Multiple Roles of Tumor Necrosis Factor - Alpha in Fracture Healing Jonathan M. Karnes, Scott D. Daffner, Colleen M. Watkins PII: DOI: Reference:
S8756-3282(15)00172-6 doi: 10.1016/j.bone.2015.05.001 BON 10718
To appear in:
Bone
Received date: Revised date: Accepted date:
20 February 2015 29 April 2015 1 May 2015
Please cite this article as: Karnes Jonathan M., Daffner Scott D., Watkins Colleen M., Multiple Roles of Tumor Necrosis Factor - Alpha in Fracture Healing, Bone (2015), doi: 10.1016/j.bone.2015.05.001
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT TITLE: Multiple Roles of Tumor Necrosis Factor - Alpha in Fracture Healing
1
RI
Jonathan M. Karnes, M.D. [email protected]
PT
AUTHORS:
1
1
NU
Colleen M. Watkins, M.D. [email protected]
SC
Scott D. Daffner, M.D. [email protected]
INSTITUTION: 1
MA
Department of Orthopaedics, West Virginia University, Morgantown, WV
CORRESPONDING AUTHOR:
TE
D
Jonathan M. Karnes, MD; Department of Orthopaedics, West Virginia University, P.O. Box 9196, Morgantown, WV 26506 Tel (304) 293-1168 Fax (304) 293-7070 Email: [email protected] KEYWORDS:
AC CE P
TNF-α, bone healing, metabolic bone disease
SOURCE OF FUNDING: None
ACCEPTED MANUSCRIPT Introduction: The healing pathway of a broken bone is primarily dictated by the mechanical relationship and position of
PT
the two fracture ends [1]. The initial stability achieved within the first days dictates the trajectory of the entire healing process and the final quality of the repair tissue [2]. Absolute stability with compression across a fracture
RI
gap induces primary bone healing, which seals the two bone fragments together using a remodeling process and
SC
occurs independently of an inflammatory reaction. Inadequate immobilization between the fracture fragments results in a higher percentage of mechanically inferior fibrous repair tissue with decreased mineralization and
NU
angiogenesis in the early stages of healing [3]. Fractures without absolute stability progress through a series of coordinated and predictable steps, known as secondary bone healing. The general paradigm for secondary bone
MA
healing follows the same pattern seen in soft tissue repair including hemostasis, inflammation, mesenchymal cell migration/proliferation, angiogenesis, and remodeling [4]. Specifically, secondary fracture healing progresses
D
through the following phases: (1) acute inflammatory response with mesenchymal stem cell recruitment,
TE
proliferation, and differentiation, (2) production of a cartilaginous fracture callus, (3) ossification and revascularization of the fracture callus, and (4) remodeling of the deposited trabecular bone [4, 5]. This process
AC CE P
also contributes to osseous recovery in surgical procedures including arthrodesis, osteotomies, and bone grafting and therefore has a wide impact on the success of most orthopaedic procedures. A significant body of basic science literature offers valuable insight into the role of the targeted chemical messengers (cytokines) in secondary bone healing. Of specific interest is Tumor Necrosis Factor-alpha (TNF-α), which is recognized as a primary mediator in the inflammatory reaction that initiates the reparative process. After injury, TNF-α is elevated in a bimodal temporal distribution within the fracture site. The peri-fracture concentrations are elevated in the initial inflammatory phase, then diminish to normal physiologic levels for the majority of the reparative process, but are later elevated in the remodeling phase [6]. Deviations from this schedule may produce inferior or more robust repair tissues [7, 8]. The interaction between TNF-α and bone is not limited to reparative states, but also extends to normal physiologic homeostasis. Poor bone integrity common in elderly, osteoporotic, diabetic, or obese patients as well as in individuals abusing alcohol or those on chronic doses of corticosteroids may arise from local and systemic deviations from the physiologic TNF-α equilibrium. The development of atrophic nonunions may be driven by interruptions in the
ACCEPTED MANUSCRIPT spatial and temporal patterns of elevated TNF-α concentrations within the fracture site. This review summarizes the role of TNF-α at specific time points along the course of fracture healing, presents evidence suggesting its role
PT
in common metabolic bone diseases, and speculates on the likely physiological responses to TNF-α modulation for both. Understanding the role of TNF-α in physiologic response to injury is important as it provides the foundation
RI
for exploration of potential treatments that facilitate recovery from injury or surgery as well potential
SC
interventions in metabolic bone disease [9]. Bone Healing Progression:
NU
The stages of bone healing and the role of TNF-α at each stage are summarized in Figure 1 and Table 1. Days 1-3: Acute Inflammatory Response and Mesenchymal Cell Recruitment (Elevated TNF-α)
MA
Disruption of the osseous cortex tears periosteal and endochondral vessels, creating a peri-fracture hematoma. Cytokines, fibrin, platelets, and cellular degradation products within the collected blood induce an
D
inflammatory reaction [6, 10] and stimulate angiogenesis [11]. Removal of the collected blood between the second
TE
and fourth day following injury results in mechanically inferior bone quality four weeks later [12], suggesting that factors within the hematoma are osteogenic. Furthermore, the hematoma itself does not contain the osteogenic
AC CE P
cell precursors, but rather stimulates cells within the periosteum to initiate endochondral ossification [13]. TNF-α is one of the initial inflammatory cytokines within the hematoma and is first produced by resident macrophages and recruited inflammatory cells [6]. Its concentration peaks at 24 hours and resolves to baseline levels within 72 hours after the injury [6, 14-17]. The TNF-α receptors p55 and p75 are found on immune cells during the early inflammatory phases and are identified on mesenchymal and osteoblastic cells during later healing stages [6]. Defining a fracture, arthrodesis, or bone grafting site as a physical space between the two osseous structures implies a relative deficiency of osteoclast and osteoblast cells within the gap to facilitate sufficient bone production [18]. Therefore, reconstituting the empty interval with sufficient concentrations of osteogenic cells and supporting vascular structures is crucial to the long-term success of the healing process. Within this context, elevated TNF-α concentrations in the early stages of fracture healing are vital as they initiate the recruitment of mesenchymal cells from surrounding soft tissue, bone cortex, bone marrow, periosteum, and blood with osteogenic and chondrogenic potential [19-21]. A major, but often forgotten, mesenchymal stem cell reserve is the surrounding muscle tissue. Fractures that are exposed to skeletal muscle demonstrate significantly better
ACCEPTED MANUSCRIPT osteosynthesis than those exposed to fasciocutaneous tissue only [22]. Further evidence to suggest a robust reserve of skeletal muscle mesenchymal stem cells comes from studies that demonstrate myoblast differentiation
PT
into osteoblasts following bone morphogenetic protein-1 (BMP-1) and bone morphogenetic protein-2 (BMP-2) [22, 23].
RI
TNF-α enhances mesenchymal cell recruitment by stimulating expression of intercellular adhesion
SC
molecule-1 (ICAM-1) and vascular cell adhesion protein-1 (VCAM-1), which optimize the intrinsic migration capacity [24]. TNF-α also induces proliferation and differentiation of mesenchymal precursor cells into
NU
chondrogenic and osteogenic phenotypes [20, 25] by inducing early BMP-2 secretion in surrounding osteoblasts, which promotes osteoblastic differentiation in mesenchymal precursor cells [17, 26-28].
MA
By stimulating mesenchymal cell recruitment and mesenchymal stem cell osteochondral differentiation, TNF-α provides the foundation for bone healing as its presence (controlling for other variables) determines the
D
quantity and phenotypic quality of the cells that initiate and maintain the healing process [19, 29]. Insufficient
TE
levels of intra-fracture recruitment cytokines likely diminish the mesenchymal stem cell population within the osseous gap and predispose the injury or surgical site to blunted callus formation ultimately resulting in inferior
AC CE P
bone healing capacity [30]. Platelets within the initial hematoma increase local transforming growth factor-beta (TGF-β) concentrations within the injury site [31, 32], which has been shown to increase mechanical strength and callus formation, [33, 34] due to stimulation of proliferation and osteochondrogenic differentiation of mesenchymal stem cells [35, 36]. All the molecular pathways involved with bone healing have not been identified or described and the activity of cytokines such as TGF- β provides an explanation for the continued, but inferior, fracture healing in the study animal subjected to TNF-α inhibition [37]. Broadly suppressing the immune response with systemic glucocorticoid administration during the inflammatory phase has been shown to diminish fracture healing in a rabbit model [38], despite having a slightly stimulatory effect on osteoblast activity [39]. This finding suggests that inflammation promotes the activation of the osseous healing process. Specific inhibition of TNF-α activity with tumor necrosis factor receptor 1 (TNFR1) and TNFR2 murine knockout models results in delayed mesenchymal cell differentiation, a greater proportion of disorganized mesenchymal tissue, and decreased early osteoblast migration into the marrow space [8, 29]. Although it seems intuitive that TNF-α inhibition may exert its effect by blunting the inflammatory response,
ACCEPTED MANUSCRIPT evidence from a murine model suggests that introduction of the TNF-α inhibitor etanercept did not decrease migration of immune cells or suppress the elevation of TNF-α or interleukin-1 beta (IL-1β) at the fracture site, but
PT
that it did suppress mesenchymal cell infiltration, proliferation, and fibroblast growth factor-2 (FGF-2) expression [20]. Therefore, it seems that TNF-α inhibition exerts its negative influence on bone healing by influencing the
RI
response of mesenchymal cells, rather than inflammatory cells to TNF-α. Conversely, supplemental TNF-α on post
SC
fracture days one and two has been demonstrated to stimulate a more robust healing response with an accelerated course and higher rates of callus mineralization [40]. This finding may be related to TNF-α stimulation
NU
of chondrogenic differentiation of progenitor cells through the nuclear factor kappa B (NF-κB)/p65 pathway, which facilitates SRY (sex determining region Y)-box 9 (Sox9) expression [41] or through other previously described TNF-α
MA
mediated mechanisms. Day 3-7: Fracture Callus Production
D
The third day following osseous injury marks the transition from a focus on mesenchymal cell precursor
TE
recruitment to production of a cartilage scaffold that bridges the fracture gap between two dissociated bone pieces. Effects from the initial inflammatory response peak at twenty-four hours following injury, but continue
AC CE P
with diminishing activity until the seventh day, overlapping with the cartilage callus formation phase [42]. Bridging osseous tissue extends from the development of hypertrophic chondrocyte populations arising from the subperiosteum as well as from differentiated mesenchymal precursors that have migrated into the injury site [6]. The production of a cartilage lattice that bridges the fracture gap is preceded by an elevation in type I and type II collagen messenger RNA (mRNA) and suppression of TNF-α and IL-1β concentrations [6]. If the TNF-α and IL-1β levels are not decreased as intended within this window, there is a synergistic interaction between the two cytokines that creates a diffuse impairment of chondrocyte proliferation and increased rates of chondrocyte apoptosis [43]. Day 7-14: Ossification and Revascularization of the Fracture Callus The relative quantity of ossified cartilage callus, callus volume, and level of type I collagen mRNA expression climaxes between days seven and fourteen following fracture, which also coincides with the initiation of cartilage resorption and the peak number of hypertrophied chondrocytes [6, 19]. The rate of longitudinal bone growth has been shown to have a linear relationship with the final relative volume of the hypertrophied
ACCEPTED MANUSCRIPT chondrocytes in both murine and porcine models [44]. Hypertrophic chondrocytes have high intra-cellular concentrations of TNF-α [45], which has been suggested to be an initiator of the hypertrophic phenotype [29].
PT
Transition from fracture callus to bone requires chondrocyte apoptosis, cartilage matrix degradation, stromal cell recruitment and osteogenic differentiation, as well as neo-vascularization [46]. TNF-α facilitates this
RI
process by inducing hypertrophic and non-hypertrophic chondrocyte apoptosis in a dose dependent fashion [47,
SC
48] by initiating a cascade of matrix metalloproteinase (MMP) activation (including MMP 9 and 14) [46] and by stimulating angiogenesis. Programmed subtraction of the temporary cartilage matrix allows endothelial cells to
NU
penetrate the callus and facilitate transition to immature bone through neovascularization [16, 49]. TNF-α also facilitates angiogenesis through enhancement of vascular endothelial growth factor (VEGF) production at the
MA
injury site [46]. Further evidence of the role of TNF-α in this stage of osteosynthesis is reported from a murine model with TNFR1 and TNFR2 knockout mice, which demonstrated fewer hypertrophic chondrocytes, delayed
D
endochondral tissue resorption secondary to impairment of chondrocyte apoptosis, and the delayed release of
TE
cytokines mediating endochondral tissue remodeling [8, 29]. Day 14 - 21: Substitution of Ossified Fracture Callus with Trabecular Bone
AC CE P
By the twenty-first day following fracture, all callus has normally been resorbed and replaced with trabecular bone. TNFR deficient mice have delayed callus resorption with very little trabecular bone formed and have a larger percentage of undifferentiated mesenchymal cells and a larger fracture callus [8] at this point in the healing process. TNF-α enhances human osteoblast proliferation in small doses applied for brief intervals, but prolonged exposure in large doses inhibits osteoblasts, likely through down-regulation of sialoprotein and osteocalcin gene expression [17, 50]. Day 28 and After: Remodeling of Deposited Trabecular Bone IL-1β and TNF-α are elevated during remodeling [31], which begins during the interval between day 21 and day 28 [6]. TNF-α has been demonstrated to increase the number of chondrocytes that secrete receptor activator of nuclear factor kappa-B ligand (RANKL) [45] and enhance trans-endothelial migration of CD14+ monocytes that can differentiate into osteoclasts [51]. These two mechanisms allow TNF-α to increase both the bone resorption capacity of established osteoclasts and potentially increase the total number of osteoclasts within a specific area to encourage the bone remodeling process.
ACCEPTED MANUSCRIPT The basic science evidence on manipulation of the peri-fracture TNF-α levels and the effect on bone healing is presented in Table 2.
PT
TNF-α and Fracture Nonunions: Fracture nonunions can be broadly classified as having either a biologic or mechanical etiology, which
RI
have characteristic radiographic presentations. Hypertrophic nonunions are identified by an excess of callus over
SC
the fracture site and occur from mechanical instability or hypermobility between the two dissociated bone ends. The innate healing response is adequate to initiate the formation of repair tissue, but the production of mature
NU
bone is never achieved. As the entire process of mesenchymal cell recruitment is initiated by the fracture hematoma, inter-fragmentary movements disrupting the emerging vascular network may generate smaller
MA
hematomas within the repair tissue that re-initiate the process of mesenchymal stem cell recruitment and differentiation. Mesenchymal cells recovered from hypertrophic nonunion calluses do demonstrate the capacity to
D
differentiate into osteogenic and chondrogenic cells, supporting the concept that the fracture callus in this
TE
situation can serve as a mesenchymal stem cell reservoir [52]. Therefore, at any time point along the course of attempted osteosynthesis, a hypertrophic nonunion may have several centers of inflammation driving multiple
AC CE P
miniature sites of mesenchymal stem cell recruitment and differentiation, resulting in a cumulatively larger than necessary immature fracture callus volume. Oligotrophic and atrophic nonunions are the result of an intrinsic deficit in the host response to injury, which is manifested by an absence or inadequate volume of repair tissue. It was previously thought that oligotrophic and atrophic nonunions were a symptom of inadequate supporting blood supply [53, 54]. This theory has been confirmed in a macroscopic context where fractures occurring in limbs with vascular injury have a higher risk of nonunion [55]; however, recent studies suggest that the effect of deficient blood supply is less significant at the cellular level. Fractures that progress to form nonunions do have fewer blood vessels infiltrating the repair tissue during the eight weeks following injury compared to tissues with normal osteosynthesis, but there is no significant difference beyond this time point [56]. This same observation was also found in patients undergoing revision surgery for hypertrophic and atrophic nonunions, which failed to demonstrate any significant difference in vessel density within the repair tissue [57]. Therefore, as the relative lack of local vascularity associated with atrophic nonunions appears only in the very early stages of fracture healing, this deficit may arise from an absence
ACCEPTED MANUSCRIPT of local angiogenesis stimulation rather than a lack of supporting vasculature external to the callus. Furthermore, evidence suggests that atrophic nonunions may have diminished growth secondary to decreased concentrations of
PT
growth factors that prime the tissue surrounding the fracture site [58]. This finding is consistent with the theory describing a regional acceleratory phenomenon (RAP), suggesting there is an activity level and cell quantity
RI
threshold that produces a sustained osteosynthesis response when achieved [30]. If the undefined critical mass is
SC
not established, the fracture will not maintain enough momentum to carry the premature repair tissue to mature trabecular bone. As the number and phenotype of chondro- and osteogenic mesenchymal cells is partially
NU
determined by the concentration of TNF-α within the fracture site during the first three days following injury, this cytokine is implicated in the development of biological nonunions. Therefore, relative TNF-α deficiency may be the
MA
principal influence in the development of oligotrophic and atrophic nonunions. Infection is another frequent nonunion etiology. This process impedes fracture healing through physical
D
disruption of the healing cascade, requiring a “resetting” of the entire twenty-eight day process for significant
TE
osteosynthesis to progress within the confines of the fracture site. Furthermore, the presence of bacteria stimulates a local immune response, which interferes with the physiologic bimodal temporal pattern of elevated
AC CE P
TNF-α. During the acute stages of osteomyelitis, the local TNF-α concentration is elevated as an immune response to bacteria within the fracture site [59-61]. This pathological elevation of TNF-α may actually stimulate the migration of mesenchymal stem cells and differentiation along the osteogenic/chondrogenic cell lineages within the first three days following fracture. However, persistently elevated TNF-α beyond the first three days decreases the cell proliferation rate and induces apoptosis in the chondrocytes that drive the process of endochondral ossification [43]. Furthermore, biofilm formation over the bone surface and the subsequently elevated TNF-α promotes osteoclastogenesis [62], which disrupts the equilibrium between bone resorption and production. This explanation could support the observation of local osteopenia that develops in cases in mature bones that are afflicted with osteomyelitis and may create a scenario where bone is resorbed as it is being synthesized with a reduced net accumulation. Metabolic Bone Disease and TNF-α: A majority of studies examining the role of TNF-α have focused on the effect of inhibition at various stages of bone healing, which is clinically relevant due to the introduction of disease-modifying anti-rheumatic
ACCEPTED MANUSCRIPT drugs (DMARDs) in rheumatoid arthritis care. A more subtle, but likely broader clinical issue for orthopaedic surgeons arises from the effect of systemic, chronic supra-physiologic TNF-α levels that are associated with
PT
obesity, diabetes, alcohol abuse, and advanced age. A murine fracture model demonstrated that persistently elevated TNF-α levels lead to a higher nonunion rate, impaired cartilage fracture callus formation, and fracture
RI
bridging with non-osteogenic fibrous tissue, which is consistent with impaired mesenchymal cell differentiation
SC
into a chondroblast phenotype [63]. Another murine study demonstrated that, given over seven days, supplemental TNF-α and IL-1β significantly impaired bone growth and that this effect was reversed with
NU
administration of anakinra (anti-IL-1) and etanercept (anti-TNF-α) [64]. Although this investigation was carried out in fetal rats, there is considerable crossover between the orchestration of fracture healing and initial long bone
MA
formation as both advance through endochondral ossification. Furthermore, after the initial phase of fracture healing, continuous application of TNF-α down-regulates the expression of sialoprotein and osteocalcin, which are
D
proxy indicators of increased bone formation [17]. From these data, it can be inferred that any process resulting in
TE
a persistently elevated TNF-α level has the potential to suppress secondary bone healing. The likely mechanism behind this observation is the diffuse impairment of chondrocyte proliferation and increased rates of chondrocyte
AC CE P
apoptosis seen with elevated TNF-α levels during the production of the bridging fracture callus between the third and seventh day following injury [43]. Osteoporosis:
-/-
Work by Li et al. [65] demonstrated that RANK nullizygous (RANK ) mice failed to demonstrate an osteoclastogenic response to exogenous TNF-α, as well as parathyroid hormone-related protein (PTHrP), 1,25 dihydroxyvitamin D [1α,25-(OH)2D3], and IL1-β, demonstrating that RANK is the central mediator of bone homeostasis. In this context, it is clear that TNF-α is a mediator of baseline bony homeostasis and is able to promote osteoclast activity through stimulation of the RANK pathway [66, 67]. Estrogen withdrawal associated with menopause prompts increased macrophage secretion of IL-12 and IL-18, inducing activation of T-cells that secrete TNF-α [68], resulting in elevated serum concentrations. Serum isolates from post-menopausal females without estrogen replacement therapy demonstrated higher levels of TNF-α and higher bone resorption activity, which was reversed with TNF-α inhibition [69]. Similar work has demonstrated that administration of etanercept in post-menopausal women significantly reduced bone turnover following withdrawal from supplemental estrogen
ACCEPTED MANUSCRIPT [70]. Further emphasis of the estrogen/TNF-α relationship arises from the observation that the bone resorption characteristics found in post-menopausal females treated with estrogen therapy were similar to premenopausal
PT
subjects. Estradiol suppresses TNF-α activity by inhibiting macrophage NF-kB signaling through kappaB-Ras2 expression [71, 72]. As osteoclasts and macrophages differentiate from the same precursor lineage and inhibition
RI
of osteoclast NF-kB pathways freezes osteoclast activity [73], estrogen depletion may remove a negative stimulus
SC
for osteoclast activity.
In addition to baseline decreases in bone mineral density, which predispose post-menopausal females to
NU
an increased fracture risk, estrogen depletion is also associated with a blunted reparative response to osseous injury. A murine model of osteoporosis demonstrated that the repair tissue had a higher ratio of undifferentiated
MA
mesenchymal cells, a less robust chondrogenic response, and a weaker callus at seven days after injury [74]. At 14 and 21 days following the injury, the osteoporotic mice had delayed bone formation and also had a greater
D
presence of immature woven, rather than lamellar bone, compared to controls. Despite the inferior response, the
TE
osteoporotic mice were also found to have an elevated concentration of BMP-2, which is likely explained by the
2 [17, 26, 27]. Obesity:
AC CE P
observation that TNF-α, which is pathologically elevated at baseline, promotes the synthesis and secretion of BMP-
Wolff’s law predicts that increased body mass augments the mechanical load across the axial skeleton and lower extremities, which would result in denser bones with superior strength. However, chronic inflammation associated with obesity may promote osteoclastogenesis through up-regulation of the NF-κB (RANK)/RANK pathway [75] via increased levels of TNF-α secreted by excess adipose tissue [76, 77]. Advanced Age: Although associated with advanced age, osteoporosis is separate from age related deficits in osteosynthesis following fracture repair, which is likely realized from an innate reduction in the robustness of the mesenchymal stem cells. A murine model comparing histological and molecular parameters in juvenile, middle age, and elderly mice demonstrated that there was a significant reduction in reparative capacity between juvenile and middle age mice, but not between middle age and elderly [78]. Specifically, the juvenile mice had earlier chondrocyte expression of collagen type II and type X, earlier osteoblast expression of osteocalcin, earlier vascular
ACCEPTED MANUSCRIPT infiltration, and earlier endochondral ossification. Furthermore, there was no significant difference in the quantity of cartilage produced between the juvenile mice and the middle age/elderly mice, but there was significantly less
PT
bone formed by the older two groups [78]. A rabbit study investigated the chondrogenic potential of explanted periosteum from subjects at different ages and found a significant decrease in the ability to stimulate periosteal
RI
progenitor cells into chondral producing phenotypes once skeletal maturity was achieved [79]. Specimens from
SC
skeletally mature subjects also had less robust cartilage production, decreased cartilage substrate uptake, and decreased cellularity within the periosteal cambium layer, but not a decreased percentage of proliferating cells
NU
[79]. Lastly, a murine model observed that older mice were found to have elevated serum concentrations of IL-6 and TNF-α at baseline when compared to younger controls [7], which may suppress the inherent bone healing
MA
ability in similar mechanisms to obesity and osteoporosis. This may explain the observation that higher proportions of osteoclasts are found within the fracture site in older subjects which likely explains the decreased callus
D
mineralization and bridging capacity following fracture [80].
TE
Diabetes:
Clinical studies have demonstrated that both insulin-dependent and non-insulin-dependent diabetics have
AC CE P
prolonged healing times following closed displaced fractures, but not in non-displaced fractures [81], suggesting that secondary bone healing is less robust in this patient population. Induced diabetes in a murine model led to an elevated TNF-α level compared to baseline, which was associated with an elevated level of chondrocyte proapoptotic mRNA and a smaller cartilage callus within an induced fracture healing site when compared to nondiabetic control subjects [47]. Diabetic mice were also shown to have enhanced chondrocyte apoptosis after the sixteenth day following injury [47], which is consistent with persistently and inappropriately elevated levels of TNFα during this stage of fracture healing. It was further demonstrated that TNF-α inhibition lead to a reversal of these findings and re-established the callus forming capacity in diabetic mice and that TNF-α inhibition in normoglycemic subjects had no impact on the callus formation [47]. Another murine model demonstrated that uncontrolled blood glucose resulted in decreased fracture callus strength, decreased collagen concentration, and decreased cellularity, but that the callus present in diabetic subjects with blood sugars controlled with insulin therapy was not significantly different than controls [82].
ACCEPTED MANUSCRIPT A similar murine model also demonstrated that diabetes was associated with an increased rate of osteoclastogenesis, which was inhibited by TNF-α inhibition and that subjects with induced diabetes continued to
PT
have significantly less fracture callus when compared to controls despite TNF-α inhibition [45] Chronic Alcohol Consumption:
RI
A murine model demonstrated that distraction osteogenesis was inhibited by ethanol consumption and
SC
that consistent administration of anti-IL-1 and anti-TNFR1 antibodies sufficiently increased the healing potential compared to control [83].
NU
Aseptic Loosening:
Prosthesis derived wear debris, titanium, and cement from total hip replacements has been shown to
MA
induce local osteoclast bone resorption in a dose dependent relationship due to their ability to increase TNF-α concentrations following particle phagocytosis by resident macrophages [84-86]. Furthermore, in vitro and in vivo
Conclusion:
TE
induced by titanium particle debris [87].
D
models have demonstrated that TNF-α inhibition with etanercept reduces osteoclast driven bone resorption
AC CE P
The morbidity and mortality associated with osteoporotic fragility fractures is staggering and emphasizes the importance of early surgical intervention and mobilization of these patients [88, 89]. As the quality of bone cannot be improved in the acute setting, enhancing the innate healing response to the fracture may reduce the incidence of nonunion and provide a more stable construct earlier in the postoperative course. Besides the benefit of an early recovery, enhancing secondary bone healing may reduce the incidence of complications and increased mortality rates associated with osteoporotic fragility fractures [90]. Despite an abundant body of basic science literature implicating its role as a primary mediator of fracture healing and metabolic bone disease, the role of TNF-α seems clinically underappreciated. The consistent themes of TNF-α regulating mesenchymal stem cell repopulation within a fracture gap and the ability of pathologically elevated TNF-α concentrations to enhance osteoclastogenesis are promising targets for TNF-α immunotherapy and may be the scientific foundation for future osteoporosis management interventions.
ACCEPTED MANUSCRIPT References:
8. 9. 10. 11. 12. 13. 14.
15. 16. 17.
18. 19. 20. 21.
22. 23. 24.
PT
RI
SC
7.
NU
5. 6.
MA
4.
D
3.
TE
2.
Schell, H., et al., The course of bone healing is influenced by the initial shear fixation stability. J Orthop Res, 2005. 23(5): p. 1022-8. Klein, P., et al., The initial phase of fracture healing is specifically sensitive to mechanical conditions. J Orthop Res, 2003. 21(4): p. 662-9. Lienau, J., et al., Initial vascularization and tissue differentiation are influenced by fixation stability. J Orthop Res, 2005. 23(3): p. 639-45. Thornton, F.J., M.R. Schaffer, and A. Barbul, Wound healing in sepsis and trauma. Shock, 1997. 8(6): p. 391-401. Marsell, R. and T.A. Einhorn, The biology of fracture healing. Injury, 2011. 42(6): p. 551-5. Kon, T., et al., Expression of osteoprotegerin, receptor activator of NF-kappaB ligand (osteoprotegerin ligand) and related proinflammatory cytokines during fracture healing. J Bone Miner Res, 2001. 16(6): p. 1004-14. Wahl, E.C., et al., Restoration of regenerative osteoblastogenesis in aged mice: modulation of TNF. J Bone Miner Res, 2010. 25(1): p. 114-23. Gerstenfeld, L.C., et al., Impaired fracture healing in the absence of TNF-alpha signaling: the role of TNFalpha in endochondral cartilage resorption. J Bone Miner Res, 2003. 18(9): p. 1584-92. Sandberg, M.M., H.T. Aro, and E.I. Vuorio, Gene expression during bone repair. Clin Orthop Relat Res, 1993(289): p. 292-312. Einhorn, T.A., et al., The expression of cytokine activity by fracture callus. J Bone Miner Res, 1995. 10(8): p. 1272-81. Street, J., et al., Is human fracture hematoma inherently angiogenic? Clin Orthop Relat Res, 2000(378): p. 224-37. Grundnes, O. and O. Reikeras, The importance of the hematoma for fracture healing in rats. Acta Orthop Scand, 1993. 64(3): p. 340-2. Mizuno, K., et al., The osteogenetic potential of fracture haematoma. Subperiosteal and intramuscular transplantation of the haematoma. J Bone Joint Surg Br, 1990. 72(5): p. 822-9. Caetano-Lopes, J., et al., Upregulation of inflammatory genes and downregulation of sclerostin gene expression are key elements in the early phase of fragility fracture healing. PLoS One, 2011. 6(2): p. e16947. Currie, H.N., et al., Spatial cytokine distribution following traumatic injury. Cytokine, 2014. 66(2): p. 112-8. Gerstenfeld, L.C., et al., Fracture healing as a post-natal developmental process: molecular, spatial, and temporal aspects of its regulation. J Cell Biochem, 2003. 88(5): p. 873-84. Lu, Z., et al., Short-term exposure to tumor necrosis factor-alpha enables human osteoblasts to direct adipose tissue-derived mesenchymal stem cells into osteogenic differentiation. Stem Cells Dev, 2012. 21(13): p. 2420-9. Frost, H.M., The biology of fracture healing. An overview for clinicians. Part I. Clin Orthop Relat Res, 1989(248): p. 283-93. Einhorn, T.A., The cell and molecular biology of fracture healing. Clin Orthop Relat Res, 1998(355 Suppl): p. S7-21. Zhou, F.H., et al., TNF-alpha mediates p38 MAP kinase activation and negatively regulates bone formation at the injured growth plate in rats. J Bone Miner Res, 2006. 21(7): p. 1075-88. Mountziaris, P.M., S.N. Tzouanas, and A.G. Mikos, Dose effect of tumor necrosis factor-alpha on in vitro osteogenic differentiation of mesenchymal stem cells on biodegradable polymeric microfiber scaffolds. Biomaterials, 2010. 31(7): p. 1666-75. Harry, L.E., et al., Comparison of the healing of open tibial fractures covered with either muscle or fasciocutaneous tissue in a murine model. J Orthop Res, 2008. 26(9): p. 1238-44. Khouri, R.K., et al., Repair of calvarial defects with flap tissue: role of bone morphogenetic proteins and competent responding tissues. Plast Reconstr Surg, 1996. 98(1): p. 103-9. Fu, X., et al., Migration of bone marrow-derived mesenchymal stem cells induced by tumor necrosis factoralpha and its possible role in wound healing. Wound Repair Regen, 2009. 17(2): p. 185-91.
AC CE P
1.
ACCEPTED MANUSCRIPT
31. 32. 33. 34. 35.
36. 37.
38. 39. 40. 41. 42.
43. 44. 45. 46.
47. 48.
PT
RI
SC
30.
NU
29.
MA
28.
D
27.
TE
26.
Zubov, D.O., [Immunoregulatory role of mesenchymal stem cells in bone reparation processes]. Fiziol Zh, 2008. 54(4): p. 30-6. Pelissier, P., et al., Induced membranes secrete growth factors including vascular and osteoinductive factors and could stimulate bone regeneration. J Orthop Res, 2004. 22(1): p. 73-9. Eguchi, Y., et al., Antitumor necrotic factor agent promotes BMP-2-induced ectopic bone formation. J Bone Miner Metab, 2010. 28(2): p. 157-64. Zhao, Y.P., et al., The promotion of bone healing by progranulin, a downstream molecule of BMP-2, through interacting with TNF/TNFR signaling. Biomaterials, 2013. 34(27): p. 6412-21. Gerstenfeld, L.C., et al., Impaired intramembranous bone formation during bone repair in the absence of tumor necrosis factor-alpha signaling. Cells Tissues Organs, 2001. 169(3): p. 285-94. Frost, H.M., The biology of fracture healing. An overview for clinicians. Part II. Clin Orthop Relat Res, 1989(248): p. 294-309. Mountziaris, P.M. and A.G. Mikos, Modulation of the inflammatory response for enhanced bone tissue regeneration. Tissue Eng Part B Rev, 2008. 14(2): p. 179-86. Bonewald, L.F. and G.R. Mundy, Role of transforming growth factor-beta in bone remodeling. Clin Orthop Relat Res, 1990(250): p. 261-76. Lind, M., et al., Transforming growth factor-beta enhances fracture healing in rabbit tibiae. Acta Orthop Scand, 1993. 64(5): p. 553-6. Nielsen, H.M., et al., Local injection of TGF-beta increases the strength of tibial fractures in the rat. Acta Orthop Scand, 1994. 65(1): p. 37-41. Iwasaki, M., et al., Regulation of proliferation and osteochondrogenic differentiation of periosteum-derived cells by transforming growth factor-beta and basic fibroblast growth factor. J Bone Joint Surg Am, 1995. 77(4): p. 543-54. Roelen, B.A. and P. Dijke, Controlling mesenchymal stem cell differentiation by TGFBeta family members. J Orthop Sci, 2003. 8(5): p. 740-8. Chen, Y.J., et al., Recruitment of mesenchymal stem cells and expression of TGF-beta 1 and VEGF in the early stage of shock wave-promoted bone regeneration of segmental defect in rats. J Orthop Res, 2004. 22(3): p. 526-34. Waters, R.V., et al., Systemic corticosteroids inhibit bone healing in a rabbit ulnar osteotomy model. Acta Orthop Scand, 2000. 71(3): p. 316-21. Malviya, A., et al., The effect of newer anti-rheumatic drugs on osteogenic cell proliferation: an in-vitro study. J Orthop Surg Res, 2009. 4: p. 17. Glass, G.E., et al., TNF-alpha promotes fracture repair by augmenting the recruitment and differentiation of muscle-derived stromal cells. Proc Natl Acad Sci U S A, 2011. 108(4): p. 1585-90. Caron, M.M., et al., Activation of NF-kappaB/p65 facilitates early chondrogenic differentiation during endochondral ossification. PLoS One, 2012. 7(3): p. e33467. Cho, T.J., L.C. Gerstenfeld, and T.A. Einhorn, Differential temporal expression of members of the transforming growth factor beta superfamily during murine fracture healing. J Bone Miner Res, 2002. 17(3): p. 513-20. Martensson, K., D. Chrysis, and L. Savendahl, Interleukin-1beta and TNF-alpha act in synergy to inhibit longitudinal growth in fetal rat metatarsal bones. J Bone Miner Res, 2004. 19(11): p. 1805-12. Breur, G.J., et al., Linear relationship between the volume of hypertrophic chondrocytes and the rate of longitudinal bone growth in growth plates. J Orthop Res, 1991. 9(3): p. 348-59. Alblowi, J., et al., High levels of tumor necrosis factor-alpha contribute to accelerated loss of cartilage in diabetic fracture healing. Am J Pathol, 2009. 175(4): p. 1574-85. Lehmann, W., et al., Tumor necrosis factor alpha (TNF-alpha) coordinately regulates the expression of specific matrix metalloproteinases (MMPS) and angiogenic factors during fracture healing. Bone, 2005. 36(2): p. 300-10. Kayal, R.A., et al., TNF-alpha mediates diabetes-enhanced chondrocyte apoptosis during fracture healing and stimulates chondrocyte apoptosis through FOXO1. J Bone Miner Res, 2010. 25(7): p. 1604-15. Aizawa, T., et al., Induction of apoptosis in chondrocytes by tumor necrosis factor-alpha. J Orthop Res, 2001. 19(5): p. 785-96.
AC CE P
25.
ACCEPTED MANUSCRIPT
55. 56. 57. 58.
59. 60. 61. 62. 63. 64.
65. 66. 67. 68. 69. 70. 71. 72. 73.
PT
RI
SC
54.
NU
53.
MA
52.
D
51.
TE
50.
Ai-Aql, Z.S., et al., Molecular mechanisms controlling bone formation during fracture healing and distraction osteogenesis. J Dent Res, 2008. 87(2): p. 107-18. Frost, A., et al., Inflammatory cytokines regulate proliferation of cultured human osteoblasts. Acta Orthop Scand, 1997. 68(2): p. 91-6. Kindle, L., et al., Human microvascular endothelial cell activation by IL-1 and TNF-alpha stimulates the adhesion and transendothelial migration of circulating human CD14+ monocytes that develop with RANKL into functional osteoclasts. J Bone Miner Res, 2006. 21(2): p. 193-206. Iwakura, T., et al., Human hypertrophic nonunion tissue contains mesenchymal progenitor cells with multilineage capacity in vitro. J Orthop Res, 2009. 27(2): p. 208-15. Rhinelander, F.W., Tibial blood supply in relation to fracture healing. Clin Orthop Relat Res, 1974(105): p. 34-81. Arany, L., et al., Arteriographic studies in delayed-union and non-union of fractures. Radiol Diagn (Berl), 1980. 21(5): p. 673-81. Dickson, K., et al., Delayed unions and nonunions of open tibial fractures. Correlation with arteriography results. Clin Orthop Relat Res, 1994(302): p. 189-93. Reed, A.A., et al., Vascularity in a new model of atrophic nonunion. J Bone Joint Surg Br, 2003. 85(4): p. 604-10. Reed, A.A., et al., Human atrophic fracture non-unions are not avascular. J Orthop Res, 2002. 20(3): p. 593-9. Reed, A.A.C., et al., DO ATROPHIC NON-UNIONS HAVE LOWER LEVELS OF GROWTH FACTORS THAN HYPERTROPHIC NON-UNIONS AND HEALING FRACTURES? Journal of Bone & Joint Surgery, British Volume, 2002. 84-B(SUPP I): p. 19. Littlewood-Evans, A.J., et al., Local expression of tumor necrosis factor alpha in an experimental model of acute osteomyelitis in rats. Infect Immun, 1997. 65(8): p. 3438-43. Klosterhalfen, B., et al., Local and systemic inflammatory mediator release in patients with acute and chronic posttraumatic osteomyelitis. J Trauma, 1996. 40(3): p. 372-8. Vallejo, J.G., C.J. Baker, and M.S. Edwards, Roles of the bacterial cell wall and capsule in induction of tumor necrosis factor alpha by type III group B streptococci. Infect Immun, 1996. 64(12): p. 5042-6. Meghji, S., et al., Surface-associated protein from Staphylococcus aureus stimulates osteoclastogenesis: possible role in S. aureus-induced bone pathology. Br J Rheumatol, 1998. 37(10): p. 1095-101. Hashimoto, J., et al., Inhibitory effects of tumor necrosis factor alpha on fracture healing in rats. Bone, 1989. 10(6): p. 453-7. Fernandez-Vojvodich, P., E. Karimian, and L. Savendahl, The biologics anakinra and etanercept prevent cytokine-induced growth retardation in cultured fetal rat metatarsal bones. Horm Res Paediatr, 2011. 76(4): p. 278-85. Li, J., et al., RANK is the intrinsic hematopoietic cell surface receptor that controls osteoclastogenesis and regulation of bone mass and calcium metabolism. Proc Natl Acad Sci U S A, 2000. 97(4): p. 1566-71. Fuller, K., et al., TNFalpha potently activates osteoclasts, through a direct action independent of and strongly synergistic with RANKL. Endocrinology, 2002. 143(3): p. 1108-18. Lam, J., et al., TNF-alpha induces osteoclastogenesis by direct stimulation of macrophages exposed to permissive levels of RANK ligand. J Clin Invest, 2000. 106(12): p. 1481-8. Roggia, C., et al., Role of TNF-alpha producing T-cells in bone loss induced by estrogen deficiency. Minerva Med, 2004. 95(2): p. 125-32. Cohen-Solal, M.E., et al., Peripheral monocyte culture supernatants of menopausal women can induce bone resorption: involvement of cytokines. J Clin Endocrinol Metab, 1993. 77(6): p. 1648-53. Charatcharoenwitthaya, N., et al., Effect of blockade of TNF-alpha and interleukin-1 action on bone resorption in early postmenopausal women. J Bone Miner Res, 2007. 22(5): p. 724-9. Murphy, A.J., P.M. Guyre, and P.A. Pioli, Estradiol suppresses NF-kappa B activation through coordinated regulation of let-7a and miR-125b in primary human macrophages. J Immunol, 2010. 184(9): p. 5029-37. An, J., et al., Estradiol repression of tumor necrosis factor-alpha transcription requires estrogen receptor activation function-2 and is enhanced by coactivators. Proc Natl Acad Sci U S A, 1999. 96(26): p. 15161-6. Soysa, N.S. and N. Alles, NF-kappaB functions in osteoclasts. Biochem Biophys Res Commun, 2009. 378(1): p. 1-5.
AC CE P
49.
ACCEPTED MANUSCRIPT
81. 82. 83. 84. 85. 86. 87. 88. 89. 90.
PT
RI
SC
NU
80.
MA
78. 79.
D
77.
TE
75. 76.
Islam, A.A., et al., Healing of fractures in osteoporotic rat mandible shown by the expression of bone morphogenetic protein-2 and tumour necrosis factor-alpha. Br J Oral Maxillofac Surg, 2005. 43(5): p. 38391. Cao, J.J., Effects of obesity on bone metabolism. J Orthop Surg Res, 2011. 6: p. 30. Pedersen, M., et al., Circulating levels of TNF-alpha and IL-6-relation to truncal fat mass and muscle mass in healthy elderly individuals and in patients with type-2 diabetes. Mech Ageing Dev, 2003. 124(4): p. 495502. Bastard, J.P., et al., Recent advances in the relationship between obesity, inflammation, and insulin resistance. Eur Cytokine Netw, 2006. 17(1): p. 4-12. Lu, C., et al., Cellular basis for age-related changes in fracture repair. J Orthop Res, 2005. 23(6): p. 1300-7. O'Driscoll, S.W., et al., The chondrogenic potential of periosteum decreases with age. J Orthop Res, 2001. 19(1): p. 95-103. Mehta, M., et al., Influences of age and mechanical stability on volume, microstructure, and mineralization of the fracture callus during bone healing: is osteoclast activity the key to age-related impaired healing? Bone, 2010. 47(2): p. 219-28. Loder, R.T., The influence of diabetes mellitus on the healing of closed fractures. Clin Orthop Relat Res, 1988(232): p. 210-6. Macey, L.R., et al., Defects of early fracture-healing in experimental diabetes. J Bone Joint Surg Am, 1989. 71(5): p. 722-33. Perrien, D.S., et al., IL-1 and TNF antagonists prevent inhibition of fracture healing by ethanol in rats. Toxicol Sci, 2004. 82(2): p. 656-60. Schwarz, E.M., et al., Tumor necrosis factor-alpha/nuclear transcription factor-kappaB signaling in periprosthetic osteolysis. J Orthop Res, 2000. 18(3): p. 472-80. Merkel, K.D., et al., Tumor necrosis factor-alpha mediates orthopedic implant osteolysis. Am J Pathol, 1999. 154(1): p. 203-10. Algan, S.M., M. Purdon, and S.M. Horowitz, Role of tumor necrosis factor alpha in particulate-induced bone resorption. J Orthop Res, 1996. 14(1): p. 30-5. Childs, L.M., et al., Efficacy of etanercept for wear debris-induced osteolysis. J Bone Miner Res, 2001. 16(2): p. 338-47. Brauer, C.A., et al., Incidence and mortality of hip fractures in the United States. JAMA, 2009. 302(14): p. 1573-9. Tsuboi, M., et al., Mortality and mobility after hip fracture in Japan: a ten-year follow-up. J Bone Joint Surg Br, 2007. 89(4): p. 461-6. Melton, L.J., 3rd, et al., Predictors of excess mortality after fracture: a population-based cohort study. J Bone Miner Res, 2014. 29(7): p. 1681-90.
AC CE P
74.
Figure 1
AC CE P
TE
D
MA
NU
SC
RI
PT
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT Table 1. Illustrates the overlapping bone healing phases by the primary cellular process as well as step in healing cascade [4-6]. Step Acute Inflammation
TNF- α Elevated
Day 3-7
Cartilage Fracture Callus Production Callus Ossification and Revascularization
Baseline
RI
Elevated
Elevated to baseline
AC CE P
TE
D
MA
Day 28 and >
Substitution with Trabecular bone Remodeling
SC
Day 14 - 28
Baseline/Rising
NU
Day 7-14
Predominant Function Mesenchymal migration, proliferation, differentiation None, inhibits cartilage production Induces hypertrophic chondrocyte phenotype, angiogenesis, and MMP production Osteoblast differentiation
PT
Time Period Days 1-3
Osteoclastogenesis
ACCEPTED MANUSCRIPT Table 2. Evidence from basic science studies that have manipulated the amount or effect of peri-fracture TNF-α. Step Acute Inflammation
Study [12]
Days 1-3
Acute Inflammation
[38]
Days 1-3
Acute Inflammation
[8, 29]
Days 1-3
Acute Inflammation
[20]
Days 1-3
Acute Inflammation
[40]
Supplemental TNF-α
Day 3-7
Fracture Callus Production
[43]
Supplemental TNF-α
Day 7-14
Callus Ossification and Revascularization
Day 14 - 21
Substitution with Trabecular Bone
Remodeling
RI
SC
Etanercept administration at fracture site
NU
MA
D
TE
AC CE P
Day 28 and >
Study Peri-fracture hematoma evacuation Glucocorticoid administration TNFR1 and TNFR2 KO (murine model)
Effect Inferior repair tissue Diminished fracture healing
PT
Time Period Days 1-3
[8, 29].
TNFR1 and TNFR2 KO (murine model)
[8]
TNFR1 and TNFR2 KO (murine model)
X
X
Delayed mesenchymal cell differentiation, decreased osteoblast numbers in the fracture site Immune cell migration was not affected, mesenchymal cell migration into fracture site was diminished Accelerated fracture healing, higher callus mineralization rates Impaired chondrocyte proliferation and increased chondrocyte apoptosis Fewer hypertrophic chondrocytes, delayed cartilage resorption due to delayed chondrocyte apoptosis Larger fracture callus with greater undifferentiated mesenchymal cells, less trabecular bone formed X
ACCEPTED MANUSCRIPT Highlights:
Secondary fracture healing is an inflammatory process that relies on immune cytokines to repopulate the
Tumor Necrosis Factor-Alpha (TNF-α) is a central mediator in this inflammatory response in combination
RI
PT
fracture gap with chondrogenic and osteogenic mesenchymal stem cells.
with the host reserve of mesenchymal stem cells. TNF-α is elevated in a bimodal temporal distribution
Disease states associated with poor bone healing such as osteoporosis and diabetes have abnormal TNF-α
NU
SC
and deviations from this schedule produce a diminished healing response.
responses to injury, which likely increases the risk of delayed and nonunion in these patient populations.
MA
Appreciation of TNF-α’s role in secondary bone healing may lead to new therapies to augment fracture
TE
D
healing and reduce the morbidity associated with protracted recovery.
AC CE P
S8756-3282(15)00172-6 doi: 10.1016/j.bone.2015.05.001 BON 10718
To appear in:
Bone
Received date: Revised date: Accepted date:
20 February 2015 29 April 2015 1 May 2015
Please cite this article as: Karnes Jonathan M., Daffner Scott D., Watkins Colleen M., Multiple Roles of Tumor Necrosis Factor - Alpha in Fracture Healing, Bone (2015), doi: 10.1016/j.bone.2015.05.001
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT TITLE: Multiple Roles of Tumor Necrosis Factor - Alpha in Fracture Healing
1
RI
Jonathan M. Karnes, M.D. [email protected]
PT
AUTHORS:
1
1
NU
Colleen M. Watkins, M.D. [email protected]
SC
Scott D. Daffner, M.D. [email protected]
INSTITUTION: 1
MA
Department of Orthopaedics, West Virginia University, Morgantown, WV
CORRESPONDING AUTHOR:
TE
D
Jonathan M. Karnes, MD; Department of Orthopaedics, West Virginia University, P.O. Box 9196, Morgantown, WV 26506 Tel (304) 293-1168 Fax (304) 293-7070 Email: [email protected] KEYWORDS:
AC CE P
TNF-α, bone healing, metabolic bone disease
SOURCE OF FUNDING: None
ACCEPTED MANUSCRIPT Introduction: The healing pathway of a broken bone is primarily dictated by the mechanical relationship and position of
PT
the two fracture ends [1]. The initial stability achieved within the first days dictates the trajectory of the entire healing process and the final quality of the repair tissue [2]. Absolute stability with compression across a fracture
RI
gap induces primary bone healing, which seals the two bone fragments together using a remodeling process and
SC
occurs independently of an inflammatory reaction. Inadequate immobilization between the fracture fragments results in a higher percentage of mechanically inferior fibrous repair tissue with decreased mineralization and
NU
angiogenesis in the early stages of healing [3]. Fractures without absolute stability progress through a series of coordinated and predictable steps, known as secondary bone healing. The general paradigm for secondary bone
MA
healing follows the same pattern seen in soft tissue repair including hemostasis, inflammation, mesenchymal cell migration/proliferation, angiogenesis, and remodeling [4]. Specifically, secondary fracture healing progresses
D
through the following phases: (1) acute inflammatory response with mesenchymal stem cell recruitment,
TE
proliferation, and differentiation, (2) production of a cartilaginous fracture callus, (3) ossification and revascularization of the fracture callus, and (4) remodeling of the deposited trabecular bone [4, 5]. This process
AC CE P
also contributes to osseous recovery in surgical procedures including arthrodesis, osteotomies, and bone grafting and therefore has a wide impact on the success of most orthopaedic procedures. A significant body of basic science literature offers valuable insight into the role of the targeted chemical messengers (cytokines) in secondary bone healing. Of specific interest is Tumor Necrosis Factor-alpha (TNF-α), which is recognized as a primary mediator in the inflammatory reaction that initiates the reparative process. After injury, TNF-α is elevated in a bimodal temporal distribution within the fracture site. The peri-fracture concentrations are elevated in the initial inflammatory phase, then diminish to normal physiologic levels for the majority of the reparative process, but are later elevated in the remodeling phase [6]. Deviations from this schedule may produce inferior or more robust repair tissues [7, 8]. The interaction between TNF-α and bone is not limited to reparative states, but also extends to normal physiologic homeostasis. Poor bone integrity common in elderly, osteoporotic, diabetic, or obese patients as well as in individuals abusing alcohol or those on chronic doses of corticosteroids may arise from local and systemic deviations from the physiologic TNF-α equilibrium. The development of atrophic nonunions may be driven by interruptions in the
ACCEPTED MANUSCRIPT spatial and temporal patterns of elevated TNF-α concentrations within the fracture site. This review summarizes the role of TNF-α at specific time points along the course of fracture healing, presents evidence suggesting its role
PT
in common metabolic bone diseases, and speculates on the likely physiological responses to TNF-α modulation for both. Understanding the role of TNF-α in physiologic response to injury is important as it provides the foundation
RI
for exploration of potential treatments that facilitate recovery from injury or surgery as well potential
SC
interventions in metabolic bone disease [9]. Bone Healing Progression:
NU
The stages of bone healing and the role of TNF-α at each stage are summarized in Figure 1 and Table 1. Days 1-3: Acute Inflammatory Response and Mesenchymal Cell Recruitment (Elevated TNF-α)
MA
Disruption of the osseous cortex tears periosteal and endochondral vessels, creating a peri-fracture hematoma. Cytokines, fibrin, platelets, and cellular degradation products within the collected blood induce an
D
inflammatory reaction [6, 10] and stimulate angiogenesis [11]. Removal of the collected blood between the second
TE
and fourth day following injury results in mechanically inferior bone quality four weeks later [12], suggesting that factors within the hematoma are osteogenic. Furthermore, the hematoma itself does not contain the osteogenic
AC CE P
cell precursors, but rather stimulates cells within the periosteum to initiate endochondral ossification [13]. TNF-α is one of the initial inflammatory cytokines within the hematoma and is first produced by resident macrophages and recruited inflammatory cells [6]. Its concentration peaks at 24 hours and resolves to baseline levels within 72 hours after the injury [6, 14-17]. The TNF-α receptors p55 and p75 are found on immune cells during the early inflammatory phases and are identified on mesenchymal and osteoblastic cells during later healing stages [6]. Defining a fracture, arthrodesis, or bone grafting site as a physical space between the two osseous structures implies a relative deficiency of osteoclast and osteoblast cells within the gap to facilitate sufficient bone production [18]. Therefore, reconstituting the empty interval with sufficient concentrations of osteogenic cells and supporting vascular structures is crucial to the long-term success of the healing process. Within this context, elevated TNF-α concentrations in the early stages of fracture healing are vital as they initiate the recruitment of mesenchymal cells from surrounding soft tissue, bone cortex, bone marrow, periosteum, and blood with osteogenic and chondrogenic potential [19-21]. A major, but often forgotten, mesenchymal stem cell reserve is the surrounding muscle tissue. Fractures that are exposed to skeletal muscle demonstrate significantly better
ACCEPTED MANUSCRIPT osteosynthesis than those exposed to fasciocutaneous tissue only [22]. Further evidence to suggest a robust reserve of skeletal muscle mesenchymal stem cells comes from studies that demonstrate myoblast differentiation
PT
into osteoblasts following bone morphogenetic protein-1 (BMP-1) and bone morphogenetic protein-2 (BMP-2) [22, 23].
RI
TNF-α enhances mesenchymal cell recruitment by stimulating expression of intercellular adhesion
SC
molecule-1 (ICAM-1) and vascular cell adhesion protein-1 (VCAM-1), which optimize the intrinsic migration capacity [24]. TNF-α also induces proliferation and differentiation of mesenchymal precursor cells into
NU
chondrogenic and osteogenic phenotypes [20, 25] by inducing early BMP-2 secretion in surrounding osteoblasts, which promotes osteoblastic differentiation in mesenchymal precursor cells [17, 26-28].
MA
By stimulating mesenchymal cell recruitment and mesenchymal stem cell osteochondral differentiation, TNF-α provides the foundation for bone healing as its presence (controlling for other variables) determines the
D
quantity and phenotypic quality of the cells that initiate and maintain the healing process [19, 29]. Insufficient
TE
levels of intra-fracture recruitment cytokines likely diminish the mesenchymal stem cell population within the osseous gap and predispose the injury or surgical site to blunted callus formation ultimately resulting in inferior
AC CE P
bone healing capacity [30]. Platelets within the initial hematoma increase local transforming growth factor-beta (TGF-β) concentrations within the injury site [31, 32], which has been shown to increase mechanical strength and callus formation, [33, 34] due to stimulation of proliferation and osteochondrogenic differentiation of mesenchymal stem cells [35, 36]. All the molecular pathways involved with bone healing have not been identified or described and the activity of cytokines such as TGF- β provides an explanation for the continued, but inferior, fracture healing in the study animal subjected to TNF-α inhibition [37]. Broadly suppressing the immune response with systemic glucocorticoid administration during the inflammatory phase has been shown to diminish fracture healing in a rabbit model [38], despite having a slightly stimulatory effect on osteoblast activity [39]. This finding suggests that inflammation promotes the activation of the osseous healing process. Specific inhibition of TNF-α activity with tumor necrosis factor receptor 1 (TNFR1) and TNFR2 murine knockout models results in delayed mesenchymal cell differentiation, a greater proportion of disorganized mesenchymal tissue, and decreased early osteoblast migration into the marrow space [8, 29]. Although it seems intuitive that TNF-α inhibition may exert its effect by blunting the inflammatory response,
ACCEPTED MANUSCRIPT evidence from a murine model suggests that introduction of the TNF-α inhibitor etanercept did not decrease migration of immune cells or suppress the elevation of TNF-α or interleukin-1 beta (IL-1β) at the fracture site, but
PT
that it did suppress mesenchymal cell infiltration, proliferation, and fibroblast growth factor-2 (FGF-2) expression [20]. Therefore, it seems that TNF-α inhibition exerts its negative influence on bone healing by influencing the
RI
response of mesenchymal cells, rather than inflammatory cells to TNF-α. Conversely, supplemental TNF-α on post
SC
fracture days one and two has been demonstrated to stimulate a more robust healing response with an accelerated course and higher rates of callus mineralization [40]. This finding may be related to TNF-α stimulation
NU
of chondrogenic differentiation of progenitor cells through the nuclear factor kappa B (NF-κB)/p65 pathway, which facilitates SRY (sex determining region Y)-box 9 (Sox9) expression [41] or through other previously described TNF-α
MA
mediated mechanisms. Day 3-7: Fracture Callus Production
D
The third day following osseous injury marks the transition from a focus on mesenchymal cell precursor
TE
recruitment to production of a cartilage scaffold that bridges the fracture gap between two dissociated bone pieces. Effects from the initial inflammatory response peak at twenty-four hours following injury, but continue
AC CE P
with diminishing activity until the seventh day, overlapping with the cartilage callus formation phase [42]. Bridging osseous tissue extends from the development of hypertrophic chondrocyte populations arising from the subperiosteum as well as from differentiated mesenchymal precursors that have migrated into the injury site [6]. The production of a cartilage lattice that bridges the fracture gap is preceded by an elevation in type I and type II collagen messenger RNA (mRNA) and suppression of TNF-α and IL-1β concentrations [6]. If the TNF-α and IL-1β levels are not decreased as intended within this window, there is a synergistic interaction between the two cytokines that creates a diffuse impairment of chondrocyte proliferation and increased rates of chondrocyte apoptosis [43]. Day 7-14: Ossification and Revascularization of the Fracture Callus The relative quantity of ossified cartilage callus, callus volume, and level of type I collagen mRNA expression climaxes between days seven and fourteen following fracture, which also coincides with the initiation of cartilage resorption and the peak number of hypertrophied chondrocytes [6, 19]. The rate of longitudinal bone growth has been shown to have a linear relationship with the final relative volume of the hypertrophied
ACCEPTED MANUSCRIPT chondrocytes in both murine and porcine models [44]. Hypertrophic chondrocytes have high intra-cellular concentrations of TNF-α [45], which has been suggested to be an initiator of the hypertrophic phenotype [29].
PT
Transition from fracture callus to bone requires chondrocyte apoptosis, cartilage matrix degradation, stromal cell recruitment and osteogenic differentiation, as well as neo-vascularization [46]. TNF-α facilitates this
RI
process by inducing hypertrophic and non-hypertrophic chondrocyte apoptosis in a dose dependent fashion [47,
SC
48] by initiating a cascade of matrix metalloproteinase (MMP) activation (including MMP 9 and 14) [46] and by stimulating angiogenesis. Programmed subtraction of the temporary cartilage matrix allows endothelial cells to
NU
penetrate the callus and facilitate transition to immature bone through neovascularization [16, 49]. TNF-α also facilitates angiogenesis through enhancement of vascular endothelial growth factor (VEGF) production at the
MA
injury site [46]. Further evidence of the role of TNF-α in this stage of osteosynthesis is reported from a murine model with TNFR1 and TNFR2 knockout mice, which demonstrated fewer hypertrophic chondrocytes, delayed
D
endochondral tissue resorption secondary to impairment of chondrocyte apoptosis, and the delayed release of
TE
cytokines mediating endochondral tissue remodeling [8, 29]. Day 14 - 21: Substitution of Ossified Fracture Callus with Trabecular Bone
AC CE P
By the twenty-first day following fracture, all callus has normally been resorbed and replaced with trabecular bone. TNFR deficient mice have delayed callus resorption with very little trabecular bone formed and have a larger percentage of undifferentiated mesenchymal cells and a larger fracture callus [8] at this point in the healing process. TNF-α enhances human osteoblast proliferation in small doses applied for brief intervals, but prolonged exposure in large doses inhibits osteoblasts, likely through down-regulation of sialoprotein and osteocalcin gene expression [17, 50]. Day 28 and After: Remodeling of Deposited Trabecular Bone IL-1β and TNF-α are elevated during remodeling [31], which begins during the interval between day 21 and day 28 [6]. TNF-α has been demonstrated to increase the number of chondrocytes that secrete receptor activator of nuclear factor kappa-B ligand (RANKL) [45] and enhance trans-endothelial migration of CD14+ monocytes that can differentiate into osteoclasts [51]. These two mechanisms allow TNF-α to increase both the bone resorption capacity of established osteoclasts and potentially increase the total number of osteoclasts within a specific area to encourage the bone remodeling process.
ACCEPTED MANUSCRIPT The basic science evidence on manipulation of the peri-fracture TNF-α levels and the effect on bone healing is presented in Table 2.
PT
TNF-α and Fracture Nonunions: Fracture nonunions can be broadly classified as having either a biologic or mechanical etiology, which
RI
have characteristic radiographic presentations. Hypertrophic nonunions are identified by an excess of callus over
SC
the fracture site and occur from mechanical instability or hypermobility between the two dissociated bone ends. The innate healing response is adequate to initiate the formation of repair tissue, but the production of mature
NU
bone is never achieved. As the entire process of mesenchymal cell recruitment is initiated by the fracture hematoma, inter-fragmentary movements disrupting the emerging vascular network may generate smaller
MA
hematomas within the repair tissue that re-initiate the process of mesenchymal stem cell recruitment and differentiation. Mesenchymal cells recovered from hypertrophic nonunion calluses do demonstrate the capacity to
D
differentiate into osteogenic and chondrogenic cells, supporting the concept that the fracture callus in this
TE
situation can serve as a mesenchymal stem cell reservoir [52]. Therefore, at any time point along the course of attempted osteosynthesis, a hypertrophic nonunion may have several centers of inflammation driving multiple
AC CE P
miniature sites of mesenchymal stem cell recruitment and differentiation, resulting in a cumulatively larger than necessary immature fracture callus volume. Oligotrophic and atrophic nonunions are the result of an intrinsic deficit in the host response to injury, which is manifested by an absence or inadequate volume of repair tissue. It was previously thought that oligotrophic and atrophic nonunions were a symptom of inadequate supporting blood supply [53, 54]. This theory has been confirmed in a macroscopic context where fractures occurring in limbs with vascular injury have a higher risk of nonunion [55]; however, recent studies suggest that the effect of deficient blood supply is less significant at the cellular level. Fractures that progress to form nonunions do have fewer blood vessels infiltrating the repair tissue during the eight weeks following injury compared to tissues with normal osteosynthesis, but there is no significant difference beyond this time point [56]. This same observation was also found in patients undergoing revision surgery for hypertrophic and atrophic nonunions, which failed to demonstrate any significant difference in vessel density within the repair tissue [57]. Therefore, as the relative lack of local vascularity associated with atrophic nonunions appears only in the very early stages of fracture healing, this deficit may arise from an absence
ACCEPTED MANUSCRIPT of local angiogenesis stimulation rather than a lack of supporting vasculature external to the callus. Furthermore, evidence suggests that atrophic nonunions may have diminished growth secondary to decreased concentrations of
PT
growth factors that prime the tissue surrounding the fracture site [58]. This finding is consistent with the theory describing a regional acceleratory phenomenon (RAP), suggesting there is an activity level and cell quantity
RI
threshold that produces a sustained osteosynthesis response when achieved [30]. If the undefined critical mass is
SC
not established, the fracture will not maintain enough momentum to carry the premature repair tissue to mature trabecular bone. As the number and phenotype of chondro- and osteogenic mesenchymal cells is partially
NU
determined by the concentration of TNF-α within the fracture site during the first three days following injury, this cytokine is implicated in the development of biological nonunions. Therefore, relative TNF-α deficiency may be the
MA
principal influence in the development of oligotrophic and atrophic nonunions. Infection is another frequent nonunion etiology. This process impedes fracture healing through physical
D
disruption of the healing cascade, requiring a “resetting” of the entire twenty-eight day process for significant
TE
osteosynthesis to progress within the confines of the fracture site. Furthermore, the presence of bacteria stimulates a local immune response, which interferes with the physiologic bimodal temporal pattern of elevated
AC CE P
TNF-α. During the acute stages of osteomyelitis, the local TNF-α concentration is elevated as an immune response to bacteria within the fracture site [59-61]. This pathological elevation of TNF-α may actually stimulate the migration of mesenchymal stem cells and differentiation along the osteogenic/chondrogenic cell lineages within the first three days following fracture. However, persistently elevated TNF-α beyond the first three days decreases the cell proliferation rate and induces apoptosis in the chondrocytes that drive the process of endochondral ossification [43]. Furthermore, biofilm formation over the bone surface and the subsequently elevated TNF-α promotes osteoclastogenesis [62], which disrupts the equilibrium between bone resorption and production. This explanation could support the observation of local osteopenia that develops in cases in mature bones that are afflicted with osteomyelitis and may create a scenario where bone is resorbed as it is being synthesized with a reduced net accumulation. Metabolic Bone Disease and TNF-α: A majority of studies examining the role of TNF-α have focused on the effect of inhibition at various stages of bone healing, which is clinically relevant due to the introduction of disease-modifying anti-rheumatic
ACCEPTED MANUSCRIPT drugs (DMARDs) in rheumatoid arthritis care. A more subtle, but likely broader clinical issue for orthopaedic surgeons arises from the effect of systemic, chronic supra-physiologic TNF-α levels that are associated with
PT
obesity, diabetes, alcohol abuse, and advanced age. A murine fracture model demonstrated that persistently elevated TNF-α levels lead to a higher nonunion rate, impaired cartilage fracture callus formation, and fracture
RI
bridging with non-osteogenic fibrous tissue, which is consistent with impaired mesenchymal cell differentiation
SC
into a chondroblast phenotype [63]. Another murine study demonstrated that, given over seven days, supplemental TNF-α and IL-1β significantly impaired bone growth and that this effect was reversed with
NU
administration of anakinra (anti-IL-1) and etanercept (anti-TNF-α) [64]. Although this investigation was carried out in fetal rats, there is considerable crossover between the orchestration of fracture healing and initial long bone
MA
formation as both advance through endochondral ossification. Furthermore, after the initial phase of fracture healing, continuous application of TNF-α down-regulates the expression of sialoprotein and osteocalcin, which are
D
proxy indicators of increased bone formation [17]. From these data, it can be inferred that any process resulting in
TE
a persistently elevated TNF-α level has the potential to suppress secondary bone healing. The likely mechanism behind this observation is the diffuse impairment of chondrocyte proliferation and increased rates of chondrocyte
AC CE P
apoptosis seen with elevated TNF-α levels during the production of the bridging fracture callus between the third and seventh day following injury [43]. Osteoporosis:
-/-
Work by Li et al. [65] demonstrated that RANK nullizygous (RANK ) mice failed to demonstrate an osteoclastogenic response to exogenous TNF-α, as well as parathyroid hormone-related protein (PTHrP), 1,25 dihydroxyvitamin D [1α,25-(OH)2D3], and IL1-β, demonstrating that RANK is the central mediator of bone homeostasis. In this context, it is clear that TNF-α is a mediator of baseline bony homeostasis and is able to promote osteoclast activity through stimulation of the RANK pathway [66, 67]. Estrogen withdrawal associated with menopause prompts increased macrophage secretion of IL-12 and IL-18, inducing activation of T-cells that secrete TNF-α [68], resulting in elevated serum concentrations. Serum isolates from post-menopausal females without estrogen replacement therapy demonstrated higher levels of TNF-α and higher bone resorption activity, which was reversed with TNF-α inhibition [69]. Similar work has demonstrated that administration of etanercept in post-menopausal women significantly reduced bone turnover following withdrawal from supplemental estrogen
ACCEPTED MANUSCRIPT [70]. Further emphasis of the estrogen/TNF-α relationship arises from the observation that the bone resorption characteristics found in post-menopausal females treated with estrogen therapy were similar to premenopausal
PT
subjects. Estradiol suppresses TNF-α activity by inhibiting macrophage NF-kB signaling through kappaB-Ras2 expression [71, 72]. As osteoclasts and macrophages differentiate from the same precursor lineage and inhibition
RI
of osteoclast NF-kB pathways freezes osteoclast activity [73], estrogen depletion may remove a negative stimulus
SC
for osteoclast activity.
In addition to baseline decreases in bone mineral density, which predispose post-menopausal females to
NU
an increased fracture risk, estrogen depletion is also associated with a blunted reparative response to osseous injury. A murine model of osteoporosis demonstrated that the repair tissue had a higher ratio of undifferentiated
MA
mesenchymal cells, a less robust chondrogenic response, and a weaker callus at seven days after injury [74]. At 14 and 21 days following the injury, the osteoporotic mice had delayed bone formation and also had a greater
D
presence of immature woven, rather than lamellar bone, compared to controls. Despite the inferior response, the
TE
osteoporotic mice were also found to have an elevated concentration of BMP-2, which is likely explained by the
2 [17, 26, 27]. Obesity:
AC CE P
observation that TNF-α, which is pathologically elevated at baseline, promotes the synthesis and secretion of BMP-
Wolff’s law predicts that increased body mass augments the mechanical load across the axial skeleton and lower extremities, which would result in denser bones with superior strength. However, chronic inflammation associated with obesity may promote osteoclastogenesis through up-regulation of the NF-κB (RANK)/RANK pathway [75] via increased levels of TNF-α secreted by excess adipose tissue [76, 77]. Advanced Age: Although associated with advanced age, osteoporosis is separate from age related deficits in osteosynthesis following fracture repair, which is likely realized from an innate reduction in the robustness of the mesenchymal stem cells. A murine model comparing histological and molecular parameters in juvenile, middle age, and elderly mice demonstrated that there was a significant reduction in reparative capacity between juvenile and middle age mice, but not between middle age and elderly [78]. Specifically, the juvenile mice had earlier chondrocyte expression of collagen type II and type X, earlier osteoblast expression of osteocalcin, earlier vascular
ACCEPTED MANUSCRIPT infiltration, and earlier endochondral ossification. Furthermore, there was no significant difference in the quantity of cartilage produced between the juvenile mice and the middle age/elderly mice, but there was significantly less
PT
bone formed by the older two groups [78]. A rabbit study investigated the chondrogenic potential of explanted periosteum from subjects at different ages and found a significant decrease in the ability to stimulate periosteal
RI
progenitor cells into chondral producing phenotypes once skeletal maturity was achieved [79]. Specimens from
SC
skeletally mature subjects also had less robust cartilage production, decreased cartilage substrate uptake, and decreased cellularity within the periosteal cambium layer, but not a decreased percentage of proliferating cells
NU
[79]. Lastly, a murine model observed that older mice were found to have elevated serum concentrations of IL-6 and TNF-α at baseline when compared to younger controls [7], which may suppress the inherent bone healing
MA
ability in similar mechanisms to obesity and osteoporosis. This may explain the observation that higher proportions of osteoclasts are found within the fracture site in older subjects which likely explains the decreased callus
D
mineralization and bridging capacity following fracture [80].
TE
Diabetes:
Clinical studies have demonstrated that both insulin-dependent and non-insulin-dependent diabetics have
AC CE P
prolonged healing times following closed displaced fractures, but not in non-displaced fractures [81], suggesting that secondary bone healing is less robust in this patient population. Induced diabetes in a murine model led to an elevated TNF-α level compared to baseline, which was associated with an elevated level of chondrocyte proapoptotic mRNA and a smaller cartilage callus within an induced fracture healing site when compared to nondiabetic control subjects [47]. Diabetic mice were also shown to have enhanced chondrocyte apoptosis after the sixteenth day following injury [47], which is consistent with persistently and inappropriately elevated levels of TNFα during this stage of fracture healing. It was further demonstrated that TNF-α inhibition lead to a reversal of these findings and re-established the callus forming capacity in diabetic mice and that TNF-α inhibition in normoglycemic subjects had no impact on the callus formation [47]. Another murine model demonstrated that uncontrolled blood glucose resulted in decreased fracture callus strength, decreased collagen concentration, and decreased cellularity, but that the callus present in diabetic subjects with blood sugars controlled with insulin therapy was not significantly different than controls [82].
ACCEPTED MANUSCRIPT A similar murine model also demonstrated that diabetes was associated with an increased rate of osteoclastogenesis, which was inhibited by TNF-α inhibition and that subjects with induced diabetes continued to
PT
have significantly less fracture callus when compared to controls despite TNF-α inhibition [45] Chronic Alcohol Consumption:
RI
A murine model demonstrated that distraction osteogenesis was inhibited by ethanol consumption and
SC
that consistent administration of anti-IL-1 and anti-TNFR1 antibodies sufficiently increased the healing potential compared to control [83].
NU
Aseptic Loosening:
Prosthesis derived wear debris, titanium, and cement from total hip replacements has been shown to
MA
induce local osteoclast bone resorption in a dose dependent relationship due to their ability to increase TNF-α concentrations following particle phagocytosis by resident macrophages [84-86]. Furthermore, in vitro and in vivo
Conclusion:
TE
induced by titanium particle debris [87].
D
models have demonstrated that TNF-α inhibition with etanercept reduces osteoclast driven bone resorption
AC CE P
The morbidity and mortality associated with osteoporotic fragility fractures is staggering and emphasizes the importance of early surgical intervention and mobilization of these patients [88, 89]. As the quality of bone cannot be improved in the acute setting, enhancing the innate healing response to the fracture may reduce the incidence of nonunion and provide a more stable construct earlier in the postoperative course. Besides the benefit of an early recovery, enhancing secondary bone healing may reduce the incidence of complications and increased mortality rates associated with osteoporotic fragility fractures [90]. Despite an abundant body of basic science literature implicating its role as a primary mediator of fracture healing and metabolic bone disease, the role of TNF-α seems clinically underappreciated. The consistent themes of TNF-α regulating mesenchymal stem cell repopulation within a fracture gap and the ability of pathologically elevated TNF-α concentrations to enhance osteoclastogenesis are promising targets for TNF-α immunotherapy and may be the scientific foundation for future osteoporosis management interventions.
ACCEPTED MANUSCRIPT References:
8. 9. 10. 11. 12. 13. 14.
15. 16. 17.
18. 19. 20. 21.
22. 23. 24.
PT
RI
SC
7.
NU
5. 6.
MA
4.
D
3.
TE
2.
Schell, H., et al., The course of bone healing is influenced by the initial shear fixation stability. J Orthop Res, 2005. 23(5): p. 1022-8. Klein, P., et al., The initial phase of fracture healing is specifically sensitive to mechanical conditions. J Orthop Res, 2003. 21(4): p. 662-9. Lienau, J., et al., Initial vascularization and tissue differentiation are influenced by fixation stability. J Orthop Res, 2005. 23(3): p. 639-45. Thornton, F.J., M.R. Schaffer, and A. Barbul, Wound healing in sepsis and trauma. Shock, 1997. 8(6): p. 391-401. Marsell, R. and T.A. Einhorn, The biology of fracture healing. Injury, 2011. 42(6): p. 551-5. Kon, T., et al., Expression of osteoprotegerin, receptor activator of NF-kappaB ligand (osteoprotegerin ligand) and related proinflammatory cytokines during fracture healing. J Bone Miner Res, 2001. 16(6): p. 1004-14. Wahl, E.C., et al., Restoration of regenerative osteoblastogenesis in aged mice: modulation of TNF. J Bone Miner Res, 2010. 25(1): p. 114-23. Gerstenfeld, L.C., et al., Impaired fracture healing in the absence of TNF-alpha signaling: the role of TNFalpha in endochondral cartilage resorption. J Bone Miner Res, 2003. 18(9): p. 1584-92. Sandberg, M.M., H.T. Aro, and E.I. Vuorio, Gene expression during bone repair. Clin Orthop Relat Res, 1993(289): p. 292-312. Einhorn, T.A., et al., The expression of cytokine activity by fracture callus. J Bone Miner Res, 1995. 10(8): p. 1272-81. Street, J., et al., Is human fracture hematoma inherently angiogenic? Clin Orthop Relat Res, 2000(378): p. 224-37. Grundnes, O. and O. Reikeras, The importance of the hematoma for fracture healing in rats. Acta Orthop Scand, 1993. 64(3): p. 340-2. Mizuno, K., et al., The osteogenetic potential of fracture haematoma. Subperiosteal and intramuscular transplantation of the haematoma. J Bone Joint Surg Br, 1990. 72(5): p. 822-9. Caetano-Lopes, J., et al., Upregulation of inflammatory genes and downregulation of sclerostin gene expression are key elements in the early phase of fragility fracture healing. PLoS One, 2011. 6(2): p. e16947. Currie, H.N., et al., Spatial cytokine distribution following traumatic injury. Cytokine, 2014. 66(2): p. 112-8. Gerstenfeld, L.C., et al., Fracture healing as a post-natal developmental process: molecular, spatial, and temporal aspects of its regulation. J Cell Biochem, 2003. 88(5): p. 873-84. Lu, Z., et al., Short-term exposure to tumor necrosis factor-alpha enables human osteoblasts to direct adipose tissue-derived mesenchymal stem cells into osteogenic differentiation. Stem Cells Dev, 2012. 21(13): p. 2420-9. Frost, H.M., The biology of fracture healing. An overview for clinicians. Part I. Clin Orthop Relat Res, 1989(248): p. 283-93. Einhorn, T.A., The cell and molecular biology of fracture healing. Clin Orthop Relat Res, 1998(355 Suppl): p. S7-21. Zhou, F.H., et al., TNF-alpha mediates p38 MAP kinase activation and negatively regulates bone formation at the injured growth plate in rats. J Bone Miner Res, 2006. 21(7): p. 1075-88. Mountziaris, P.M., S.N. Tzouanas, and A.G. Mikos, Dose effect of tumor necrosis factor-alpha on in vitro osteogenic differentiation of mesenchymal stem cells on biodegradable polymeric microfiber scaffolds. Biomaterials, 2010. 31(7): p. 1666-75. Harry, L.E., et al., Comparison of the healing of open tibial fractures covered with either muscle or fasciocutaneous tissue in a murine model. J Orthop Res, 2008. 26(9): p. 1238-44. Khouri, R.K., et al., Repair of calvarial defects with flap tissue: role of bone morphogenetic proteins and competent responding tissues. Plast Reconstr Surg, 1996. 98(1): p. 103-9. Fu, X., et al., Migration of bone marrow-derived mesenchymal stem cells induced by tumor necrosis factoralpha and its possible role in wound healing. Wound Repair Regen, 2009. 17(2): p. 185-91.
AC CE P
1.
ACCEPTED MANUSCRIPT
31. 32. 33. 34. 35.
36. 37.
38. 39. 40. 41. 42.
43. 44. 45. 46.
47. 48.
PT
RI
SC
30.
NU
29.
MA
28.
D
27.
TE
26.
Zubov, D.O., [Immunoregulatory role of mesenchymal stem cells in bone reparation processes]. Fiziol Zh, 2008. 54(4): p. 30-6. Pelissier, P., et al., Induced membranes secrete growth factors including vascular and osteoinductive factors and could stimulate bone regeneration. J Orthop Res, 2004. 22(1): p. 73-9. Eguchi, Y., et al., Antitumor necrotic factor agent promotes BMP-2-induced ectopic bone formation. J Bone Miner Metab, 2010. 28(2): p. 157-64. Zhao, Y.P., et al., The promotion of bone healing by progranulin, a downstream molecule of BMP-2, through interacting with TNF/TNFR signaling. Biomaterials, 2013. 34(27): p. 6412-21. Gerstenfeld, L.C., et al., Impaired intramembranous bone formation during bone repair in the absence of tumor necrosis factor-alpha signaling. Cells Tissues Organs, 2001. 169(3): p. 285-94. Frost, H.M., The biology of fracture healing. An overview for clinicians. Part II. Clin Orthop Relat Res, 1989(248): p. 294-309. Mountziaris, P.M. and A.G. Mikos, Modulation of the inflammatory response for enhanced bone tissue regeneration. Tissue Eng Part B Rev, 2008. 14(2): p. 179-86. Bonewald, L.F. and G.R. Mundy, Role of transforming growth factor-beta in bone remodeling. Clin Orthop Relat Res, 1990(250): p. 261-76. Lind, M., et al., Transforming growth factor-beta enhances fracture healing in rabbit tibiae. Acta Orthop Scand, 1993. 64(5): p. 553-6. Nielsen, H.M., et al., Local injection of TGF-beta increases the strength of tibial fractures in the rat. Acta Orthop Scand, 1994. 65(1): p. 37-41. Iwasaki, M., et al., Regulation of proliferation and osteochondrogenic differentiation of periosteum-derived cells by transforming growth factor-beta and basic fibroblast growth factor. J Bone Joint Surg Am, 1995. 77(4): p. 543-54. Roelen, B.A. and P. Dijke, Controlling mesenchymal stem cell differentiation by TGFBeta family members. J Orthop Sci, 2003. 8(5): p. 740-8. Chen, Y.J., et al., Recruitment of mesenchymal stem cells and expression of TGF-beta 1 and VEGF in the early stage of shock wave-promoted bone regeneration of segmental defect in rats. J Orthop Res, 2004. 22(3): p. 526-34. Waters, R.V., et al., Systemic corticosteroids inhibit bone healing in a rabbit ulnar osteotomy model. Acta Orthop Scand, 2000. 71(3): p. 316-21. Malviya, A., et al., The effect of newer anti-rheumatic drugs on osteogenic cell proliferation: an in-vitro study. J Orthop Surg Res, 2009. 4: p. 17. Glass, G.E., et al., TNF-alpha promotes fracture repair by augmenting the recruitment and differentiation of muscle-derived stromal cells. Proc Natl Acad Sci U S A, 2011. 108(4): p. 1585-90. Caron, M.M., et al., Activation of NF-kappaB/p65 facilitates early chondrogenic differentiation during endochondral ossification. PLoS One, 2012. 7(3): p. e33467. Cho, T.J., L.C. Gerstenfeld, and T.A. Einhorn, Differential temporal expression of members of the transforming growth factor beta superfamily during murine fracture healing. J Bone Miner Res, 2002. 17(3): p. 513-20. Martensson, K., D. Chrysis, and L. Savendahl, Interleukin-1beta and TNF-alpha act in synergy to inhibit longitudinal growth in fetal rat metatarsal bones. J Bone Miner Res, 2004. 19(11): p. 1805-12. Breur, G.J., et al., Linear relationship between the volume of hypertrophic chondrocytes and the rate of longitudinal bone growth in growth plates. J Orthop Res, 1991. 9(3): p. 348-59. Alblowi, J., et al., High levels of tumor necrosis factor-alpha contribute to accelerated loss of cartilage in diabetic fracture healing. Am J Pathol, 2009. 175(4): p. 1574-85. Lehmann, W., et al., Tumor necrosis factor alpha (TNF-alpha) coordinately regulates the expression of specific matrix metalloproteinases (MMPS) and angiogenic factors during fracture healing. Bone, 2005. 36(2): p. 300-10. Kayal, R.A., et al., TNF-alpha mediates diabetes-enhanced chondrocyte apoptosis during fracture healing and stimulates chondrocyte apoptosis through FOXO1. J Bone Miner Res, 2010. 25(7): p. 1604-15. Aizawa, T., et al., Induction of apoptosis in chondrocytes by tumor necrosis factor-alpha. J Orthop Res, 2001. 19(5): p. 785-96.
AC CE P
25.
ACCEPTED MANUSCRIPT
55. 56. 57. 58.
59. 60. 61. 62. 63. 64.
65. 66. 67. 68. 69. 70. 71. 72. 73.
PT
RI
SC
54.
NU
53.
MA
52.
D
51.
TE
50.
Ai-Aql, Z.S., et al., Molecular mechanisms controlling bone formation during fracture healing and distraction osteogenesis. J Dent Res, 2008. 87(2): p. 107-18. Frost, A., et al., Inflammatory cytokines regulate proliferation of cultured human osteoblasts. Acta Orthop Scand, 1997. 68(2): p. 91-6. Kindle, L., et al., Human microvascular endothelial cell activation by IL-1 and TNF-alpha stimulates the adhesion and transendothelial migration of circulating human CD14+ monocytes that develop with RANKL into functional osteoclasts. J Bone Miner Res, 2006. 21(2): p. 193-206. Iwakura, T., et al., Human hypertrophic nonunion tissue contains mesenchymal progenitor cells with multilineage capacity in vitro. J Orthop Res, 2009. 27(2): p. 208-15. Rhinelander, F.W., Tibial blood supply in relation to fracture healing. Clin Orthop Relat Res, 1974(105): p. 34-81. Arany, L., et al., Arteriographic studies in delayed-union and non-union of fractures. Radiol Diagn (Berl), 1980. 21(5): p. 673-81. Dickson, K., et al., Delayed unions and nonunions of open tibial fractures. Correlation with arteriography results. Clin Orthop Relat Res, 1994(302): p. 189-93. Reed, A.A., et al., Vascularity in a new model of atrophic nonunion. J Bone Joint Surg Br, 2003. 85(4): p. 604-10. Reed, A.A., et al., Human atrophic fracture non-unions are not avascular. J Orthop Res, 2002. 20(3): p. 593-9. Reed, A.A.C., et al., DO ATROPHIC NON-UNIONS HAVE LOWER LEVELS OF GROWTH FACTORS THAN HYPERTROPHIC NON-UNIONS AND HEALING FRACTURES? Journal of Bone & Joint Surgery, British Volume, 2002. 84-B(SUPP I): p. 19. Littlewood-Evans, A.J., et al., Local expression of tumor necrosis factor alpha in an experimental model of acute osteomyelitis in rats. Infect Immun, 1997. 65(8): p. 3438-43. Klosterhalfen, B., et al., Local and systemic inflammatory mediator release in patients with acute and chronic posttraumatic osteomyelitis. J Trauma, 1996. 40(3): p. 372-8. Vallejo, J.G., C.J. Baker, and M.S. Edwards, Roles of the bacterial cell wall and capsule in induction of tumor necrosis factor alpha by type III group B streptococci. Infect Immun, 1996. 64(12): p. 5042-6. Meghji, S., et al., Surface-associated protein from Staphylococcus aureus stimulates osteoclastogenesis: possible role in S. aureus-induced bone pathology. Br J Rheumatol, 1998. 37(10): p. 1095-101. Hashimoto, J., et al., Inhibitory effects of tumor necrosis factor alpha on fracture healing in rats. Bone, 1989. 10(6): p. 453-7. Fernandez-Vojvodich, P., E. Karimian, and L. Savendahl, The biologics anakinra and etanercept prevent cytokine-induced growth retardation in cultured fetal rat metatarsal bones. Horm Res Paediatr, 2011. 76(4): p. 278-85. Li, J., et al., RANK is the intrinsic hematopoietic cell surface receptor that controls osteoclastogenesis and regulation of bone mass and calcium metabolism. Proc Natl Acad Sci U S A, 2000. 97(4): p. 1566-71. Fuller, K., et al., TNFalpha potently activates osteoclasts, through a direct action independent of and strongly synergistic with RANKL. Endocrinology, 2002. 143(3): p. 1108-18. Lam, J., et al., TNF-alpha induces osteoclastogenesis by direct stimulation of macrophages exposed to permissive levels of RANK ligand. J Clin Invest, 2000. 106(12): p. 1481-8. Roggia, C., et al., Role of TNF-alpha producing T-cells in bone loss induced by estrogen deficiency. Minerva Med, 2004. 95(2): p. 125-32. Cohen-Solal, M.E., et al., Peripheral monocyte culture supernatants of menopausal women can induce bone resorption: involvement of cytokines. J Clin Endocrinol Metab, 1993. 77(6): p. 1648-53. Charatcharoenwitthaya, N., et al., Effect of blockade of TNF-alpha and interleukin-1 action on bone resorption in early postmenopausal women. J Bone Miner Res, 2007. 22(5): p. 724-9. Murphy, A.J., P.M. Guyre, and P.A. Pioli, Estradiol suppresses NF-kappa B activation through coordinated regulation of let-7a and miR-125b in primary human macrophages. J Immunol, 2010. 184(9): p. 5029-37. An, J., et al., Estradiol repression of tumor necrosis factor-alpha transcription requires estrogen receptor activation function-2 and is enhanced by coactivators. Proc Natl Acad Sci U S A, 1999. 96(26): p. 15161-6. Soysa, N.S. and N. Alles, NF-kappaB functions in osteoclasts. Biochem Biophys Res Commun, 2009. 378(1): p. 1-5.
AC CE P
49.
ACCEPTED MANUSCRIPT
81. 82. 83. 84. 85. 86. 87. 88. 89. 90.
PT
RI
SC
NU
80.
MA
78. 79.
D
77.
TE
75. 76.
Islam, A.A., et al., Healing of fractures in osteoporotic rat mandible shown by the expression of bone morphogenetic protein-2 and tumour necrosis factor-alpha. Br J Oral Maxillofac Surg, 2005. 43(5): p. 38391. Cao, J.J., Effects of obesity on bone metabolism. J Orthop Surg Res, 2011. 6: p. 30. Pedersen, M., et al., Circulating levels of TNF-alpha and IL-6-relation to truncal fat mass and muscle mass in healthy elderly individuals and in patients with type-2 diabetes. Mech Ageing Dev, 2003. 124(4): p. 495502. Bastard, J.P., et al., Recent advances in the relationship between obesity, inflammation, and insulin resistance. Eur Cytokine Netw, 2006. 17(1): p. 4-12. Lu, C., et al., Cellular basis for age-related changes in fracture repair. J Orthop Res, 2005. 23(6): p. 1300-7. O'Driscoll, S.W., et al., The chondrogenic potential of periosteum decreases with age. J Orthop Res, 2001. 19(1): p. 95-103. Mehta, M., et al., Influences of age and mechanical stability on volume, microstructure, and mineralization of the fracture callus during bone healing: is osteoclast activity the key to age-related impaired healing? Bone, 2010. 47(2): p. 219-28. Loder, R.T., The influence of diabetes mellitus on the healing of closed fractures. Clin Orthop Relat Res, 1988(232): p. 210-6. Macey, L.R., et al., Defects of early fracture-healing in experimental diabetes. J Bone Joint Surg Am, 1989. 71(5): p. 722-33. Perrien, D.S., et al., IL-1 and TNF antagonists prevent inhibition of fracture healing by ethanol in rats. Toxicol Sci, 2004. 82(2): p. 656-60. Schwarz, E.M., et al., Tumor necrosis factor-alpha/nuclear transcription factor-kappaB signaling in periprosthetic osteolysis. J Orthop Res, 2000. 18(3): p. 472-80. Merkel, K.D., et al., Tumor necrosis factor-alpha mediates orthopedic implant osteolysis. Am J Pathol, 1999. 154(1): p. 203-10. Algan, S.M., M. Purdon, and S.M. Horowitz, Role of tumor necrosis factor alpha in particulate-induced bone resorption. J Orthop Res, 1996. 14(1): p. 30-5. Childs, L.M., et al., Efficacy of etanercept for wear debris-induced osteolysis. J Bone Miner Res, 2001. 16(2): p. 338-47. Brauer, C.A., et al., Incidence and mortality of hip fractures in the United States. JAMA, 2009. 302(14): p. 1573-9. Tsuboi, M., et al., Mortality and mobility after hip fracture in Japan: a ten-year follow-up. J Bone Joint Surg Br, 2007. 89(4): p. 461-6. Melton, L.J., 3rd, et al., Predictors of excess mortality after fracture: a population-based cohort study. J Bone Miner Res, 2014. 29(7): p. 1681-90.
AC CE P
74.
Figure 1
AC CE P
TE
D
MA
NU
SC
RI
PT
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT Table 1. Illustrates the overlapping bone healing phases by the primary cellular process as well as step in healing cascade [4-6]. Step Acute Inflammation
TNF- α Elevated
Day 3-7
Cartilage Fracture Callus Production Callus Ossification and Revascularization
Baseline
RI
Elevated
Elevated to baseline
AC CE P
TE
D
MA
Day 28 and >
Substitution with Trabecular bone Remodeling
SC
Day 14 - 28
Baseline/Rising
NU
Day 7-14
Predominant Function Mesenchymal migration, proliferation, differentiation None, inhibits cartilage production Induces hypertrophic chondrocyte phenotype, angiogenesis, and MMP production Osteoblast differentiation
PT
Time Period Days 1-3
Osteoclastogenesis
ACCEPTED MANUSCRIPT Table 2. Evidence from basic science studies that have manipulated the amount or effect of peri-fracture TNF-α. Step Acute Inflammation
Study [12]
Days 1-3
Acute Inflammation
[38]
Days 1-3
Acute Inflammation
[8, 29]
Days 1-3
Acute Inflammation
[20]
Days 1-3
Acute Inflammation
[40]
Supplemental TNF-α
Day 3-7
Fracture Callus Production
[43]
Supplemental TNF-α
Day 7-14
Callus Ossification and Revascularization
Day 14 - 21
Substitution with Trabecular Bone
Remodeling
RI
SC
Etanercept administration at fracture site
NU
MA
D
TE
AC CE P
Day 28 and >
Study Peri-fracture hematoma evacuation Glucocorticoid administration TNFR1 and TNFR2 KO (murine model)
Effect Inferior repair tissue Diminished fracture healing
PT
Time Period Days 1-3
[8, 29].
TNFR1 and TNFR2 KO (murine model)
[8]
TNFR1 and TNFR2 KO (murine model)
X
X
Delayed mesenchymal cell differentiation, decreased osteoblast numbers in the fracture site Immune cell migration was not affected, mesenchymal cell migration into fracture site was diminished Accelerated fracture healing, higher callus mineralization rates Impaired chondrocyte proliferation and increased chondrocyte apoptosis Fewer hypertrophic chondrocytes, delayed cartilage resorption due to delayed chondrocyte apoptosis Larger fracture callus with greater undifferentiated mesenchymal cells, less trabecular bone formed X
ACCEPTED MANUSCRIPT Highlights:
Secondary fracture healing is an inflammatory process that relies on immune cytokines to repopulate the
Tumor Necrosis Factor-Alpha (TNF-α) is a central mediator in this inflammatory response in combination
RI
PT
fracture gap with chondrogenic and osteogenic mesenchymal stem cells.
with the host reserve of mesenchymal stem cells. TNF-α is elevated in a bimodal temporal distribution
Disease states associated with poor bone healing such as osteoporosis and diabetes have abnormal TNF-α
NU
SC
and deviations from this schedule produce a diminished healing response.
responses to injury, which likely increases the risk of delayed and nonunion in these patient populations.
MA
Appreciation of TNF-α’s role in secondary bone healing may lead to new therapies to augment fracture
TE
D
healing and reduce the morbidity associated with protracted recovery.
AC CE P
