The Role of Inflammation in Osteoarthritis

NIH Campus, Building 31, Room 6C10
9:00 a.m. – 4:00 p.m.

Background

The overall goal of all NIAMS roundtables is to discuss scientific and clinical needs, and to listen to concerns and challenges facing the scientific community. These sessions provide a valuable source of input for the NIAMS planning process. This specific roundtable focused on recent advances in inflammation research to enhance understanding of the early process involved in the initiation and onset of osteoarthritis (OA). The discussion prompted possible avenues for interventions to prevent or delay disease development. Part of the Institute’s scientific planning process, the full-day roundtable addressed topics including the impact of inflammation, the evidence for activation of inflammation in the development of OA, and the relationship of inflammation to other pathogenic pathways in OA. Additionally, the group discussed how interventions during the initial injury of a joint and the ensuing healing process might influence the later development of OA.

Approach

Scientists from across the country with expertise in OA were invited to participate in this roundtable discussion. In advance of the meeting, participants were encouraged to consult with colleagues on several questions, including:

How does the emerging data with regard to the link between OA and the inflammatory process impact the current view of the pathophysiology of disease?
How can these findings on pathways, mechanisms and mediators of inflammation be prioritized for their role(s) in joint tissue damage?
What is the role of inflammation in the development of OA after joint trauma and/or injury?
What impact would inhibition of inflammation have in human disease? At what point of the inflammatory pathway would such an intervention be important?
What new or existing biomarkers might predict the point at which an intervention could/should be made, and how that intervention might influence the development of OA?
What are the limitations of currently used animal and in vitro models of OA with regard to translation of new treatments to human disease?
What new or re-purposed treatments for OA might be envisioned based on these specific views of the disease? Can pharmacological treatment of inflammation, without simultaneously addressing other risk factors, prevent disease onset or progression?
What are the next steps in regard to translating this information to improve prevention, diagnosis, and treatment of OA?

These topics and responses served as the basis for the discussion. Although not all of these topics were discussed in depth at the roundtable, NIAMS leadership and the appropriate program staff read each comment. The NIAMS greatly appreciates the community’s input on these questions.

Overview, Perspectives, New Visions and Challenges

Knee OA has been viewed as being a more mechanically-driven, rather than inflammatory, disease. However, it has long been known that low grade inflammation plays some role in the processes of cartilage degeneration and repair at many stages of the disease. In the past several years, numerous scientific papers have been published in both animal models and human studies indicating an increasingly important role for both inflammation of the synovium, and activation of the inflammatory complement system (the membrane attack complexes) in the pathogenesis of knee OA. Additionally, there is increasing evidence that major risk factors for knee OA are associated with alterations in systemic and local (at the articular cartilage chondrocyte level) cytokines and inflammatory mediators. Several circumstances where inflammation may be involved in the genesis and/or progression of OA are described below.

Obesity

While the mechanical impact of obesity is associated with an increased risk of OA, there is evidence from animal studies that body weight itself may not be the only causative factor. If body weight is high, but pro-inflammatory cytokines are low, joint damage is mitigated. Additionally, human studies have shown that there is an increased risk of disease in both weight-bearing and non-weight-bearing joints of obese individuals. The close link between obesity and OA may be explained, in part, by the high levels of circulating pro-inflammatory cytokines, including those released from adipose tissue.

Post-Joint Injury

Injury to the joint has been shown to result in the production of anabolic growth factors and catabolic cytokines at the site of the lesion, which over time can lead to broad regions of damage. Recently, complement protein has been shown to play a major role in the development of OA in a mouse model of joint destabilization. Numerous other molecules have been suggested as possible targets to prevent cartilage damage post injury, such as reactive oxygen species, tumor necrosis factor ?, caspases and interleukin-1 (IL-1). However, the precise role and timing of these inflammatory processes are not well-described or understood. While in vitro and animal data exist demonstrating the cartilage response to inhibition of these inflammatory mediators, only short-term studies have been carried out in humans, with relatively weak indicators of long-term clinical outcomes. OA related to joint injury will be referred to as post-traumatic OA (PTOA) in this summary.

Aging

Clearly, there are interactions between the described risk factors for OA and the aging process. While inflammatory mediators have been implicated in OA, particularly in the early stages of disease, under the circumstances noted above, questions remain about how these other risk factors interact with the aging process in all the related tissues. Several mechanisms that initiate inflammatory responses, such as deficiencies in cellular homeostasis mechanisms, cellular senescence, extensive post-translational modification of the extracellular matrix, and crystal deposition, are specific to aging joint tissues. Understanding these inflammatory stimuli may provide unique opportunities for therapeutic interventions, especially in OA first manifesting in older individuals.

A better understanding of the interactions between the inflammatory process and pathophysiology of OA may provide possible treatment options that could improve the outcomes for patients at various progressive stages — including pre-symptomatic. The remainder of this document summarizes the discussion points for each of the previous questions, but does not indicate consensus.

How does the emerging data with regard to the link between OA and the inflammatory process impact the current view of the pathophysiology of disease?

OA has long been considered to be a largely degenerative disease, although numerous studies over the past few decades indicate an important role for inflammation in the progression of OA and its severity. Inflammation in OA is very different than the inflammation seen in rheumatoid arthritis (RA). Inflammation in OA is characterized as an innate immune response, typically comprised of the cells and mechanisms that defend us from infection by other organisms in a non-specific manner. Recent studies indicate that a part of the innate immune system called the complement pathway is involved in OA. The complement system is a biochemical cascade of the immune system that helps the body clear pathogens and other toxic materials. It is thought that the inflammatory pathways in OA are activated in response to joint injury, rather than an infection. However, some still consider that inflammation is a secondary feature of OA, occurring as a result of tissue injury and the reparative process which follows. In contrast, the inflammation in RA is characterized as an autoimmune response, in which the body attacks its own tissues.

When inflammation occurs in the development of OA, its role in the continued progression of the disease provides critical information for the understanding of the disease process. However, it is also necessary to recognize that, while inflammation can lead to tissue damage, it is also central to the repair process.

A complete understanding of OA requires consideration of the joint as a whole organ, including the synovium, cartilage, bone, menisci, adipose tissue, and muscle. There is evidence that inflammatory processes and mediators may play important roles in the initiation and progression of OA. The development of synovitis, an inflammation of the synovial membrane which lines the joints, may occur as a response to joint injury, but could also be the initiating event that leads to the development of OA. Emerging data suggests that all components of the joint, including the cartilage, synovium, menisci, and subchondral bone, participate in the inflammatory process. Inflammation of subchondral bone, located just below the cartilage, has been revealed as bone bruises, bone marrow lesions, or cysts on MRI. In the development of OA, subchondral bone becomes thickened, even before there is loss of articular cartilage. The cysts are lined with inflammatory cells that release molecules into the joint space which drive bone remodeling and the development of synovitis. In a circular fashion, factors released from the bone and synovium can also promote the degradation of joint cartilage. A variety of proteins, including fibronectin, hyaluronic acid, and collagen can activate and stimulate the inflammatory pathways, which can then set the stage for the initiation of cartilage breakdown and the onset of OA.

The role of cytokines in the initiation and progression of disease has been well studied. Cytokines can produce both pro-inflammatory and tissue destructive responses, but they can also promote the repair process in tissues. The role of IL-1 is not clear, since only a subpopulation of individuals with OA have elevated levels in their synovial fluid, and those are not drastically higher than in normal individuals. Studies have suggested that IL-1 may have a role at the very first stages of triggering a rapid destruction of cells in the joint cartilage, but it is not clear how far its effect is extended. Other biological factors, like matrix fragments, Toll-like receptors, IL-15/IL-17, and reactive oxygen species may also be important in the development of OA. Chondrocytes, or cartilage cells, from PTOA and in the OA lesions in older individuals manifest higher levels of TNF-alpha, IL-1 beta, and IL-6 and higher levels of expression of specific receptors for these cytokines, suggesting they too may play a role.

Inflammation may be mediated by the chondrocytes once they become damaged due to abnormal wear and tear or trauma. There is some evidence that chondrocytes make specific pro-inflammatory molecules that help to initiate and perpetuate the low level inflammation in the OA joint. Chondrocytes are also susceptible to mechanical stress which can trigger the repair response. Stimulation of this repair response is necessary; however, excessive or extended inflammation can also lead to more tissue damage. Anti-inflammatory agents may help protect the cartilage from further damage by inhibiting an excessive inflammatory response.

Inflammation may mediate the pathophysiology in a subset of individuals with OA, especially those with PTOA. The inflammatory processes initiated as a rapid response to a joint or cartilage injury may be very different from the low level, chronic inflammation caused by obesity or other inflammatory mechanisms. However, the role of inflammation in idiopathic or obesity-associated OA is less clear, especially with regard to systemic inflammatory mediators. Adipokines are cell-to-cell signaling proteins secreted by adipose tissue. The role of several adipokines, including leptin, adiponectin, visfatin, and resistin in the development of OA are beginning to be clarified. Studies in this area may also explain the development of OA in obesity, independent of the mechanical stresses on the joints.

Precise delineation of the contribution of inflammation to OA pathogenesis, clinical manifestations, clinical course, and outcome is an important area for scientific inquiry. Recognizing the role of inflammation in the development and progression of OA will open up a host of therapeutic avenues focusing on the inflammatory pathways.

How can these findings on pathways, mechanisms and mediators of inflammation be prioritized for their role(s) in joint tissue damage?

The initiation of OA, both its timing and cause, may be more difficult to determine in humans than in animal models. Many factors coalesce for the development of disease, including an inflammatory episode, acute damage, age, obesity, and malalignment. It is important to use information gained from animal models to guide prioritization of further research. However, most animal models replicate trauma or instability of the joints. There is a need to develop more animal models with "spontaneous" OA because disease development may be different when the joint is injured compared to OA that spontaneously develops due to other factors. Also, animal models tend to only look at OA in the early stages of disease development, not over the long-term. Specific examples include using animal models to determine the major inflammatory pathways in PTOA, and then validating these pathways in humans with PTOA.

The mechanisms involved in all stages of OA and degenerative joint disease are not well identified, specifically the interactions between mechanics and biology and the role of inflammation in these interactions. There is a need to learn more about the normal biological processes in the joints to be able to interpret findings from damaged joints. The role of chronic, low level inflammation is also an important factor to consider.

The conduct of clinical studies in OA is a significant challenge; populations are heterogeneous in terms of stage of disease and risk factor profiles. It will be difficult to find a single target or a single therapy that will work for everyone. However, there may be subsets of individuals where proof-of-principle studies can be done to focus on specific pathways and mechanisms. Phenotyping of individuals is very common in the cancer community to determine the exact nature of the disease and the best way to treat it. It would be helpful to be able to phenotype patients with OA to determine their inflammatory profiles.

What is the role of inflammation in the development of OA after joint trauma and/or injury?

Joint trauma has two aspects. The first involves cell and tissue damage occurring shortly after injury and sometimes persisting for weeks. This mechanical damage or stress stimulates biologic processes that, if unchecked, may promote progressive joint degeneration. For example, joint injuries result in the release of mediators, such as reactive oxygen species, tumor necrosis factors, interleukins, and matrix protein fragments that can cause progressive tissue damage. The role of joint inflammation in the progression of PTOA is supported by studies showing that leukocyte-derived cytokines and oxidants are damaging to the articular cartilage, and therapies targeting individual cytokines and reactive oxygen species protect the cells of the cartilage. The second aspect is the biomechanical disruption of joint function that includes both short- and long-term adaptive and remodeling responses in the bone, cartilage, ligaments, and joint capsule to adapt to the alterations in the biomechanical demands.

Injury to the anterior cruciate ligament (ACL), regardless of whether reconstruction is performed or not, leads to a higher risk for developing OA. Individuals with injury to the ACL could be an important study group to assess the role of inflammation and to follow the long-term development of OA. Some individuals experience a rapid onset of PTOA following injury to a joint. In others, OA takes many years to develop. It would be useful to develop imaging techniques and to identify biomarkers to determine which individuals are at risk for developing OA more rapidly. A paper published recently showed that 100 percent of individuals with severe knee injuries went on to develop OA. Most of these knee injuries involved fractures with other soft tissue damage.

A significant barrier to progress in the treatment of joint injuries is our incomplete knowledge of the origins and propagation of post-traumatic inflammation. How and when does minimal and localized tissue damage within a joint trigger a strong, organ-wide inflammatory response? It is important to recognize that inflammation may be beneficial to early joint repair. Suppressing this beneficial inflammation may do more damage than good.

Activation of the inflammatory process is important not only in the joint, but also in the surrounding tissues, including adipose, bone, synovium, cartilage, and meniscus. One must also consider how the development of inflammation differs depending on the joint involved (e.g., ankle, knee, and hip).

What impact would inhibition of inflammation have in human disease? At what point of the inflammatory pathway would such an intervention be important?

The initiators, mediators, and pathways associated with inflammation differ during the development and progression of OA. Knowledge of the disease stage and the acute or chronic state of inflammation in the joint is critical in the selection of therapeutic strategies to target inflammation. The early therapeutic targeting of inflammation could potentially prevent or attenuate its progression to a chronic inflammatory state. However, it is important to not only treat the inflammation, but to also treat the symptoms, such as pain and swelling. Further research may help to identify what parts of the inflammatory cascade should be targeted for interventions. Consideration should be given to the effects of inhibiting inflammation in the whole body over the long-term. For example, non-steroidal anti-inflammatory agents may have the potential to retard the development of OA; however, their long-term clinical use is limited by severe side effects.

OA is a multifactorial disease, so targeting a single inflammatory pathway may not be successful without addressing other factors. However, timing of the intervention in the general population becomes more difficult when OA arises due to other factors, such as obesity. It may be easier to identify a specific time to intervene when OA develops as a result of injury to the joint.

What new or existing biomarkers might predict the point at which an intervention could/should be made, and how that intervention might influence the development of OA?

While the cost of developing and validating biomarkers can make them difficult to utilize, this is an area of opportunity for OA research. The Osteoarthritis Initiative (OAI) partnership with the Foundation for the NIH (FNIH) and the Biomarkers Consortium could yield critical information in the search for both imaging and biochemical biomarkers of OA. The OAI was started with the intent to identify and follow a cohort of individuals at high risk for developing knee OA. However, individuals were also enrolled who already had established OA. The OAI has continued to follow these individuals for progression of disease for almost 8 years. The current partnership with the FNIH’s Biomarkers Consortium focuses on the search for biomarkers that might predict the progression of disease.

Standard biomarkers of inflammation include IL-1, TNF-alpha, chemokines, IL-8, IL-6, adipokines, and collagen derivatives of nitrous oxide; however, the biochemical mediators of inflammation may differ in OA depending on the stage of disease. Imaging biomarkers, such as bone marrow lesions and other changes in subchondral bone, can also provide insight into the development of OA. Multiple sources of biomarkers, including urine, serum, and synovial fluid, could also be considered. Inflammation is a common thread in obesity, cardiovascular disease, and metabolic syndrome, and it is important to differentiate these general serum markers from those that may be more specific for OA. Markers of synovial proliferation or activation could also be further investigated in OA. Biomarkers to measure these various stages of disease may provide valuable information for the most appropriate treatment, and for the measurement of response to treatment.

Aging is a major risk factor for a number of diseases, including OA. There are ongoing efforts to address the contribution of aging in the development of OA in the absence of injury. It is important for the research community to collaborate on the study of common risk factors in the development of a number of diseases. A recent meeting focused on aging as a risk factor for the development of disease; a workshop last year focused on inflammation as a risk factor in age-related disease development.

What are the limitations of currently used animal and in vitro models of OA with regard to translation of new treatments to human disease?

While animal models are necessary to identify translational therapies, the models currently available are not sufficient. Many of the results obtained in animal models do not translate into the same results in human studies. Animal models are best used when trying to understand particular aspects of OA, rather than the full disease process as it occurs in humans. Some of the limitations of animal models include a lack of aged animals, genetic diversity, and disease comorbidities often associated with OA. There is a lack of models to optimally study the effects of aging and obesity, for example, on the development of OA. However, these same issues can also be beneficial when studying OA in animal models because it allows one to remove sources of variation when looking at disease mechanisms. For example, surgical destabilization of a joint in an animal mimics the trauma seen in humans in that a normal, healthy joint is injured and then exposed to altered loading, which then leads to adaptive and reparative responses. This tests the ability of the joint to repair itself, regain normal function, and prevent further progressive damage.

Biomechanics of the joint is an important factor in the development of OA, and more evidence indicates that abnormal biomechanics may also alter inflammatory pathways. The biomechanics of the knee joint are very different between rodents and humans, and it is difficult to translate measurements of biomechanics from rodents to humans. In addition, differences in locomotion between rodents and humans have been an understudied area of research. Improvements in technology would enable better characterization of joint mechanobiology in animal models. OA research in animal models is almost entirely carried out on young animals. Natural development of OA is often a long and slow progression in humans; it is difficult to replicate this in an animal model with a relatively short lifespan. In addition, the reparative processes and degenerative processes differ as we age.

To date, there has been some success in treating OA in animal models; however, this has not translated to successful treatments for humans. Although studies in animal models have been successful, their predictive value remains uncertain. The slow disease progression over many years also presents a challenge to the design of clinical studies. The current treatment guidelines for OA include the use of acetaminophen, glucosamine, chondroitin sulfate, non-steroidal anti-inflammatories, COX 2 inhibitors, steroids, and opioids, which are not disease-modifying agents. Currently, some ongoing trials of disease-modifying drugs are in progress. However, there are limitations in the ongoing and completed trials for agents to treat and/or prevent OA, including defining endpoints, recruiting the appropriate patient populations given the heterogeneity of OA and the variety of risk factors for its development, and the advanced stage of disease.

The variety of risk factors and the stage of disease are very important when looking at the impact of clinical trials outcomes. Risk factors for OA include genetics, misalignment, obesity, joint overuse, injury, age, and meniscus, ligament or tendon abnormalities. Additionally, individuals with only age as a risk factor may not have symptoms of disease; individuals with many risk factors may not respond to drug treatment; and individuals with joint injury may show early signs of disease. To date, trials of disease-modifying drugs have been carried out in individuals with advanced disease but who have progressed slowly to that point. It is important to determine not only how OA can best be treated, but also the optimal time for intervention.

What new or re-purposed treatments for OA might be envisioned based on these specific views of the disease? Can pharmacological treatment of inflammation, without simultaneously addressing other risk factors, prevent disease onset or progression?

Therapeutic interventions need to take into consideration several factors, including treatment of the acute joint injury if present, treatment of the resulting inflammation, and treatment of any co-morbidity, such as obesity or abnormal biomechanics, which may increase the likelihood of developing OA.

It is important to develop novel ways to treat inflammation resulting from trauma to the joint. Aspiration of synovial fluid from the joint immediately following injury may reduce inflammation. Additionally, it is important to determine if the administration of intra-articular injections of anti-inflammatory agents after trauma can reduce the development of PTOA, and if so, the appropriate time course for these injections. Anti-inflammatory agents in tandem with surgical repair would be beneficial when there is an acute joint injury. Anti-inflammatory agents would likely not be very effective in individuals with end-stage OA or severely misaligned joints. Anti-inflammatory agents might not be the choice for treatment over a long period of time due to their side effects. The use of systemic anti-inflammatory agents, including biologics, should be fully studied in clinical trials.

The development of true disease-modifying agents for OA may be difficult, since it is generally a localized disease with good treatment options available, including total joint replacement for end-stage disease. Disease-modifying agents would have to be taken over a period of several years or decades, be proven to be safe for long-term administration, and be effective at improving pain and function.

It is also important to take into account the patient’s perspective when treating a long-term condition such as OA. Individuals with OA are likely to be maintained on anti-inflammatory agents for many years and a number of the treatment options have severe side effects. It may be difficult to recruit them to volunteer for a new drug treatment trial which may or may not provide relief. The identification of potential therapeutic targets that slow down progression or improve pain and function is important.

The National Center for Advancing Translational Sciences (NCATS) at NIH is facilitating the re-purposing of therapeutics by using information from both successful and unsuccessful clinical trials to inform other studies.

What are the next steps in regard to translating this information to improve prevention, diagnosis, and treatment of OA?

The ability to stratify the disease and identify OA subtypes would facilitate development of new disease modifying interventions. Stratifications might include mechanisms of disease onset, stage of disease, number and types of involved joints, and presence/absence of inflammation. Biomarkers may help to identify different subsets of OA that may not share the same pathogenic mechanisms and enable the development of new, targeted treatments.

Translation of basic findings to human health requires pre-clinical testing in practical, reproducible, and clinically relevant animal models. However, with OA, there are not only differences related to species, but also those related to biomechanics and anatomy. These differences limit the translation of findings from smaller species such as rodents to humans. There are opportunities to use valid models to identify context-specific (in this case inflammation-related) disease mechanisms that may identify new treatment targets.

The link between inflammation and obesity has been well characterized. Individuals who are obese and have metabolic disease are at great risk for the development of OA. It is essential to incorporate dietary factors and exercise as variables when developing animal models. Prevention of obesity will have a large impact on the development of OA. Mechanical damage/stress stimulates biologic processes that promote progressive joint degeneration. It is critical to educate the population about both the risk factors for the development of OA and ways to potentially prevent it, such as weight loss and regular, moderate exercise.

Further studies are needed to improve the assessment of inflammation-related changes caused by OA in joint tissues, particularly the articular cartilage, subchondral bone, and the synovial membrane, using both currently available and novel imaging techniques. It is important to establish the relationship between synovial inflammation, subchondral bone changes, and cartilage damage. The development of both biochemical and imaging markers is also needed. The development of a safe, accurate, reproducible, non-invasive, and non-contrast method to assess synovitis would be beneficial.

Investigators are currently working to better define inflammatory targets in OA. Questions remain with regard to the source of the inflammation (i.e., mesenchymal cells or classic inflammatory cells) and the specific mediators. Interventions to prevent deleterious inflammation (e.g., anti-inflammatory medications, chemokine inhibitors, tumor necrosis factor, interleukin antagonists) could be investigated in pre-clinical models. Local intra-articular cell stress, necrosis, and inflammatory mediators may play a critical role in post-traumatic cartilage loss.

One final area of importance, which could play a role in the clinical management of OA but was not discussed in detail at this roundtable, is the role of inflammation in pain, and in particular, the role of inflammation in pain sensitization. TNF-alpha and IL-1 beta both are thought to be contributors to the transition from acute to chronic pain in OA. The transition from acute, activity-related pain to chronic, persistent pain is not well understood. Targeting inflammation to prevent sensitization and transition to chronic, persistent pain would have a tremendous public health impact. Clinical trials could determine the effects of anti-inflammatory agents in individuals with both early primary and post-traumatic OA.

This discussion yielded many areas of interest for future research endeavors.

Participants

ABRAMSON, Steven B., M.D., New York University School of Medicine
BUCKWALTER, Joseph A., M.S., M.D., University of Iowa
CHU, Constance R., M.D., University of Pittsburgh
EVANS, Christopher H., D.SC., Ph.D., Beth Israel Deaconess Medical Center
GOLDRING, Steven R., M.D., Hospital for Special Surgery
GRIFFIN, Timothy, Ph.D., Oklahoma Medical Research Foundation
JIMENEZ, Sergio A., M.D., Thomas Jefferson University
LOESER, Richard F., Jr., M.D., Wake Forest School of Medicine
LOTZ, Martin, M.D., The Scripps Research Institute (Co-Chair)
MALFAIT, Anne-