some concern still exists about adverse long-term toxicity, neither calcitonin nor bisphosphonates are yet considered standard therapy. If they are used at all, they should be used intermittently, perhaps on a rotating cycle.

Osteonecrosis of hips, knees, and occasionally shoulders is a known risk of receiving glucocorticoids. Its pathogenesis and treatment are both controversial [91], but compromised vascular supply to trabecular bone or the cumulative effects of multiple fatigue fractures in osteoporotic bone are theorized. Although short-term, high-dose glucocorticoids have been associated with osteonecrosis, steady long-term doses seem to pose the highest risk. Unfortunately, joint replacement is usually required to control pain and restore function [92].

Novel Pharmacologic Approaches

Alterations of the glucocorticoid molecule could decrease steroid complications. Deflazacort is a synthetic glucocorticoid (not currently available in the United States) that is nearly as efficacious as prednisone and seems to have fewer side effects [93, 94]. Compared with prednisone or betamethasone, deflazacort causes a smaller decrease in intestinal calcium absorption [86] and causes less renal calcium excretion [95]. Preliminary studies on bone mineral density also support the benefit of deflazacort compared with prednisone, but more clinical trials are necessary to define whether or not deflazacort has a higher overall therapeutic index than do standard glucocorticoids.

Addendum

Recently, Sambrook and colleagues [97] conducted a prospective study involving randomization of patients treated with long-term glucocorticoids to three groups: 1) calcium plus calcitriol plus calcitonin; 2) calcium plus calcitriol; or 3) calcium alone for 1 year. The combination of calcium, calcitriol, and calcitonin was associated with the best maintenance of lumbar spine bone density during and 1 year after experimental therapy. No differences were observed among the treatment groups in femoral head bone density. Hypercalcemia occurred in one quarter of patients receiving calcitriol. As stated in the accompanying editorial by Meunier [98], the optimal regimen for prevention of glucocorticoid-induced osteoporosis has not been found.