Mechanisms that drive bone pain across the lifespan

Abstract

Disorders of the skeleton are frequently accompanied by bone pain and a decline in the functional status of the patient. Bone pain occurs following a variety of injuries and diseases including bone fracture, osteoarthritis, low back pain, orthopedic surgery, fibrous dysplasia, rare bone diseases, sickle cell disease and bone cancer. In the past 2 decades, significant progress has been made in understanding the unique population of sensory and sympathetic nerves that innervate bone and the mechanisms that drive bone pain. Following physical injury of bone, mechanotranducers expressed by sensory nerve fibres that innervate bone are activated and sensitized so that even normally non‐noxious loading or movement of bone is now being perceived as noxious. Injury of the bone also causes release of factors that; directly excite and sensitize sensory nerve fibres, upregulate proalgesic neurotransmitters, receptors and ion channels expressed by sensory neurons, induce ectopic sprouting of sensory and sympathetic nerve fibres resulting in a hyper‐innervation of bone, and central sensitization in the brain that amplifies pain. Many of these mechanisms appear to be involved in driving both nonmalignant and malignant bone pain. Results from human clinical trials suggest that mechanism‐based therapies that attenuate one type of bone pain are often effective in attenuating pain in other seemingly unrelated bone diseases. Understanding the specific mechanisms that drive bone pain in different diseases and developing mechanism‐based therapies to control this pain has the potential to fundamentally change the quality of life and functional status of patients suffering from bone pain.

Introduction

The study of chronic bone pain usually focuses on common diseases such as osteoarthritis, low back pain and osteoporosis induced bone fracture. All of these diseases increase with the age‐related decline in the mass, quality and strength of the skeleton 1. However, there are a host of other adult diseases that generate chronic bone pain as well as over 500 human genetic disorders of bone and cartilage (Figure 1) that frequently present when the patients are young 2. Genetic diseases of the bone/joint that are accompanied by chronic bone pain include; osteogenesis imperfecta, Engelmann disease, Danlos syndrome, fibrous dysplasia, Paget's disease, sickle cell disease and juvenile arthritis 3, 4. While many of the 500 adult and genetic diseases of bone and joint are individually rare disorders, when combined they become a very significant number of patients who suffer from chronic bone pain throughout their life 2, 5.

Figure 1.

A partial list of human disorders across the lifespan that are frequently accompanied by bone pain. For an extensive list of diseases that are frequently accompanied by bone pain: http://www.rightdiagnosis.com/symptoms/bone_pain/common.htm; in children, http://www.nof.org/articles/5; and for a list of rare (orphan) bone diseases of bone and joint http://www.usbji.org/projects/RBDPN_op.cfm?dirID=252

Currently, the most common classes of pharmacological agents used to treat bone pain are nonsteroidal anti‐inflammatory drugs (NSAIDs) and opiates 5, 6, 7, 8. However, while NSAIDs (including ibuprofen, COX‐2 inhibitors, naproxen, and diclofenac) can be effective in the short‐term relief of bone pain, when used over an extended period they can have unwanted and severe renal, hepatic, and gastrointestinal side effects 9. In light of these issues with NSAIDs, it has now become more common for opiates to be used to control long term moderate‐to‐severe bone pain. However, recent data have suggested that although opiates can be useful in controlling nonmalignant bone pain for 2–3 months, long‐term use (>2–3 months) is associated with reduced functional status and decreased likelihood of returning to work, as well as potential development of dependence, constipation and respiratory depression 10, 11, 12. In older individuals, opiates are also more likely to induce dizziness, vertigo and cognitive clouding all of which increases the likelihood of falling which can result in bone fracture 13, 14.

In light of the side effect profile of NSAIDs and opiates, new mechanism‐based analgesics that relieve bone pain with a lower side effect profile are clearly needed. To develop such analgesics, rodent models of malignant and nonmalignant bone pain were developed 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28. These models of bone pain were then utilized to define the mechanisms that drive bone pain, test whether new mechanism‐based therapies could relieve bone pain and translate promising candidates into human clinical trials 29. The present review summarizes these results from preclinical and human studies. These data suggest that developing a deeper understanding of the mechanisms of one type of bone pain frequently provides unexpected insight and analgesic therapies for a variety of disorders of the skeleton.

The innervation of normal bone

There is a very tight regulation of the sensory and sympathetic innervation of the normal skeleton 30, 31, 32, 33, 34, 35. For example, whereas the articular cartilage is completely lacking in any blood vessels or nerve fibres 31, 32, 35 the periosteum has a remarkably dense sensory and sympathetic innervation (Figure 2). The bone marrow and mineralized bone are also innervated by sensory and sympathetic nerve fibres with the approximate density (per unit area) in the periosteum, bone marrow and cortical bone being 100:2:0.1 31, 35. While much of this analysis was performed on the young, adult and aging mouse femur, previous studies have noted a similar density, morphology and general organization of sensory and sympathetic nerve fibres in the rat calvaria and mandible 36, tibia 37 as well as human bones 38, 39, 40.

Figure 2.

The sensory and sympathetic innervation of normal bone. Micro‐computed tomography image of the adult (4‐month‐old) mouse femur (A) and a schematic (B) showing the location of sensory nerve fibres in the normal bone. Confocal microscopic images showing the sensory and sympathetic innervation in the periosteum (C) and the cortical bone (D) of the mouse femur in young (10‐day‐old), adult (4‐month‐old) and aging (24‐month‐old) mice. Note that in normal bone, sensory and sympathetic nerve fibres in the periosteum, cortical bone and marrow each have a unique organization and all contain sensory neurons that can transmit nociceptive stimuli from bone. Abbreviations: C, cambium layer of the periosteum; CB, cortical bone; CD31, platelet endothelial cell adhesion molecule that is expressed by endothelial cells; CGRP, calcitonin gene related peptide (labels small diameter sensory nerve fibers); DAPI, 4′,6‐diamidino‐2‐phenylindole (a counterstain that labels the nucleus of all cells); DIC, differential interference contrast that illustrates cortical bone; HC, cortical pores in mice that correspond to Haversian canals in humans; TH, tyrosine hydroxylase (labels sympathetic nerve fibers)

Within each compartment of bone there is also tight regulation of nerves in terms of density, phenotype and morphology. For example, in the periosteum of the mouse femur, > 90% of nerve fibres are present in the cambium layer with fewer than 10% being present in the fibrous layer 31. Previous studies have suggested that most of the nerve fibres in bone are tropomyosin receptor kinase A positive (TrkA+), are thinly or unmyelinated, and have conduction velocity characteristics of A‐delta and C‐fibres 16, 17, 18, 37, 41, 42, 43, 44, 45. In the periosteum, the A‐delta and C‐sensory nerve fibres are arranged in a fishnet‐like pattern, which appears to be designed to act as a neural net to detect mechanical injury or distortion of the underlying cortical bone 31, 46. In contrast, sympathetic nerve fibres in the periosteum usually have a characteristic corkscrew‐like morphology and are almost invariably tightly associated with blood vessels 31, 46.

In the cortical bone, the great majority of sensory and sympathetic nerve fibres are largely confined to the vascularized cortical pores in mice 31, 47 and vascularized Haversian canals in humans 38, 39, 40. In the bone marrow, the sensory nerve fibres are linear in appearance, whereas the sympathetic nerve fibres again have a corkscrew shaped pattern as they tightly wrap around blood vessels 31, 39. Thus, in terms of location, density and morphology, both sensory and sympathetic nerve fibres are tightly regulated in the normal, uninjured bone. However, as discussed below, injury or disease of the bone can induce a remarkable and highly ectopic reorganization of both sensory and sympathetic nerve fibres.

Acute activation of mechano‐ and chemo‐sensitive nociceptors in bone

Following acute physical injury to bone, mechanosensitive nociceptors that innervate bone are activated and sensitized (Figure 3). Normally, these mechanosensitive nociceptors in the bone are silent 48, 49, 50, 51. However, when these sensory nerve fibres are mechanically distorted (such as occurs following bone fracture) nociceptors that innervate bone respond within seconds 52, 53 and signal this mechanical distortion from the bone to the spinal cord and higher centres of the brain 48, 49. These mechanically sensitive sensory nerve fibres will continue to rapidly discharge until they are returned to their original position, which is usually when the bone is reset and stabilized by a rod or cast 49. Unfortunately, in bone diseases where there is inherent instability of the bone (such as in unhealed fractures in the elderly or bone cancer) these mechanosensitive sensory nerve fibres in bone will be activated whenever the weakened bone is moved or loaded 7, 49. Previous reports have also suggested that increased intraosseous pressure in the bone marrow can drive bone pain by stimulating mechanosensitive nociceptors that innervate the bone 18, 41, 49, 50.

Figure 3.

Mechanisms that contribute to acute and chronic bone pain. An evolving set of mechanisms drives acute and chronic bone pain. Bone pain involves nociceptor activation, sensitization, transcriptional changes, ectopic sprouting and central sensitization in the spinal cord and brain. If rapid and effective bone healing occurs (such as following bone fracture in the young), these events return to pre‐fracture levels with resolution of the bone pain. However, if rapid and effective healing does not occur (such as in bone cancer or non‐healed fractures in the aged) these mechanisms can synergistically increase with time, resulting in chronic bone pain. Abbreviations: ASIC‐3, acid sensing ion channel‐3; BDNF, brain‐derived neurotrophic factor; BR2, bradykinin receptor‐2; CGRP, calcitonin gene related peptide; Ca, calcium channels; K, potassium channels; Na, sodium channels; p75, low‐affinity nerve growth factor receptor; SP, substance P; TrkA, Tropomyosin receptor kinase A (high affinity nerve growth factor receptor; TRPV‐1, transient receptor potential vanilloid‐1. Adapted from Enomoto et al. (2018) 120

Bone pain induced by trauma (i.e. bone fracture) is at least initially driven by mechanical distortion of mechanosensitive sensory nerve fibres in the bone. However, in other diseases such as osteogenesis imperfecta or bone cancer, osteoclast induced acidosis can also play a significant role in driving bone pain. Studies have shown that many sensory nerve fibres that innervate bone express acid sensing ion channels including TRPV1, ASIC1 and ASIC3 16, 54 that will be readily activated by a pH of 3–4, which is the extracellular pH generated by osteoclasts when resorbing bone. One of the best examples of therapies initially used to treat one bone pain and then relieving another bone pain are bisphosphonates and Denosumab which both target osteoclasts. Both were originally developed to treat osteoporosis and were later shown to also relieve bone cancer pain 55, 56, 57, 58, 59, 60. Since this original observation, bisphosphonates or Denosumab have also been shown to attenuate bone pain in human patients with; osteogenesis imperfecta, juvenile Paget's disease, fibrous dysplasia, aneurysmal bone cyst and complex regional pain syndrome 5, 59, 61, 62, 63, 64, 65. Mechanistically, what these diseases share in common is excessive osteoclast activity which generates a low pH that will stimulate acid sensing ion channels expressed by sensory nerve fibres that innervate bone 54, 59, 60. These results emphasize the point that therapies that attenuate one type of bone pain that is driven by a specific mechanism (here acidosis) may also be effective in attenuating pain in other bone diseases, which are driven by similar mechanisms.

Sensitization of nociceptors innervating bone

A second mechanism that plays a significant role in amplifying both acute and chronic bone pain (Figure 3) is sensitization of bone nociceptors (i.e. sensory neurons that detect noxious stimuli in bone). Agents that have been reported to sensitize bone nociceptors include bradykinin, endothelins, epidermal growth factor, glial cell line‐derived neurotrophic factor, histamine, nerve growth factor (NGF), prostaglandins, tumour necrosis factor and vascular endothelial growth factor 25, 66, 67, 68, 69, 70, 71, 72, 73. Probably the best studied factor that has been examined in bone pain in both preclinical and human clinical studies is NGF. Once NGF is released in the injured bone, it binds to TrkA+ receptors that are expressed by many bone nociceptors 43, 74, 75, 76. Binding of NGF to TrkA+ results in the phosphorylation and sensitization of a variety of receptors and ion channels expressed by nociceptors 9, 16, 77, 78, 79, 80, 81, 82, 83. These include ion channels that respond to acid (TRPV1 and ASIC3), receptors that bind prostaglandins and bradykinin, and mechanotransducers expressed by sensory nerve fibres 84. Following this NGF induced sensitization, there is an enhanced response of nociceptors to even small amounts of acid released by osteoclasts as well as prostaglandins and bradykinin released by injured tissues 84, 85. In addition, excitation and firing of mechanosensitive nociceptors is magnified following low dose treatment of NGF in both animals and humans, thus increasing the signalling of noxious input to the spinal cord and brain 26, 77.

Transcriptional changes in sensory neurons

It is now well established from preclinical and clinical studies that the peripheral release of NGF and activation of TrkA play important roles in driving bone pain 9, 19, 23, 26, 84, 86, 87, 88. Once NGF is released in bone, it binds to its cognate receptor TrkA and is retrogradely transported as a complex back to the cell body of the sensory neuron located in the dorsal root ganglion 84. When the NGF‐TrkA complex reaches the nucleus of the sensory neuron, there is significant alterations in several genes that are important for the detection and signalling of noxious bone stimuli 16, 44, 84, 89. Thus, NGF: induces upregulation of the ion sensing channels TRPV1 and ASIC3; and increases the expression of neurotransmitters including substance P, calcitonin gene‐related peptide, brain‐derived neurotrophic factor, and a variety of sodium and calcium channels that regulate the excitability of bone nociceptors 84, 90. Recent data also suggest that even low doses of NGF can increase the electrophysiological excitability of human sensory neurons for weeks 77 suggesting that long term transcriptional mechanisms may be involved in these NGF induced changes.

Therapies targeting NGF (Figure 5) have been shown to relieve pain in a variety of human skeletal pathologies including osteoarthritis 86, 91, 92, 93, low back pain 87, 94 and bone cancer pain 95, 96. Presumably, blockade of NGF will be contraindicated in young patients with skeletal pain as NGF is involved in the growth and survival of the developing sensory and sympathetic nervous system in the young 9. However, many individuals with genetic disorders of the bone and joint (such as fibrous dysplasia) not only have significant pain when they are young but continue to have chronic skeletal pain throughout their adult life 65. Whether targeting blockage of NGF or its cognate receptor TrkA will block pain in adults with genetic disorders of the bone, complex regional pain syndrome and other painful bone diseases, has yet to be determined.

Figure 5.

Data showing that blocking osteoclast activity or nerve growth factor can attenuate human bone pain. Data from a human trial (A) showing a reduction in bone pain and an increase in functional status of patients with breast cancer metastases to bone following administration of a bisphosphonate (ibandronate), which targets osteoclasts and reduces bone acidosis. Recent data have demonstrated that bisphosphonates can also reduce pain in osteogenesis imperfect, Paget's disease, fibrous dysplasia and complex regional pain syndrome. Results from a human clinical trial (B) showing that anti‐nerve growth factor (tanezumab) can reduce chronic osteoarthritis pain. Anti‐NGF therapy has also been shown to attenuate pain in patients with low back pain and in bone cancer patients with high pain and low opiate use. These data suggest therapies that attenuate one type of bone pain may be effective in reducing pain in seemingly unrelated bone diseases if the diseases share a common mechanism(s) that is driving the pain. Data in (A) are from Ringe et al. (2007) and Body et al. (2004) 59, 60 and in (B) from Lane et al. (2010) 121. VAS, visual analogue scale

Ectopic sprouting in sensory and sympathetic nerve fibres

The fourth mechanism that may be involved in driving skeletal pain is ectopic nerve sprouting (Figure 4). Following injury or disease, several neurotrophic factors including NGF, glial cell line‐derived neurotrophic factor, vascular endothelial growth factor and epidermal growth factor are released by stromal and inflammatory cells and can induce nerve sprouting, resulting in hyperinnervation of the marrow, mineralized bone and periosteum 68, 96, 97, 98, 99, 100, 101, 102, 103. It should be noted that nerve sprouting has been observed even in bone fractures that are undergoing normal healing 104. This nerve sprouting can be observed in the normal callus and probably is important in causing the patient to guard the injured limb and refrain from excessive use of the fractured bone until it heals. As the fractured bone normally heals and the callus is resorbed, these newly sprouted nerve fibres are pruned back so that, when the bone fully heals, the innervation of bone returns to its normal state.

Figure 4.

Images of nerve sprouting in malignant and non‐malignant bone pain. As seen in (A) sensory nerve fibres (bright yellow) in the normal bone marrow have a regular and linear appearance. Following invasion of the bone marrow (B) by prostate cancer cells (green) there is marked increase in the density due to the ectopic sprouting of sensory nerve fibres. Prior administration of anti‐nerve growth factor (NGF) before tumour invasion of the bone largely blocks this ectopic nerve sprouting (C). Similar sprouting of sensory nerve fibres is also observed in non‐malignant pain states following injury of bone. (D, E) Sensory nerve fibres (red) in the normal marrow and periosteum respectively. (F, G) Increased sensory innervation of the marrow (F) and periosteum (G) in a bone at 6 months postfracture when effective bone healing has not occurred. Abbreviations: anti‐NGF, a monoclonal antibody that binds to nerve growth factor (NGF); CGRP, calcitonin gene related peptide (labels small diameter sensory nerve fibers); DAPI, 4′,6‐diamidino‐2‐phenylindole (a counterstain that labels the nucleus of all cells); GFP, green fluorescent protein that is expressed by prostate cancer cells

However, in cases of injured or diseased bone where normal or rapid bone healing does not occur, this ectopic nerve sprouting is not pruned back and now the injured or diseased aspects of the bone are hyperinnervated 28, 68, 96, 98, 99, 100, 101, 102, 103. Now, normally non‐noxious loading or movement of the bone may be perceived as a noxious event as the bone is not only weaker than healthy bone but the weak bone is hyperinnervated making detection of any mechanical distortion of the weakened bone more likely. Skeletal diseases where ectopic sprouting has been observed include bone cancer 27, 99, 100, unhealed bone fracture 7, osteoarthritis 28, 101, 105 and the degenerated vertebral disc 102, 103, 106, 107. Whether significant and unwanted sprouting of sensory and sympathetic nerve fibres also occurs in patients with genetic disorders of the skeleton, has yet to be determined. However, if ectopic nerve sprouting in bone does occur in these conditions, it may provide insight into the mechanisms that drive the transition from acute to chronic skeletal pain in these patients 9, 108.

Central sensitization

The fifth mechanism that participates in driving bone pain is changes that occur in the spinal cord and brain, which cause central sensitization that amplifies the perception and severity of pain 89, 109, 110, 111, 112. Central sensitization is thought to occur when the chemical, electrophysiological and pharmacological systems that transmit and modulate pain are altered in both the spinal cord and higher centres of the brain 113, 114, 115, 116, 117, 118, 119. These changes cause an exaggerated perception of painful stimuli so that normally mild painful stimuli are perceived as highly painful (hyperalgesia) and normal nonpainful stimuli such as normal loading or use of joint or bone is now perceived as a painful event (allodynia). Additionally, central sensitization also contributes to the phenomenon of referred pain, where areas adjacent to the initial injury become hypersensitive. A common example of this is in whiplash injuries to the neck where, following injury to one cervical vertebra, nearby areas such as the shoulder, arm and back also become hypersensitive 116.

While we do not yet know the specific mechanisms that generate central sensitization following injury to bone, what is known is that injury to skeleton seems to be much more effective at inducing central sensitization as compared to injury to skin or muscle 49, 110, 118. As noted by Woolf and Wall in 1986 “ …a twisted ankle invokes relatively little destruction of tissue and elicits an abrupt localized stabbing pain that dies down quickly but is followed by a prolonged period of spreading, poorly localized deep pain, and tenderness that affects reflexes and gait. In contrast, localized skin damage produces an acute burst of pain that gradually dies down over minutes but is associated with a spatially restricted response of flair, wheal and surrounding tenderness” 110. Just why injury to the skeleton is so effective at producing central sensitization remains unclear 49, 117, 118, 119, 120, 121 although the skeleton is innervated by a very distinct population of primary afferent neurons that may be uniquely effective at inducing central sensitization in the spinal cord and higher centres of the brain.

Conclusions and limitations

In the past 2 decades there has been a remarkable increase in our understanding of the specific mechanisms that drive bone pain. The specific nerve fibres that innervate the bone have begun to be defined as well as the remarkable changes these nerves can undergo in injury and disease. Mechanisms that drive bone pain are now known to include: activation of mechanotranducers and acid sensing ion channels expressed by sensory nerve fibres; sensitization, transcriptional changes and ectopic sprouting of sensory nerve fibres; and central sensitization involving both the spinal cord and brain. What is also clear is that insight into mechanisms that generate bone pain in one disease can be very useful in understanding the mechanisms that generate pain in other bone diseases (Figure 5).

While significant progress has been made, we are only beginning to understanding how nerves and bone interact and modulate each other. Importantly, it can still be very challenging to fully control most chronic bone pain. However, if new mechanism‐based therapies that better control bone pain can be developed, they have the potential to fundamentally change the quality of life and functional status of patients suffering from skeletal pain.

Competing Interests

There are no competing interests to declare.

Research supporting this manuscript was funded by NIH grants CA154550, CA157449, and NS023970 to Patrick Mantyh. Dr Mantyh has served as a consultant and/or received research grants from Abbott (Abbott Park, IL), Adolor (Exton, PA), Array Biopharma (Boulder, CO), Johnson and Johnson (New Brunswick, NJ), Merck (White Plains, New York), Pfizer (New York, NY), Plexxikon (Berkley, CA), Rinat (South San Francisco, CA, and Roche (Palo Alto, CA).

Mantyh P. W. (2019) Mechanisms that drive bone pain across the lifespan, Br J Clin Pharmacol, 85, 1103–1113. doi: 10.1111/bcp.13801.