Glucocorticoid-Induced Bone Loss in Dermatologic Patients: Title and subTitle BreakAn Update FREE

Systemic corticosteroids are extremely useful agents in the treatment of immunobullous diseases and collagen vascular diseases in dermatology. Dermatologists, along with rheumatologists and pulmonologists, are among the leading prescribers of systemic glucocorticoids in the world. The efficacy of glucocorticoids is well known, but they are not without adverse effects, which unfortunately are frequent. The numerous adverse effects are directly related to both the dose and duration of treatment.¹

The most predictable and debilitating of these adverse effects is rapid bone loss with an increase in osteoporotic fracture risk, particularly in cases of long-term use.² It has been demonstrated that more than one third of individuals treated with corticosteroids for 5 to 10 years will have an osteoporotic fracture.^{3 - 4} Fractures of the spine and hip are associated with marked increases in morbidity and mortality. Hip fracture is the most severe and costly complication of osteoporosis, with a 1-year mortality rate as high as 35% in elderly patients and an additional 30% left bedridden or wheelchair bound.^{3 - 5} Prevention of osteoporosis is a very important public health concern; the estimated national expenditure for osteoporosis and associated fractures was $18 billion in 2002, and the cost is rising.⁶

Glucocorticoid-induced osteoporosis is second only to postmenopausal osteoporosis as the most common form. Epidemiologic data demonstrate that approximately 0.5% to 1% of the population at any time is using systemic glucocorticoids.^{7 - 8} Observational studies suggest that 30% to 50% of patients requiring long-term oral steroids will eventually experience bone fractures.^{9 - 12}

Despite these data and the availability of effective preventive and therapeutic options for corticosteroid-induced osteoporosis, studies have shown that less than 50% of patients receiving long-term treatment with oral steroids have been evaluated for osteoporosis, and less than 25% have been treated.^{8 ,13} In a recent report of 37 dermatologist-treated patients undergoing long-term corticosteroid therapy, only 7 (19%) had received bisphosphonates from their primary dermatologist.¹⁴ This discrepancy exists because it is difficult for clinicians to know the dose required for effective control of disease, the length of time that therapy will be needed, and whether repeated administration will be needed to treat recurrence. However, studies also indicate a great variability among clinicians in both the awareness of glucocorticoid-induced osteoporosis and the importance of prevention and treatment as the standard of care.¹⁵

Patients with osteoporosis may remain asymptomatic for many years, and bone loss occurs much sooner than the conspicuous signs of steroid use such as thinning of the skin, increased ecchymoses, proximal muscle wasting, truncal obesity, decreased vertebral length, and upper thoracic kyphosis. Vertebral compression fractures may develop with little warning or with minimal or no trauma.^{16 - 17} It is therefore of paramount importance to prevent osteoporosis by treating steroid-dependent patients before such complications manifest.

Since the last review by our group¹⁸ of this important topic, there have been many refinements in existing classes of medicines, more aggressive recommendations for patient management, and development of novel agents to prevent and treat glucocorticoid-induced osteoporosis. With ever-increasing evidence emerging from trials regarding glucocorticoid-induced osteoporosis, the recommendations for managing patients undergoing long-term steroid therapy need periodic updating. The present review aims to raise awareness in the dermatology community of these recent developments and to provide a clear treatment approach to aid in the care of patients treated with systemic corticosteroids.

Although this review focuses mainly on osteoporosis induced by oral corticosteroids, it is important for dermatologists to remember that highly potent topical steroids also pose a potential threat. Case reports have demonstrated suppression of the hypothalamic-pituitary-adrenal axis following cessation of topical clobetasol treatment,¹⁹ which indicates that topical corticosteroids can induce supraphysiologic levels of glucocorticoids. Any opportunity to use topical steroids in exchange for oral preparations should be endorsed, but it is important to remember that potent topical agents are still potentially harmful to bone.

A comprehensive discussion of the pathogenesis of steroid-induced osteoporosis is beyond the scope of this article. However, several aspects are important to mention to properly explain the mechanism of this disease and the rationale for various treatment options. The predominant effect of glucocorticoids on the skeleton is a loss of trabecular bone induced by several mechanisms. Simply stated, the causes of bone loss are multifactorial, including direct effects on bone metabolism and indirect (systemic) effects as well.^{20 - 22} The combination of these mechanisms leads to increased bone resorption and greatly diminished bone formation (Figure 1), so ideal treatment would prevent or diminish resorption and promote bone anabolism.

The direct effects of corticosteroids on bone metabolism involve reducing the number and function of osteoblasts (cells responsible for bone formation) while increasing the function of osteoclasts (cells responsible for bone resorption). The number of osteoblasts is decreased by inhibiting osteoblastogenesis as well as promoting osteoblast apoptosis. The function of osteoblasts is abated by a decrease in the translation of collagen type I, osteocalcin, fibronectin, and insulinlike growth factor, all of which contribute to the anabolic effects of the osteoblast cell. The function of the osteoclast is increased by an increase of receptor activator of nuclear factor κB ligand (RANKL) and a reduction in osteoprotegrin (OPG). Produced by osteoblasts, RANKL acts on an osteoclast receptor (RANK) to induce osteoclastogenesis, whereas OPG is a competitive inhibitor to the same receptor site. Glucocorticoids increase RANKL expression and decrease OPG levels, leading to an uninhibited signal promoting osteoclast proliferation.²³

The indirect (systemic) effects of corticosteroids include an increase in urinary calcium excretion, a decrease in intestinal absorption of calcium, secondary hyperparathyroidism, and a decrease in sex steroid production from actions on both the pituitary and adrenal glands. All of these effects ultimately lead to increased osteoclast-mediated bone resorption and subsequent bone loss.^{24 - 25}

The earliest changes of corticosteroid-induced bone loss are seen in bones with high trabecular content such as the lumbar spine and ribs, but bone loss can occur at any site. The standard method for assessing bone mass is dual-energy x-ray absorptiometry (DEXA) of the spine and hip (both sites, not just 1). This is a low-risk, noninvasive imaging technique that offers advantages such as lower radiation dose, lower cost, and more rapid examination time than other evaluation methods. The bone mineral density (BMD) is expressed as a T score, which basically defines the number of standard deviations away from a mean BMD score derived from a control group of normal individuals of the same sex between the ages of 20 and 35 years.

The World Health Organization classification of bone densiometry²⁶ defines normal bone density as no more than 1 SD below the mean value of the control group (T score, 0 to −1). Osteopenia, or low bone mass, is between 1 and 2.5 SDs below normal (T score, −1 to −2.5). Osteoporosis is diagnosed when the T score is −2.5 or lower. Some laboratories also provide Z scores in a BMD report, and these scores compare a patient's BMD with that of his or her age-matched peers.²⁷ The use of this measurement is controversial because T scores are the standard to define osteoporosis and guide overall therapy for this condition.

The most detailed analysis to date of the relationship between oral glucocorticoid use and fracture risk was included in the General Practice Research Database of the United Kingdom.²⁸ This retrospective cohort study compared nearly 250 000 individuals undergoing long-term glucocorticoid treatment with an equal number of age- and sex-matched controls in regard to risk of fracture. The risk of fracture increased within 3 months of initiating treatment with oral corticosteroids. The risk was more closely related to daily dose than to cumulative dose, with a significant increase in vertebral fracture seen with a dose as low as 2.5 mg of prednisone daily. Every-other-day dosing also proved ineffective in reducing bone loss. Results were categorized into the following 3 prednisone (or equivalent) dose categories: less than 2.5 mg/d; 2.5 to 7.5 mg/d; or more than 7.5 mg/d. The relative risk for vertebral fractures was 1.55, 2.59, and 5.18, respectively. The relative risk for hip fracture was 0.99, 1.77, and 2.27, respectively. There was a linear relationship between steroid dose and clinical fractures up to a dose of 20 mg/d; at doses greater than 20 mg/d, a more exponential increase in fractures was observed. In addition, increased age (most important), female sex, low body mass index (calculated as weight in kilograms divided by the square of height in meters), and low bone density also correlated with an increase in fracture risk. These data refute the notion of a “safe” prednisone dose: all doses correlated with increased risk.

Another study recognized that patients with glucocorticoid-induced osteoporosis appeared to have fracture rates at similar BMD levels as patients with age-related osteoporosis,²⁹ although some other studies have demonstrated that patients with glucocorticoid-induced osteoporosis are in fact more likely to experience fractures at BMDs much higher than those found in traditional osteoporosis.^{30 - 32} Although this point remains unclear, fracture risk with steroid use seems to rise when BMD T scores fall below −1.5. Since it can be expected that the BMD T score for an individual undergoing steroid therapy may drop 10% or 1 SD in the first year, initiation of treatment should be considered at BMD levels greater than levels recommended for treatment of traditional osteoporosis (T score of −2.5). Therefore, current recommendations suggest treatment should be initiated for any patient with a T score of −1.5 or below.³³

Glucocorticoid-induced bone loss can be prevented and treated. Because the bone loss is most dramatic in the first 3 to 6 months of treatment, preventive measures must be implemented the same day the steroid is prescribed.²⁸ Delay in implementing primary prevention measures in the appropriate patients (patients without bone loss) can have devastating and avoidable consequences. In addition, patients already undergoing steroid treatment will, without exception, have some degree of bone loss and will greatly benefit from treatment (secondary prevention).

Interventions should include both nonpharmacologic and pharmacologic therapies. Pharmacologic therapies include calcium and cholecalciferol (hereinafter “vitamin D”); antiresorptive therapies such as bisphosphonates, salmon calcitonin (calcitonin), estrogen replacement, and selective estrogen receptor modulators; and the first anabolic agent, recombinant parathyroid hormone.³⁴

For every patient regardless of age and dose and duration of steroid therapy, certain general measures should be implemented (Table 1).³⁵ The first is for the clinician to use the lowest effective dose possible, because every decrease in dose is beneficial to bone loss, and also to use topical corticosteroids instead of systemic when appropriate. Next, all patients should be instructed to engage in both weight-bearing and cardiovascular exercises, which appear to have small beneficial effects on bone density.³⁶ Most patients should be encouraged to strengthen the proximal musculature of the abdomen, back, and limbs. Despite the lack of data to support the premise that exercise is beneficial at reducing fractures, exercise remains a recommendation for all patients. When appropriate, patients should be referred to physiatrists or physical therapists who specialize in treating osteoporosis and creating exercise programs for patients with the disease. Another important recommendation is to reduce modifiable risk factors such as smoking, alcohol abuse, and caffeine intake. Patients should also be encouraged to maintain a nutritional diet including adequate calcium and vitamin D.³³

Calcium in the diet may be the limiting factor for bone growth and maintenance. The calcium intake of the average Western diet is estimated to be 500 to 1000 mg/d, and there is evidence in postmenopausal osteoporosis that increasing this amount by an additional 1000 mg/d slows bone loss by one third. Extensive data on calcium intake in corticosteroid-induced osteoporosis indicates that increased calcium intake by itself is probably insufficient to prevent bone loss, but it should be regarded as an adjunctive therapy in all patients.^{35 ,37} However, in a study of calcium in combination with vitamin D, a significant decrease in bone loss was observed in patients undergoing long-term treatment with low-dose corticosteroids.³⁷ We recommend prescribing 1500 mg/d of calcium and 800 IU/d of vitamin D for all patients treated with corticosteroids. The only restrictions are for patients with a history of nephrolithiasis or hypercalcemia. It has been suggested that patients treated with calcium and vitamin D should have urine tests for hypercalciuria (>300 mg of calcium excreted in 24 hours) and those who test positive should be treated with thiazide diuretics,³⁸ which decrease urinary excretion of calcium by causing resorption in the kidney, thus increasing serum calcium concentrations. However, no strong evidence exists to recommend thiazide diuretics in all patients.

Vitamin D metabolites (calcitriol and alfacalcidol) are newer formulations and quite distinct in action from plain vitamin D. While several studies have shown these medications to be beneficial in reducing bone loss,³⁹ there are no data to suggest that they are superior to plain vitamin D.⁴⁰ One potential problem with these medications is the development of hypercalcemia, which was observed in 25% of patients treated concomitantly with calcium. Therefore, until more convincing studies are available, vitamin D with calcium should be used before the vitamin D metabolites.

Bisphosphonates are currently the preferred agents for both treatment and prevention of glucocorticoid-induced osteoporosis, with several excellent studies demonstrating their effectiveness in both men and women. They act by binding hydroxyapatite on the surface of bone and preventing osteocyte-mediated bone resorption while also simulating production and preventing apoptosis of osteoblasts and also increasing levels of osteoprotegrin (the competitive inhibitor of RANKL).⁴¹ The largest studies to date have been of oral alendronate sodium and risedronate sodium, both of which were shown to prevent bone loss and reduce new fractures. Since our group's last update,¹⁸ which recommended 10 mg/d of alendronate sodium^{42 - 43} and 5 mg/d of risedronate sodium,^{44 - 46} newer formulations of once weekly doses of 70 mg/wk and 35 mg/wk, respectively, have been released.⁴⁷ Once-weekly dosing has improved patient compliance with these treatments. Oral bisphosphonates have very poor bioavailability, so they must be taken in the morning, fasting, and with only water to be absorbed at all. Furthermore, because oral bisphosphonates can cause upper gastrointestinal irritation, patients must not lie flat for 60 minutes after ingesting the dose. With these restrictions, it is clear why once-weekly dosing would meet with far better compliance by patients.⁴⁸ In addition, canine studies have shown that once-weekly bisphosphonate dosing causes less gastrointestinal irritation than once-daily dosing and has similar efficacy.⁴⁹ Another potential adverse effect of bisphosphonates, albeit rare, is osteonecrosis of the jaw. Most cases were associated with intravenous (IV) bisphosphonates and high doses of oral bisphosphonates.^{50 - 51}

Currently, alendronate and risedronate are the most commonly prescribed oral bisphosphonates. However, another drug has recently been approved by the US Food and Drug Administration for treatment of postmenopausal osteoporosis: ibandronate sodium, an oral bisphosphonate. It is dosed at 150 mg orally, just once per month, with similar BMD increases and fracture efficacy as weekly alendronate.⁵² Studies of ibandronate in steroid-induced osteoporosis are pending, but it would seem logical to extrapolate data from postmenopausal osteoporosis research.

Prior to recent developments, patients who are unable to remain upright for 60 minutes after dosing, have difficulty swallowing, or have reflux esophagitis or other upper gastrointestinal motility disorders were unable to receive the most effective medications available for corticosteroid-induced osteoporosis. Deconditioned elderly people (like many patients with bullous pemphigoid), already at high risk for bone loss, represent a cohort less able to remain upright following oral bisphosphonate therapy. However, a recent study suggests that the IV bisphosphonate pamidronate disodium may be used as an alternate to oral bisphosphonates to prevent glucocorticoid-induced osteoporosis.⁵³ Currently, IV pamidronate is not approved by the Food and Drug Administration for this indication, but in patients unable to tolerate oral bisphosphonates it has been widely used off label. The current recommendation of dosing for IV pamidronate disodium is 90 mg once and then 30 mg every 3 months. Recently, weekly intramuscular injection of another bisphosphonate, clodronate (clodronic acid) (100 mg), was shown to prevent bone loss and fractures in patients taking glucocorticoids, although this preparation is not widely available.⁵⁴ Another new and extremely potent IV bisphosphonate dosed once a year, zoledronate (zoledronic acid), has been shown (in postmenopausal osteoporosis) to produce changes in bone density similar to those seen with standard oral bisphosphonates.⁵⁵ A case report of a young girl with lupus erythematosus with multiple vertebral collapses due to glucocorticoids revealed significant clinical and densitometric improvement when treated with IV zolendronate.⁵⁶ This report shows effectiveness of zolendronate for glucocorticoid-induced bone loss and in children. Fracture data are still pending (phase 3 trial), but this medication offers tremendous promise.

Studies in premenopausal women have shown the risk of fracture to be much lower than in postmenopausal women. Bisphosphonates should be used with caution in this subset of patients because they cross the placenta and have teratogenic effects in animal studies. In addition, bisphosphonates have extremely long half-lives, and the risks to the developing fetus, even years after termination of bisphosphonate treatment, are yet unknown. Until long-term data are available, these medications should be avoided in individuals who wish to conceive.

Another subset of patients less commonly studied is children. One study demonstrated that alendronate was effective therapy in children older than 4 years, with a bone density increase of 15% vs 2% among controls.⁵⁷ In addition, all the children demonstrated a height increase. Another study showed that once-weekly oral alendronate therapy is well tolerated in children taking glucocorticoids. It suppresses bone resorption and increases volumetric bone density at the lumbar spine and mechanical strength of the femoral shaft.⁵⁸ There have been some reports of increased fractures in children treated with bisphosphonates secondary to osteoporosis and decreased flexibility of the bone; however, this has not been consistently observed and is not associated with adults.⁵⁹

While calcitonin is a potent antiresorptive agent in vitro, clinical studies have repeatedly shown relatively weak antiresorptive actions in vivo. Numerous studies have evaluated the efficacy of calcitonin in glucocorticoid-induced osteoporosis. A meta-analysis of 9 such randomized controlled studies indicated that calcitonin treatment for 6 to 12 months was more effective than calcium alone in maintaining spinal bone density.^{60 - 61} However, calcitonin failed to demonstrate decreases in fracture risk in the spine or femoral neck. Calcitonin should be regarded as, at best, a second-line agent, and its attractiveness is even further diminished by need for intranasal administration.

For years, estrogen preparations have been very popular for the treatment and prevention of postmenopausal osteoporosis.⁶² The Women's Health Initiative (WHI) study was a large controlled trial that demonstrated significantly decreased incidence of spine and hip fracture among women treated with estrogen and progesterone over 5 years.⁶³ Unfortunately, the WHI also reported an excess of cardiovascular and cerebrovascular disease in these women and increased breast cancer risk. These risks outweigh the potential benefit on bone loss, and thus estrogen and progesterone should not be a first-line treatment of osteoporosis. It is unclear whether low-dose estrogen without progesterone has the same risk profile as the estrogen and progesterone combination, but until studies confirm a decreased risk, these agents should be used judiciously.

With the concern generated over estrogen therapy, more attention has focused on the use of selective estrogen receptor modulators (SERMs), which exert tissue-selective estrogenic actions while having anti-estrogenic actions on other tissues as well, therefore potentially conferring the beneficial skeletal effects of estrogen without producing the adverse actions on the breast and vascular systems. Raloxifene hydrochloride, the most commonly used SERM, was found to increase BMD in the lumbar spine and hip.^{64 - 65} When compared directly with estrogen in treating postmenopausal women, raloxifene was found to be inferior in increasing bone density. The role of raloxifene in preventing corticosteroid-induced osteoporosis remains to be established. The effects of SERMs on cardiovascular and cerebrovascular disease, in light of the WHI results, will also be important to rationalizing its use in treating osteoporosis.⁶⁴

Testosterone levels are also known to decrease in men undergoing long-term glucocorticoid therapy. One study of 15 men undergoing long-term glucocorticoid treatment for asthma revealed that all 15 were testosterone deficient.⁶⁶ This study also revealed that after 1 year of testosterone replacement therapy (250 mg intramuscularly per month), bone density was increased at the lumbar spine but not at other places. Serum testosterone levels should be measured in patients undergoing long-term corticosteroid treatment and testosterone replacement therapy initiated if the levels are found to be low. Contraindications include prostate cancer or benign prostatic hypertrophy.

Since glucocorticoid-induced bone loss is primarily a disorder marked by decreased bone formation, it makes sense to find therapies that elicit anabolic effects on bone. Until the introduction of teriparatide, fragmented recombinant parathyroid hormone, no such therapy was available. Teriparatide has been shown to considerably increase trabecular bone mass, improve bone microstructure, and prevent fractures, thus providing benefits that cannot be provided by current antiresorptive drugs.⁶⁷ Teriparatide appears to be superior to antiresorptive therapy (alendronate) in improving BMD at the lumbar spine. No direct comparisons of antifracture efficacy between these 2 classes have been done,⁶⁸ but in 1 study in glucocorticoid-induced osteoporosis, teriparatide led to dramatic increases in BMD.⁶⁹

Currently, the cost of teriparatide is a limiting factor in its use, particularly without compelling evidence that it is more effective than bisphosphonates. However, a recent review⁷⁰ pointed out 3 clinical indications for potential use of teriparatide: history of osteoporotic fractures, severely diminished BMD (T score, <−3.0), and failed bisphosphonate therapy, all cases where bone anabolism would be advantageous.

In patients with glucocorticoid-induced osteoporosis with fractures, teriparatide should be considered a first-line treatment.⁷⁰ Currently, it is not recommended to use both teriparatide and bisphosphonates in combination, and some studies indicate bisphosphonates may inhibit the effectiveness of teriparatide.⁷¹ However, there has been some compelling evidence that initiating bisphosphonate therapy following the cessation of teriparatide treatment may aide in maintaining the bone formation created by parathyroid hormone.⁷² Future clinical studies will refine the appropriate use of teriparatide and other parathyroid hormone formulations on the horizon.

Strontium ranelate is a divalent cationic drug that has recently emerged as a therapeutic option for osteoporosis. Early studies have revealed that bone density increases were substantial with strontium treatment, approaching those seen with bisphosphonates. A large randomized clinical trial confirmed these early reports, demonstrating that 2 g of strontium ranelate taken daily reduced the risk for vertebral fractures by 40% over a 3-year period.⁷³ No data in glucocorticoid-induced bone loss with strontium exists, and this agent is not available in the United States.

Statins are widely used in the world to treat dyslipidemia. Recently, statins were found to stimulate bone formation in vitro and in vivo.⁷⁴ However, reports of a negative association of statins with bone formation have also been published.⁷⁵ This inconsistency may be attributable to the fact that statins have very limited access to bone and are mainly targeted to the liver. In animal models, the potency of statins in bone anabolism appears high, and a goal of developing a bone-targeted statin is now in progress. Development of a transdermal statin would be of interest because this administration method would bypass first-order metabolism in the liver.

As the understanding of the cellular and molecular mechanisms of bone metabolism are further elucidated, new targets of interest in the treatment of bone loss will come to light. Current treatments under study include a recombinant osteoprotegrin molecule, antibodies against the endogenous peptide RANKL, inhibitors of osteoclast enzymes such as cathepsins K and L, matrix metalloproteinases, and Src tyrosine kinase. Integrin antagonists that block osteoclast attachment to bone are also of potential value. Molecules that stimulate the endogenous product of parathyroid hormone are being explored, as well as agonists to osteoblast-activating receptors like LRP5. Almost all therapeutic options share the final common pathway of either inhibiting osteoclast resorption or stimulating osteoblast bone formation.⁷⁶

The available treatments for glucocorticoid-induced bone loss are certainly efficacious; however, whether this aggressive approach is cost-effective in all patients undergoing systemic corticosteroid treatment remains to be established. One recent study evaluated the role of alendronate, calcium with vitamin D, and placebo in a cost-effectiveness analysis and found alendronate to be more cost-effective than placebo over a 10-year follow-up period.⁷⁷ Another study of postmenopausal osteoporosis similarly looked at the cost-effectiveness of alendronate vs placebo. This study found that the average direct cost to patients with hip fractures was $25 027, with indirect costs of $200 to $6000 depending on age.⁷⁸ With the rising cost of health care, especially in the inpatient setting, the overall cost of treating a fracture will only escalate. While many of the available medications are quite expensive (Table 2), compared with the hospital charges secondary to a fracture, they remain a bargain in high-risk patients such as those on a long-term corticosteroid regimen.

Dermatologic patients receiving prolonged corticosteroid therapy, defined as longer than 3 months, should be evaluated for prophylactic treatment to decrease the risk of bone loss. For duration of steroid use shorter than 3 months, current literature does not suggest treatment.^{33 ,79} All patients starting glucocorticoid treatment should be given instruction on general measures for prevention of bone loss (Table 1). Clinicians should assess each patient's risk of osteoporosis and fracture based on known risk factors (Table 3). The total length of treatment with an oral corticosteroid should also be carefully estimated.

The simplest cases to manage are those from high-risk populations such as postmenopausal women, men older than 65 years, patients with a history of low-trauma fractures in any bone, patients with a history of frequent falls, or patients taking more than 20 mg/d of prednisone. Patients who meet any 1 of these criteria should begin treatment with an oral bisphosphonate immediately when starting oral glucocorticoid therapy, without the need for BMD testing (DEXA scan). Patients who will require 3 months of corticosteroid treatment but who do not meet these criteria should have BMD testing (DEXA of the hip and spine) prior to beginning treatment. Patients with T scores lower than −1.5 should also begin treatment with an oral bisphosphonate. In addition to patients starting a course of corticosteroids, patients already in the midst of a long-term corticosteroid regimen should also be screened with a DEXA scan of the hip and spine. T score results in these patients should be evaluated by exactly the same criteria as the scores of de novo users. Figure 2 shows an algorithm for the management of patients treated with systemic glucocorticoids. Bone mineral density should be reassessed every 6 to 12 months until glucocorticoid therapy is terminated. At this time, patients should then be treated until BMD normalizes in a manner similar to those not using glucocorticoids.

The first choice for prevention and treatment of glucocorticoid-induced osteoporosis should be a potent oral bisphosphonate such as alendronate sodium (70 mg/wk) or risedronate sodium (35 mg/wk). Patients who are unable to sit upright for 60 minutes after taking oral bisphosphonate or who have esophagitis should be treated with an intravenous bisphosphonate such as pamidronate disodium or zolendronate (zolendronic acid). Premenopausal women interested in becoming pregnant should not receive bisphosphonates. Options for this group include vitamin D metabolites or calcium and plain vitamin D. If bisphosphonates are used, strict contraceptive measures should be addressed and documented. In patients with severely diminished bone density (T score, <−3.5) or patients with active osteoporotic fractures, the anabolic agent teriparatide should be considered as a first-line option for up to 2 years. When considering IV bisphosphonates or teriparatide, dermatologists should work in concert with patients' primary care physicians and endocrinologists to have these agents administered. Endocrinologists should certainly be involved in the treatment of children and premenopausal women undergoing long-term treatment with glucocorticoids. Furthermore, in postmenopausal women, estrogen replacement is no longer recommended for long-term prevention of osteoporosis.

In conclusion, every physician prescribing systemic corticosteroids must be aware of this terrible and preventable complication, osteoporosis, and ensure that every patient is receiving general measures and, if applicable, preemptive pharmacotherapy. Dermatologists are among the leading prescribers of oral glucocorticoids; therefore, this topic is of utmost importance. Not only can failure of treatment lead to serious morbidity for dermatologic patients, it may also have medicolegal ramifications for the clinician. Current approaches have enabled physicians to greatly reduce the number of fractures as a result of corticosteroid use. As the knowledge of these approaches become widespread and novel agents are developed, hopefully the incidence will continue to decline.

Correspondence: Gil Yosipovitch, MD, Department of Dermatology, Wake Forest University School of Medicine, Medical Center Boulevard, Winston Salem, NC 27157-1071 (gyosipov@wfubmc.edu or summey@alumni.unc.edu).

Financial Disclosure: None.

Accepted for Publication: August 10, 2005.

Author Contributions: Both authors contributed equally to the design and prepartion of this article.