Principles of Therapy Against Equine Joint Disease

Osteoarthritis is the most common chronic joint disease causing lameness in mature equine (60%) and human (21%) populations.160-162 Pain associated with chronic joint disease is the major contributing factor associated with lameness and decreased performance in companion and sport horses alike.163 As understanding of the complex etiology and pathogenesis of this disease have grown over the past several decades, so too have the number of treatment options available to equine veterinarians. Therapies directed against osteoarthritis (OA) aim at mitigating both pain and progression of this degenerative disease. Simply put, the mechanisms of any given therapy may be separated into symptom-modifying (e.g. decrease in pain or lameness grade) and disease-modifying (e.g. prevention of further cartilage degradation). The majority of recent literature refers to therapies utilized against OA as symptom-modifying osteoarthritis drugs (SMOAD) and disease-modifying osteoarthritis drugs (DMAOD), where any given treatment may exert characteristics of one, or both. This concept is particularly important in the equine athlete, as career longevity is paramount in the successful treatment of equine joint disease. As such, the clinician should always consider the downstream SMAOD and DMOAD effects of a chosen treatment against equine joint disease.

Today’s market provides an ever-growing number of options available to the equine clinician. Treatments, including non-steroidal anti-inflammatory drugs (NSAID), corticosteroids, hyaluronan, polysulfated glycosaminoglycan, pentosan polysulfate, biologic therapies including stem cells, and a number of supplements are often administered via either intra-articular, systemic, or oral routes, and the scientific and clinical evidence of each will be discussed in the following review. The majority of presented information has been adapted from the second edition of Joint Disease in the Horse.159

Non-Steroidal Anti-inflammatory Drugs (NSAID)

The term non-steroidal anti-inflammatory drugs (NSAIDs) refers to those agents that inhibit some component of the enzyme system responsible for converting arachidonic acid into prostaglandins and thromboxane (arachidonic acid cascade). Of importance, is inhibition

Osteoarthritis is the most common chronic joint disease causing lameness in mature equine (60%) and human (21%) populations.160-162 Pain associated with chronic joint disease is the major contributing factor associated with lameness and decreased performance in companion and sport horses alike.163 As understanding of the complex etiology and pathogenesis of this disease have grown over the past several decades, so too have the number of treatment options available to equine veterinarians. Therapies directed against osteoarthritis (OA) aim at mitigating both pain and progression of this degenerative disease. Simply put, the mechanisms of any given therapy may be separated into symptom-modifying (e.g. decrease in pain or lameness grade) and disease-modifying (e.g. prevention of further cartilage degradation). The majority of recent literature refers to therapies utilized against OA as symptom-modifying osteoarthritis drugs (SMOAD) and disease-modifying osteoarthritis drugs (DMAOD), where any given treatment may exert characteristics of one, or both. This concept is particularly important in the equine athlete, as career longevity is paramount in the successful treatment of equine joint disease. As such, the clinician should always consider the downstream SMAOD and DMOAD effects of a chosen treatment against equine joint disease.

Today’s market provides an ever-growing number of options available to the equine clinician. Treatments, including non-steroidal anti-inflammatory drugs (NSAID), corticosteroids, hyaluronan, polysulfated glycosaminoglycan, pentosan polysulfate, biologic therapies including stem cells, and a number of supplements are often administered via either intra-articular, systemic, or oral routes, and the scientific and clinical evidence of each will be discussed in the following review. The majority of presented information has been adapted from the second edition of Joint Disease in the Horse.159

Non-Steroidal Anti-inflammatory Drugs (NSAID)

The term non-steroidal anti-inflammatory drugs (NSAIDs) refers to those agents that inhibit some component of the enzyme system responsible for converting arachidonic acid into prostaglandins and thromboxane (arachidonic acid cascade). Of importance, is inhibition of the prostaglandin E (PGE) series of prostanoids, as they are intimately involved in pain, modifications in cartilage metabolism, and persistent inflammation in diseased joints. Specifically, elevations in synovial concentrations of prostaglandin E2 (PGE2) are recognized to be associated with synovial inflammation and cartilage matrix diminution in horses with OA. Currently, NSAIDs persist as a mainstay of treatment in cases of acute injury, due mostly to their potent, but variable, mitigation of pain and inflammation locally and at the level of the spinal cord.3 However, the potential for well-recognized undesired effects including renal and gastrointestinal toxicity prevent their use as a long-term treatment of joint disease. This may be in the clinicians favor however, as recent research suggests that inhibition of PGE2 may have long-term unfavorable effects on cartilage metabolism.4 Despite this, controversy based on current knowledge, and a lack of evidence in the horse make it unlikely that clinical use of NSAIDs should be altered.

All NSAIDs inhibit cyclooxygenase (COX) to some degree, and their action may be non-specific, or limited to inhibition of a single isoenzyme of the COX pathway (COX-1 and COX-2).1, 2 COX-1 is constitutively produced, and functions to help maintain normal physiologic parameters of the gastrointestinal and renal systems. COX-2 is inducible, and is primarily associated with mononuclear and synovial cell-mediated inflammation in equine joint disease. It is not however, as simple as COX-1 is good and COX-2 is bad. Constitutive COX-2 production has been shown in normal physiologic processes in multiple organ systems including the brain, kidney, pancreas, and bone. Also, in a murine model, COX-2 suppression delayed gastric ulcer healing.

Phenylbutazone and flunixin meglumine are the most commonly used NSAIDs in equine veterinary medicine, and may be given either intravenously or orally. Phenylbutazone, often given at a dose of 4.4 mg/kg once or 2.2 mg/kg twice daily, has a non-COX selective mechanism of action, and is considered to be one of the more potent NSAIDs for its symptom-modifying effects. Thus, it is commonly chosen as a treatment for acute equine musculoskeletal disease. This is supported by research showing decreased synovial fluid concentrations of PGE2 following experimental induction of OA, though results have been variable in treatment of naturally occurring OA.5, 6 Generally, it is recommended that phenylbutazone be administered at a dose of 4.4 mg/kg once daily. This is primarily the result of one study showing the elimination half-life of phenylbutazone to be 24 hours in exudative material, though its plasma half-life has been shown to be 5.5 hours in horses and ponies.9-12 Due to its non-COX selective action, long term phenylbutazone administration has been associated with the development of undesired effects in vivo, such as elevated serum creatinine levels, and diarrhea. Data investigating whether or not phenylbutazone is harmful to articular cartilage is contradictory. Two in vitro studies found no evidence that phenylbutazone was harmful to cartilage, However, one recent in vitro study showed decreased proteoglycan synthesis in cartilage explants exposed to serum from horses that were previously administered phenylbutazone for 14 days.7, 8, 21 These undesired effects have led to the development of more COX-2 selective NSAIDs, such as firocoxib. It has been approved for use in horses for the control of pain and inflammation associated with OA, with the thought being that the selective action should improve the safety profile of long-term administration. Secondarily, there is now a topical NSAID available in the United States that contains 1% diclofenac sodium. Extrapolating from the human literature, it is thought that topical NSAIDs could be clinically beneficial while reducing systemic side effects. In regards to joint disease, recent work has demonstrated DMOAD effects of this drug in equine OA, characterized by increased proteoglycan staining and cartilage GAG content.5

Anecdotal and objective assessment alike show the clinical ability of NSAIDs to improve lameness. The proposed mechanism by which NSAIDs provide pain reduction involves reversing alterations in the nociceptive threshold, both locally and peripherally, via inhibition of prostaglandin production.14 Multiple studies testing the efficacy of 4.4 mg/kg of phenylbutazone for pain mitigation in varying models have shown significant reductions in lameness ranging from 2 to 24 hours post-treatment.13, 15 Interestingly, one study comparing 4.4 mg/kg to 8.8 mg/kg showed no advantage of the higher dose over the lower.16 Flunixin Meglumine, often administered at a dose of 1.1 mg/kg, may provide analgesia as rapidly as 2 hours post-administration, and persist for up to 30 hours.17 When phenylbutazone and flunixin meglumine were compared in horses with navicular syndrome, both provided improvement in lameness grade and peak vertical force, with no significant difference between the two.18 However, when administered simultaneously over 5 days (oral administration of phenylbutazone and IV administration of flunixin) in 29 horses with naturally occurring lameness, the combination showed better clinical improvement 12 hours after the last dose compared to phenylbutazone alone.6 It is important to note that one horse in the group receiving both NSAIDs died of necrotizing colitis, highlighting the importance of clinical discretion.

Firocoxib is the only COX-2 selective inhibitor licensed for use in horses in the United States. When compared with oral phenylbutazone in 253 horses in a randomized controlled trial, there was no difference in clinical lameness scores between groups.19 These findings were supported by another study that showed that approximately 80% of 390 horses with OA had improved lameness scores after 14 days of oral firocoxib. Improvement was most rapid over the initial 7 days, though continued at a slower rate until day 14.20

NSAIDs are commonly used to lessen inflammation in sport horses. As such, it is important to consider their potential analgesic effects with regards to withdrawal from competition. Thorough reviews of the literature have been conducted with varying interpretations. Though there are reports arguing an association between NSAID use and musculoskeletal injury, further studies must be carried out to determine whether horses with higher plasma concentrations of NSAIDs have an altered risk of musculoskeletal injury compared with other horses.22 This is due mostly to the fact that no studies investigating this association have normalized for other possible risk factors (age, discipline, surface, length of performance, gender, training program, preexisting pathologic conditions, etc.) contributing to musculoskeletal injuries in horses. Thus the real role of NSAIDs as a possible risk factor for musculoskeletal injuries remains up for debate.

In summary, NSAIDs remain the standard of care for first-line treatment of traumatically induced inflammation, with phenylbutazone and flunixin meglumine being the most commonly used. Newer COX-2 preferential inhibitors are available, but should not be viewed as replacements to phenylbutazone and flunixin meglumine. Though NSAIDs and corticosteroids work to prevent or reduce the inflammatory response, they are not one in the same, as they exert their effects at different levels in the inflammatory cascade. COX-1 is mainly responsible for protective prostaglandins, while COX-2 plays some accessory role but is more important than previously thought. These facts, however, still might not outweigh the beneficial effects of selective COX-2 inhibition in joint disease.

Intra-articular Corticosteroids

The intra-articular (IA) use of corticosteroids for the treatment of musculoskeletal conditions has been reported since as early as 1955, and remains an important tool for practitioners caring for the equine athlete.23 Corticosteroids are potent anti-inflammatory agents via inhibition of inflammatory processes at all levels. The anti-inflammatory effect of glucocorticoids occurs through alteration of cytoplasmic receptors, while pain relief is attributed to inhibition of the phospholipase A2 enzyme and cyclooxygenase 2 (COX-2) expression in the arachidonic acid cascade.31 Other well recognized effects include reductions in capillary dilation, margination migration and accumulation of inflammatory cells, and inhibition of soluble mediators including IL-1 and tumor necrosis factor alpha (TNFα).32, 33

Hydrocortisone was the earliest reported (1955) corticosteroid used in horses, and showed profound improvements is clinical signs in most cases.23 In the following decades, corticosteroid use for musculoskeletal conditions increased, along with reports indicating corticosteroids may be harmful in horses.24 Specifically, they were indicted as producing laminitis, catastrophic breakdown, and steroid arthropathy. However, to the author’s knowledge, there has never been any scientific demonstration of a comparable response associated with corticosteroid usage in horses, but some authors continue to perpetuate the concept. Instances of degenerative joint disease caused by corticosteroids were presented without proof of such a pathogenesis.25 Likewise, data on the potential for TA to produce laminitis concludes that there is no association between the occurrence of laminitis and the IA use of TA.45 In general, the literature supports the use of total body doses of 45 mg as safe in an otherwise healthy horse. As it pertains to catastrophic breakdown injury, a role for corticosteroids in the pathogenesis has never been proven. This is supported by work showing that TA and MPA are not harmful to subchondral bone, but that it is more likely “that microdamage in the subchondral bone occurs early in the exercising athlete and this microdamage can lead to pathologic fractures.” 38, 43, 47, 48

Today, several corticosteroids are available to the modern practitioner. Those most commonly used in equine practice include methylprednisolone acetate (MPA), dexamethasone, betamethasone, and triamcinolone acetonide (TA). Long-standing debate surrounds the use of IA corticosteroids, as some individuals argue that their pain-modifying effects may lead to overuse and subsequent cartilage degradation. Early research on the intra-articular use of MPA investigated doses from 80-120 mg/joint (high by clinical standards) in normal horses and those with induced OA, both with/without subsequent exercise.26-30 These studies focused primarily on the middle carpal and antebrachiocarpal joints. Each, to a varying degree, and based largely on the selected dose and severity of the model, showed negative effects on the joint and cartilage metabolism. Deleterious results included decreased glycosaminoglycan (GAG) and proteoglycan staining, chondrocyte necrosis, hypocellularity, and cartilage fibrillation. As many of the selected models were drastically more severe than those used currently, it is likely that some of these effects were compounded by induced trauma and instability. Currently, an osteochondral fragment model (OFM) of traumatic osteoarthritis serves as the gold standard for in vitro investigation of SMOAD and DMOAD effects of joint therapy in the horse.35-37

More recent studies looking at the effect of corticosteroids on equine cartilage have shown more beneficial results by attenuating cartilage degradation. In equine cartilage explants, this action was achieved with dexamethasone and TA through decreases in interleukin 1 (IL-1) and activator protein C (APC) driven degradation. Previous work has shown that IL-1 and APC (synthesized by chondrocytes at sites of cartilage fibrillation) synergize to promote degradation of articular cartilage.34 Controlled in vivo studies further clarify the therapeutic response of intra-articular corticosteroids in the horse. A foundational study to the body of work investigating joint therapies in the osteochondral fragment model (OFM) found that two injections of 15 mg of betamethasone 21 days apart in the middle carpal joint caused no deleterious effects to the articular cartilage when compared to saline.35 Interestingly, this study also compared exercise versus non-exercise in injected joints, and concluded that there are no harmful effects of exercise in the presence of corticosteroids. Although not significant, beneficial effects were noted as is often reported clinically.

Over time, the OFM was altered to allow for one joint to operate as an internal control, by creating a fragment in only one of the two middle carpal joints on an individual. When MPA was investigated in the OFM (the only controlled in vivo study in the horse), lameness, synovial PGE2 concentrations, and synovial membrane staining were mildly improved. However, cartilage histopathologic scores were significantly worse, confirming the deleterious effects of MPA on articular cartilage exceed the benefits from inhibiting inflammation.37 These findings are supported by multiple other investigators showing that multiple injections of clinical concentrations of IA MPA inhibit development and maturation of repair tissues, are detrimental to the synovial membrane, and degrade articular cartilage.39-41 It is important to point out, however, that in one study, a single dose of MPA did not cause long term detrimental effects in the quality of cartilage repair tissues.39 In equine practice, IA corticosteroids and local anesthetics are often used together (or on the same day). Current evidence from a recent study investigating the effects of MPA and lidocaine on bovine articular cartilage showed that when chondrocytes were exposed to both simultaneously, no cells survived. The combined use of local anesthetics and corticosteroids in vivo in the horse have not been investigated.

In contrast to the results using MPA, when the effects of TA were investigated in the OFM the outcome was quite different. IA administration of 12 mg of TA produced decreased lameness scores, lower synovial fluid total protein (TP), and higher hyaluronan (HA) and GAG concentrations versus controls. Interestingly, TP, HA, and GAG concentrations were improved regardless of the joint treated (fragment vs. no fragment) indicating a remote DMOAD effect. It is important to mention that this remote effect also applies when using MPA, thus the deleterious effects of repeated use of this medication may still cause negative effects on joints other than the one directly being treated. Overall, the results support favorable effects of TA on degree of clinically detectable lameness, synovial fluid, synovial membrane, and articular cartilage, without deleterious effects on subchondral bone.36

A 2009 survey of 831 members of the American Association of Equine Practitioners (AAEP) revealed that MPA and TA are the most commonly used corticosteroids to treat equine musculoskeletal disease in the United States.46 These medications are often combined with hyaluronic acid (HA) for intra-articular use, and are utilized by convention in low and high-motion joints respectively. The most current research does not provide clear evidence of the beneficial effect for the addition of HA to either MPA or TA for intra-articular injection. In fact, a recent publication investigating reduction in lameness grade following IA treatment with either HA plus TA or TA alone was conducted. It concluded that the success rate of IA TA three weeks after treatment was 87.8% while that of TA plus HA was 64.1%. This study did, however, only investigate the SMOAD (lameness) effects of each therapy, and provided no insight into the inflammatory state of the joint following treatment. In other studies looking at the combination of MPA plus HA versus no treatment on cartilage explants, there was mild elevation in proteoglycan synthesis. Whether or not this is beneficial is unclear, as increases in proteoglycan synthesis can be an indicator of early OA.44

Does treatment with one corticosteroid last longer than another? Traditionally, duration of response to corticosteroid injection was thought to have an inverse correlation with its water solubility, however others have proposed that duration of action is more a reflection of the drug’s rate of hydrolysis and receptor binding affinity in the joint.49, 50 Though the complete mechanism driving duration of action remains unsolved, it is likely a combination of these processes, as well as total dose administered, duration of treatment, and the crystal size of the suspension.32, 50, 53 Insoluble compounds may persist in the joint for a longer period of time, however they do not incite their anti-inflammatory properties until being hydrolyzed to their biologically active forms (prodrug) for binding to the appropriate receptors. Historically, MPA has been recognized as having the most rapid onset of action compared to other commonly selected corticosteroids. In fact, MPA was shown to be hydrolyzed to methylprednisolone in as little as 2 hours, and persist as the prodrug in synovial fluid for up to 39 days.51 A clinical SMOAD effect was shown to persist for up to 42 days following IA administration of MPA.37 In contrast, TA is very rapidly absorbed from the joint into the bloodstream. Following IA administration of 6mg TA, synovial fluid concentrations peaked at 1 day following administration, and were undetectable by day 15.52 In the author’s opinion, the onset of clinically relevant improvement in pain occurs faster with TA as compared to MPA, and the duration of improvement is slightly in favor of MPA (i.e. longer duration with MPA). In practice, veterinarians often recommend varying periods of convalescence following IA administration of corticosteroids. In human patients receiving this treatment, multiple sources report a prolonged clinical response when rest from exercise is allowed after injection. Thus, as one author mentions “a period of restricted joint motion would likely reduce clearance of the medication and enable better penetration of IA tissues,” though other research exists. 31 The authors typically use 24 hours of rest following IA injection of TA, and 4-7 days when using MPA.

In summary, when utilized intra-articularly, corticosteroids provide potent anti-inflammatory effects. Between products, differences in their beneficial versus deleterious effect profiles exist. In the author’s opinions, betamethosone has no deleterious side effects, TA can promote cartilage health, and MPA has been shown to have some deleterious effects. The duration of clinical response to each product is multifactorial, though prolonged long-term effects have been documented, and are likely due to downstream effects of interaction with cytoplasmic receptors. A period of rest following IA administration may increase local tissue absorption, though exercise has not shown to be detrimental. There is no good evidence linking corticosteroids to induction of laminitis in otherwise healthy horses. Lastly, there is no evidence that HA in combination with corticosteroids mitigates negative effects, however research showing the chondroprotective nature of HA reasons that combination therapy with TA or betamethasone is appropriate.

Hyaluronan

Hyaluronan (HA) is a disaccharide molecule composed of d-glucuronic acid and N-acetyl-d-glucosamine, and is produced by synoviocytes and chondrocytes. It functions to form the backbone of the polyglycan aggrecan, and is integral to the proper function of both synovial fluid and articular cartilage. HA, like other disaccharides is comprised of repeating units, and thus may be produced in various lengths. As such, the length dictates the molecular weight (degree of polymerization), and in conjunction with overall concentration, is responsible for normal function. HA operates to lubricate the joint, and to form a barrier at the level of the synovial membrane to regulate fluid exchange.53, 54 In a diseased joint, this barrier formation minimizes white blood cell (WBC), free radical, and pro-inflammatory cytokine infiltration into the joint.55

Much research has been conducted on the effectiveness of HA as a treatment of joint disease. Multiple products exist in today’s market, and though they contain relatively the same active ingredient, they differ in several key aspects, including protein concentration, cross-linking, and molecular weight. Based on extensive review of the literature, HA with a molecular weight of greater than about 500,000 Daltons consistently shows benefit.56 The majority of commercial products available have a molecular weight of somewhere between 1-3 million Daltons, but as mentioned earlier, may vary in cross-linking and protein content. Those products that are synthetically cross-linked, are marketed as providing “viscosupplementation,” in an attempt to return the viscoelastic properties of synovial fluid in a diseased joint to normal levels. In turn, as the synthetic cross-linking of an HA product increases, so too does the molecular weight, resistance to free radicals, and retention time in the synovial space.55

Increased molecular weight does not, however, directly correlate with clinical efficacy. In one study comparing multiple treatments with mid to low molecular weight HA formulations (mid 0.8 to 1.5 million Daltons product to a low 0.5 to 0.7 million Dalton product), the mid-weight product showed statistical superiority by reducing pain in patients with knee OA by up to 50% for 6 months post-injection.57 It may then, be reasonable to recommend use of HA products with a molecular weight of 1-3 million Daltons (mid-molecular weight), as it provides lower monetary cost to the patient, and shows clinical superiority. In the horse, a dose of at least 20 mg in the middle carpal joint was needed in order to evoke clinical improvement based on force plate analysis.58 Extrapolation of this dose to other joints should be treated with caution, as each joint has a different total surface area and synovial fluid volume.

Does HA have SMAOD and DMOAD effects in the equine joint? The limited number and quality of studies investigating this question leave the answer unresolved. However, in human medicine, a comprehensive study which compiled published human trials showed a 28% to 54% reduction in pain and a 9% to 32% improvement in function for up to 18 months compared with baseline values.59 These SMOAD effects have not been confirmed in the horse. DMOAD effects in the equine joint have been confirmed, characterized by a decrease in histologic articular cartilage fibrillation, and improvements in synovial membrane parameters.62 This is supported by evidence from human literature showing that mid-weight HA provides preservation of cartilage volume and no significant cartilage loss versus controls.60 In a second human study, patients received 4 courses of mid-molecular weight HA treatments with 5 weekly injections at each treatment course. Responders showed improvement in knee OA symptoms and a significant long-term effect that lasted for at least 1 year following the final treatment.61 Human clinical trials, as well as the previously mentioned equine study, provide good evidence that DMOAD action of HA administered weekly for three or more treatments exists.55

When HA is compared to other joint therapies, such as corticosteroids, the results are variable. A comparison of high molecular weight HA versus betamethasone in human knee OA patients showed no significant difference between groups.63 Interestingly, when low-molecular-weight HA was compared to MPA, short-term results showed no difference between the two, however at 45 days post-treatment, pain scores were lower in the HA treated group versus MPA.64 Likewise, high molecular weight HA compared to triamcinolone hexotonide revealed more rapid pain relief with corticosteroid treatment, but OA and pain scores were significantly lower at long-term follow-up (12 and 26 weeks) in joints receiving only HA.65 Nearly 60% of equine practitioners in one survey reported that they routinely combine HA and corticosteroids for intra-articular administration.66 In vivo research supports this notion, as one study in a rabbit model of OA showed an 88% reduction in pathology with the combination of corticosteroid and HA, compared to 53% and 72% respectively with HA or TA alone.67 Again, in humans receiving a combination of IA HA and TA for knee OA, conclusions revealed that combination therapy provided a more rapid improvement in pain, had beneficial effects during 1-year post-treatment and showed no signs of deleterious effects on joint structure.68 Combination therapy, based on the current peer-reviewed literature, supports the notion that some synergistic action between HA and TA may exist, and routine use may be warranted. In the population of equine athletes today, HA is also often administered intravenously as a prophylactic effort. When 40mg of HA administered intravenously once weekly for three weeks in a model of equine OA, clinical lameness, synovial membrane histology, and synovial fluid parameters were improved over controls at 42 days following the last treatment.69

HA has been shown to moderately decrease OA-associated pain in humans, and is not refuted by work conducted in equine models. There are reports of beneficial effects from intra-articular administration of combined HA and TA, and a guideline for use based on the literature is 20 to 22 mg of a mid-molecular weight HA with 3 to 5 mg of triamcinolone acetonide in a 10 to 15 mL joint as a single injection.55 Evidence for beneficial effects of IA HA alone does exist (2 serial injections 1 week apart), and intravenous administration of HA for prophylaxis may be advantageous.

Polysulfated Glycosaminoglycan

Polysulfated glycosaminoglycan (PSGAG) is a polysulfated polysaccharide with DMOAD activity, consisting mainly of the GAG chondroitin sulfate. It can be administered IA or IM, and is primarily used in joint disease when damage to articular cartilage is suspected, and aims to prevent or reverse cartilage degeneration. Recently evidence also supports its application for mitigation of synovitis.70

Early in vitro studies on PSGAG showed contradictory results, with one study finding inhibition of matrix metalloproteinase 3 (MMP-3) production, and others showing increased collagen and GAG synthesis with a dose-dependent inhibition of proteoglycan synthesis.71, 72 In vivo in the horse, IA PSGAG provided significant reduction in cartilage fibrillation, erosion, chondrocyte death, and GAG staining in a carpal synovitis model using sodium mono iodoacetate, but did nothing to improve pre-existing articular cartilage lesions.73 When applied intra-articularly in the equine OFM of OA, PSGAG significantly reduced synovial vascularity and subintimal fibrosis compared to HA or saline.74 When PSGAG was combined with low dose TA, some ill-effects were noted and thus this combination is not recommended based on experimental in vivo work in the horse.75

A 2011 survey of equine practitioners revealed that the majority of individuals (84.1%) using PSGAG administer it intramuscularly (IM). However, when comparing IA versus IM administration, it was concluded that greater potency was achieved via the IA route.76 Because of early reports around IA PSGAG potentiating joint infection, the IM route likely has more general clinical acceptance. It is important to note that with the concurrent use of IA antibiotics (amikacin), the IA route for PSGAG has been shown to be an acceptable administration route, and is in fact the route of preference for the authors.77 Though PSGAGs are often used by practitioners prophylactically, there is minimal scientific evidence of efficacy in this way.

Pentosan Polysulfate

Pentosan polysulfate (PPS) is isolated from beech-wood hemicellulose of the beech tree, and has a chemical structure of repeating units of f (1-4)-linked β-Dxylano-pyranoses.80 When the equine formulation is administered IM or subcutaneously (SQ) blood concentrations peak at approximately 2 hours. Initial animal model studies proposed several beneficial mechanisms of PPS for treatment of OA, including promotion of proteoglycan synthesis, inhibition of proteoglycan degradation, and increased synthesis of chondorprotective signaling molecules.81-83 Other early studies also highlighted the anticoagulant effects of PPS, and hypothesized that it could be beneficial in the treatment of joint disease via improved perfusion to subchondral bone.84 Due to these vascular effects, it has been proposed that PPS could delay the rate of subchondral bone necrosis and sclerosis.80 However, in a study investigating the anticoagulant effects of IV PPS in the horse, a dose-dependent increase in partial thrombin time that persisted for up to 24 hours was shown.87 Thus some have recommended that doses of 3 mg/kg or greater should not be administered to horses within 24 hours of intense activities, or where physical injury may be a risk.

The mechanisms of action of PPS pertinent to joint disease are speculated to be two-fold, in that it has been shown to promote HA synthesis in osteoarthritic joints, and to inhibit induction of articular cartilage matrix degeneration via MMP inhibition and cytokine receptor modulation, though other research exists.85, 86 The first in vivo study in the horse utilized the OFM model to compare 3mg/kg PPS IM to an equivalent volume of saline. Results showed minor DMOAD effects, including reduced articular cartilage fibrillation, increased markers of chondroitin sulfate (CS) synthesis in arthritic and non-arthritic joints of the same individual, and trends toward improved cartilage histology scores.88 Based on the remote effect on CS synthesis, a systemic upregulation of aggrecan synthesis may occur.159 When results of IM administration of a recommended dose (3 mg/kg) of PPS are compared to the IM administration of the recommended dose (500 mg) of PSGAG using the same experimental model in a separate study, PPS produced more favorable outcomes.89 Interestingly, a second study using the OFM model of OA showed beneficial DMOAD effects of IV PPS given in conjunction with N-acetyl glucosamine (NAG) and HA. Results were, however, inferior to those demonstrated with PPS alone in the same model.90 Though there is no published confirmation of SMOAD effects of PPS, clinical application in 39 horses with OA showed that 3 mg/kg of PPS IM improved lameness scores faster and maintained this improvement for longer when compared to 500mg of IM PSGAG.159

Clinical reports indicate that 3 mg/kg given IM once a week for 4 weeks is anecdotally beneficial. Repeated evaluation confirms that compared to IM PSGAG, Pentosan Polysulfate is the only systemically administered DMOAD available to the equine veterinarian.

Biologic Therapies

Biologic therapy has become mainstream in the treatment of equine joint disease. Commercially available options include autologous conditioned serum (ACS), platelet-rich plasma (PRP), and stem cells. Various attributes are prescribed to each, and equine veterinarians tend to utilize that biologic therapy with which they are most comfortable applying based on experience.

ACS production was first described in 2003, and involves incubating whole blood with medical-grade glass beads with specific surface characteristics.91 This process is meant to increase the production of anti-inflammatory cytokines, however cannot be achieved without also allowing production of pro-inflammatory cytokines. What is quite important, as has been shown in the horse, is the ratio of anti to pro-inflammatory cytokine production.92 IRAP, as ACS products have come to be called, refers to interleukin 1 receptor antagonist protein, and is the target molecule of upregulation in ACS processing. It is recognized, however, that ACS products contain many proteins other than solely IRAP, and that during incubation many of these (at least 35 different proteins) are at least doubled in concentration (D. Frisbie, unpublished data). In fact, “using equine blood IRAP, IL-10, insulin-like growth factor-1, transforming growth factor-β, tumor necrosis factor (TNF)-α, and IL-1β have all been significantly upregulated using a commercial ACS kit compared with baseline serum.” 92 This highlights the fact that IRAP is not the only protein in ACS, but that it is more of a “soup” of biologically active molecules.

There is evidence to support the positive role of ACS in the treatment of joint disease. Results from a human study comparing ACS to HA and saline showed significantly superior results when ACS was used to treat knee OA versus the other two in 376 patients. It also showed that ACS treated joints had a lower incidence of adverse reactions (23%) compared to saline (28%), and HA (23%).93 In another study, some clinical OA parameters were decreased up to 1 year following ACS treatment, and IL-1β concentrations were lower in synovial fluid for up to 10 days.94

There is one in vivo equine study assessing ACS in induced equine OA. It demonstrated improvement in lameness, synovial membrane parameters, and cartilage fibrillation, indicating DMOAD and SMOAD effects.95 Interestingly, IRAP concentrations increased and remained elevated throughout the study period, likely indicating a prolonged beneficial effect stimulated endogenous production. Clinical use of ACS in the horse to treat joint disease however is anecdotal, but according to a survey of 791 equine practitioners, the most common indication for use is in steroid-unresponsive joints.

PRP has multiple recognized uses in medicine today, and optimization for a specific application is likely to be a focus of future research. In equine practice, it is often utilized for treatment of musculoskeletal conditions relating to tendinous, ligamentous, and articular injury. Not all PRP products are created equally, and variability can be seen in platelet concentration, platelet activators, and white blood cell (WBC) concentration. For musculoskeletal conditions, platelet concentrations of 2-6 times that of serum have been shown to produce positive results, however concentrations in excess of 6 times suggest negative effects.96 In addition, platelet activation may be important to stimulate production and release of biologically active molecules. Three mechanisms of platelet activation have been proposed, including endogenous activation, calcium chloride, and thrombin. In trials comparing each method, only calcium chloride and endogenous activation were found not to be detrimental.96, 97 Secondarily, one author has proposed platelet activation via post-injection shockwave therapy. In one in vitro study where platelets were exposed to extracorporeal shockwave therapy, growth factor concentrations were increased when compared to unactivated PRP.104 It is important to note that none of the inflammatory cytokines known to be produced by platelets were analyzed, and it is not known if platelets were truly activated, or simply destroyed as no comparison to platelet lysate was investigated.

The appropriate WBC concentration in PRP products is a constant cause for debate. Currently, studies have shown deleterious effects of “high” WBC preparations of PRP though, to date, no evidence of the same effect has been published on “low” WBC preparations.98, 99 The presence of WBCs could increase the presence of catabolic enzymes in vivo, though this work has not been done. Also important, is the understanding that platelets contain over 200 different pre-formed biologically active proteins stored within their α-granules. Some of these are growth factors, while others may be pro-inflammatory. The clinical manifestation may therefore be dictated by the ratio of beneficial to deleterious molecules.

Two systematic reviews have been conducted in the human literature. The first contained 6 studies, and compared IA HA or saline to varying preparations of PRP in a total of 653 patients with knee OA. Overall, the results indicated significant improvement in patient’s functional outcomes, with no significant effects on pain measures.100, 101 The second review was larger, and included 1543 patients with knee OA, and again reported significant functional improvements.101 Differences in effectiveness based on centrifugation methods or activation were not identified, however results from a single centrifugation did not appear as compelling as those using a double-centrifugation technique. This is in contrast to a second study which revealed an increased incidence of pain and swelling following a double compared to a single-centrifugation technique.102

Little conducted work has been published on the IA use of PRP in the horse. One study again confirmed that no activation, or activation via calcium chloride yielded the least clinical reaction, lowest intra-articular WBC concentration, and best growth factor profile.97 Also, these authors note that ““in normal joints, intra-articular PRP induces a mild to moderate inflammatory response in synovial fluid, which lasts about 1 day,” and that platelets became activated by simply being mixed with synovial fluid. Thus the addition of activators may not be necessary. A second study investigating what the author refers to as autologous protein solution (APS) in 40 horses with naturally occurring OA, showed significant improvement in lameness at 14 following treatment.103 All reports, both human and equine suggest that case selection is likely important, and that there is a greater likelihood of beneficial outcomes in milder versus severe cases on OA. In summary, there is more published clinical evidence in humans than in horses, and the level of evidence supporting use in the horse is currently greater for that of ACS versus PRP.

Stem Cells

Mesenchymal stem cells (MSCs) are those cells that have the ability to replicate and differentiate into the various mesenchymal tissues.105 Their use for the treatment of orthopedic disease is rapidly becoming more common. Continued use has brought to light several important considerations, including source, dose, timing of treatment, and indications for specific types of musculoskeletal disease/injury.

Stem cells may be harvested from a number different tissue sources in the body, but currently, bone-marrow-derived cells harvested from the ilium show significantly better results based on cellular matrix production following chondrocyte differentiation.107 When utilizing them therapeutically, doses in human studies with successful outcomes, and anecdotal evidence in horses have utilized cell numbers ranging from 10 to 50 million.2 When used to treat IA or tendon injury, significantly better outcomes are reported when treatment is delayed beyond the resolution of the inflammatory phase of injury.108

Equine practitioners utilize stem cells for a variety of different applications, though their therapeutic use in joint disease has been most rigorously studied in cartilage resurfacing, OA, and to treat of intra-articular soft tissue injury. Focal cartilage defects are reported to occur in over 50% of human knee and equine stifle arthroscopies.109-112 In an effort to regenerate functional articular cartilage using MSCs, they are either held over a defect using a matrix, or simply injected freely into a joint compartment. The feasibility of MSC implantation using a matrix has been demonstrated in multiple human clinical trials.113, 114 Most experimental studies using this technique in the horse have been relatively unsuccessful, and have gained little traction, as cells must have been previously harvested and expanded prior definitive diagnosis of cartilage erosion at arthroscopy. One study, however, showed the superiority of autologous-derived cells over allogeneic, based on the fact that more radiographic pathology was noted in allogeneic-treated defects compared with those treated with autologous cells. Interestingly too, are the findings that MSC’s may have a proclivity to bone formation when combined with a PRP/fibrin matrix over cartilage defects.115 The combination of these results seem to indicate that direct injection may currently be a superior technique to direct matrix implantation.

When MSCs are injected freely into a joint compartment, they have been shown to inhabit multiple articular tissues, including articular cartilage and the synovial membrane.116 This is noted in the human literature via two randomized trials, where multiple doses of either blood or bone marrow-derived stem cells, were combined with HA and injected into the knee of patients following arthroscopic debridement. In the first study, both imaging and biopsy results showed significant improvement compared to HA alone, however functional scores were not significantly improved.119 The second, on the other hand, showed both imaging and functional evidence of improvement for 1 to 2 years following the last injection.118 Evidence in a focal cartilage defect equine model in the stifle showed improved aggrecan staining (indicating a better repair) at 12 months following a single injection of 20 million bone-marrow-derived MSCs 4 weeks after surgery.117

For the treatment of OA with MSCs, results vary between models. In one study investigating IA bone-marrow-derived MSC in rabbits with transected anterior cruciate ligaments, results revealed significantly less cartilage degeneration, osteophyte formation, and subchondral sclerosis compared with control animals.120 In the horse, however, results confirmed safety, but were not as compelling for the use of MSCs in a model of stifle OA. The author reports that the timing of injection (14 days following surgery) may have been inappropriate, and could explain the results. Also important to note, as it addresses clinical concern, is that a combination of studies utilizing allogeneic or bone-marrow-derived cells report reaction rates (flare) of no greater than 9% in the horse.121-123

Intra-articular soft tissue lesions may provide the most convincing evidence for the use of IA MSCs. In 2003, a study where caprine stifles were destabilized via ACL transection and complete medial meniscectomy, showed that 6 million IA MSCs provided not only protective effects to the articular cartilage, but regeneration of 50-70% of initial meniscal volume.124 The authors termed it a “neomeniscus,” and commented that decreased OA scores in joints treated with the MSCs were likely due to partial restabilization of the joint by the newly formed tissue. Subsequent studies in human and rabbit models have supported this early work, and show superior tissue quality with more rapid tissue regeneration following meniscectomy.125 With positive evidence in other species, research was then conducted in horses with clinical lameness localized to the stifle. Initial pilot data in 15 cases of meniscal injury, showed 67% of horses with meniscal injury that received IA MSCs were able to return to their previous level of work. This lead to expansion of the project including more cases and follow up as long as 36 months, and results confirmed again, that arthroscopic meniscal debridement combined with HA plus IA MSC administration allowed 76% of horses to return to work, with 43% returning to their previous level of performance.121 It is important to note that all individuals included in this study were refractory to previous medical management.

Overall, autologous bone-marrow-derived culture-expanded MSCs are used most commonly in IA equine orthopedic research. There are again, multiple applications for MSCs in the equine patient, including cartilage injury, OA, and intra-articular soft tissue injury. A period of at least 4 weeks between surgery/injury and MSC treatment appears to be appropriate. Focal articular cartilage defects may be seen to improve with arthroscopic debridement and MSC plus HA implantation, there is little evidence supporting efficacy in cases of equine OA, and in cases with meniscal injury, long-term improvement can be accomplished.

Oral Joint Supplements in Equine Joint Disease

Due to the massive cost associated with treatment and prevention of osteoarthritis, much research and development has gone into production of easily administered oral supplements. It is estimated that nearly 50% of horse owners purchase and administer oral dietary supplements, thus the equine veterinarian must be familiar with the myriad of options.126 The majority are aimed at supplementation with building blocks of articular cartilage, or mitigation of the inflammatory cascade responsible of osteoarthritis progression.

The majority of joint supplements contain glucosamine (GU) and chondroitin sulfate (CS). Both are components of articular cartilage, and are thought to counteract degradation via enzymatic inhibition and provision of cartilage precursors.127-140 Commercial products are often manufactured from bovine tracheal tissue, and the Perna canaliculus sea mussel. In vitro evidence on equine cartilage explants showed a minor reduction in cartilage matrix (GAG) degredation, but only at higher doses. Interestingly, doses from 6.5 to 25 mg/mL have been reported to have detrimental effects of cartilage.142 Obviously, in vitro studies do not accurately represent clinical application of oral supplements, as efficacy is highly susceptible to bioavailability of active ingredients in the gastrointestinal tract. Absorption must efficient enough to raise blood and tissues levels to a therapeutic level in order to have any effect. In horses, oral absorption of CS and GU varies from 22% to 32% and 2.5% to 6.5% respectively.141, 144 Though these numbers are low, concentrations of GU have been shown to increase in synovial fluid following oral administration of clinically selected doses.141 Also, in the face of joint inflammation, evidence has been produced that suggests increased delivery of oral GU into synovial fluid compared to normal joints.145 However, amounts measured in synovial fluid from in vivo studies in the horse have never achieved concentrations equal to those shown to be therapeutic in in vitro models. It appears that the type of GU and CS used, dose, and inflammatory status of the joint may dictate delivery to the desired site. Multiple in vivo studies have been conducted in the horse, and results are variable. A few, however, support the use of CS and GU for treatment of OA based on subjective improvements in lameness grade, objective increases in ground reaction forces, and decreases in joint injection frequency.146, 147 None of these studies were without flaw, and evidence for their use is relatively weak.

A proprietary blend of New Zealand green-lipped mussel, shark cartilage, abalone, and Biota orientalis lipid extract is also commercially available. Individually, extracts of each of these components show anti-inflammatory and chondroprotective effects on inflamed equine cartilage explants.143 In vivo, oral administration of doses as high as 75g/day to healthy horses for 84 days showed no adverse effects (indicating safety). Also, when synovitis was induced in the fetlock joint, those individuals receiving the oral supplement showed significantly lower synovial fluid PGE2 and GAG concentrations indicating prevention of inflammation associated with synovitis.148 Secondarily, when 25 mg/kg/day of an oral extract of the New Zealand green-lipped mussel was tested compared to placebo in horses with lameness attributed to fetlock OA, clinical evaluation showed a significant reduction in severity of lameness, improved response to joint flexion, and reduced joint pain.149 It was concluded that green-lipped mussel extract may have in vitro anti-inflammatory effects, and that it can significantly alleviate the severity of lameness and joint pain in horses with fetlock OA.

Three other oral supplements, avocado soy unsuponafiables (ASU), polysaturated fatty acids (PUFAs), and cetyl myristoleate (CM) have gained popularity in recent years. ASU is produced by extracting the oils from avocados and soy beans. In vitro studies have shown anabolic, anticatabolic, and anti-inflammatory effects on human chondrocytes.150 In vivo evidence in the horse has not shown its ability to decease pain or lameness associated with OA, but cartilage scores and synovial inflammation were significantly reduced when ASU was administered orally for 70 days.151 However, the only commercially available ASU product for horses in the United States has not shown in vivo efficacy. PUFAs are found in fish and fish oils. Via incorporation their analogs into the cell wall of tissues, they have anti-inflammatory effects by prevention of initiation of the arachidonic acid cascade and subsequent eicosanoid production. Though there is currently little evidence supporting the use of PUFA for OA treatment or prevention in the horse, some evidence does exist in human, canine, and guinea pigs with OA.152-156 CM is another fatty acid used in joint supplements, and much like PUFAs, is thought to inhibit inflammatory mediators produced in the arachidonic acid cascade.157 When a product containing CM, GU hydrochloride, methylsulfonylmethane, and hydrolyzed collagen was administered orally in 39 lame horses, significant improvements was seen in lameness at a walk, response to joint flexion and quality of life, but not in lameness at a trot when compared to placebo.158

Use of oral joint supplements in equine athletes is widespread. Though some evidence from in vitro studies suggest DMOAD effects, evidence of in vivo efficacy, especially in the horse, is limited. Oral supplements are likely to be the focus of continued research and debate.