The authors state that they have no conflicts of interest.
A closed femur fracture pain model was developed in the C57BL/6J mouse. One day after fracture, a monoclonal antibody raised against nerve growth factor (anti-NGF) was delivered intraperitoneally and resulted in a reduction in fracture pain-related behaviors of ∼50%. Anti-NGF therapy did not interfere with bone healing as assessed by mechanical testing and histomorphometric analysis.
Introduction: Current therapies to treat skeletal fracture pain are limited. This is because of the side effect profile of available analgesics and the scarcity of animal models that can be used to understand the mechanisms that drive this pain. Whereas previous studies have shown that mineralized bone, marrow, and periosteum are innervated by sensory and sympathetic fibers, it is not understood how skeletal pain is generated and maintained even in common conditions such as osteoarthritis, low back pain, or fracture.
Materials and Methods: In this study, we characterized the pain-related behaviors after a closed femur fracture in the C57BL/6J mouse. Additionally, we assessed the effect of a monoclonal antibody that binds to and sequesters nerve growth factor (anti-NGF) on pain-related behaviors and bone healing (mechanical properties and histomorphometric analysis) after fracture.
Results: Administration of anti-NGF therapy (10 mg/kg, days 1, 6, and 11 after fracture) resulted in a reduction of fracture pain-related behaviors of ∼50%. Attenuation of fracture pain was evident as early as 24 h after the initial dosing and remained efficacious throughout the course of fracture pain. Anti-NGF therapy did not modify biomechanical properties of the femur or histomorphometric indices of bone healing.
Conclusions: These findings suggest that therapies that target NGF or its cognate receptor(s) may be effective in attenuating nonmalignant fracture pain without interfering with bone healing.
Skeletal fracture pain can have a significant impact on quality of life and functionality of the affected individual.(1) In the young population, fracture pain frequently results from sports injuries, car accidents, improvised exploding devices, or landmines.(2,3) In this patient population, fracture pain can be difficult to control, interfere with effective rehabilitation, and may lead to the development of complex regional pain syndrome.(4,5) Skeletal fracture pain also impacts the elderly; painful osteoporotic fractures of the vertebrae, hip, or femur are common in both postmenopausal women and older men.(6–10) Pain caused by osteoporotic fractures of the hip and femur greatly reduces mobility resulting in decreased bone and muscle mass, which leads to a marked increase in morbidity and mortality.(11–13)
The relative paucity of therapies to treat fracture pain is in large part caused by the scarcity of animal models that closely mirror the human condition, which would allow the development of mechanism-based therapies to treat this pain. Several studies have shown mineralized bone, marrow, and periosteum receive a significant innervation by primary afferent sensory neurons(14) and postganglionic sympathetic neurons.(15–17) However, what remains unknown is how sensory neurons that innervate the skeleton are activated in common nonmalignant skeletal pain conditions such as osteoarthritis, low back pain, or fracture.
The closed femur fracture model was originally developed in the rat(18) and subsequently in the mouse(19) to explore the factors that contribute to bone healing.(20–22) In this study, we adapted this model in the C57BL/6J (C57) mouse strain and simultaneously examined indices of skeletal pain and bone healing after fracture. One reason the C57 mouse was chosen is that many genetically modified mice have been developed in a C57 background,(23) and these mice may be useful in understanding the factors that generate and maintain skeletal pain. Additionally, C57 mice have a greater reduction in BMD with age compared with many other strains of mice,(24) which suggests they may be an effective strain for examining pain and bone healing after osteoporotic fracture.(25,26) In this study, we show a stereotypic progression of skeletal remodeling and pain-related behaviors after fracture of the femur in young adult C57 mice. Administration of anti-nerve growth factor (NGF) therapy attenuated fracture-induced pain without negatively influencing bone healing.
MATERIALS AND METHODS
Experiments were performed on a total of 6 adult male C3H/HEJ (C3H) mice and 42 adult male C57BL/6J (C57) mice (Jackson Laboratories, Bar Harbor, ME, USA), weighing 20–25 g. Mice were housed in accordance with the NIH guidelines under specific pathogen-free conditions in partial autoclaved cages maintained at 22°C with a 12-h alternating light and dark cycle and were given food and water ad libitum. All procedures were approved by the Institutional Animal Care and Use Committee at the University of Minnesota.
Surgical and fracture procedure
Before femoral pin placement, mice received an intraperitoneal injection of 100 mg/kg ketamine and 10 mg/kg xylazine to provide anesthesia. An incision of ∼6 mm was made in the skin and the proximal patellar ligament of the left femur was severed, revealing the synovial space of the knee joint as previously described.(19,27) A 30-gauge needle was used to core between the condyles into the medullary canal of the left femur. Mice were immediately radiographed to ensure proper coring; any mice with needles outside of the medullary canal were killed. A 0.5-mm dental burr was used to prime the opening of the hole, and a 29-gauge needle was used as a pilot to widen the hole. A precut 0.011-in-diameter stainless steel wire (pin; Small Parts, Miami Lakes, FL, USA) was inserted into the medullary space for fracture stabilization, and dental amalgam was used to secure the pin and close the hole. Wound clips (MikRon Precision, Gardena, CA, USA) were used to close the incision and were removed 5 days after pin placement.
A closed mid-diaphyseal fracture of the left femur was produced 14 days after pin placement in mice under anesthesia (100 mg/kg ketamine and 10 mg/kg xylazine, IP) using a three-point bending device (BBC Specialty Automotive Center, Linden, NJ, USA) as previously described.(28) The methodology used to induce a closed fracture of the femur was originally described for the rat by Bonnarens and Einhorn(18) and later adapted for mice by Manigrasso and O'Connor.(19) Immediately after fracture, mice were radiographed to ensure localization of fracture to mid-diaphysis of the femur. Mice that met any of the exclusion criteria were immediately killed and were not used for this study. Exclusion criteria were adapted from Gerstenfeld et al.(29) and included fractures located too far from the mid-diaphyseal region of the femur, dislodged pins, nonvisible fracture after impact, and fragmentation of the bone. After recovery from anesthesia after fracture, mice were allowed unrestricted movement and limb weight bearing.
Mice were behaviorally analyzed before fracture (day 0) and at days 1, 2, 4, 7, 10, 14, 18, and 24 after fracture to assess ongoing (spontaneous) fracture pain-related behaviors, as previously described.(32–42) Briefly, the number of hind paw flinches and time spent guarding over a 2-min observation period were recorded as measures of ongoing pain, because these endpoints are similar to observations in patients who protect their fractured limb.(43) The number of total vertical stands requiring the use of both hind limbs (used as a measure of loading on the fractured limb) was recorded during a 2-min observation period at each behavioral time-point. Total vertical stands were defined as the number of times the animal stands on both hind limbs supporting their entire body weight.(44,45) The experimental protocol consisted of three different groups: pin + vehicle (n = 7), fracture + vehicle (n = 2), and fracture + anti-NGF (n = 13). All C57 mice were behaviorally analyzed (guarding, flinching, and rearing) in all of the previously stated time-points. To monitor the general health of the mice, body weights were recorded throughout the experiment, and mice were observed for any potential side effects caused by the therapy including ataxia, illness, or lethargy.
Treatment with anti-NGF therapy
The anti-NGF antibody, (mAb 911; Rinat Laboratories, PGRD, Palo Alto, CA, USA) is effective in sequestering NGF. Thus, NGF is unable to bind to its receptors (trkA and p75) and exert its actions.(46) It binds to both human and rodent NGF and does not appreciably bind to other members of the neurotrophin family.(46) Previous studies have shown anti-NGF antibody possess a plasma half-life of ∼5–6 days in the mouse and it does not appreciably cross the blood-brain barrier.(47)
Before behavioral and bone analysis, C57 mice were divided into three treatment groups, as previously stated, receiving either vehicle (pin + vehicle and fracture + vehicle) or anti-NGF antibody (fracture + anti-NGF). Vehicle (sterile saline, IP) or anti-NGF antibody were administered to mice on days 1, 6, and 11 after fracture (10 mg/kg, IP, each time-point). Baseline behavioral analysis was conducted before anti-NGF administration on day 1. Ten milligrams per kilogram was selected as this dose has been shown to be efficacious in bone cancer pain(27,37) and osteoarthritis(47) rodent models.
Radiographical and μCT analysis
Radiograph (Specimen Radiography System Model MX-20; Faxitron X-ray Corp., Wheeling IL, USA; Kodak film Min-R 2000; Kodak, Rochester, NY, USA) images of fractured femurs were obtained immediately after fracture and at all behavior time-points (days 1, 2, 4, 7, 10, 14, 18, and 24), as well as day 30 and day 42. Fracture-induced callus formation was radiographically assessed at ×4 magnification.(48,49) The images were scanned (Scanjet XPA; Hewlett Packard, Palo Alto, CA, USA) at 600-dpi resolution, and the area of the callus was determined by a blinded investigator using Image ProPlus v 6.0 software (Media Cybernetics, Silver Spring, MD, USA). Additionally, the images were used to evaluate bone bridging of the fractured femurs at all time-points. The mean bone bridging score was evaluated on a 0–4 scale based on Bergenstock et al.(28) Briefly, points were given for bridging across the two cortices of the fractured femur (one point each) and across the two peripheral sides of the callus (one point each) by a blinded investigator.
μCT images of naïve femurs of C3H and C57 mice were obtained with an eXplore Locus SP μCT (GE Healthcare, London, Ontario, Canada). The conebeam CT scanner used a 2300 × 2300 CCD detector with current and voltage set at 80 μA and 80 KVp, respectively. A 360° scan was performed with a 3000-ms integration time with images reconstructed at 16-μm3 resolution. 3D images were created using the MicroView analysis software (GE Healthcare)
The time-point to mechanically test the structural strength and stiffness of the fractured femurs was based on radiographic analysis that showed that day 42 was the first time that fractured femurs of our C57 mice treated with vehicle and anti-NGF had a score greater than 3 (where 0 is no bridging, and 4 is complete bridging across the callus). This bone bridging scoring system is based on that devised by Bergenstock et al.(28) for assessing fracture healing outcomes. This time for healing agrees with the study by Taguchi et al.,(50) who also used a closed femur fracture in 8-wk-old male mice. In these mice, mature bone was evident at 5 wk but not 3 wk after fracture. Additionally, Manigrasso and O'Connor(19) reported biomechanical properties of the healing of mice femurs are not reached until 6 wk after fracture. For these reasons, biomechanical testing was performed at day 42 (6 wk) in this study.
Briefly, mice were killed with compressed CO2 and perfused intracardially with 12 ml of 0.1 M PBS. Pins were removed, and femurs were loaded to failure in four-point bending at a displacement rate of 0.03 mm/s using a servohydraulic material test machine (Instron 5548 Microtester; Instron Corp., Norwood, MA, USA).(51) The femurs were loaded at mid-diaphysis by two load points 5.4 mm apart on the posterior surface and resting on two supports 9.0 mm apart on the anterior surface. Load and displacement data were collected every 0.1 s using Bluehill 2 software (Instron Corp.) and exported to Microsoft Excel for analysis. Ipsilateral femurs were tested for all groups (n = 6 for each group), pin + vehicle, fracture + vehicle, and fracture + anti-NGF. Maximum load and stiffness were derived from the load-displacement curves.(51,52)
Histomorphometric analysis of the fracture + vehicle and fracture + anti-NGF mice were performed as previously described.(53–55) Briefly, unfixed femurs were decalcified in 10% EDTA for 1 wk at 4°C, paraffin-embedded, and serially sectioned in the sagittal plane using a Leica Microsystems RM2135 microtome (Wetzlar, Germany) at 7 μm thickness. Five sections at least 150 μm apart spanning at least 0.75 mm of the center of the fracture callus of each animal were stained with hematoxylin and eosin (H&E).(54) Sections were photographed at ×20 using a SPOT II digital camera with SPOT image capture software (Diagnostic Instruments, Sterling Heights, MI, USA) attached to an Olympus BX51 microscope. Primary histomorphometric measurements included the area and perimeter of the total callus (T.Ar and T.Pm), void space (Vd.Ar and Vd.Pm), and trabeculae (Tb.Ar and Tb.Pm). The T.Ar and T.Pm were obtained from the outer most boundary of the callus.(53) The Vd.Ar and Vd.Pm contained all uncalcified areas, including marrow cavity, within the callus.(57) Tb.Ar and Tb.Pm were directly measured from the cancellous bone contained within the marrow.(58) All measurements were performed by a blinded investigator using Image ProPlus v6.0.
Total callus area (TOT.Ar; TOT.Ar = T.Ar − Vd.Ar + Tb.Ar), total callus perimeter (TOT.Pm; TOT.Pm = T.Pm + Vd.Pm + Tb.Pm), and area of the cortical bone in the callus (Ct.Ar; Ct.Ar = T.Ar − Vd.Ar) were derived as described by Parfitt et al.(53,59) Standardized 3D histomorphometric parameters of volumes and surfaces were extrapolated from 2D areas and perimeters. These included total volume (TV), total bone volume (BV), bone surface (BS), cortical bone volume (CBV), and trabecular bone volume (TBV; T.Ar = TV, TOT.Ar = BV, TOT.Pm = BS, Ct.Ar = CBV, Tb.Ar = TBV).(53) The parameters BV, CBV, and TBV were expressed as the percent of the total callus volume, BV (%), CBV (%), and TBV (%), respectively.(60–62) Microarchitecture analysis of the fracture callus was calculated from the aforementioned 3D measures to assess trabecular thickness (Tb.Th; Tb.Th = 2BV/BS), trabecular number [Tb.N; Tb.N = (BV/TV)/Tb.Th], and trabecular separation [Tb.Sp; Tb.Sp = Tb.Th × TV/(BV-1)].(30,53,58,63,64)
The SPSS v. 11 statistics package (SPSS, Chicago, IL, USA) was used to perform statistical analyses. Group differences in spontaneous guarding, spontaneous flinching, and spontaneous vertical stands were assessed with an ANOVA for the individual samples collected at each postfracture time-point. Significant ANOVA results were followed by individual two-group comparisons using Fisher's least significant difference (LSD) pairwise multiple comparison test. For bone healing data analysis, each dependent outcome variable was compared across groups using a nonparametric ANOVA (Kruskal-Wallis) followed by nonparametric posthoc comparisons between pairs of groups using the Mann-Whitney U-test. Results were considered statistically significant at p < 0.05. In all cases, the investigator was blind to the experimental status of each animal.
Comparison of femurs from C57 and C3H
Cortical bone thickness differences in two common strains of mice used in bone research, C57 and C3H, are shown in Fig. 1. C57 mice have thinner cortical walls compared with C3H mice as shown by multiple imaging techniques: cross-sectional μCT (Fig. 1A), cross-sectional H&E (Fig. 1B), longitudinal μCT (Fig. 1C), and radiographs (Fig. 1D).
Fracture production in C57 mice
The three-point fracture protocol resulted in reproducible transverse or slightly oblique mid-diaphyseal femoral fractures (Fig. 2). In this study, 9 of 34 mice that were fractured met the exclusion criteria. Mice were killed immediately after fracture if any of the previously defined exclusion criteria were met,(29) which included fractures located too far from the mid-diaphyseal region of the femur, dislodged pins, nonvisible fracture after impact, or fragmentation of the femur at the site of fracture.
Assessment of ongoing pain-related behaviors afrer femoral fracture
Ongoing (spontaneous guarding and flinching) pain-related behaviors were analyzed over a 2-min period in pin + vehicle, fracture + vehicle, and fracture + anti-NGF mice. Fracture + vehicle mice exhibited a significantly greater time spent guarding and a significantly increased number of flinches compared with pin + vehicle mice (Figs. 3A and 3B) from day 1 through day 18 after fracture. These ongoing pain-related behaviors peaked at day 4 after fracture and decreased gradually through day 18 after fracture. At day 24 after fracture pain-related behaviors in fracture + vehicle mice were not significantly different to those in pin + vehicle mice (data not shown).
Treatment with anti-NGF (10 mg/kg, administered at days 1, 6, and 11 after fracture) significantly reduced fracture-induced ongoing guarding and flinching pain-related behaviors when compared with fracture + vehicle mice. The analgesic effect was evident at the first time-point examined (day 2), which was 24 h after initial administration of anti-NGF and this analgesic effect continued through day 18 after fracture (Figs. 3A and 3B). There were no significant side effects, such as loss of weight, reduced luster of fur, ataxia, fever, or lethargy, observed in fracture + anti-NGF mice (data not shown).
Spontaneous exploratory vertical stands after femoral fracture
Before fracture all mice were not significantly different in the total number of spontaneous vertical stands in a 2-min period (Fig. 3C). Fracture + vehicle mice displayed a significant decrease in the number of spontaneous vertical stands on their injured hind limb compared with pin + vehicle mice. The fracture-induced reduction in vertical stands was maximum day 1 after fracture and returned gradually until day 10 after fracture. After day 10, there were no significant differences observed between pin + vehicle and fracture + vehicle mice (data not shown). Treatment with anti-NGF did not have an effect on the total number of spontaneous vertical stands in a 2-min period because there was no significant difference between fracture + anti-NGF mice and fracture + vehicle mice at all time-points analyzed (Fig. 3C).
Radiographic analysis of callus formation and bridging after femoral fracture
In this model, callus formation and bridging in the fracture + vehicle and fracture + anti-NGF was examined at days 7, 10, 14, 18, 24, 30, and 42. Total callus area, defined by the total radiopaque area within the outermost boundary of the fracture callus (Fig. 4), was measured in fracture + vehicle and fracture + anti-NGF groups. A radiopaque callus was first evident at day 7 after fracture in the fracture + vehicle and the fracture + anti-NGF mice and peak callus area occurred at day 14 for fracture + vehicle and day 18 for fracture + anti-NGF mice (Fig. 5). Callus area was significantly increased in the fracture + anti-NGF compared with the fracture + vehicle mice at days 14, 18, 24, 30, and 42 after fracture (Fig. 5).
For bridging, the values (using the bone bridging scoring system devised by Bergenstock et al.,(28) where 0 is no bridging and 4 is complete bridging across the callus and femur) at days 7, 10, 14, 18, 24, 30, and 42 after fracture for fracture + vehicle and fracture + anti-NGF were 0 ± 0, 0.08 ± 0.08; 0.75 ± 0.22, 0.54 ± 0.24; 2.25 ± 0.22, 1.85 ± 0.10; 2.25 ± 0.18, 2.33 ± 0.19; 2.73 ± 0.24, 2.58 ± 0.19; 3.09 ± 0.16, 2.82 ± 0.23l and 3.55 ± 0.21, 3.18 ± 0.23, respectively (no significant difference was observed between groups at any time-point). In light of these data, along with previous studies in mice by Taguchi et al.(50) and Manigrasso et al.,(19) in this report, the biomechanical properties of the mouse bone were assessed using a four-point bending device at day 42 after fracture.
Biomechanical analysis of bone after femoral fracture
Biomechanical properties of femurs from age-matched pin + vehicle, fracture + vehicle, and fracture + anti-NGF mice were evaluated at day 42 after fracture. At this time-point, fracture + vehicle mice showed no significant differences in values of maximum load and stiffness compared with pin + vehicle mice. Anti-NGF treatment did not seem to modify the biomechanical properties of the bone after fracture as femurs of fracture + anti-NGF mice have no significant difference in values of maximum load and stiffness compared with those of fracture + vehicle mice (Table 1).
Table Table 1.. Biomechanical Testing Analysis of Bone Healing 42 Days After Fracture
Histomorphometric analysis of callus area after femoral fracture
Histomorphometric analysis examining total callus parameters and microarchitecture measurements of the fracture callus was performed on day 42 after fracture. Total callus parameters included BV (%), CBV (%), and TBV (%). Microarchitecture of the fracture callus was assessed using Tb.Th, Tb.Sp, and Tb.N. At day 42 after fracture, the total callus parameter TBV (%) of bones from fracture + anti-NGF mice were significantly higher compared with bones of fracture + vehicle mice (Table 2). At the same time-point, bones of fracture + anti-NGF had a lower trabecular separation (Tb.Sp) compared with bones of fracture + vehicle mice. No other significant differences were found in the total callus parameters (BV [%] and CBV [%]) and in the microarchitecture measurements (Tb.Th and Tb.N) used in this study (Table 2).
Table Table 2.. Histomorphometric Analysis of Bone Healing 42 Days After Fracture
Pain and bone healing after femoral fracture in the C57 mouse
In this study, we developed a reproducible closed femur fracture model in C57 mice to simultaneously measure pain and bone healing. There are several advantages of using C57 mice to examine bone healing and pain after fracture. One of the universal characteristics of long bones and spines of middle-age and older mammals is loss in bone mass (osteopenia). If the bone loss is severe enough in humans, the result is osteoporosis, a skeletal disorder characterized by an increased incidence of fractures with sequelae that frequently include pain, loss of mobility, and increased loss of muscle and bone mass.(65) In general C57 mice have a relatively thin cortical wall, low BMD, and greater propensity to develop osteoporosis as they age(25) compared with inbred strains of mice such as the C3H mice,(66) which have been used in our laboratory to examine malignant bone pain.(27,67)
A second advantage of using C57 mice is that a significant fraction of all genetically modified mice has been developed in this strain.(23,68) Genetically modified mice including knockouts, knock-ins, and cell-specific expression of fluorescent markers or proteins have revolutionized our ability to define the specific molecules and cells involved in generation of a variety of neurological diseases including chronic pain.(23,68,69) Generating a pain fracture model in the C57 background will allow investigators to use genetically modified C57 mice without having to perform extensive backcrossing to simultaneously examine the role that specific molecules play in the pain and bone healing that follows fracture.(68) Additionally, the behavioral characteristics of C57 mice have been extensively studied in variety of inflammatory and neuropathic pain models,(70,71) which will allow investigators to compare and contrast the analgesic efficacy of novel therapies in a commonly used strain of mice in different types of acute and chronic pain states. The reproducibility and robust endpoints for both fracture-induced pain behaviors and bone healing in the C57 model should thus allow investigators to examine the effects of novel, mechanism-based therapies on skeletal pain and bone healing in both young and aged mice.
Current analgesics used to treat fracture pain
A major reason fracture pain remains a significant health problem is the limited repertoire of analgesics available to treat this pain without negatively influencing fracture healing and/or the ability of the patient to participate in effective rehabilitation. For example, nonsteroidal anti-inflammatory drugs (NSAIDS), which are effective in reducing a variety of musculoskeletal pains,(72,73) have been shown to inhibit fracture healing in both mice(74) and rats.(75) Thus, studies in rodents have shown that NSAIDS and selective cyclo-oxygenase-2 (COX-2) inhibitors hinder callus formation and effective bridging of the fracture site that results in delayed bone healing, increased incidence of nonunion of bone, and decreased bone strength.(31,76,77) These findings have been supported by similar studies where NSAIDS and COX-2 inhibitors appear to delay, but not completely inhibit, bone healing and bridging after fracture.(4,78) These data, together with recent reports that show selective prostaglandin agonists of the EP2 receptor accelerate bone healing after fracture(79,80) indeed suggest that the use of NSAIDS and COX-2 inhibitors may delay fracture induced bone healing.
There are relatively few studies that have examined the direct effect that opioids or their receptors may have on bone healing. However, opioids do have a variety of nonskeletal side effects that can inhibit bone healing. Opioids as a class cause increased somnolence, agitation, constipation, dizziness, and cognitive impairment that can reduce mobility, resulting in loss of bone and muscle mass.(81) In young individuals with severe fractures, long-term opiate use can result in dependence and a reduced ability to promptly and fully participate in effective musculoskeletal rehabilitation that is necessary for early and effective bone healing.(82,83) In elderly patients, opioid side effects tend to be more pronounced.(84) After osteoporotic fractures in the elderly, minimum bed rest is desired to minimize inactivity induced loss of bone and muscle mass.(85) However, administration of strong opiates will, in general, reduce the ability of these patients to effectively engage in the exercise and rehabilitation necessary for more rapid bone healing. Together, these data highlight the need for the development of novel, mechanism based therapies to treat skeletal pain that have negligible or a positive effect on bone healing.
NGF, pain, and bone healing after fracture
Using this C57 model of fracture pain and bone healing we show that, after femoral fracture, anti-NGF therapy reduced pain related behaviors, while having either no effect in some measures or a positive effect in other measures of bone healing. Concerning the timing of the administration of anti-NGF therapy, it should be emphasized that anti-NGF was first administered 1 day after fracture, which is a time-point when a human would begin to receive long-term analgesic therapy for fracture pain. The analgesic effect of anti-NGF therapy in reducing skeletal pain was statistically significant at the first time-point examined (which was day 2 after fracture) and the analgesic efficacy of the therapy remained as long as pain-related behaviors were present. Importantly, the analgesic effect of anti-NGF did not result in inappropriate loading of the fractured bones as assessed by the number of spontaneous vertical stands that require mechanical loading of the two hind limbs. Additionally, anti-NGF therapy did not result in longer healing times that may occur when the fracture site is loaded or stressed inappropriately as it does in both rodents and humans.(86,87)
In this study, we targeted anti-NGF for reducing nonmalignant fracture pain because our previous results showed that anti-NGF was efficacious in reducing malignant pain in mouse models of sarcoma- and prostate-induced bone cancer pain.(27,37) These results showed that anti-NGF therapy resulted in a 50% reduction in fracture-induced pain behaviors. We believe that anti-NGF is exerting its analgesic effect by binding to and sequestering NGF, which prevents NGF binding to its cognate receptors trkA or p75.(88)
It has been reported that a variety of inflammatory, immune, and stromal cells upregulate the expression of NGF and NGF receptors after fracture(89) and tissue injury.(90) In the adult, NGF can directly activate and sensitize sensory neurons involved in the conduction of pain originating from the skin(91–93) and visceral organs.(94,95) NGF is thought to excite and sensitize sensory neurons by binding to its cognate receptors trkA and p75 that are expressed by a subpopulation of mostly unmyelinated and thinly myelinated sensory neurons.(88,96) NGF binding to trkA has been shown to directly depolarize trkA expressing nociceptors in vivo and in vitro and that binding of NGF to trkA directly lowers the threshold for depolarization in these neurons.(88,97,98) Additionally, NGF has been shown to modulate and/or sensitize a variety of neurotransmitters, receptors, ion channels, and structural molecules expressed by nociceptors including: neurotransmitters (substance P, brain derived neurotrophic factor, and CGRP), receptors (bradykinin, P2X3), channels (TRPV1, ASIC-3, and sodium channels), transcription factors (ATF-3), and structural molecules (neurofilaments and the sodium channel anchoring molecule p11).(88,96,99–101) It has also been shown that NGF lowers the threshold and enhances the response of nociceptors to mechanical stimuli,(102) suggesting that NGF may play a role in activating/sensitizing mechanotransducers in sensory fibers that innervate bone. Together, the above results suggest that NGF may be an upstream regulator of a variety of neurotransmitter/receptors that are expressed by nociceptors that innervate the bone and that are involved in sensing and transmitting pain after fracture of the skeleton.
One rather unique aspect of the sensory innervation of bone, which may partially explain why anti-NGF therapy is effective in relieving both malignant(27,37) and nonmalignant skeletal pain (this study), is that the majority of nerve fibers innervating the bone seem to be CGRP-expressing fibers, nearly all of which co-express trkA.(103,104) Thus, in both humans and rodents, the primary afferent sensory fibers that innervate the human(105) and rodent vertebral discs(106) and bone,(15,107–109) seem to be CGRP/trkA expressing fibers with few unmyelinated nonpeptidergic IB4/RET+ nerve fibers being present in these tissues.(15) Thus, because bone seems to largely lack the redundancy of the nonpeptidergic IB4/RET+ nerve fibers that are present in skin,(110) anti-NGF may be particularly efficacious in relieving bone pain where nociceptors that express CGRP/trkA are the primary nerve fiber population that is driving the pain.
We also examined whether anti-NGF therapy delays fracture healing. As assessed radiographically and histomorphometrically, chronic administration of anti-NGF therapy increased the callus area around the fracture site in all time-points examined. Whereas callus size may not be a reliable indicator of proper bone healing,(111,112) callus formation is required for the initial stabilization of the bone that decreases fracture pain(113) and permits the proper destruction and formation of mineralized bone at the fracture site to begin to occur.(63,114,115) Based on this, it is possible to suggest that the anti-NGF analgesic effect could be partially explained by an increase in the callus formation after fracture. To examine bone healing, we performed biomechanical and histomorphometric analysis of the bones 42 days after fracture. When anti-NGF was administered at 1, 6, and 11 days after fracture, anti-NGF therapy either had no effect or a slightly positive effect on bone healing. Previous results using topical application of NGF to the fracture site in the rat suggested that NGF could promote bone formation after fracture.(116) However, it is difficult to compare the penetration and pharmacokinetics of NGF when topically applied to the rat bone(116) versus this study where the monoclonal antibody raised against NGF had a defined plasma half-life of 6 days(47) and was systemically available to all sites outside the blood-brain barrier. Given the time it takes to heal fractures of the femur in the mouse versus human (weeks versus months) and the greater amount of time the patient would be receiving anti-NGF therapy, further studies of the potential osteogenic effects of anti-NGF after fracture are clearly warranted.
Current therapies to treat skeletal fracture pain are limited. This is largely because of the side effects of currently available analgesics and the scarcity of animal models to study the mechanisms that generate and maintain this pain. In this study, we developed a fracture pain model in C57 mice where pain and bone healing can be simultaneously assessed. One day after fracture, anti-NGF was delivered systemically and resulted in an ∼50% reduction in fracture-related pain behaviors. Anti-NGF therapy did not interfere with the mechanical strength of the newly formed bone and only had a small but significant effect on bone healing as assessed by callus formation and histomorphometric analysis. Understanding the mechanisms that drive skeletal pain to develop mechanism-based therapies will have significant potential to improve rehabilitative outcomes in both young and old patients with skeletal fractures.
This work was supported by National Institutes of Health Grants NS23970 and NS048021 and a Merit Review from the Veterans Administration. We thank Therese Schachtele for excellent administrative assistance.