INTRODUCTION — The management of osteoarthritis (OA) consists of nonpharmacologic, pharmacologic, and surgical approaches. Conventional pharmacologic therapy for OA has mainly targeted symptomatic relief, with agents such as nonsteroidal antiinflammatory drugs (NSAIDs), acetaminophen, opioid analgesia, and intraarticular injections.
As we gain a better understanding of the pathogenetic mechanisms in OA, there has been a growing interest in the development of drugs whose mechanism of action are directed towards different pain pathways, as well as the inhibition of catabolic processes or stimulation of anabolic processes in the OA joint. These drugs are referred to as disease-modifying OA drugs (DMOADs) or structure-modifying OA drugs (SMOADs).
To date, no pharmacologic agents have been approved of by regulatory authorities for disease modification in OA. Existing trial strategies have failed to detect clinical efficacy of potential DMOADs based on the lack of responsive outcome measures to record treatment-related improvements in pain, function, or joint structure, or just due to lack of efficacy of these agents. While there has been progress with potential new analgesics, including targeting nerve growth factor, the availability of a potential structural modification agent appears distant.
A review of investigational approaches to the pharmacotherapy of OA will be discussed here. Established therapies for OA and surgical interventions are discussed elsewhere. (See "Overview of the management of osteoarthritis" and "Management of knee osteoarthritis" and "Management of moderate to severe knee osteoarthritis" and "Management of hand osteoarthritis" and "Management of hip osteoarthritis" and "Overview of surgical therapy of knee and hip osteoarthritis".)
TARGETS FOR TREATMENT — The major goals of osteoarthritis (OA) management are pain control, functional improvement, education about the disease, self-management, and prevention or slowing of structural changes to the joints. However, developing agents to address these goals are challenging given the complexity of pain pathways and the etiology of structural changes.
As is discussed elsewhere, the etiology of OA is multifactorial and affects the composition and structure of all tissues of the joints, including articular cartilage, bone, ligaments, synovium, and synovial fluid. In addition, there is a poor correlation between symptoms and structural changes, as is often demonstrated on plain radiographic imaging, reinforcing the concept that the pain experienced in OA is likely due to a combination of nociceptive (including inflammatory) and/or neuropathic pain. It is also likely that there are multiple local structural sources of pain including subchondral bone marrow lesions, bone, soft tissue around the joint, and bursae. (See "Pathogenesis of osteoarthritis" and "Epidemiology and risk factors for osteoarthritis".)
In the drug development process, the application of preclinical models of OA to human models has been filled with challenges. Due to difficulties in demonstrating clinical efficacy, many potentially promising disease-modifying OA drug (DMOAD) trials have been terminated. Some of the major categories of investigational agents for the treatment of OA are presented below.
Pain modification — Investigational agents targeting pain in OA have focused on inhibiting nerve growth factor.
Inhibition of nerve growth factor — There is evidence supporting the role of nerve growth factor in the development of sensory neurons such as nociceptors, which are involved in sensing tissue damage. Expression of nerve growth factor is increased in injured or inflamed tissue and is associated with increased pain. However, safety concerns related to a potential increased risk of rapidly progressive OA have limited the development of agents targeting this factor.
Tanezumab, a humanized monoclonal antibody that blocks nerve growth factor (anti-NGF), was shown to reduce knee pain while walking compared with placebo in a randomized trial including 450 patients with moderate to severe knee OA . However, trials of tanezumab were put on partial clinical hold by the US Food and Drug Administration (FDA) in 2010 due to secondary osteonecrosis . A subsequent study found that despite initial reports, tanezumab was not associated with an increase in osteonecrosis but rather rapidly progressive OA . There were additional concerns regarding peripheral nervous system effects .
In a trial including 696 patients with moderate to severe hip or knee OA, tanezumab-treated patients (2.5 mg and 5 mg groups) showed greater improvements in scores assessing pain and physical function compared with patients in the placebo group after 16 weeks . However, rapidly progressive OA occurred only in tanezumab-treated patients, and the incidence of total joint replacements was also higher in the tanezumab-treated groups. Furthermore, the improvements in pain and function were relatively modest. Due to safety concerns regarding the risk of rapidly progressive OA, tanezumab is no longer under development .
Another anti-NGF agent, fasinumab, appeared to reduce pain and improve function in a trial including 421 patients with moderate to severe knee or hip OA . Adverse effects were higher in the treatment group, particularly with higher rates of nervous system and musculoskeletal symptoms. A dose-dependent increase in arthropathies (7 percent of fasinumab-treated patients and 1 percent of placebo-treated patients) was also observed during the study. Trials of fulranumab, which showed similar efficacy and adverse events to other anti-NGF agents, were reportedly discontinued during phase III development due to the company's portfolio strategy rather than concerns about the drug's safety .
Structural modification — Drugs targeting structural changes in OA include agents that modify inflammatory pathways, cartilage catabolism and anabolism, and bone remodeling.
Tumor necrosis factor-alpha inhibitors — Tumor necrosis factor (TNF)-alpha, predominantly produced by activated synoviocytes, phagocytic mononuclear cells, and articular cartilage, is one of the major cytokines involved in the destruction of articular matrix. Together with interleukin (IL) 1beta, it activates inflammatory signaling pathways that end in the activation of NF-kappa-beta, a transcription factor that regulates the expression of chemokines, cytokines, adhesion molecules, and matrix-degrading enzymes. (See "Pathogenesis of osteoarthritis", section on 'Inflammatory mediators'.)
The concept of TNF-alpha inhibition, which has been effective for the treatment of rheumatoid and psoriatic arthritis, has been applied to OA with mixed results with different TNF-alpha inhibitors.
Infliximab, a chimeric monoclonal antibody directed against TNF-alpha, appears to have a potential effect on disease progression. In a 12-month open-label pilot study in erosive hand OA, patients who received monthly intraarticular infliximab injections into the affected joints of the more severely affected hand were found to have a reduction in the anatomical lesion radiological score when compared with joints injected with saline . In another cohort of rheumatoid arthritis patients, intraarticular infliximab injections reduced the radiological progression of incident secondary hand OA after three years of follow-up .
The data for adalimumab, a fully human monoclonal antibody also directed against TNF-alpha, have shown mixed results in the treatment of OA. Several studies have failed to demonstrate improvements in hand OA pain with subcutaneous adalimumab injections [11,12]. However, another randomized trial including 60 patients with erosive hand OA did show a significant decrease in the number of new erosions with subcutaneous adalimumab injections after 12 months of follow-up .
TNF-alpha blockade may not be the appropriate agent in a low-level inflammatory condition. Despite limited effects on pain in OA patients, there may be a potential positive effect on disease progression and the development of erosions.
Interleukin 1beta inhibitors — IL-1 is capable of inducing expression of matrix metalloproteinases (MMPs) and other catabolic genes, which stimulate the turnover of extracellular matrix remodeling. IL-1beta antagonists have been shown to prevent cartilage breakdown in a multitude of animal models [14,15]. However, translation to human models failed to demonstrate a reduction in pain in patients with knee OA. As an example, in a randomized trial with 170 patients with symptomatic knee OA, intraarticular injections with anakinra, a recombinant human IL-1 receptor antagonist, was well tolerated but was not associated with improvements in OA symptoms compared with placebo . Another trial with a monoclonal antibody that binds type 1 IL-1 receptor also failed to elicit a statistically significant difference in Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) pain score in knee OA after three months of treatment .
Findings from an exploratory analysis suggest that there may be a role for IL-1 inhibition in large joint OA; however, these results are limited by the fact that the parent trial was not designed to examine the efficacy of IL-1beta inhibitors for OA . The analysis included 10,061 patients with elevated C-reactive protein (CRP) levels and a history of myocardial infarction who were randomly assigned to receive placebo or canakinumab, a human anti-IL-1beta monoclonal antibody. After a median follow-up of 3.7 years, the incidence rates for total hip and knee arthroplasty were lower among the patients receiving canakinumab compared with the placebo group (0.31 versus 0.54 events per 100 person-years, respectively; hazard ratio [HR] 0.58 [95% CI 0.61-0.87]). However, direct trial evidence is needed before conclusions can be made regarding a potential role for canakinumab in the management of OA. Also, of note, there were more deaths attributed to infection or sepsis in the canakinumab groups compared with the placebo group in the parent trial .
Initial studies show good tolerability of anti-IL-1 blockade in healthy individuals, but a trend towards decreased absolute neutrophil count was observed . Further studies of an anti-IL-1alpha/beta dual variable domain-immunoglobulin are underway, targeting a specific patient population in whom there may be benefit.
Additional information on the role of IL-1 in the pathogenesis of OA can be found separately. (See "Pathogenesis of osteoarthritis", section on 'Inflammatory mediators'.)
Inducible nitric oxide synthase inhibition — Inducible nitric oxide synthase (iNOS), an isoform of nitric oxide synthase, is highly expressed in chondrocytes. It activates MMPs, inhibits collagen and proteoglycan synthesis, and induces chondrocyte cell death .
Studies of cindunistat hydrochloride maleate, an orally administered irreversible inhibitor of iNOS, did not demonstrate reduction in the rate of joint space narrowing in knee OA. A post-hoc analysis showed a beneficial effect at 48 weeks in patients with radiographically less advanced OA, but this effect was not sustained at the 96-week follow-up .
Bradykinin receptor B2 antagonist — The OA joint synovium generates bradykinin, a vasodilator and inflammatory peptide. Bradykinin, through the B2 receptor, via the activation of chondrocytes and transcription factor NF-kappa-beta, induces the release of proinflammatory cytokines. It also induces angiogenesis and promotes vessel permeability, growth, and remodeling.
Icatibant, an intraarticular bradykinin receptor B2 antagonist, showed some initial potential analgesic effects . However, further drug trials were suspended due to a possible lack of efficacy . Additional studies of the effect of fasitibant, another bradykinin receptor B2 antagonist, on arthritis pain is also underway .
Cartilage catabolism and anabolism
Bone morphogenetic protein-7 — Bone morphogenetic protein (BMP)-7, previously termed osteogenic protein-1, is expressed by human chondrocytes and contributes to the repair of mature tissues . BMP-7 also suppresses catabolic activities of cytokines, including IL-1, IL-6, IL-8, MMP-1, and MMP-13 .
An initial randomized trial including 33 patents with knee OA showed promise with the use of weekly intraarticular recombinant human BMP-7 injections . At 12-week follow-up, there was a trend towards symptomatic improvement in those who received different doses of recombinant BMP-7 compared with placebo. In the treatment groups, similar trends of improvement were observed in WOMAC pain and function subscales. Further studies of evaluating the use of recombinant BMP-7 in the treatment of OA are underway.
Fibroblast growth factor — Based on preclinical study evidence, fibroblast growth factor (FGF)-18 (sprifermin) is thought to be a potential therapeutic agent for OA. In animal models of injury-induced OA, FGF-18 has been shown to have an anabolic effect on cartilage, stimulating proteoglycan synthesis and cartilage matrix formation [29,30]. However, the available evidence for its use in patients with knee OA is limited and still of uncertain clinical importance.
Two phase I trials of intraarticular recombinant human FGF-18 for knee OA demonstrated a statistically significant dose-dependent improvement in tibiofemoral cartilage volume and a reduction in joint space narrowing after 12 months [31,32]. In a larger phase II trial including 549 patients with symptomatic knee OA, intraarticular administration of 100 micrograms of sprifermin every 6 or 12 months was associated with an increase in total femorotibial cartilage thickness after 2 years when compared with placebo . However, there were no improvements in total WOMAC scores among those receiving sprifermin. Although the difference in cartilage thickness was statistically significant, the clinical relevance of these findings remains uncertain.
Further clinical studies are warranted to gain a better understanding of the effects of sprifermin at optimal doses for possible disease modification in OA.
Matrix metalloproteinase inhibitors — MMP inhibition was thought to be a potential target for OA as MMP genes are expressed in all aspects of the articular joint, and one of its main roles is the enzymatic collagen breakdown in articular cartilage. Of all the MMPs identified to date, collagenases MMP-1, MMP-8, MMP-13, and MMP-14 are associated with cartilage collagen destruction [34,35]. (See "Pathogenesis of osteoarthritis", section on 'Proteases'.)
Thus far, MMP inhibition for treating OA has been relatively unsuccessful due to intolerable musculoskeletal side effects, including musculoskeletal pain, Dupuytren's contracture, and frozen shoulder [36,37].
ADAMTS — Aggrecanase activity increase is a known trigger for cartilage degradation. A disintegrin and metalloproteinase with thrombospondin motifs (ADAMTS)-4 and ADAMTS-5 are two key aggrecanases. Collagen and aggrecan are the primary structural components of the cartilage extracellular matrix, and the progression of OA correlates with their degradation [38,39]. Thus, collagenases and aggrecanases are prospective targets for the development of DMOAD.
Mitogen-activated protein kinase inhibitor — Alternate signal transduction pathways, which regulate MMP, are being explored, based on the side effect profile of direct MMP inhibition. Mitogen-activated protein kinases (MAPK) are central regulators of cell signaling pathways that control cell proliferation, survival, and matrix synthesis  and regulate chondrocyte production of MMPs .
Several MAPK inhibitors have been investigated in clinical trials, with only a few progressing to phase II trials due to associated liver and central nervous system adverse effects .
Cathepsin-K inhibitor — Cathepsin-K (CatK) is expressed in osteoclasts and is a key enzyme involved in bone resorption. It is also expressed in chondrocytes in cartilage, where it can cleave type II collagen and aggrecan, which are main components of the cartilage matrix . Inhibition of CatK has been shown to reduce biomarkers of bone resorption and cartilage degradation in animal models [44,45]. Clinical trials with CatK inhibitors have shown variable outcomes. Balicatib did not reduce cartilage volume loss, and morphea-like skin reactions have been reported . In another trial including 244 patients with symptomatic and radiographic knee OA, pain scores at 26 weeks among patients taking MIV-711 (a potent CatK inhibitor) were similar to those in the placebo group . However, a reduction in bone and cartilage progression was observed in the MIV-711 group. The safety profile was reassuring and was without an increased risk of morphea-like skin reactions.
Wnt signaling pathway inhibitors — The Wnt pathway has a prominent role in OA joints, particularly with cartilage catabolism and regeneration through modulation of the differentiation of osteoblasts and chondrocytes and production of catabolic proteases [48,49]. SM04690 (lorecivivint) is a novel small-molecule Wnt pathway inhibitor proposed to have multiple effects on the articular cartilage, including slowing down cartilage degradation and reducing inflammation . A proof-of-concept trial consisting of 455 patients with unilateral symptomatic knee OA did not meet the primary endpoint of change from baseline in WOMAC pain score in the target knee . However, in an analysis of secondary endpoints, symptomatic and medial joint space width improvements were observed in patients in the lorecivivint group at 52 weeks. Findings from a phase IIb randomized trial that included 695 patients with moderate to severe knee OA who were randomized to receive either lorecivivint or placebo were more promising . Overall, statistically significant improvements were observed in patient-reported outcomes that included WOMAC pain, function, and patient global assessments compared with placebo .
Platelet-rich plasma — Platelet-rich plasma (PRP) is derived from autologous blood, with platelets being the main constituent. The mechanism of action is not well-understood, but it is thought to provide high concentrations of growth factors including tissue growth factor and platelet-derived growth factors, which can mediate the proliferation of mesenchymal stem cells and increase matrix synthesis and collagen formation . PRP is thought to reduce inflammation in an OA joint by enhancing the expression of NF-kappa-beta inhibitor, and thus reduce NF-kappa-beta signaling and dampen its downstream inflammatory cytokine activation . It has shown the ability to switch off the inhibition of type II collagen and aggrecan gene expression in IL-1beta-activated NF-kappa-beta in chondrocytes . Thus far, PRP has predominantly been used to treat musculoskeletal conditions, particularly tendon-related pathologies. In general, no major adverse events have been reported, with only minor and transient risks and adverse effects of PRP seen, which include pain/bleeding, tenderness, and swelling and bruising at the site of injection . However, general risks associated with intraarticular injections should be applied in this context, which includes risk of local infection. (See "Joint aspiration or injection in adults: Complications" and "Elbow tendinopathy (tennis and golf elbow)", section on 'Platelet-rich plasma and other biologic injections'.)
Increasing evidence from well-designed trials do not support the use of PRP for the management of OA [57,58]. Prior data suggested that PRP may provide short-term (up to 12 months) symptomatic relief for knee and hip OA , but the majority of the trials revealed a high likelihood of biases with low to moderate methodologic quality and variability in their PRP protocols, with lack of longer-term follow-ups [59,60]. A subsequent randomized trial including 100 patients with ankle OA found that intraarticular PRP injections, compared with placebo injections, did not demonstrate a statistically significant improvement in a composite score of pain and function during the 26-week follow-up period . Similarly, another trial including patients with knee OA did not demonstrate a benefit in terms of pain or volume loss . Additional information regarding trials of PRP for knee OA is presented separately (see "Management of knee osteoarthritis", section on 'Platelet-rich plasma'). Factors that may contribute to differences in results among trials include the lack of standardization of preparations, with different preparation protocols with varying concentrations of platelets, presence/absence of white cells, and mode of activation (ie, mechanically with freeze-thawing cycles, chemically, or endogenously).
Tissue gene therapy — Transforming growth factor-beta (TGF-beta) is part of cellular differentiation and growth synthesis of extracellular matrix and chondrogenesis, and TGF-beta signaling pathways are thought to be involved in early cartilage development and maintenance of cartilage over time. TissueGene-C is a novel gene and cell therapy consisting of human allogeneic chondrocytes and irradiated GP2-293 cells overexpressing TGF-beta1 . Results from two small phase II trials [62,63] and one phase III trial  demonstrated statistically significant improvements in pain and function in patients. Trends of structural improvement were observed but were not statistically significant.
Calcitonin — The main role of calcitonin is to regulate calcium homeostasis. It is produced primarily by the parafollicular C cells in the thyroid gland. It binds to calcitonin receptors on osteoclasts and promotes the inhibition of bone resorption, and thus increases the activity of osteoblasts. In preclinical studies, calcitonin has been shown to induce inhibition of MMP activity and cartilage degradation, both in vivo and ex vivo . It also may have an effect on inhibition of toll-like receptors (TLRs), which releases cytokines and inflammatory mediators .
Pilot studies of salmon calcitonin comparing two doses of salmon calcitonin formulated with a 5-CNAC carrier against placebo demonstrated no significant differences in pain scores . However, a considerable improvement was seen in the function score among patients receiving oral salmon calcitonin.
Bisphosphonates — Commonly used as the treatment for osteoporosis, the therapeutic basis of bisphosphonates is based on their direct inhibitory effect on the function of osteoclasts (see "Bisphosphonate therapy for the treatment of osteoporosis"). They have the potential to retard subchondral bone remodeling. Patients with progressive OA have been found to have higher urinary levels of C-terminal telopeptide of type II collagen (CTX-II), a marker of cartilage degradation ; thus, the application of bisphosphonates seems ideal in such cases. Bisphosphonates can also exert a slight immunomodulating effect via the inhibition of proinflammatory cytokines, with etidronate in vitro, showing some inhibitory effect on MMPs when they binds to human cartilage .
The use of bisphosphonates in OA has yielded mixed results. The British study of risedronate in structure and symptoms of knee osteoarthritis (BRISK) trial was a randomized control trial of patients with mild to moderate OA of the medial compartment of the knee . Patients were randomized to once daily risedronate (5 or 15 mg) or placebo. Those who received 15 mg of risedronate showed improvement in the WOMAC index, significant improvement in the patient global assessment, and decreased used of walking aids.
In another large study, the Knee OA Structural Arthritis (KOSTAR) study, 2483 patients with medial compartment knee OA from North America and the European Union were given risedronate or placebo and followed for two years . The risedronate-treated groups did not show significant improvements in signs and symptoms of OA compared with placebo, and there was no significant reduction in radiographic progression.
In a hip OA study with alendronate versus placebo, significant improvements in WOMAC pain score were seen with alendronate, but there were no significant differences in structural OA progression . Another trial including 223 patients with knee OA found that yearly zoledronic infusions compared with placebo did not result in statistically significant differences in cartilage volume loss over the two-year follow-up period .
Zoledronic acid, a once-yearly bisphosphonate infusion, may be of potential promise with the initial trial of a single infusion of zoledronic acid compared with placebo, showing a significant reduction in areal bone marrow lesion size at six months with an ongoing trend after 12 months . A significant reduction in Visual Analogue Scale pain score was also seen at six months, which was not sustained by 12 months. Additional studies are needed to establish the potential benefits of bisphosphonates in the treatment of OA.
Strontium — Strontium ranelate, a treatment for osteoporosis in postmenopausal women, is classified as a dual-action bone agent given its ability to both stimulate cartilage matrix formation and inhibit subchondral bone resorption , thereby improving the overall architecture of the bone. It has been shown to reduce the progression of radiographic features of spinal OA and back pain in females with concomitant osteoporosis and spinal OA . (See "Overview of the management of osteoporosis in postmenopausal women", section on 'Other'.)
A three-year phase III randomized trial of patients randomly allocated to strontium ranelate 1 g/day, 2 g/day, or placebo found that treatment with strontium ranelate was associated with smaller degradations in joint space width, improved WOMAC total score, and pain subscores in the group receiving the higher dose of strontium . MRI assessing cartilage volume loss and bone marrow lesions was further done in a subset of patients. Strontium ranelate dosed at 2 g/day was associated with significant reduction in both cartilage volume loss in the plateau and bone marrow lesion progression in the medial compartment at 36 months.
Adverse effects of strontium ranelate use, including myocardial infarction, will affect its consideration for future treatment of OA. (See "Overview of the management of osteoporosis in postmenopausal women", section on 'Other'.)
SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Osteoarthritis".)
●A number of investigational agents have been evaluated in patients with osteoarthritis (OA). The major goals of treatment are pain control and prevention or slowing of structural changes to the joints. (See 'Targets for treatment' above.)
●Investigational agents targeting pain have focused on inhibiting nerve growth factor. (See 'Inhibition of nerve growth factor' above.)
●Drugs targeting structural changes in OA include agents that modify inflammatory pathways, cartilage catabolism and anabolism, and bone remodeling. (See 'Inflammatory pathways' above and 'Cartilage catabolism and anabolism' above and 'Bone remodeling' above.)
●At present, none of the investigational agents discussed in this topic are recommended for the treatment of OA.
ACKNOWLEDGMENT — The UpToDate editorial staff acknowledges Kenneth Kalunian, MD, who contributed to an earlier version of this topic review.