Background
Venous thromboembolism is a significant potential complication following orthopaedic surgery and an important cause of morbidity and mortality in adults. In the absence of thromboprophylaxis, the rate of deep vein thrombosis following major lower extremity orthopaedic surgery is between 40-60%, while the risk of developing fatal pulmonary embolism is between 1-2% [
1,
2]. Current treatments for the prevention of venous thromboembolism include heparin and low-molecular weight heparins (LMWHs). While heparin and LMWH therapy are effective measures for the prevention of thromboembolism, a number of studies have suggested that long term administration may negatively affect bone and some have associated their use with the risk of developing osteoporosis [
3,
4]. Although heparin-induced osteoporosis is a rare adverse effect, its incidence is thought to be in the range of 2-5% [
5]. Studies have shown that prolonged unfractionated heparin treatment is associated with bone loss and an increased risk of fracture [
6,
7]. Studies have also shown that women who receive extended heparin therapy during pregnancy, have significant reduction of bone density in their lumbar spines [
8]. There is some evidence to suggest that LMWH may have a reduced incidence of osteopenia and osteoporosis when compared with unfractionated heparin therapy [
3].
Rivaroxaban or Xarelto™ (Bayer Schering Pharma AG) is an anti-thrombotic drug that was granted marketing approval by the European Commission (Enterprise & Industry/Pharmaceuticals) in 2008 for the prevention of venous thromboembolism in adult patients undergoing elective hip and knee replacement surgery [
9]. A direct Factor Xa inhibitor, this drug represents an attractive alternative to heparin and LMWH for the purposes of prophylactic anti-coagulation as it is administered orally and thus removes the need for daily injections.
Rivaroxaban therapy is recommended for 5 weeks post hip replacement surgery and 2 weeks post knee replacement surgery [
10]. In light of the fact that heparin and LMWHs have been associated with adverse effects on bone and a risk of developing osteoporosis, the aim of this study was to investigate the effects of rivaroxaban, compared with enoxaparin, on human osteoblasts
in vitro. The effect of each drug was evaluated in terms of its effect on osteoblast viability, function and gene expression.
Discussion
Long term administration of heparin and LMWH has been associated with a negative effect on bone and an increased risk of developing osteoporosis [
3,
4]. Xarelto™ (Rivaroxaban) is a new anti-thrombotic drug that was recently licensed for the prevention of venous thromboembolism in adult patients undergoing elective hip and knee replacement surgery. At present, the effects of rivaroxaban on bone and osteoblasts are unknown. In this study, we investigated the effect of rivaroxaban on human osteoblasts in terms of its effect on viability, function and gene expression. Acute and chronic changes in gene expression were measured after 1 and 7 days treatment whereas the functional markers (alkaline phosphatase) were measured following 7 days of treatment. In addition, the effects induced by rivaroxaban were compared to those induced by enoxaparin, a commonly used LMWH. The principal finding of this study was that, rivaroxaban treatment leads to a reduction in osteoblast function, as measured by alkaline phosphate activity, and that this was associated with reduced expression of the bone marker, osteocalcin, the major osteoblast factor, Runx2, and the osteogenic factor, BMP-2.
Clinical trials comparing the efficacy and side effects of enoxaparin and rivaroxaban therapy on the outcome of thrombosis following joint replacement surgery favour the use of rivaroxaban for the prevention of thrombotic events post-arthroplasty (RECORD trials) [
10]. This, coupled with the fact that rivaroxaban is available orally (eliminating the need for invasive administration), makes the use of rivaroxaban an attractive option for both orthopaedic surgeons and patients alike. Enoxaparin is generally administered for 5 to 7 days in the immediate post-operative period while the patient remains in hospital. In the case of rivaroxaban, therapy is recommended for 5 weeks post hip replacement surgery and 2 weeks post knee replacement surgery [
10]. As long term administration of heparin and LMWH has been associated with reduced bone mineral density [
8,
15] and increased fracture rates in pregnant women [
16,
17], we investigated the effect of rivaroxaban on human osteoblasts.
In this study, neither enoxaparin nor rivaroxaban treatment caused a reduction in osteoblast cell viability indicating that both drugs do not show cytotoxic effects to osteoblasts studied. Our findings in relation to enoxaparin are in accordance with those of other studies, which have shown that enoxaparin does not have a cytotoxic effect on osteoblast viability [
18]. Similarly, other low molecular weight heparins, such as fondaparinux, do not induce cytotoxic effects on osteoblast proliferation [
19]. However, while rivaroxaban and enoxaparin treatment did not reduce osteoblast proliferation, both drugs caused a reduction of an established marker of osteoblast function: alkaline phosphatase. The bone-specific isoform of alkaline phosphatase is a tetrameric glycoprotein found on the surface of osteoblast cells and is believed to have a significant function in the mineralization of bone matrix [
20]. Previous studies have shown that heparin and low-molecular-weight heparins, such as enoxaparin, exert a negative effect on alkaline phosphatase expression [
21,
22]. In the present study, both enoxaparin and rivaroxaban treatment caused a significant reduction in osteoblast alkaline phosphatase activity, with rivaroxaban causing a more negative effect. Osteocalcin is another commonly used marker of bone formation/turnover [
23]. Osteocalcin expression was also reduced by enoxaparin and rivaroxaban treatment. This decrease was found to be statistically significant following treatment with rivaroxaban for 7 days indicating that prolonged rivaroxaban therapy may have a negative impact on osteoblast function.
Our findings of reduced alkaline phosphatase activity and osteocalcin expression in the absence of reduced cell viability suggested that enoxaparin and rivaroxaban treatment may negatively affect bone through a reduction in osteoblast function rather than a decrease in osteoblast proliferation. Therefore, to determine if rivaroxaban and enoxaparin influence osteoblast function through changes in osteoblast signalling, the expression of the transcription factor, Runx2, and the pro-osteogenic growth factor, BMP-2 were investigated. The exposure of human osteoblasts to enoxaparin and rivaroxaban resulted in a significant reduction of both Runx2 and BMP-2 expression in this study. Runx2 is the main transcription factor responsible for the development and maintenance of the osteoblast phenotype [
24] and targeted disruption of this gene in mice leads to a complete lack of ossification [
25]. The expression and activation of the Runx2 transcription factor is regulated by a number of bone-derived growth factors, including BMP-2 [
26]. BMP2 has previously been shown to play a crucial role in bone formation and repair [
27], and also in the regulation of osteoclastogenesis [
28].
These findings suggest that enoxaparin and rivaroxaban may affect osteoblast function by reducing BMP-2 induced bone formation. Since the exposure of human osteoblasts to both drugs did not result in a cytotoxic affect in this study, a downregulation of BMP-2 signalling may explain the reduction in osteoblast functionality. Runx2 regulates the expression of several major extracellular matrix genes expressed by osteoblasts including alkaline phosphatase and osteocalcin [
29,
30]. Thus, a reduction in BMP-2 and Runx2 signalling, could lead to a reduction in alkaline phosphatase activity and osteocalcin expression, and ultimately lead to an impairment of osteoblast function. Studies have demonstrated that heparin can inhibit BMP-2 osteogenic activity by binding to both the BMP-2 and BMP receptor, and this effect has been associated with reduced Runx2, osteocalcin and alkaline phosphatase expression [
31]. However, this type of BMP-2 mediated repression has not been reported in relation to enoxaparin or rivaroxaban.
Bone metabolism is a continuous remodelling process involving both bone formation by osteoblasts and bone resorption by osteoclasts. Heparin has been found to negatively affect bone formation by promoting the activation of osteoclasts and decreasing bone volume in rats, by inducing bone resorption in rat osteoclasts
in vitro[
32,
33] and enhancing osteoclastic bone resorption through an inhibition of osteoprotegerin activity [
34]. While the effect of enoxaparin and rivaroxaban on osteoclast function was not investigated as part of this study, our study confirmed that both thromboprophylactic agents negatively affect osteoblast function. Such findings indicate that long term therapy with enoxaparin or rivaroxaban could potentially translate into clinical effects on bone homeostasis.
Conclusions
In conclusion, rivaroxaban and enoxaparin treatment led to a reduction in alkaline phosphatase activity and a reduction in BMP-2, osteocalcin and Runx2 mRNA expression, indicating that treatment with both drugs leads to a general negative effect on osteoblast activity. Due to the increased duration of therapy (as compared to enoxaparin), rivaroxaban could potentially be more detrimental than enoxaparin in terms of its effects on osteoblast function. While the findings of this study indicate that rivaroxaban negatively affects osteoblast function, further clinical investigations are required to elucidate whether these effects of rivaroxaban therapy translate into demonstrable effects on bone homeostasis in vivo. Given that these agents are typically promoted for use following hip and knee arthroplasty, and that many modern hip and knee implants are 'uncemented' and depend on effective bone ingrowth for fixation, these findings are of potential clinical importance.
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
GNS conceived this study, carried out all experiment work and drafted the manuscript. PMW participated in the co-ordination of the study and the drafting of the manuscript. KJM participated in the design and coordination of this study, and participated in drafting the manuscript. All authors have read and approved the final manuscript.