The Genesis of Pain in Osteoarthritis: Inflammation as a Mediator of Osteoarthritis Pain

1. Osteoarthritis pain: A worldwide unmet medical need.

In the United States, approximately 20% of adults suffer from chronic pain, creating a substantive personal and societal burden and costing the economy up to $635 billion per year^1,2. The musculoskeletal system is one of the leading sources of chronic pain, with osteoarthritis (OA) as the leading cause. As the most prevalent form of arthritis, OA affects over 32.5 million US adults. Worldwide, 240 million individuals live with symptomatic OA^3–6, and its prevalence is increasing as the world’s population is ageing⁷. It has been recognized that OA is associated with an increased risk of all-cause mortality, often because of reduced mobility⁸.

Chronic pain and disability define clinical OA, and activity-limiting pain has been shown to be the greatest burden driving patients to seek medical attention in older adults⁹. Chronic pain can have a profound negative impact on patients’ physical lives as well as their mental health, and knee OA is significantly correlated with poor mental health and depressive symptoms^10–12.

Disease-modifying and pain-alleviating therapies for OA remain elusive. Currently available analgesics are largely inadequate in addition to having adverse side effects, resulting in a major unmet medical need for patients. Therefore, furthering our understanding of the origins and mechanisms of OA pain will be critical to developing effective treatments.

2. Pain in Osteoarthritis

OA pathology is characterized by progressive joint damage, including cartilage erosion, subchondral bone sclerosis, synovitis, bone remodeling with osteophyte formation, and meniscal damage, all of which may contribute to pain. There is no simple correlation between pain and structural pathology, and in knee OA, studies have only found weak correlations between pain and radiographic findings^13–15. In contrast, MRI studies, which image soft tissues, have uncovered that synovitis and bone marrow lesions are correlated with pain. This suggests that specific aspects of joint pathology may indeed cause OA pain, offering the prospect for targeted treatment strategies, although this has not yet led to agents that have been approved by regulatory agencies¹⁶. Attempts have also been made to stratify patients based on pain phenotypes, but translating this into therapeutic strategies has so far been unsuccessful¹⁷.

Pain in OA is complex and multi-factorial, modified by genetic, psychological, and environmental determinants. Patients struggle to self-describe pain with just ‘intensity’, and describe numerous types of pain that vary in duration, depth, type of occurrence, impact, and rhythm¹⁸. Several recent reviews have thoughtfully discussed different complex aspects of the genesis of pain in OA, including peripheral mechanisms of pain and sensitization^{19 20}, central pain pathways²¹, the role of innate immune responses²², and structural correlates of OA pain²³. When considering the genesis of joint pain, we should bear in mind that pain is one of the cardinal signs of inflammation. While the role of low-level inflammation in the pathogenesis of OA joint damage has been extensively studied in recent years²⁴, it is not at all clear how exactly inflammation contributes to the accompanying pain. Targeting inflammation in OA can provide pain relief for some patients, but the effect size for nonsteroidal anti-inflammatory drugs (NSAIDs), a first-line treatment for OA, is moderate at best²⁵. Likewise, intra-articular corticosteroid injections – while providing acute pain relief – have not shown long-term analgesic effects^26,27. This illustrates that joint inflammation alone is not enough for sustaining pain in OA. Inflammatory mediators can act on many cells, including neurons, and inflammation comes in many guises, operating at many levels, including in the joint, in the nervous system (“neuro-inflammation”), and systemically. In addition, inflammation can both induce pain, as well as promote the resolution of pain. In this narrative review, we discuss recent discoveries that clarify how these different levels of inflammation may initiate, maintain, and modify pain in OA. We searched PubMed for the following terms: “osteoarthritis”, “pain”, “inflammation”, “cytokines”, “NGF”, “synovium”, “DRG”, “obesity”, and principally focused on studies published in the last 5 years.

3. Local inflammation in the OA joint: contributions to pain.

The pain pathway can be broken down into signal detection and neuron stimulation, signal transmission to the dorsal horn via the peripheral nervous system, and transmission to the brain via the central nervous system (Figure 1). A range of mechanical, thermal, and chemical stimuli - both noxious and non-noxious - are sensed in peripheral tissues via sensory neurons. This generates action potentials that carry these signals to the dorsal horn via the dorsal root ganglia (DRG), where the cell bodies of sensory neurons are located. Sensory nerves are either thick, myelinated Aβ fibers which are activated by joint movement, or thinly myelinated Aδ and unmyelinated C fibers, which are activated by noxious mechanical, thermal and chemical stimuli through specialized receptors²⁸. These high-threshold pain-sensing neurons are called nociceptors (from the Latin nocere – to harm). For example, noxious heat is detected by the capsaicin receptor, transient receptor potential vanilloid type 1 (TRPV1)²⁹, generating action potentials that invade the DRG. Voltage-gated sodium channels are essential for the generation of action potentials. From the DRG, the painful signals are then transmitted to the dorsal horn in the spinal cord, where the first synapse occurs²⁸. Second-order neurons, activated by the release of glutamate, project to supraspinal regions, and from there, the signal is relayed to higher regions of the brain, where pain perception occurs, and information is generated on the location and intensity of the pain, while affective aspects modify the pain experience³⁰.

Figure 1: Pain pathway and sites of inflammation and neuroplasticity in osteoarthritis.

Nociceptors (Aδ or C- fibers) detect stimuli in joint tissues, generating action potentials that are transmitted to the dorsal horn via the dorsal root ganglia (DRG), and to the brain via the central nervous system. Inflammatory states and neuroimmune interactions can occur at multiple points along this pathway. Structural neuroplasticity occurs in the OA joint, with free nerve endings sprouting in the subchondral bone and synovium. Joint nociceptors become sensitized by mediators released in OA pathology, including nerve growth factor (NGF), inflammatory cytokines and chemokines, and damage-associated molecular patterns (DAMPs). Immune cell infiltration in synovium and DRG can include macrophages, T- and B- lymphocytes, and mast cells, which can further amplify pain mechanisms. Neuroimmune interactions also occur in the DRG and the dorsal horn. Finally, factors such as obesity, age, and sex may alter systemic inflammatory inputs and modify OA pain.

Below, we will discuss how inflammatory mediators generated in the OA joint may modify the nociceptive pathway through direct effects on sensory neurons that innervate joint tissues.

3.1. Local inflammatory mediators in the joint may affect neuronal activity and neuronal sprouting.

Inflammation and tissue injury are associated with sensitization of pain pathways, which results in augmented pain responses. For example, patients with OA are sensitized to mechanical stimuli, and quantitative sensory testing (QST) has revealed hyperalgesia, manifesting as lowered pain pressure thresholds when a force is applied to the joint, as well as allodynia, where non-noxious stimuli are perceived as painful. QST measures are more closely associated with pain severity than radiographic severity, and may aid in identifying people susceptible to chronic pain (reviewed in³¹). Sensitization to mechanical stimuli is also a feature of all animal models of OA³². Thus, it can be expected that elucidating the mechanisms underlying sensitization of joint nociceptors may be a key step in understanding the genesis of chronic OA pain³³. Here, we will briefly discuss inflammatory mediators present in the joint that may promote pain through direct effects on joint nociceptors.

3.1.1. Inflammatory cytokines and chemokines -

Inflammatory mediators, including cytokines, chemokines, neurotrophins, and prostaglandins can all activate nociceptors through direct binding of their cognate receptors expressed by these cells³⁴. OA joint tissues and the synovial fluid contain many such pro-algesic factors, including tumor necrosis factor (TNF)-α, interleukin (IL)1-β, IL-6, and chemokine (C-C motif) ligand 2 (CCL2)³⁵, and in vitro studies have shown that OA synovial fluid induces hyperexcitability in DRG neurons³⁶. In subjects with OA, synovial fluid levels of IL-7, IL-12, interferon (IFN)-γ, IL-10 and IL-13 have been correlated with knee pain³⁷. Clinical trials with antibodies neutralizing traditional pro-inflammatory cytokines, including IL-1 and TNF-α, have been largely disappointing³⁸, and more recent work in animal models deals with cytokines that were previously not a focus in OA research, including IL-17 and IL-33. IL-17A causes long-lasting sensitization of joint nociceptors to mechanical stimuli when injected into the rat knee³⁹. In an inflammatory model (antigen-induced arthritis), Il17 null mice have less mechanical hyperalgesia, even though joint damage is as severe as in wild type mice, suggesting a direct contribution of IL-17A to the genesis of pain⁴⁰. Na et al. reported that IL-1 receptor antagonist (IL-1Ra) null mice have increased pain and cartilage destruction in the monosodium iodoacetate (MIA) model of OA, but attenuation of increased IL-17 levels using IL-17/IL-1RA double-deficient mice resulted in significantly decreased pain scores⁴¹. Another cytokine that was recently studied in the context of OA is IL-33, a member of the IL-1 family that plays a role in innate and adaptive immunity⁴². He et al. reported increased concentrations of IL-33 in the synovial fluid of OA patients, and increased expression of IL-33 and its receptor, suppression of tumorigenicity 2 (ST2), in OA chondrocytes⁴³. IL-33 expression is also increased in mice after DMM surgery, and intra-articular administration of recombinant IL-33 caused mechanical allodynia. Neutralizing monoclonal antibodies against IL-33 and ST2 attenuated both OA progression and pain in the DMM model.

There is also continued interest in the chemokine, CCL2, as a target for OA pain. CCL2 levels are increased in OA synovial fluid, and this has been correlated with symptoms^44,45. Several laboratories have found that CCL2-CCR2 signaling is essential for the initiation and maintenance of pain-related behaviors in the DMM model (reviewed in⁴⁶). This signaling pathway may contribute to OA pain through multiple mechanisms, including macrophage recruitment to the joint and to the DRG, as well as through direct neuronal activation^47,48. CCR2 (the main receptor for CCL2) is expressed on intra-articular nociceptors, both in naïve mice and after destabilization of the medial meniscus (DMM)⁴⁹. Furthermore, intra-articular injection of CCR2 receptor antagonist into the knee joint reduced knee hyperalgesia after DMM surgery, suggesting a local role for CCL2-CCR2 signaling in sensitization of joint nociceptors, and providing a possibility for therapeutic targeting of this pathway in early disease.

Finally, targeting the transcription factor, NF-κB, which is activated by both mechanical stress as well as cytokines and chemokines, remains an area of active investigation. In particular, recent work has focused on developing intra-articular therapeutic strategies that target this pathway in the context of post-traumatic OA, revealing analgesic effects⁵⁰.

3.1.2. Nerve Growth Factor (NGF) -

The neurotrophin, NGF, has received much attention in the field in recent years, because clinical trials with monoclonal antibodies that neutralize NGF showed promising analgesic effects in OA of the knee⁵¹. However, the ill-understood risk for rapidly progressive OA in a subset of patients receiving the antibodies has prevented FDA approval of this biological therapy⁵².

During embryonic development, NGF is crucial for the growth and survival of neurons. Postnatally, NGF retains key functions, especially in inflammation, when levels are dramatically increased⁵³. Pain production is a major biological effect of NGF, and local injection of NGF produces immediate and often long-lasting sensitization and pain responses in animals and humans, including when injected into the joint^54–57. The mechanisms by which NGF exerts its pain-producing actions have been extensively reviewed, and involve both rapid effects and long-term neuroplastic changes⁵⁸. The high-affinity receptor for NGF, tropomyosin kinase A (TrkA), is expressed by nociceptors, and NGF-TrkA signaling results in excitatory effects that lead to functional sensitization of nociceptors, both through transactivation of ion channels such as TRPV1, and through retrograde transport of the NGF/TrkA complex to the cell bodies in the DRG, where it initiates gene expression changes that lead to synthesis of pro-algesic peptides (substance P, CGRP) and ion channels such as Na_V1.8.

An important property of NGF that may contribute to its pain-promoting activity, is that it can stimulate neuronal sprouting at an injured site, as has been described in models of bone fracture, bone cancer, and inflammation induced by intra-articular (i.a.) complete Freund’s adjuvant, where anti-NGF antibodies blocked neuronal sprouting and pain⁵⁸. This biological property may be of key importance in the context of OA, since it is becoming increasingly clear that profound neuronal plasticity of nociceptors occurs as part of OA joint pathology. Most strikingly, in human OA knees as well as in rodent models, osteochondral channels breach the tidemark between the subchondral bone and the articular cartilage, and these channels contain neurons and blood vessels^59–62. At this time, it is not clear to which extent these newly sprouted nociceptors contribute to pain in human OA knees, but the presence of CGRP-immunoreactive nociceptors in these channels has been associated with pain⁶¹. Neuronal sprouting has also been reported in the medial synovium in a surgical murine model of OA⁶².

Clearly, it will be important to identify the key drivers of this neuroplasticity in order to enhance our understanding of the genesis of OA joint pain. Candidate molecules have been identified, such as netrin, an osteoclast-derived molecule that has been linked to neuronal sprouting in the subchondral bone in a surgical murine model⁶³. Importantly, NGF levels are elevated in the synovial fluid of OA joints⁶⁴, and expression of NGF is increased in OA chondrocytes^64,65, synovial fibroblasts⁶⁶ (associated with pain) and osteochondral channels⁶⁷ (also associated with pain). Therefore, it is highly likely that NGF plays a key role in the observed anatomical neuroplasticity observed in OA joints, although this has not been directly investigated.

3.1.3. Damage associated molecular patterns (DAMPs) -

Products produced as part of the joint degradation process have the potential to further amplify ongoing joint damage, and innate immune mechanisms are the focus of intense research in OA pathogenesis⁶⁸, including OA pain (recently reviewed²²). Indeed, joint remodeling in OA generates a host of DAMPs, such as fragments of proteoglycans or inflammation-associated alarmins (e.g., S100 proteins) that can bind receptors, including Toll-like receptors (TLRs) and NOD-like receptors (NLRs), which are expressed by many joint tissue cells, including nociceptors. This leads to direct activation of nociceptors, resulting in pain. Tissue damage products can also signal to TLRs on innate immune cells, causing them to produce chemokines and cytokines that also act on pain-sensing neurons, further amplifying and maintaining pain responses.

3.2. The role of synovium in osteoarthritis pain

OA pathology involves all joint tissues, and inflammatory cytokines,chemokines, DAMPs, and NGF can exert their biological effects in all these tissues, and sensitize nociceptors. As we have discussed, neuronal sprouting has been observed in the subchondral bone of OA joints, as well as in the synovium (Figure 1). This begs the question as to how synovitis may contribute to OA pain. The synovial tissue lines the joint cavity and secretes synovial fluid, aiding in joint movement. Inflammation of the synovial tissue is observed in more than half of OA patients throughout disease progression, including early stages of disease^69,70. Synovitis, as detected by MRI, is associated with pain in knees with and without radiographic OA, as well as with sensitization in knee OA^71,72. Of interest, different anatomical patterns of synovitis have been described, and a pattern including patellar sites of synovitis have been particularly associated with pain⁷³. Synovitis in the OA joint is characterized by an influx of a variety of leucocytes, primarily macrophages and T-lymphocytes, but mast cells, B-lymphocytes and plasma cells can also be present⁷⁴. The quantity of activated macrophages has been associated with OA pain severity and radiographic OA severity⁷⁵. Correlation studies have also suggested a role for lymphocytes in OA pain; for example, Zhu et al. showed a positive correlation between circulating CD4+ T cells and WOMAC pain scores⁷⁶. In the synovial tissue, proinflammatory T helper cell type 1 (Th1) infiltration has been recently reported in early OA patients⁷⁷. Similarly, Nees et al. recently investigated T cell subsets in the synovial tissue of OA patients, with a focus on pain and disability, and found correlations between levels of both Th1 and Th2 T cells with OA-related knee pain⁷⁸. Finally, mast cell numbers have also been found to be elevated in OA synovium⁷³, and an interesting recent study in murine MIA reported that NGF-stimulated synovial mast cells produce prostaglandin D₂, which can signal to nociceptors, thus linking NGF-TrkA signaling in mast cells to OA pain⁷⁹.

In order to unravel how infiltrating immune cells contribute to OA pain, it will be critical to elucidate the bi-directional crosstalk between the nervous system and the immune system. For example, targeting granulocyte macrophage colony stimulating factor (GM-CSF) prevents pain in murine collagenase-induced osteoarthritis (CiOA)^80,81. Tewari et al. showed the absence of functional GM-CSF receptors on mouse nociceptors, but found that conditioned medium from GM-CSF treated macrophages could drive nociceptor transcriptional changes⁸². This exemplifies again how neuroimmune crosstalk may operate in the knee and offer avenues for targeted intervention.

Recent development of technologies such as bulk and single cell RNA sequencing has allowed a broader approach to mapping the immune cell environment in the synovial tissue. While most studies have not yet examined a direct link of the synovial microenvironment to OA pain, they highlight that the advancement of these technologies will allow delineation of the relationship between immune cells and OA pain in the future. In particular, Chou et al. performed single cell RNAseq on synoviocytes of three patients undergoing total knee replacement for medial compartment knee OA. This allowed them to identify 12 synovial cell types, including synovial fibroblasts, endothelial cells, and various types of immune cells. In addition, they were able to molecularly define the synovial cells producing particular cytokines, for example, inflammatory macrophages and dendritic cells expressing HLA-DQA1, HLA-DQA2, oxidized low density lipoprotein receptor 1 (OLR1), or TLR2 were linked to production of IL-1β⁸³. Another study used bulk RNAseq to characterize macrophages purified from the synovial tissue of subjects with OA or with inflammatory arthritis⁸⁴. Gene expression analyses and phenotyping revealed a heterogeneous population of macrophages within the OA samples, one of which were “inflammatory-like” macrophages that had a proliferative phenotype and aligned closely with inflammatory arthritis macrophages. While neither of these studies incorporated pain as a variable, they clearly illustrate how such approaches may aid the identification of specific cell types and their interactions with other cell types such as neurons, and how they contribute to particular aspects of the pathology, including pain. To illustrate this point, a recent single cell RNAseq study uncovered that synovial fibroblasts from painful sites of the patellar synovium were distinct from fibroblasts from non-painful areas, with an altered gene signature promoting fibrosis, inflammation and growth of neurons⁸⁵, supporting the idea that synovitis is linked to neuroplasticity, and thus pain. Finally, a bulk RNAseq study analyzed synovial biopsies collected from knee OA patients with either low (0–3) or high (7–10) visual analogue scale (VAS) pain levels at the time of surgery⁸⁶. Interestingly, while no major differences in the cellular composition were noted between pain groups, their analysis identified neuronal genes such as NTRK2, encoding the receptor for the neurotrophin BDNF, as differently upregulated in the high vs. low pain groups. The detailed identification and characterization of synovial tissue immune cells and molecular profiles, and their interactions with sensory neurons hold great promise for unraveling precise mechanisms of pain generation in the joint.

4. Neuroinflammation: An emerging player in OA pain

Inflammatory responses in the central and peripheral nervous system are considered “neuroinflammation”, a process increasingly recognized to contribute to chronic pain⁸⁷, including in OA⁸⁸. A hallmark of neuroinflammation is the activation of glial cells, such as microglia, as well as infiltrating macrophages in the DRG and the dorsal horn, which release chemokines and other mediators, and through neuroimmune interactions, modify pain signaling. The description of neuroinflammation in OA is still its infancy, as briefly summarized below.

4.1. DRG

Neuroinflammation has been described in several rodent models of OA. For example, in a rat CiOA model, DRG satellite glial cells become activated, alongside microglia activation in the dorsal horn – coinciding with pain development⁸⁹. In addition to activation of resident glia, a growing body of recent research implicates immune cell infiltration of the DRG, in particular macrophages, as a contributor to pain in OA^47,90. Raoof et al. showed in two rodent models of OA that M1-like macrophages accumulate in the DRG away from injury site. Furthermore, inhibition of these M1-like macrophages in the DRG by intrathecal injection of IL-4/IL-10 fusion protein or M2-like macrophages resolved persistent pain⁹⁰. Microarray analysis of DRGs from a murine OA model distinguished genes differentially expressed in early vs. persistent OA pain, and identified genes that contribute to the initiation and maintenance of OA pain and neuroinflammation canonical pathway, including Cx3cl1, Ccl2, Tlr1, and Ngf⁹¹.

4.2. Dorsal horn

Ongoing C-fiber input in the dorsal horn markedly increases second-order neuron excitability, causing activation of microglia and astrocytes, which produce a host of cytokines and chemokines that further amplify the signal³⁰. In experimental models of OA, these processes are only beginning to be studied, and activated microglia have been described in the rat MIA model⁹², as well as in a surgical model of OA, where microgliosis was a feature of late-stage OA and was associated with persistent pain⁹³. At this time, it is not clear if targeting microglia activation can modify pain. A recent study in a rat MIA model reported that inhibition of glia with minocycline and fluorocitrate alleviated mechanical allodynia. This activation state was present alongside increased joint nociceptor innervation and DRG expression of Activating Transcription Factor 3 (ATF3), a neuronal stress marker. The authors hypothesized that microglial activation induced proinflammatory changes that contributed to OA pain⁹⁴.

Nuclear factor-kappa B (NF-κB) is a transcription factor with well described roles in inflammatory responses. Previous studies have reported the association of NF-κB activity with chronic pain in the complete Freund’s adjuvant (CFA) model of inflammation⁹⁵, vincristine-induced neuropathy model⁹⁶, and an RM-1 induced bone cancer pain model⁹⁷. Li et al. sought to identify the mechanism NF-κB, astrocyte proliferation and pain in the rat MIA OA model. They found that both the number of astrocytes, and expression of NF-κB/p65 in astrocytes in the dorsal horn increased significantly. Furthermore, upon spinal inhibition of NF-κB/p65, mechanical hyperalgesia was alleviated and significant decreases were measured in the expression inflammatory cytokines IL-1β, TNF-α and IL-33 in the dorsal horn⁹⁸.

While we know very little about the precise mechanisms operating in OA, neuroimmune interactions clearly modify pain pathways, not just in the joint but also in the peripheral and central nervous systems. Information on microglial and astrocyte activation in the brain is scant, but we identified one recent study that reported supraspinal astrocyte activation as a mechanism underlying anxiety-augmented pain behaviors in rat MIA⁹⁹. Furthermore, microglia activation was recently shown in a murine model of rheumatoid arthritis, illustrating the wide impact of joint inflammation on the nervous system¹⁰⁰. Hence, it can be expected that the study of neuroinflammation holds great potential for identifying specific pathways that may be targeted to attenuate pain in chronic arthritis.

5. Systemic inflammatory inputs to pain

Chronic inflammation can be associated with increased systemic levels of cytokines and other mediators that may have a profound effect on the body’s homeostasis, including pain. Below, we briefly discuss factors that have been associated with systemic inflammation, including obesity, and how they may affect nociception, sensitization, and chronic pain in the context of OA.

5.1. Obesity

A correlation between obesity and OA has been historically recognized, with research showing that in as many as 24% of OA patients, disease can be attributed to being overweight or obese¹⁰¹. In addition, being overweight/obese is also associated with increased OA pain¹⁰². Mechanical overload of the joint may contribute to this, but obesity has also been associated with OA in non-weight bearing joints, suggesting that biomechanical loading may not be the sole cause of OA pain in obese or overweight patients and that other factors such as systemic and/or local inflammation may contribute^103,104. In support of this, the association between metabolic syndrome and knee OA is greater for symptomatic OA compared to radiographic OA¹⁰⁵. Links between obesity and non-mechanical painful disorders such as headaches and migraines have also been reported, suggesting a direct role for diet and intake of essential fatty acids in chronic pain, with suggested links to altered inflammatory states¹⁰⁶.

Several studies in experimental models have attempted to unravel these complex relationships between obesity, adiposity, loading, inflammation, and pain. In mice with high-fat induced OA, wheel-running is protective against joint damage rather than damaging, further suggesting that increased joint loading is not sufficient to explain OA¹⁰⁷. The authors found that exercised mice had improved glucose tolerance and disrupted proinflammatory cytokine networks in the absence of weight loss. Song et al. evaluated pain behaviors in two strains of rats subjected to high-fat diets, one strain in which obesity is induced and one strain in which obesity is not induced. Although no changes in circulating inflammatory cytokines were measured, an infiltration of macrophages into the DRG was observed, suggesting a link between high-fat diets and the nervous system in chronic pain¹⁰⁸. Finally, a recent study showed that mice with lipodystrophy were protected from joint damage and knee hyperalgesia after DMM, even under a high fat diet. Adding back a mixture of subcutaneous and visceral adipose tissue from wildtype mice could overcome this protection¹⁰⁹. This study suggests an important role for adipose tissue as a mediator of joint degeneration and associated pain in mice, independently of body weight. Although the link between obesity and OA progression may not be entirely due to overloading of joints, physical activity has long been a recommendation for the treatment of OA. The benefits of exercise for OA patients are clear, with studies showing clear improvements in disease burden and pain scores^110–112. Conversely, during Covid-19 lockdowns, researchers found that VAS and WOMAC scores increased significantly, correlating with a reduction in physical activity¹¹³. A recent study demonstrated that diet and exercise intervention in overweight and obese people suffering from knee OA resulted in an improvement of pain and function that was independent from BMI¹¹⁴. This was attributed partially to a change in serum levels of inflammatory factors, including IL-6, TNF-α, soluble IL-1 receptor and C-reactive protein (CRP). The association of obesity and osteoarthritis is clear, however understanding the relative contribution of mechanical, inflammatory, and metabolic factors to OA pain requires further investigation.

5.2. Other systemic factors that can modify the inflammatory aspects of OA pain

Several modifiable and unmodifiable factors profoundly affect the inflammatory process. Inflammation is considered one of the mechanisms that contribute to age-related diseases, and chronic age-related stimulation of the innate immune system is termed “inflamm-aging”¹¹⁵. Aging is one of the major risk factors for radiographic as well as symptomatic OA, and research is increasingly uncovering how systemic and local age-related inflammation can contribute to OA joint damage¹¹⁶. With age, levels of inflammatory mediators such as IL-6 and TNFα increase systemically and in adipose tissues, while in the joint, senescent cells accumulate and the senescence-associated secretory phenotype (SASP) creates a locally damaging environment¹¹⁷. However, the relationship of inflamm-aging and OA pain remains understudied. It has been shown that systemic levels of CRP, TNFα and IL6 are associated with knee pain in older adults¹¹⁸. Interestingly, the authors found that when corrected for the presence of radiographic OA, these associations of inflammatory markers and knee pain remained.

Sex is a major determinant in the inflammatory process, and while differences between males and females in both the disease progression and pain burden of OA have been widely reported¹¹⁹, it is not at all yet clear how this relates to mechanisms of pain. Females have a higher symptom burden and are more likely to use health care, resulting in a higher prevalence of clinical diagnosis^120,121. This gender discrepancy becomes particularly apparent in older adults, after 50 years of age¹²⁰. Mun et al. hypothesized that an underlying mechanism of sex-based risk differences in developing OA might be due to exaggerated inflammatory responses to pain in women compared to men. The authors measured serum levels of IL-6 over time following laboratory-induced pain and found that IL-6 levels were significantly increased in women compared to men after painful stimulus¹²². Similarly, Perruccio et al. explored sex-based differences of systemic inflammatory markers in OA patients scheduled for knee arthroplasty, and found that IL-6 was positively correlated with knee pain in females only. Furthermore, they found positive correlations of IL-1β and IL-18 with knee pain in males, but these same cytokines were negatively correlated with knee pain in females¹²³. Clearly, this will be an important area of study for the successful development of analgesics in OA.

6. Conclusions and future directions

In summary, recent research has generated crucial novel insights into the generation of OA pain. Notably, OA joint pathology is characterized by profound neuroplasticity, with the sprouting of nociceptors in the subchondral bone and the synovium. Secondly, innate immune and inflammatory mechanisms mediate OA pathology and pain at multiple levels, and interact with sensory neurons in the joint and also in the nervous system. It can be expected that the precise molecular elucidation of these neuroimmune interactions will offer a rich substrate for targeted interventions. Several modifying systemic factors will need to be taken into account when studying these mechanisms, including sex, age, and obesity.

Finally, it should be considered that OA is strongly mechanically driven, promoted by joint overload. ‘Mechanoinflammation’ has been suggested as a mechanical signaling pathway driving inflammatory signaling and pain in OA¹²⁴. One potential pathway that may be of particular interest is the mechanosensitive channel, Piezo2, expressed by nociceptors. Piezo2 becomes sensitized under inflammatory conditions and plays a role in mediating mechanical allodynia^125,126. Interestingly, a link between NGF and Piezo2 has also been proposed as a mechanism of activation of ‘silent nociceptors’ – nociceptors that begin to sense mechanical stimuli after sensitization¹²⁷. Therefore, the interactions between inflammatory pathways and mechanical stress on the joints may well be a unique feature of OA pathogenesis and the genesis of OA pain.

Key Points:

• Chronic pain is the most burdensome symptom for OA patients, yet pain is often inadequately managed in OA.

• Nerve growth factor (NGF), inflammatory cytokines and chemokines, and DAMPs are pro-algesic through direct activation of nociceptors.

• Neuronal sprouting occurs in the synovium and subchondral bone of OA joints, and neuroimmune interactions contribute to the genesis of pain.

• Neuroinflammation in the dorsal root ganglia (DRG) and the dorsal horn further amplify pain signaling and may contribute to pain chronification.

• Factors such as obesity and sex can alter systemic levels of cytokines contributing to chronic inflammation in OA, and affect sensitization and pain.

Synopsis:

Chronic pain is a substantial personal and societal burden worldwide. Osteoarthritis (OA) is one of the leading causes of chronic pain and is increasing in prevalence in accordance with a global aging population. In addition to affecting patients’ physical lives, chronic pain also adversely affects patients’ mental wellbeing. However, there remain no pharmacological interventions to slow down the progression of OA, and pain-alleviating therapies are largely unsuccessful. The presence of low-level inflammation in OA has been recognised for many years as a major pathogenic driver of joint damage. Inflammatory mechanisms can occur locally in joint tissues, such as the synovium, within the sensory nervous system, as well as systemically, caused by modifiable and unmodifiable factors. Understanding how inflammation may contribute to, and modify pain in OA will be instrumental in identifying new druggable targets for analgesic therapies. In this narrative review we discuss recent insights into inflammatory mechanisms in OA pain. We discuss how local inflammation in the joint can contribute to mechanical sensitization and to structural neuroplasticity of joint nociceptors, through pro-inflammatory factors such as nerve growth factor, cytokines and chemokines. We consider the role of synovitis, and the amplifying mechanisms of neuroimmune interactions. We then explore emerging evidence around the role of neuroinflammation in the dorsal root ganglia and the dorsal horn. Finally, we discuss how systemic inflammation associated with obesity may modify OA pain, and suggest future research directions.