Are OPG and RANKL Involved in Human Fracture Healing? Julia Ko¨ttstorfer,1 Anita Thomas,2 Markus Gregori,1 Mathias Kecht,1 Georg Kaiser,1 Stefan Eipeldauer,1 Kambiz Sarahrudi1 1 Medical University Vienna, University Clinic for Trauma Surgery, Wa¨hringer Gu¨rtel 18-20 1090, Vienna, Austria, 2Department of Internal Medicine III, Division of Endocrinology and Metabolism, Medical University Vienna, Gender Medicine Unit, Wa¨hringer Gu¨rtel 18-20 1090, Vienna, Austria
Received 5 April 2014; accepted 24 July 2014 Published online 12 September 2014 in Wiley Online Library (wileyonlinelibrary.com). DOI 10.1002/jor.22723
ABSTRACT: Human fracture healing is a complex interaction of several cytokines that regulate osteoblast and osteoclast activity. By monitoring OPG (osteoprotegerin) and sRANKL we aimed to possibly predict normal or impaired fracture healing. In 64 patients with a fracture of a long bone serum level of sRANKL and OPG were evaluated with respect to bony union (n ¼ 57) or pseudarthrosis (n ¼ 7). Measurements were carried out at admission and at 1, 2, 4, 6, 8, 12, 24, and 48 weeks after the injury. Patients’ serum levels were compared to 33 healthy controls. Fracture hematoma contained significantly higher sRANKL and OPG concentrations compared to patients serum (p ¼ 0.005, p ¼ 0.028). OPG level in fracture hematoma was higher compared to the unions serum level (p ¼ 0.028). sRANKL was decreased in unions during the observation period. In non-unions sRANKL and OPG levels showed a variable course, with no statistical significance. This is the first study to document the course of OPG and sRANKL in normal and delayed human fracture healing emphasizing its local and systemic involvement. We provide evidence of strongly enhanced OPG levels in patients with a long bone fracture compared to healthy controls. Further, levels of free sRANKL were decreased during regular fracture repair. ß 2014 Orthopaedic Research Society. Published by Wiley Periodicals, Inc. J Orthop Res 32:1557–1561, 2014. Keywords: fracture healing; RANKL; OPG
BACKGROUND Fracture healing in humans is a cascade of intramembranous bone formation and enchondral ossification. This physiologic process is regulated by complex interactions of hormones, growth factors, and extracellular matrix proteins.1 Within the last decade the bone remodeling process has been intensively investigated. Numerous cytokines, angiogenic factors, proteases, and morphogens with significant roles in fracture healing were identified.2,3 Among these, cytokines that interact with the TNFa (tumor necrosis factor a) receptor family seem to be crucial in osteoclastogenesis. The cellular interactions of osteoblasts and osteoclasts are critical and mainly regulated upon the RANKL/OPG/RANK complex.4,5 RANKL (receptor activator of N-kB ligand) is a potent physiological inducer of osteoclast activity. RANKL, by interacting with the RANK (receptor activator of N-kB), stimulates the preosteoclast differentiation and increases osteoclast activity. Membrane RANKL is expressed by many cell types especially osteoblasts. The cell-bound form of RANKL is the most common and has been suggested to be somewhat more potent than soluble RANKL in stimulating osteoclastogenesis in vitro. But the soluble form evolves by splitting a truncated ectodomain by a TNF-converting enzyme-like protease and is measurable in circulation.5,6 OPG (osteoprotegerin), a glycoprotein mainly synthesized by osteoblasts is the corresponding decoy receptor. OPG, by binding to RANKL is a competitive inhibitor of RANKL action and causes its anti-resorp-
Conflict of interest: None. Grant sponsor: Medical Scientific Fund of the Mayor of the City of Vienna; Grant number: 12069. Correspondence to: Kambiz Sarahrudi (T: þ43 1 40400 5901; F: þ43 1 40400 5949; E-mail: [email protected]) # 2014 Orthopaedic Research Society. Published by Wiley Periodicals, Inc.
tive effect. Thereby, OPG suppresses bone resorption and increases bone mass.7 Thus, the OPG/RANKL ratio reflects the rate of bone resorption in general and its role is well recognized in osteoporosis.8 Interestingly, it has been shown, that an increased RANKL/OPG ratio in iliac bone samples is accompanied by increased fracture susceptibility.8 Comparable findings were reported by Mizuno et al. who identified a relative OPG deficiency causing bone loss in a murine model.9 On the other hand, OPG administration has been shown to be effective in preventing bone resorbtion.10 Quantification of serum RANKL and OPG concentration during the course of human fracture healing after long bone fracture is still lacking in current literature. We therefore observed the expression of these cytokines in patients with physiologic and delayed fracture healing.
METHODS This study was approved by the Ethic Committee of the Medical University of Vienna and conducted in accordance with the declaration of Helsinki. Patients gave informed written consent to be enrolled in the study. The recruitment parameters, sample collection schedule, matching process, patient demographics as well as exclusion criteria of this study have been previously published in detail.11 In brief, between April 2006 and 2008 a consecutive series of 113 patients with meta-/diaphyseal fractures of long bone (humerus, femur, lower leg, and forearm) and surgical treatment were included. In order to have a homogenous study group and due to the strict selection criteria 49 patients with incomplete data were excluded from further investigation. Finally the data of 64 patients were analyzed (33 males and 31 females). Mean age of the patients was 59.3 years (range, 17–90 years). Patients’ serum was collected following a standardized time schedule. The follow up examination was based on clinical and radiological examination at 1, 2, 4, 6, 8, 12, 24, and 48 weeks after trauma. All patients were followed up for at least six months after the operation. The diagnosis of bony consolidation or delayed union was based on exercise-induced JOURNAL OF ORTHOPAEDIC RESEARCH DECEMBER 2014
1557
1558
¨ TTSTORFER ET AL. KO
Table 1. Patients Demographics
Number Mean age (range) Gender (men:women)
Unions
Non-Unions
Healthy Volunteers
57 61.3 yrs (17–90) 31:26
7 51.4 yrs (17–73) 2:5
33 37.1 yrs (25–49) 16:17
Table 2. Operative Procedure and Incidence of Non-unions Nailing Humerus Forearm Femur Lower Leg Total
Non-union After Nailing
Plating
Non-union After Plating 2
2 2
11 4 2 4 21
3 3 11 19 36
pain, and X-rays or computed tomography. Delayed union was defined as failed fracture healing without radiological signs of bony consolidation after 4 months postoperatively. Non-union was defined as the absence of complete consolidation at 6 months after surgery. Mean serum OPG and sRANKL levels were evaluated dependant on whether bony union was achieved (n ¼ 57) or not (n ¼ 7) (Table 2). Demographic data of the patients are presented in Table 1. A total of 33 healthy volunteers (16 males, 17 females, mean age: 37.1 years) formed the control group and donated one blood sample each. Blood Samples Fracture hematoma was obtained at surgery within 24 hours after the accident. Peripheral venous blood was obtained from each patient at 1, 2, 4, 6, 8, 12, 24, and 48 weeks after trauma. The specimens were centrifuged immediately and the resulting supernatant was stored at 80 ˚C until assayed. Measurement of OPG and sRANKL OPG and sRANKL concentrations were quantified using commercial enzyme-linked immunosorbent assay (ELISA) kits. (Biomedica Medizinprodukte, Vienna, Austria) Details on design and performance of the essay for sRANKL have been previously published, as the assays detects specifically the biologic active form of the protein.12 All analytical steps were performed according to the manufacturers recommended protocol. Concentrations are presented as mean of duplicate measurements. Statistical Analysis Comparisons between independent groups of continuous variables were performed by nonparametric Mann–Whitney U-test. For comparison of serum with fracture hematoma of union patients and sRANKL and OPG concentrations between different timepoints non-parametric Wilcoxon-test for paired samples was used. Correlation between sRANKL and OPG was analyzed by Spearman’s correlation coefficient. Statistical analyses were performed using IBM SPSS for Windows 19. Data are presented as box plot displaying the outliners, the extremes, the upper and lower quartile and the median. The statistical significance level was set at p 0.05. JOURNAL OF ORTHOPAEDIC RESEARCH DECEMBER 2014
2 1 5
RESULTS OPG Levels Significantly Elevated in Fracture Hematoma Compared to OPG Serum Concentration The median OPG concentration in fracture hematoma was significantly higher compared to union patients serum immediately after fracture (49.33 pmol/l (range, 3.28–265.24) vs. 5.09 pmol/l (range, 1.68–110.15) (p ¼ 0.028)). The systemic OPG concentration in union patients immediately after fracture was more than the systemic OPG concentration measured in healthy controls (5.09 pmol/l (range, 1.68–110.15) vs. 3.94 pmol/ l (range, 2.36–7.8)). This difference was not significant (p ¼ 0.148). Figure 1 demonstrates the local and serum OPG concentrations immediately after fracture. Serum OPG Concentration in Patients with Physiological Fracture Healing Shows Constant Development Median serum OPG concentration was 5.09 pmol/l (range, 1.68–110.15) immediately after fracture. After
Figure 1. Box plot displaying the outliners, the extremes, the upper and lower quartile and the median of the OPG concentration on a logarithmic scale in control serum (n ¼ 35), in serum of union patients and in fracture hematoma of union patients (n ¼ 7).
RANKL/OPG IN FRACTURE HEALING
1559
Figure 3. Box plot displaying RANKL concentrations in control serum, in serum of union patients and in fracture hematoma of union patients (n ¼ 11). Figure 2. Box plot displaying the OPG concentrations in controls and in patients. No significant differences.
this initial OPG peak, serum concentration fell within the first week and OPG level measured 4.3 pmol/l one week after the injury (range, 1.89–12.67). The OPG level in the serum of union patients fairly remained at this level over the next weeks. During the whole observation period up to week 48 serum levels of union patients approximated the level of healthy controls (Fig. 2). Systemic OPG Concentrations in Patients with Impaired Fracture Healing Follow Different Time-Course OPG concentrations in patients with impaired healing evolved differently compared to patients with normal fracture healing. OPG levels measured a median concentration of 6.2 pmol/l at week 1 and further decreased to the lowest concentration at week 6 (3.6 pmol/l). This was followed by another increase between week 8 and week 48, with an OPG level rising up to 6.1 pmol/l in week 48. The direct comparison between the OPG level of patients with impaired fracture healing and the control serum was performed at weeks 1, 2, 4, 12, and 24 after the operation. At theses time points no statistically significant differences were observed between the two groups.
0–0.35)) the sRANKL levels steadily increased and two peaks at week 6 and 12 were observed. The peak at week 12 (0.03 pmol/l) was succeeded by a fall to week 24 (0.004 pmol/l). Serum sRANKL Concentration in Patients with Non-union are Found to Progress Differently Compared to Regular Fracture Healing A rather irregular course of sRANKL concentrations was found in non-union patients. An initial peak was observed in week 1 (0.015 pmol/l), after which the sRANKL level decreased upon week 2 (0.011 pmol/l). sRANKL concentration increased again during week 4, 6, and 8 (0.135 pmol/l). Thereafter, sRANKL level decreased to low levels ranging between zero and 0.253 pmol/l at week 12 and increased again until week 24 (0.021 pmol/l) (Fig. 4). A comparison between the sRANKL level of patients with impaired fracture healing and the serum of healthy controls was only possible for weeks 1, 2, and 4 after fracture stabilization. At theses time points no statistically significant differences were found between the groups.
Significantly Elevated RANKL Levels in Fracture Hematoma Compared to sRANKL Serum Concentration The local RANKL concentration measured in fracture hematoma was similar to the systemic concentration measured at the same time in patients as well as in control serum. The median RANKL concentration in fracture hematoma was 0.18 pmol/l compared to 0.03 pmol/l in patients serum (p ¼ 0.005). Figure 3 shows the local and soluble serum RANKL concentration immediately after the fracture. Low Serum sRANKL Concentration Directly After Trauma Might Reflect sRANKL Recruitment to the Osteoclasts After an initial nadir of the free sRANKL concentration at the day of the trauma (0.04 pmol/l (range,
Figure 4. Box plot displaying sRANKL concentrations in controls and in patients. Asterisks indicate significant differences in sRANKL concentrations between patients and controls. JOURNAL OF ORTHOPAEDIC RESEARCH DECEMBER 2014
1560
¨ TTSTORFER ET AL. KO
Figure 5. Box plot showing the RANKL/OPG ratio of patients and controls.
The RANKL/OPG Ratio Develops Rather Different Comparing Union and Non-union Patients The RANKL/OPG ratio in patients with regular fracture healing showed an initial nadir one week after the fracture compared to the RANKL/OPG ratio of healthy controls (0.007 pmol/l). This decrease was found to be significant (p ¼ 0.02). Over the next weeks the RANKL/ OPG ratio of union patients approached the level of healthy controls. The RANKL/OPG course in nonunion patients showed a rather inverse development when compared to union patients with decreased levels over the whole observation period (Fig. 5).
DISCUSSION Bone resorbtion and bone formation are coupled and mediated by an osteoblast osteoclast crosstalk. This interaction is characterized by a mutual dependence of RANKL and OPG.4,5 The critical role of these proinflammatory cytokines is well investigated in murine fracture healing.13 Moreover, OPG and RANKL were tracked in 36 elderly patients after an intertrochanteric fracture over a period of 12 weeks by Wang et al.14 Their study population consisted of 36 elderly patients who sustained an intertrochanteric fracture and 30 healthy controls. The levels of RANKL and OPG in the fracture group were significantly higher compared to controls immediately and 4 weeks after the injury. By monitoring OPG and RANKL for 48 weeks after sustaining a fracture, we aimed to identify individuals who might be at high risk for pseudarthrosis. The results presented here reveal notable discrepancies between regular union and non-union patients. In our collective OPG concentrations were significantly elevated in both, fracture hematoma and peripheral serum compared to the OPG serum level of healthy controls. The distinct OPG peak at the very first time point in the union-group goes with the phase of maximal cartilage formation during the early course of callus consolidation. OPG levels at all further time points were found to be above baseline but rather JOURNAL OF ORTHOPAEDIC RESEARCH DECEMBER 2014
declining as enchondral resorbtion progressed. A similar temporal relation is reported from a fracture model by Kon et al. who measured mRNA expression in the fracture callus of mice.13 Thereby, two OPG peaks were observed: the first 24 h after the fracture and the second at day 7, which reflects the phase of utmost cartilage formation. Our data reveal elevated OPG levels during the whole observation period when compared to healthy controls. This is in contrast to findings by Abdallah et al. who measured OPG levels in the iliac crest of postmenopausal women.8 These patients were either operated for an acute hip fracture or underwent elective arthroplasty. Interestingly, biopsy specimens from patients with hip fractures had a significantly lower OPG mRNA content.8 This might express the skeletal fragility that ultimately causes osteoporotic fractures. These results nicely point out the molecular differences between osteoporotic and non-osteoporotic individuals. In our collective OPG levels in non-unions showed a rather inconstant course with a final peak at week 48. This might reflect the organisms’ continuous effort to induce osteogenesis at the fracture site. These data suggest that OPG could be an important paracrine mediator of bone metabolism during fracture healing. Interestingly, the concentration of sRANKL one week after the injury measured an all time low in union patients in our collective. (Fig. 3) In patients with regular fracture healing a continuous increase until week 12 was observed. After the peak in week 12, sRANKL levels decreased again, which might reveal an increased resorbtion effort within the fracture callus. Initially very low sRANKL levels might reflect the recruitment of sRANKL as major activator binding to osteoclasts during the early course of fracture healing. This finding is supported by results published by Kon et al. who detected very low mRNA levels of RANKL in unfractured murine bone but observed a very fulminate peak immediately after fracture.13 Interestingly, patients in our collective who did not achieve union after 6 months showed a rather variable course of sRANKL compared to regular unions. Unfortunately, the relatively small number of non-unions limits the power of these results. RANKL is meant to be a central and potent inducer of osteoclastogenesis.15 OPG is the corresponding decoy receptor, acting as a competitive inhibitor of RANKL action. Despite the direct dependency of these cytokines, serum levels did not evolve concomitantly during the course of physiologic fracture healing. This might be due to the fact that a variety of cytokines are tightly coupled during bone remodeling.11,16 The RANKL/OPG ratio is suggested to provide additional information regarding the fracture healing progress.14 The balance of these key regulators reflects the osteoclast activation and might be a possible predictor of impaired fracture healing. Our results show interesting differences between union and non-unions regarding the development of the RANKL/OPG balance over 48 weeks.
RANKL/OPG IN FRACTURE HEALING
Nevertheless, the significance of our data is limited by the relatively small number of patients. Moreover, the differences between RANKL and OPG might be due to their various functions besides their role in bone homeostasis and fracture healing. This study provides for the first time evidence of the presence of sRANKL and OPG in fracture hematoma during the phase of immediate response as well as in peripheral serum during the entire period of human fracture healing. We provide evidence of strongly enhanced OPG levels in patients with fracture of a long bone compared to healthy controls. Further, levels of free sRANKL were decreased during regular fracture healing. The results indicate local and systemic involvement of sRANKL and OPG in human fracture healing.
ACKNOWLEDGMENTS This study was supported by the Medical Scientific Fund of the Mayor of the City of Vienna (grant 12069 issued to Julia Ko¨ttstorfer).
REFERENCES 1. Einhorn TA. 1995. Enhancement of fracture-healing. J Bone Joint Surg Am 77:940–956. 2. Einhorn TA. 1998. The cell and molecular biology of fracture healing. Clin Orthop Relat Res 355:S7–21. 3. Gerstenfeld LC, Cullinane DM, Barnes GL, et al. 2003. Fracture healing as a post-natal developmental process: molecular, spatial, and temporal aspects of its regulation. J Cell Biochem 88:873–884. 4. Wong BR, Josien R, Lee SY, et al. 1997. TRANCE (tumor necrosis factor [TNF]-related activation-induced cytokine), a new TNF family member predominantly expressed in T cells, is a dendritic cell-specific survival factor. J Exp Med 186:2075–2080.
1561
5. Lacey DL, Timms E, Tan HL, et al. 1998. Osteoprotegerin ligand is a cytokine that regulates osteoclast differentiation and activation. Cell 93:165–176. 6. Yasuda H, Shima N, Nakagawa N, et al. 1998. Osteoclast differentiation factor is a ligand for osteoprotegerin/osteoclastogenesis-inhibitory factor and is identical to TRANCE/ RANKL. Proc Natl Acad Sci USA 95:3597–3602. 7. Schneeweis LA, Willard D, Milla ME. 2005. Functional dissection of osteoprotegerin and its interaction with receptor activator of NF-kappaB ligand. J Biol Chem 280:41155– 41164. 8. Abdallah BM, Stilgren LS, Nissen N, et al. 2005. Increased RANKL/OPG mRNA ratio in iliac bone biopsies from women with hip fractures. Calcif Tissue Int 76:90–97. 9. Mizuno A, Amizuka N, Irie K, et al. 1998. Severe osteoporosis in mice lacking osteoclastogenesis inhibitory factor/osteoprotegerin. Biochem Biophys Res Commun 247:610–615. 10. Bekker PJ, Holloway D, Nakanishi A, et al. 2001. The effect of a single dose of osteoprotegerin in postmenopausal women. J Bone Miner Res 16:348–360. 11. Sarahrudi K, Thomas A, Mousavi M, et al. 2011. Elevated transforming growth factor-beta 1 (TGF-beta1) levels in human fracture healing. Injury 42:833–837. 12. Hawa G, Brinskelle-Schmal N, Glatz K, et al. 2003. Immunoassay for soluble RANKL (receptor activator of NFkappaB ligand) in serum. Clin Lab 49:461–463. 13. Kon T, Cho TJ, Aizawa T, et al. 2001. Expression of osteoprotegerin, receptor activator of NF-kappaB ligand (osteoprotegerin ligand) and related proinflammatory cytokines during fracture healing. J Bone Miner Res 16:1004–1014. 14. Wang XF, Zhang YK, Yu ZS, et al. 2013. The role of the serum RANKL/OPG ratio in the healing of intertrochanteric fractures in elderly patients. Mol Med Rep 7:1169–1172. 15. Fuller K, Wong B, Fox S, et al. 1998. TRANCE is necessary and sufficient for osteoblast-mediated activation of bone resorption in osteoclasts. J Exp Med 188:997–1001. 16. Sarahrudi K, Mousavi M, Thomas A, et al. 2010. Elevated levels of macrophage colony-stimulating factor in human fracture healing. J Orthop Res 28:671–676.
JOURNAL OF ORTHOPAEDIC RESEARCH DECEMBER 2014
Received 5 April 2014; accepted 24 July 2014 Published online 12 September 2014 in Wiley Online Library (wileyonlinelibrary.com). DOI 10.1002/jor.22723
ABSTRACT: Human fracture healing is a complex interaction of several cytokines that regulate osteoblast and osteoclast activity. By monitoring OPG (osteoprotegerin) and sRANKL we aimed to possibly predict normal or impaired fracture healing. In 64 patients with a fracture of a long bone serum level of sRANKL and OPG were evaluated with respect to bony union (n ¼ 57) or pseudarthrosis (n ¼ 7). Measurements were carried out at admission and at 1, 2, 4, 6, 8, 12, 24, and 48 weeks after the injury. Patients’ serum levels were compared to 33 healthy controls. Fracture hematoma contained significantly higher sRANKL and OPG concentrations compared to patients serum (p ¼ 0.005, p ¼ 0.028). OPG level in fracture hematoma was higher compared to the unions serum level (p ¼ 0.028). sRANKL was decreased in unions during the observation period. In non-unions sRANKL and OPG levels showed a variable course, with no statistical significance. This is the first study to document the course of OPG and sRANKL in normal and delayed human fracture healing emphasizing its local and systemic involvement. We provide evidence of strongly enhanced OPG levels in patients with a long bone fracture compared to healthy controls. Further, levels of free sRANKL were decreased during regular fracture repair. ß 2014 Orthopaedic Research Society. Published by Wiley Periodicals, Inc. J Orthop Res 32:1557–1561, 2014. Keywords: fracture healing; RANKL; OPG
BACKGROUND Fracture healing in humans is a cascade of intramembranous bone formation and enchondral ossification. This physiologic process is regulated by complex interactions of hormones, growth factors, and extracellular matrix proteins.1 Within the last decade the bone remodeling process has been intensively investigated. Numerous cytokines, angiogenic factors, proteases, and morphogens with significant roles in fracture healing were identified.2,3 Among these, cytokines that interact with the TNFa (tumor necrosis factor a) receptor family seem to be crucial in osteoclastogenesis. The cellular interactions of osteoblasts and osteoclasts are critical and mainly regulated upon the RANKL/OPG/RANK complex.4,5 RANKL (receptor activator of N-kB ligand) is a potent physiological inducer of osteoclast activity. RANKL, by interacting with the RANK (receptor activator of N-kB), stimulates the preosteoclast differentiation and increases osteoclast activity. Membrane RANKL is expressed by many cell types especially osteoblasts. The cell-bound form of RANKL is the most common and has been suggested to be somewhat more potent than soluble RANKL in stimulating osteoclastogenesis in vitro. But the soluble form evolves by splitting a truncated ectodomain by a TNF-converting enzyme-like protease and is measurable in circulation.5,6 OPG (osteoprotegerin), a glycoprotein mainly synthesized by osteoblasts is the corresponding decoy receptor. OPG, by binding to RANKL is a competitive inhibitor of RANKL action and causes its anti-resorp-
Conflict of interest: None. Grant sponsor: Medical Scientific Fund of the Mayor of the City of Vienna; Grant number: 12069. Correspondence to: Kambiz Sarahrudi (T: þ43 1 40400 5901; F: þ43 1 40400 5949; E-mail: [email protected]) # 2014 Orthopaedic Research Society. Published by Wiley Periodicals, Inc.
tive effect. Thereby, OPG suppresses bone resorption and increases bone mass.7 Thus, the OPG/RANKL ratio reflects the rate of bone resorption in general and its role is well recognized in osteoporosis.8 Interestingly, it has been shown, that an increased RANKL/OPG ratio in iliac bone samples is accompanied by increased fracture susceptibility.8 Comparable findings were reported by Mizuno et al. who identified a relative OPG deficiency causing bone loss in a murine model.9 On the other hand, OPG administration has been shown to be effective in preventing bone resorbtion.10 Quantification of serum RANKL and OPG concentration during the course of human fracture healing after long bone fracture is still lacking in current literature. We therefore observed the expression of these cytokines in patients with physiologic and delayed fracture healing.
METHODS This study was approved by the Ethic Committee of the Medical University of Vienna and conducted in accordance with the declaration of Helsinki. Patients gave informed written consent to be enrolled in the study. The recruitment parameters, sample collection schedule, matching process, patient demographics as well as exclusion criteria of this study have been previously published in detail.11 In brief, between April 2006 and 2008 a consecutive series of 113 patients with meta-/diaphyseal fractures of long bone (humerus, femur, lower leg, and forearm) and surgical treatment were included. In order to have a homogenous study group and due to the strict selection criteria 49 patients with incomplete data were excluded from further investigation. Finally the data of 64 patients were analyzed (33 males and 31 females). Mean age of the patients was 59.3 years (range, 17–90 years). Patients’ serum was collected following a standardized time schedule. The follow up examination was based on clinical and radiological examination at 1, 2, 4, 6, 8, 12, 24, and 48 weeks after trauma. All patients were followed up for at least six months after the operation. The diagnosis of bony consolidation or delayed union was based on exercise-induced JOURNAL OF ORTHOPAEDIC RESEARCH DECEMBER 2014
1557
1558
¨ TTSTORFER ET AL. KO
Table 1. Patients Demographics
Number Mean age (range) Gender (men:women)
Unions
Non-Unions
Healthy Volunteers
57 61.3 yrs (17–90) 31:26
7 51.4 yrs (17–73) 2:5
33 37.1 yrs (25–49) 16:17
Table 2. Operative Procedure and Incidence of Non-unions Nailing Humerus Forearm Femur Lower Leg Total
Non-union After Nailing
Plating
Non-union After Plating 2
2 2
11 4 2 4 21
3 3 11 19 36
pain, and X-rays or computed tomography. Delayed union was defined as failed fracture healing without radiological signs of bony consolidation after 4 months postoperatively. Non-union was defined as the absence of complete consolidation at 6 months after surgery. Mean serum OPG and sRANKL levels were evaluated dependant on whether bony union was achieved (n ¼ 57) or not (n ¼ 7) (Table 2). Demographic data of the patients are presented in Table 1. A total of 33 healthy volunteers (16 males, 17 females, mean age: 37.1 years) formed the control group and donated one blood sample each. Blood Samples Fracture hematoma was obtained at surgery within 24 hours after the accident. Peripheral venous blood was obtained from each patient at 1, 2, 4, 6, 8, 12, 24, and 48 weeks after trauma. The specimens were centrifuged immediately and the resulting supernatant was stored at 80 ˚C until assayed. Measurement of OPG and sRANKL OPG and sRANKL concentrations were quantified using commercial enzyme-linked immunosorbent assay (ELISA) kits. (Biomedica Medizinprodukte, Vienna, Austria) Details on design and performance of the essay for sRANKL have been previously published, as the assays detects specifically the biologic active form of the protein.12 All analytical steps were performed according to the manufacturers recommended protocol. Concentrations are presented as mean of duplicate measurements. Statistical Analysis Comparisons between independent groups of continuous variables were performed by nonparametric Mann–Whitney U-test. For comparison of serum with fracture hematoma of union patients and sRANKL and OPG concentrations between different timepoints non-parametric Wilcoxon-test for paired samples was used. Correlation between sRANKL and OPG was analyzed by Spearman’s correlation coefficient. Statistical analyses were performed using IBM SPSS for Windows 19. Data are presented as box plot displaying the outliners, the extremes, the upper and lower quartile and the median. The statistical significance level was set at p 0.05. JOURNAL OF ORTHOPAEDIC RESEARCH DECEMBER 2014
2 1 5
RESULTS OPG Levels Significantly Elevated in Fracture Hematoma Compared to OPG Serum Concentration The median OPG concentration in fracture hematoma was significantly higher compared to union patients serum immediately after fracture (49.33 pmol/l (range, 3.28–265.24) vs. 5.09 pmol/l (range, 1.68–110.15) (p ¼ 0.028)). The systemic OPG concentration in union patients immediately after fracture was more than the systemic OPG concentration measured in healthy controls (5.09 pmol/l (range, 1.68–110.15) vs. 3.94 pmol/ l (range, 2.36–7.8)). This difference was not significant (p ¼ 0.148). Figure 1 demonstrates the local and serum OPG concentrations immediately after fracture. Serum OPG Concentration in Patients with Physiological Fracture Healing Shows Constant Development Median serum OPG concentration was 5.09 pmol/l (range, 1.68–110.15) immediately after fracture. After
Figure 1. Box plot displaying the outliners, the extremes, the upper and lower quartile and the median of the OPG concentration on a logarithmic scale in control serum (n ¼ 35), in serum of union patients and in fracture hematoma of union patients (n ¼ 7).
RANKL/OPG IN FRACTURE HEALING
1559
Figure 3. Box plot displaying RANKL concentrations in control serum, in serum of union patients and in fracture hematoma of union patients (n ¼ 11). Figure 2. Box plot displaying the OPG concentrations in controls and in patients. No significant differences.
this initial OPG peak, serum concentration fell within the first week and OPG level measured 4.3 pmol/l one week after the injury (range, 1.89–12.67). The OPG level in the serum of union patients fairly remained at this level over the next weeks. During the whole observation period up to week 48 serum levels of union patients approximated the level of healthy controls (Fig. 2). Systemic OPG Concentrations in Patients with Impaired Fracture Healing Follow Different Time-Course OPG concentrations in patients with impaired healing evolved differently compared to patients with normal fracture healing. OPG levels measured a median concentration of 6.2 pmol/l at week 1 and further decreased to the lowest concentration at week 6 (3.6 pmol/l). This was followed by another increase between week 8 and week 48, with an OPG level rising up to 6.1 pmol/l in week 48. The direct comparison between the OPG level of patients with impaired fracture healing and the control serum was performed at weeks 1, 2, 4, 12, and 24 after the operation. At theses time points no statistically significant differences were observed between the two groups.
0–0.35)) the sRANKL levels steadily increased and two peaks at week 6 and 12 were observed. The peak at week 12 (0.03 pmol/l) was succeeded by a fall to week 24 (0.004 pmol/l). Serum sRANKL Concentration in Patients with Non-union are Found to Progress Differently Compared to Regular Fracture Healing A rather irregular course of sRANKL concentrations was found in non-union patients. An initial peak was observed in week 1 (0.015 pmol/l), after which the sRANKL level decreased upon week 2 (0.011 pmol/l). sRANKL concentration increased again during week 4, 6, and 8 (0.135 pmol/l). Thereafter, sRANKL level decreased to low levels ranging between zero and 0.253 pmol/l at week 12 and increased again until week 24 (0.021 pmol/l) (Fig. 4). A comparison between the sRANKL level of patients with impaired fracture healing and the serum of healthy controls was only possible for weeks 1, 2, and 4 after fracture stabilization. At theses time points no statistically significant differences were found between the groups.
Significantly Elevated RANKL Levels in Fracture Hematoma Compared to sRANKL Serum Concentration The local RANKL concentration measured in fracture hematoma was similar to the systemic concentration measured at the same time in patients as well as in control serum. The median RANKL concentration in fracture hematoma was 0.18 pmol/l compared to 0.03 pmol/l in patients serum (p ¼ 0.005). Figure 3 shows the local and soluble serum RANKL concentration immediately after the fracture. Low Serum sRANKL Concentration Directly After Trauma Might Reflect sRANKL Recruitment to the Osteoclasts After an initial nadir of the free sRANKL concentration at the day of the trauma (0.04 pmol/l (range,
Figure 4. Box plot displaying sRANKL concentrations in controls and in patients. Asterisks indicate significant differences in sRANKL concentrations between patients and controls. JOURNAL OF ORTHOPAEDIC RESEARCH DECEMBER 2014
1560
¨ TTSTORFER ET AL. KO
Figure 5. Box plot showing the RANKL/OPG ratio of patients and controls.
The RANKL/OPG Ratio Develops Rather Different Comparing Union and Non-union Patients The RANKL/OPG ratio in patients with regular fracture healing showed an initial nadir one week after the fracture compared to the RANKL/OPG ratio of healthy controls (0.007 pmol/l). This decrease was found to be significant (p ¼ 0.02). Over the next weeks the RANKL/ OPG ratio of union patients approached the level of healthy controls. The RANKL/OPG course in nonunion patients showed a rather inverse development when compared to union patients with decreased levels over the whole observation period (Fig. 5).
DISCUSSION Bone resorbtion and bone formation are coupled and mediated by an osteoblast osteoclast crosstalk. This interaction is characterized by a mutual dependence of RANKL and OPG.4,5 The critical role of these proinflammatory cytokines is well investigated in murine fracture healing.13 Moreover, OPG and RANKL were tracked in 36 elderly patients after an intertrochanteric fracture over a period of 12 weeks by Wang et al.14 Their study population consisted of 36 elderly patients who sustained an intertrochanteric fracture and 30 healthy controls. The levels of RANKL and OPG in the fracture group were significantly higher compared to controls immediately and 4 weeks after the injury. By monitoring OPG and RANKL for 48 weeks after sustaining a fracture, we aimed to identify individuals who might be at high risk for pseudarthrosis. The results presented here reveal notable discrepancies between regular union and non-union patients. In our collective OPG concentrations were significantly elevated in both, fracture hematoma and peripheral serum compared to the OPG serum level of healthy controls. The distinct OPG peak at the very first time point in the union-group goes with the phase of maximal cartilage formation during the early course of callus consolidation. OPG levels at all further time points were found to be above baseline but rather JOURNAL OF ORTHOPAEDIC RESEARCH DECEMBER 2014
declining as enchondral resorbtion progressed. A similar temporal relation is reported from a fracture model by Kon et al. who measured mRNA expression in the fracture callus of mice.13 Thereby, two OPG peaks were observed: the first 24 h after the fracture and the second at day 7, which reflects the phase of utmost cartilage formation. Our data reveal elevated OPG levels during the whole observation period when compared to healthy controls. This is in contrast to findings by Abdallah et al. who measured OPG levels in the iliac crest of postmenopausal women.8 These patients were either operated for an acute hip fracture or underwent elective arthroplasty. Interestingly, biopsy specimens from patients with hip fractures had a significantly lower OPG mRNA content.8 This might express the skeletal fragility that ultimately causes osteoporotic fractures. These results nicely point out the molecular differences between osteoporotic and non-osteoporotic individuals. In our collective OPG levels in non-unions showed a rather inconstant course with a final peak at week 48. This might reflect the organisms’ continuous effort to induce osteogenesis at the fracture site. These data suggest that OPG could be an important paracrine mediator of bone metabolism during fracture healing. Interestingly, the concentration of sRANKL one week after the injury measured an all time low in union patients in our collective. (Fig. 3) In patients with regular fracture healing a continuous increase until week 12 was observed. After the peak in week 12, sRANKL levels decreased again, which might reveal an increased resorbtion effort within the fracture callus. Initially very low sRANKL levels might reflect the recruitment of sRANKL as major activator binding to osteoclasts during the early course of fracture healing. This finding is supported by results published by Kon et al. who detected very low mRNA levels of RANKL in unfractured murine bone but observed a very fulminate peak immediately after fracture.13 Interestingly, patients in our collective who did not achieve union after 6 months showed a rather variable course of sRANKL compared to regular unions. Unfortunately, the relatively small number of non-unions limits the power of these results. RANKL is meant to be a central and potent inducer of osteoclastogenesis.15 OPG is the corresponding decoy receptor, acting as a competitive inhibitor of RANKL action. Despite the direct dependency of these cytokines, serum levels did not evolve concomitantly during the course of physiologic fracture healing. This might be due to the fact that a variety of cytokines are tightly coupled during bone remodeling.11,16 The RANKL/OPG ratio is suggested to provide additional information regarding the fracture healing progress.14 The balance of these key regulators reflects the osteoclast activation and might be a possible predictor of impaired fracture healing. Our results show interesting differences between union and non-unions regarding the development of the RANKL/OPG balance over 48 weeks.
RANKL/OPG IN FRACTURE HEALING
Nevertheless, the significance of our data is limited by the relatively small number of patients. Moreover, the differences between RANKL and OPG might be due to their various functions besides their role in bone homeostasis and fracture healing. This study provides for the first time evidence of the presence of sRANKL and OPG in fracture hematoma during the phase of immediate response as well as in peripheral serum during the entire period of human fracture healing. We provide evidence of strongly enhanced OPG levels in patients with fracture of a long bone compared to healthy controls. Further, levels of free sRANKL were decreased during regular fracture healing. The results indicate local and systemic involvement of sRANKL and OPG in human fracture healing.
ACKNOWLEDGMENTS This study was supported by the Medical Scientific Fund of the Mayor of the City of Vienna (grant 12069 issued to Julia Ko¨ttstorfer).
REFERENCES 1. Einhorn TA. 1995. Enhancement of fracture-healing. J Bone Joint Surg Am 77:940–956. 2. Einhorn TA. 1998. The cell and molecular biology of fracture healing. Clin Orthop Relat Res 355:S7–21. 3. Gerstenfeld LC, Cullinane DM, Barnes GL, et al. 2003. Fracture healing as a post-natal developmental process: molecular, spatial, and temporal aspects of its regulation. J Cell Biochem 88:873–884. 4. Wong BR, Josien R, Lee SY, et al. 1997. TRANCE (tumor necrosis factor [TNF]-related activation-induced cytokine), a new TNF family member predominantly expressed in T cells, is a dendritic cell-specific survival factor. J Exp Med 186:2075–2080.
1561
5. Lacey DL, Timms E, Tan HL, et al. 1998. Osteoprotegerin ligand is a cytokine that regulates osteoclast differentiation and activation. Cell 93:165–176. 6. Yasuda H, Shima N, Nakagawa N, et al. 1998. Osteoclast differentiation factor is a ligand for osteoprotegerin/osteoclastogenesis-inhibitory factor and is identical to TRANCE/ RANKL. Proc Natl Acad Sci USA 95:3597–3602. 7. Schneeweis LA, Willard D, Milla ME. 2005. Functional dissection of osteoprotegerin and its interaction with receptor activator of NF-kappaB ligand. J Biol Chem 280:41155– 41164. 8. Abdallah BM, Stilgren LS, Nissen N, et al. 2005. Increased RANKL/OPG mRNA ratio in iliac bone biopsies from women with hip fractures. Calcif Tissue Int 76:90–97. 9. Mizuno A, Amizuka N, Irie K, et al. 1998. Severe osteoporosis in mice lacking osteoclastogenesis inhibitory factor/osteoprotegerin. Biochem Biophys Res Commun 247:610–615. 10. Bekker PJ, Holloway D, Nakanishi A, et al. 2001. The effect of a single dose of osteoprotegerin in postmenopausal women. J Bone Miner Res 16:348–360. 11. Sarahrudi K, Thomas A, Mousavi M, et al. 2011. Elevated transforming growth factor-beta 1 (TGF-beta1) levels in human fracture healing. Injury 42:833–837. 12. Hawa G, Brinskelle-Schmal N, Glatz K, et al. 2003. Immunoassay for soluble RANKL (receptor activator of NFkappaB ligand) in serum. Clin Lab 49:461–463. 13. Kon T, Cho TJ, Aizawa T, et al. 2001. Expression of osteoprotegerin, receptor activator of NF-kappaB ligand (osteoprotegerin ligand) and related proinflammatory cytokines during fracture healing. J Bone Miner Res 16:1004–1014. 14. Wang XF, Zhang YK, Yu ZS, et al. 2013. The role of the serum RANKL/OPG ratio in the healing of intertrochanteric fractures in elderly patients. Mol Med Rep 7:1169–1172. 15. Fuller K, Wong B, Fox S, et al. 1998. TRANCE is necessary and sufficient for osteoblast-mediated activation of bone resorption in osteoclasts. J Exp Med 188:997–1001. 16. Sarahrudi K, Mousavi M, Thomas A, et al. 2010. Elevated levels of macrophage colony-stimulating factor in human fracture healing. J Orthop Res 28:671–676.
JOURNAL OF ORTHOPAEDIC RESEARCH DECEMBER 2014
