Glucocorticoid-Induced Osteoporosis: A Review
- First Online:
- Cite this article as:
- Bouvard, B., Legrand, E., Audran, M. et al. Clinic Rev Bone Miner Metab (2010) 8: 15. doi:10.1007/s12018-009-9051-9
- 199 Views
Glucocorticoid (GC)-induced osteoporosis is the main cause of secondary osteoporosis. Fractures, which are often asymptomatic, can occur in as many as 50% of patients receiving chronic GC therapy. GCs have direct and indirect effects on bone cells (osteoblasts, osteocytes, and osteoclasts) with a suppression of bone formation and an increased bone resorption. The management of patients exposed to GCs should include the use of the minimal effective dose of GC, general health measures, and adequate intakes of calcium and vitamin D. Bisphosphonates are nowadays largely used in GC-induced osteoporosis and teriparatide has proved its efficiency as well.
KeywordsGlucocorticoid-induced osteoporosisFracturesBone mineral densityMicroarchitectureBone formationBone resorptionBisphosphonates
Tissular effects of GCs
Hypothyroidism and hypogonadism, menstrual irregularity
Dyslipemia, hypertension, thrombosis
Combination of visceral obesity, systemic arterial hypertension, impairment of glucose tolerance, and/or dyslipemia. Hyperphagia and weight gain
Involution of lymphoid tissue mass and lymphopenia
Increase susceptibility to infectious diseases
Hirsutism, acne, easy bruising, centripetal obesity, moon facies, buffalo hump, supraclavicular fat pads, and violaceous abdominal striae, lipodystrophy, altered wound healing
Nephrolithiasis and impairment of renal clearance, hypercalciuria, hyperuricosuria, increased urinary levels of oxalate, and decreased urinary levels of citrate
Central nervous system
Intellective deficiency, psychological impairment, anxiety, depression, atrophy of specific cerebral areas
Pancreatitis, epigastric pain
Direct and indirect effects of GCs on bone
↓ bone formation, ↑ bone resorption
↓ differentiation and function
↑ span life
Muscular weakness, proteolysis of myofibrils
↑ myostatin, muscle cramps
↓ sex steroids, ↓ LH pulse frequency
↓ intestinal absorption
↑ renal excretion
Molecular Effects of GCs
The GC Receptor
In bone cells, cellular effects of GCs are primarily mediated by the GC receptor which is a member of the nuclear receptor superfamily. In the absence of ligand, the receptor is retained in the cytosol (cGCR) as part of a chaperone-containing multiprotein complex . The human cGCR gene, located at chromosome 5q31.3, is widely expressed in a number of bone cells including osteoblasts (OB), osteocytes, and chondrocytes . The GCR transcripts are also detected in stromal-like tumor cells, macrophage-like cells [putative osteoclast (OC) precursor], and multinucleated osteoclast-like cells . Different studies have suggested that the polymorphism of the cGCR gene is correlated with bone mineral density (BMD) variation and may explain the heterogeneity to GC-associated bone loss and fractures [9–11].
Molecular effects of glucocorticoids
Transactivaction: synthesis of anti-inflammatory proteins through positive GREs (lipocortin 1, IκB)
Suppression of transcription of inflammatory genes through negative GREs
Transrepression: monomers of GC/GCR complex interact with transcription factors which are involved in regulating expression of pro-inflammatory genes (AP-1, NF-κB) thereby suppressing synthesis of inflammatory cytokines (IL-1, IL-2, TNF-α, IFN-γ)
Competition for nuclear coactivators between the GC/cGCR complex and transcription factors
Interactions of GC with cellular membranes
Non-genomic effects mediated by the cGCR
Specific interactions with a membrane-bound GCR
Tissue responses to GCs are partly determined at a prereceptor level through expression of the 11β-hydroxysteroid dehydrogenase enzymes (11β-HSDs), which catalyze the interconversion of the hormonally inactive cortisone into active cortisol [17–19]. Two isozymes, 11β-HSD1 and 11β-HSD2, have been shown to regulate GC and mineralocorticoid hormone action. 11β-HSD1 acts predominantly as a reductase converting inactive cortisone to active cortisol. In contrast, 11β-HSD2 inactivates cortisol to cortisone. 11β-HSD1 transcripts are increased by GCs and agents that block 11β-HSD1 in bone may reduce GC effects on bone . IL-1 and TNF-α inhibit the expression of 11β-HSD2 and induce mRNA expression and activity of 11β-HSD1 in human OB, suggesting that IL-1 and TNF-α may sensitize skeletal tissue to the action of GC . Variation in 11β-HSD isoenzyme expression and activity may explain individual susceptibility to GIOP.
Cellular Effects of GCs
Cellular effects of GCs
Induction of the adipogenetic transcription factors (PPARγ and CCAAT enhancer binding protein family)
Opposing Wnt/β-catenin signaling pathway: ↑ Dkk-1 ↑ SFRP1, ↓ PI3K/Akt/GSK3beta/beta-catenin/LEF axis and ↑ histone deacetylase 1
↓ Krox 20
↑ MKP-1 and Notch-1
↓ synthesis and function of IGF-I
↓ type I collagen synthesis
↓ synthesis of IGFBP-3, -4 and -5 and ↑ IGFBP-6
↑ collagenases 1 and 3
↑ caspase 3, ↑ Bim, ↓ Bcl-2/Bax
↑ caspase 3, ↑ Bim, ↓ Bcl-2/Bax
Reduction in the elastic modulus and mineral to matrix ratio around the osteocyte lacuna
↑ M-CSF, RANK-L, ↓ osteoprotegerin
↑ IL-6, ↓ IFN-β
Impact on the cytoskeleton
Effects of GCs on OB Function
GCs reduce the expression of insulin-like growth factors (IGF) I and II known to increase OB differentiation [32, 33], type I collagen synthesis, and bone formation, and increase bone collagen degradation by decreasing the synthesis of collagenases 1 and 3 . GCs also reduce the expression of non-collagenous proteins such as osteocalcin, osteopontin, bone alkaline phosphatase, and tissue inhibitor of metalloproteinase 1 [25, 35].
GCs also inhibit the synthesis of IGFBP-3, -4, and -5 (which are binding proteins that can stimulate bone cell growth) and increase the expression of IGFBP-6 (a binding protein that selectively blocks the effects of IGF-II on OBs) . Taken together, these effects induce a marked reduction in the number of OBs and in their capacity to synthesize bone matrix.
In vivo and in vitro studies indicate that GCs favor the differentiation of bone marrow stromal cells toward the adipocytic pathway by the induction of the adipogenetic transcription factors (PPARγ and CCAAT enhancer binding protein family). This may occur at the expense of OB cell differentiation [37–40]. GCs also inhibit OB differentiation by opposing Wnt/β-catenin signaling, a key regulator of osteoblastogenesis. In the absence of the Wnt protein, β-catenin is phosphorylated by glycogen-synthase kinase-3β (GSK-3β), and then degraded by ubiquitination. In the presence of Wnt, the protein binds to frizzled receptors and then to the co-receptors LRP-5 or LRP-6 (low-density lipoprotein receptor-related proteins) leading to inhibition of GSK-3β activity. When GSK-3β is not active, β-catenin is stabilized and translocates to the nucleus, where it associates with transcription factors to regulate gene expression. Deletions of either Wnt or β-catenin result in the absence of osteoblastogenesis and increased osteoclastogenesis. The Wnt pathway can be inactivated by Dickkopf-1 (Dkk-1), an antagonist that prevents Wnt binding to its receptor complex. Dexamethasone (Dex) markedly induces the expression of Dkk-1 mRNA in cultured human OB . The addition of anti-Dkk-1 antibodies partially restores the transcriptional activity suppressed by Dex ; in contrast knocking down Dkk-1 expression in bone tissue abrogates GC impairment of osteogenic activities, microarchitecture, and bone mass . Furthermore, a supraphysiological level of GC can enhance SFRP-1 expression, another antagonist of the Wnt/β-catenin signaling pathway; knocking down SFRP1 abrogates this effect [44, 45]. GCs also inhibit osteoblast differentiation through the repression of growth hormone (GH) and bone morphogenic protein 2 (BMP-2) which enhances OB transcription factors . Other less known mechanisms have been identified to explain the impact of GCs on OB differentiation involving Krox 20/EGR2 , MPK-1 , Notch , and the TGF-β-Smad pathway .
Apoptosis of Cell of the Osteoblastic Lineage
Effects of GCs on Bone Resorption
In vitro, GCs increase the expression of M-CSF  and RANK-L [13, 55] in stromal and OBs (two cytokines involved in OC differentiation), and decrease osteoprotegerin expression (the soluble decoy receptor of RANK-L) [56, 57]. In vivo, GCs act directly on OCs to delay apoptosis of mature cells explaining the enhanced and prolonged bone resorption observed in GIOP [58, 59]. However, GCs seem to decrease their resorption capacities by disturbing cytoskeleton organization .
Tissular Effects of GCs
Bone Histomorphometry and Microarchitecture
BMD and Fractures
Synthetic GCs are widely used in clinical practice . Since the osteopenic effects of GCs are greater in trabecular than in cortical bone, bone loss especially affects the axial skeleton (vertebral bodies, pelvis, and ribs). Strong correlations have been found between the cumulative GC dose and the reduction in spine and hip BMDs [67–72]. The onset of bone loss is rapid, within the first months of starting GC therapy and slows down after about 1 year of treatment, while remaining higher than normal [73, 74]. In GCs-treated patients, the underlying disease (such as rheumatologic diseases, chronic pulmonary disorders, inflammatory bowel diseases, or post-transplantation status) also constitutes a risk of osteoporosis by itself [75–78].
The use of GCs significantly increases the risk of any fracture at all ages . It is estimated that as many as 50% of patients requiring long-term GCs (>3 months) will ultimately suffer from fractures [80, 81]. The risk of fracture starts rapidly after the onset of GC treatment and even low doses of GCs are able to induce fractures. For the same BMD level, the risk of all fractures is substantially greater in GIOP than in post-menopausal osteoporosis , with an increased incidence of asymptomatic vertebral fractures . Therefore, GC may have an effect on bone quality that leads to an increased risk of fracture. The different pharmacological types of GCs do not induce the same risk of fracture: oral GCs (such as prednisolone or prednisone) are more potent than hydrocortisone and are associated with a dose-dependent increase in the overall fracture risk. No increase in fracture risk seems to be associated with corticosteroids used as intra-articular or local use. The effect of inhaled GCs on vertebral fractures risk remains controversial. No bone loss is associated with inhaled corticosteroids, except at daily dosages above 7.5 mg of prednisolone (equivalent to 1875 μg of budesonide or beclomethasone) [84–86]. Furthermore, there is no evidence that long-term treatment of children with inhaled GCs at low doses is associated with a reduction in BMD or with an increased risk of osteoporosis or fracture [87, 88]. Current evidence shows that treatment with inhaled corticosteroids (medium or high doses) can delay growth rate at the start of treatment with beclomethasone or budesonide. However, this is transitory, since the adult height reached by asthmatic children with inhaled corticosteroids treatment is not different than that reached by non-asthmatic children . The excess risk of fracture (mainly vertebral fractures) seems to return to baseline values within 1 year after withdrawal of GCs [73, 86].
The effects of GCs on bone metabolism are reflected in marked changes in biochemical markers of bone turnover [90–92]. Markers of bone formation such as serum osteocalcin fall within a few hours of treatment and parallel the GC dose. Results for bone resorption markers are more variable: they depend on the marker used and their evolution upon time is variable [56, 93, 94].
Others Actions of GCs (Table 2)
In addition to the direct effects of GCs on bone cells, the catabolic effects of GCs on muscle may contribute to fracture risk by inducing muscular weakness which, in turn, increases the incidence of falls . Reduction of gonadal hormones is also an important mechanism through the inhibitory effects of GCs on pituitary gonadotropins . GCs also inhibit calcium absorption in the gastrointestinal tract because of a secondary resistance to vitamin D and enhance urinary loss of calcium. Despite the decrease in serum calcium, hyperparathyroidism does not appear nowadays to be a determinant of bone resorption or skeletal loss in acute or chronic use of GCs .
Experimental Models of GIOP
Animal models of GIOP
Action in vivo
↓ BMD, ↓ bone formation, ↑ bone resorption
↑ BMD or ↓ BMD
↑ BV or ↓ BV
Inhibition of bone formation and bone resorption
↓ BV, ↓ mineralizing surfaces, ↑ eroded surfaces, ↑ OB, and osteocytes apoptosis
Heterogeneity of bone loss
↓ number of OC, ↑ eroded surfaces, ↓ osteoblastic surfaces, ↓ bone formation rates
↓ BMD, ↓ BV, ↓ markers of bone formation and bone resorption
Marked bone loss
Treatment of GIOP
Preventive Counter Measures and Lifestyle Recommendations
Preventive measures should be considered for all patients once the decision has been made to initiate a long-term GC therapy; they should be started at the beginning of GC therapy. Optimal management strategies to prevent bone loss should include the use of the lowest effective GC dose . Good nutrition with adequate levels of dietary calcium should be recommended to all patients on GCs. Smoking and excessive alcohol intakes are both detrimental to bone health. Weight-bearing and muscle strengthening exercises are recommended.
Calcium and Vitamin D
Calcium and vitamin D supplementation is advised in all GIOP management guidelines. The benefit of calcium and vitamin D supplementation in the prevention of GIOP appears to be modest. Nevertheless, randomized clinical trials and meta-analysis have shown that treatment with calcium and vitamin D, calcitriol (1,25 dihydroxyvitamin D3), and alphacalcidol (1-α-hydroxyvitamin D3), is more effective in preventing GIOP than placebo or calcium alone with a relative preservation of BMD. The effect is more pronounced at the spine than at the hip [106–108].
Two large preventive trials using etidronate p.o., dosed at 400 mg, given cyclically every 3 months have shown that vertebral BMD improved significantly in the etidronate group compared to calcium alone and calcium and vitamin D groups [109, 110]. Analyses of pooled data from these trials subsequently reported a reduction in vertebral fractures .
Studies using alendronate 5 and 10 mg/day p.o. have shown a significant increase in lumbar spine and femoral neck BMD compared with placebo . There were fewer new vertebral fractures in the alendronate groups than in the placebo group and markers of bone turnover decreased significantly in the alendronate groups.
Two 12-month randomized controlled trials have been performed with risedronate (2.5 or 5 mg/day) versus placebo in pre- and post-menopausal women and men aged 18–85, taking an average of 12 to 15 mg of prednisone daily. Significant increases in vertebral and femoral BMD have been observed in the risedronate-treated groups . In both studies, there was a 70% decrease in the relative risk of new vertebral fractures in patients taking 5 mg/day of risedronate compared to placebo. No statistically significant effects on non-vertebral and hip fractures were found in GC users treated with risedronate.
Intravenous ibandronate was shown to decrease the rate of vertebral fractures in an open study on 115 patients with established GIOP. After 3 years, the frequency of new vertebral fractures was significantly lower in the ibandronate group compared to the alfacalcidol group (8.6 vs. 22.8%) .
Intravenous pamidronate was shown to significantly increase BMD at the lumbar spine, femoral neck, and total hip when compared to calcium alone in a randomized, placebo-controlled trial of 32 patients beginning chronic GC therapy .
Intravenous zoledronic acid (5 mg) is more efficient than oral 5 mg/day risedronate to increase lumbar spine BMD from baseline to 12 months in men and women, for both treatment and prevention. A single infusion of zoledronic acid 5 mg suppresses bone turnover (β-CTx, P1NP) significantly more than daily oral risedronate for up to 1 year [116, 117].
Recombinant Human Parathyroid Hormone
In a 12-month, randomized clinical trial in post-menopausal osteoporotic women who were taking GCs and hormone replacement therapy (HRT), a marked improvement of BMD at the vertebral and femoral sites was demonstrated after daily injections of human parathyroid hormone 1–34 (teriparatide) given subcutaneously for 12 months. The BMD gain was maintained during the follow-up period of 12 months off treatment. However, no difference in the fracture rate was found between both groups. In a 18-month randomized, double-blind, controlled trial  comparing teriparatide (20 μg s.c. once daily) with alendronate (10 mg p.o. once daily) in GCs users, the mean BMD increase at the lumbar spine was significantly larger in the teriparatide group than in the alendronate group. New vertebral fractures were less frequent in the teriparatide group than in the alendronate group and the incidence of non-vertebral fractures was similar in the two groups.
Hormone Replacement Therapy
The substitution of gonadal sex hormones is mentioned as an option for patients using GCs but there are limited data to support the use of HRT to treat osteoporosis in patients receiving prolonged therapy with GCs. In a subgroup analysis of 42 women from a study of 200 post-menopausal women with rheumatoid arthritis who were receiving low-dose GCs, HRT recipients had a significant increase in lumbar spine BMD relative to controls although femoral neck BMD was not significantly changed in either group . Twelve months of treatment with testosterone injections significantly increased mean lumbar spine BMD by 5% relative to controls in 15 men with asthma receiving GCs .
Several studies have concerned the use of subcutaneous and nasal spray calcitonin in GIOP. None showed a significant decrease in fracture risk after 12–24 months of follow-up. A meta-analysis showed that calcitonin prevented bone loss at the spine and forearm after the first year of therapy, but had no effect on bone loss at the hip. Calcitonin was not statistically different from placebo at preventing fractures of the spine or hip fractures .
Vitamin K2 (menatetrenone) is approved in Japan for both prevention and treatment of osteoporosis. Studies have shown that vitamin K2 reduces vertebral and hip fractures and improves bone quality .
Specific studies in GIOP have not been done yet but effects of denosumab on structural damage, BMD, and bone turnover in rheumatoid arthritis are promising .
Studies with strontium ranelate on GIOP are underway but results have not been presented yet. Data using the selective estrogen receptor modulator raloxifene to treat or prevent GIOP are lacking.
GCs have outstanding therapeutic effects and are some of the most important drugs in clinical use today. Unfortunately, GC therapy is sometimes accompanied by severe and/or irreversible side effects like osteoporosis. The different molecular mechanisms of GC actions are better understood and support a certain number of starting points for the development of optimized and/or new GCs and GC receptor ligands such as selective GC receptor agonists that could improve clinical medicine by demonstrating a better benefit–risk ratio compared with conventional GCs [3, 124, 125]. Abnormalities in bone quality associated with GCs can occur even before deleterious effects on BMD are observed. For this reason, GIOP guidelines differ in the use of BMD criteria to recommend treatment for GC users (the intervention threshold of the Royal College of Physicians is a T score of −1.5  and those of the American College of Rheumatology is a T score of –1 ). Nevertheless, every patient receiving GC therapy should have a bone assessment in order to receive calcium, vitamin D, bisphosphonates, or anabolic agents if necessary.
This study was made possible by grants from the Bioregos “Pays de la Loire” program and INSERM and was also supported by a grant from the French Society for Rheumatology (SFR).