: Philip N Sambrook, MD, LLB, FRACP, Professor of Rheumatology, Institute of Bone & Joint Research, University of Sydney, Sydney, Australia. Email: email@example.com
Introduction: Glucocorticoids have major effects on bone metabolism, leading to accelerated osteoporosis and fracture.
Methods: This review will attempt to summarize current knowledge about their effects in light of new information and important remaining questions, especially with respect to management of this common clinical problem.
Results: Glucocorticoids affect bone through multiple pathways, influencing both bone formation and bone resorption. Evidence from randomised trials suggests that postmenopausal women receiving glucocorticoids are at greatest risk of rapid bone loss and consequent fracture and should be actively considered for prophylaxis. Based upon available evidence, the rank order of choice for prophylaxis would be a bisphosphonate followed by vitamin D. For established glucocorticoid induced osteoporosis, parathyroid hormone followed by a bisphosphonate appears useful.
Conclusions: Glucocorticoid-induced bone loss can be effectively prevented or reversed.
Glucocorticoids are known to affect bone through multiple pathways (Fig. 1), influencing both bone formation and bone resorption.1–3 The most important effects appear to be on bone formation due to direct effects on cells of the osteoblastic lineage,3 although indirect effects related to sex steroid production are also important. The predominant effect of glucocorticoids is inhibitory towards osteoblasts and differentiation of osteoblast precursors. Increased osteoctye apoptosis has also been implicated as an important mechanism of glucocorticoid osteoporosis.4 Glucocorticoids have been shown to reduce the birth rate of osteoblasts and osteoclasts and cause earlier death of osteoblasts.4 The effect of glucocorticoids to promote apoptosis in osteocytes as well as in osteoblasts could account for the increase in fracture susceptibility. Glucocorticoids have been shown to decrease intestinal calcium absorption and increase urinary phosphate and calcium loss by direct effects on the kidney, contributing to secondary hyperparathyroidism and hence increased bone resorption.1–3 Increased bone resorption is probably only a temporary phenomenon during the first 6 or 12 months of glucocorticoid therapy and is in part due to the effects of glucocorticoid on osteoprotegerin and its ligand, RANKL (receptor activator of NFκB ligand). Glucocorticoids increase the expression of RANKL and decrease osteoprotegerin expression in osteoblasts leading to postponement of osteoclast apoptosis.5 At the organ level, direct inhibitory effects of glucocorticoids on bone formation have also been documented in histomorphometric studies.6
GLUCOCORTICOIDS, BONE LOSS AND FRACTURES
When high-dose glucocorticoids are used, bone loss can be rapid, ranging between 5–15% per annum.7,8 However, not all patients on glucocorticoids lose bone. The reasons for this large degree of individual variation are unknown but probably include differences in dose, underlying disease and genetic susceptibility to glucocorticoids.9 Glucocorticoid bone loss is dose-dependent,10 although ‘low-dose’ glucocorticoids may still cause rapid initial bone loss in some patients. In patients on chronic low-dose therapy, it is uncertain whether there is continuing slow loss, although the few longitudinal studies and clinical trials suggest it may stabilize in some individuals.11–13 The degree of bone loss due to glucocorticoids varies markedly according to skeletal site and appears greater at trabecular-rich bone sites. Glucocorticoid-induced bone loss may also be affected by the underlying disease for which the steroid is prescribed. For example, spinal bone loss in rheumatoid arthritis is related to disease activity14 and greater loss is seen in patients treated with daily prednisolone doses ranging between 1–5 mg than doses of more than 5 mg because of poorer disease control in the former group.14
Inhaled steroids are less likely to have systemic effects than oral glucocorticoids, but in higher doses, result in adrenal suppression and reduced bone density. Analysis of inhaled glucocorticoid users from the general practice registry in the United Kingdom matched against controls observed the relative risk of vertebral, hip and non-vertebral fracture was 1.5, 1.2 and 1.2, respectively, but no differences were found between inhaled and bronchodilator groups.15
The major clinical manifestation of the effects of glucocorticoids on bone are fractures most commonly in the vertebrae and ribs, regions of the skeleton with high amounts of trabecular bone. In an early study of glucocorticoid-dependent asthmatics, 11% of 128 patients surveyed had sustained 58 fractures and 42% of 19 patients (mean duration of glucocorticoid use, 8 years) followed prospectively sustained fractures, predominantly vertebral and to a lesser extent ribs.16 Some studies have suggested that glucocorticoid-treated patients suffer vertebral fractures at higher threshold levels of spinal bone density than observed in involutional osteoporosis, suggesting a qualitative as well as quantitative defect of bone,17–19 likely to reflect changes in microarchitecture or trabecular connectivity.6
The prevalence of vertebral fracture with glucocorticoids has been estimated at 28%.20 A large cohort study of subjects on glucocorticoid identified from a general practice registry in the United Kingdom observed the relative risk of clinical vertebral fracture during oral glucocorticoid therapy was 2.6, with the relative risk for hip fracture being 1.6 and for non-vertebral fracture 1.3.10 Fracture risk increased with increasing daily doses of glucocorticoids and when discontinued, the fracture risk appeared to return to baseline.
TREATMENT OF GLUCOCORTICOID BONE LOSS
Over the last 15 years there have been an increasing number of studies investigating the efficacy of various treatments in glucocorticoid bone loss. As bone loss is most rapid soon after starting glucocorticoids, the studies are best classified as either primary prevention (in patients starting glucocorticoids) or treatment of established loss (in patients on chronic therapy).
Several trials which have examined the efficacy of bisphosphonates on glucocorticoid-induced bone loss.13,21–25 The two most commonly used bisphosphonates in clinical practice currently are alendronate and risedronate. Saag etal.13 reported the combined results of two trials in 477 glucocorticoid treated subjects who received prophylaxis with alendronate or calcium/vitamin D (800–1000 mg daily plus 250–500 IU daily, respectively). Patients were stratified according to whether they had received glucocorticoids for less than 4 months, 4–12 months or greater than 12 months. Over 12 months of follow-up, the mean change in lumbar spine bone mineral density (BMD) in patients in the primary prevention group was +3.0% for alendronate 10 mg/day compared to –1% in the placebo group. In those who had received chronic glucocorticoids for greater than 12 months, the increase with alendronate was +2.8% compared to +0.2% for calcium. A post hoc analysis of incident vertebral fractures determined semiquantitatively, significantly favoured alendronate in postmenopausal women only (13%vs 4.4%). Risedronate has also been examined in glucocorticoid osteoporosis.23 In a primary prevention trial in 224 glucocorticoid-treated subjects who received prophylaxis with either residronate or calcium 500 mg daily, residronate 5 mg per day prevented spinal bone loss (+0.6%) compared to calcium (–2.8%) over 12 months. Incident vertebral fracture rates were 17.3% with calcium and 5.7% for residronate 5 mg (P = 0.072). Vertebral fractures were only seen in postmenopausal women and men, not in premenopausal women. In 290 patients receiving chronic glucocorticoid treatment (prednisone ≥ 7.5 mg/day for ≥ 6 months), treatment with risedronate in a dose of 5 mg significantly increased lumbar spine (+2.9%) and femoral neck (+1.8%) BMD. Treatment with risedronate 2.5 mg daily showed a similar, although non-significant trend. Although not powered to show fracture efficacy, 15% of patients in the control group versus 5% in the risedronate groups sustained new vertebral fractures, indicating a 70% reduction in fracture rate. A recent study in 833 glucocorticoid-treated subjects comparing a single injection of zoledronate with daily risedronate observed zoledronate was superior to risedronate in increasing BMD at most sites and in reducing bone turnover markers.25 This was seen in both prevention and treatment subpopulations over 12 months.25
Although the term vitamin D is sometimes used to encompass both the calciferols and active vitamin D metabolites, they have quite distinct therapeutic effects. The most commonly used active hormonal forms of vitamin D, have been calcitriol (1,25 dihydroxy vitamin D) and alfacalcidol (1α hydroxy vitamin D). A number of studies have suggested beneficial effects of both the calciferols and active vitamin D metabolites in glucocorticoid osteoporosis,12,26,27 although in head-to-head comparisons, the effect of bisphosphonates appears superior to vitamin D.24,28,29
The anabolic actions of teriparatide include direct stimulation of osteoblastogenesis and inhibition of osteoblast apoptosis, reversing two of the key inhibitory effects of glucocorticoids on bone formation.4 A recent study by Saag etal.30 compared teriparatide with alendronate in 428 patients on chronic glucocorticoids with established bone loss defined as a T-score < –2 (or < –1 in the presence of a fragility fracture). Saag etal.30 report that treatment with teriparatide had significantly greater effects on bone density than alendronate at both the lumbar spine (7.2%vs 3.4%, P < 0.001) and hip (3.8%vs 2.4%, P = 0.005) over 18 months. Importantly, teriparatide-treated patients had a significant reduction the incidence of new vertebral fractures compared to alendronate over 18 months (0.6%vs 6.1%) although not on non-vertebral fractures.
To summarize, evidence from recent trials suggests that postmenopausal women receiving glucocorticoids are at greatest risk of rapid bone loss and consequent fracture and should be actively considered for prophylactic measures. In men and premenopausal women receiving glucocorticoids, the decision to use prophylaxis is less clear and will depend upon a number of factors including BMD, anticipated dose and duration of glucocorticoids and other risk factors. Based upon available evidence the rank order of choice for prophylaxis would be a bisphosphonate followed by a vitamin D metabolite. Since fracture risk is a function of multiple factors including the severity of low bone density as well as the duration of exposure, treatment with therapy to increase bone density will further reduce fracture risk in patients receiving chronic low dose glucocorticoids. For established glucocorticoid-induced osteoporosis in the treatment setting, parathyroid hormone appears to be particularly useful (Fig. 2).