Emed Pain Management Blog

Emed Pain Management Blog

Evolving Management Strategies for Osteoarthritic Pain

More targeted analgesics and disease-modifying drugs are in high demand for future management of this growing and painful joint condition.By Don L. Goldenberg, MDPages 20-27

Osteoarthritis (OA) is the most common form of arthritis, affecting approximately 27 million Americans,1 and serving as a leading cause of lower extremity disability worldwide. The prevalence of knee and hip OA, in particular, are rising, attributed to increased aging and obesity in the general population. Therefore, this review will focus on new management strategies for OA of the knee and hip. It is worth noting that some studies have been limited to OA of the knee and hip whereas others combine OA of those weight-bearing joints with generalized OA, including the hands, feet, shoulder and spine. This review will not include surgical options, but rather, will discuss emerging trends in pharmacological and non-pharmacological management. Currently, there are no approved therapies that reverse the progression of knee or hip OA, although some are on the horizon. Until then, the treatment of knee and hip OA is largely based on pain reduction and improving function.

The Root of Osteoarthritic Pain: New Understandings

Pain is the primary symptom in OA and the primary reason that patients eventually undergo total joint replacement. Until recently, OA pain was considered to be caused solely by peripheral, structural joint damage. As cartilage degraded and bony hypertrophy further narrowed the joint space, pain and joint mobility worsened. However, the degree of radiological OA joint damage does not often correlate well with pain severity.2 Some OA patients with minimal knee OA on X-ray report incapacitating pain, whereas others with end-stage radiologic joint findings have little discomfort.

Osteoarthritic pain is now known to be driven by both structural and central components;3,4 the peripheral pathway has been better defined. Bone marrow lesions, noted in the subchondral bone on magnetic resonance imaging (MRI), correlate with OA knee pain much better than cartilage narrowing. Low levels of synovial inflammation also correlate with OA pain. Cytokines, including interleukin-1 (IL-1), chemokines, and metalloproteinases released intraarticularly stimulate, both peripheral and central pain.5 Nerve growth factor (NGF) is a neurotrophin released from OA tissue that sensitizes peripheral nociceptors. NGF levels are increased in synovial fluid and NGF expression correlates well with OA knee and hip pain.5 Neuropeptides, including calcitonin gene-related peptide (CGRP), have been implicated in OA pain as well.5

Persistent peripheral pain from the OA joint drives central pain. In studies, central sensitization in knee OA was demonstrated by evidence of lowered pain threshold at multiple anatomical sites.3 Functional brain imaging (fMRI) revealed significantly greater activation in the brainstem of OA patients in response to punctate stimulation of referred pain areas compared to controls.4 Osteoarthritis patients who complained of widespread pain described neuropathic pain, reported depression, sleep disturbances, and fatigue and were likely to have a prominent central component to their pain.3,4 The recognition of the importance of central pain in OA led to the approval of duloxetine in the treatment of OA and the incorporation of biopsychological therapies such as acupuncture, yoga, Tai-Chi, and meditation-mindfulness in early OA management.

Structural Damage

In rheumatoid arthritis (RA), structural damage has been directly related to the immune/inflammatory response, and disease-modifying anti-rheumatic drugs (DMARDs) have greatly lessened joint destruction while uniformly improving joint pain and function. In osteoarthritis, the mechanism of joint damage is less understood. DMARDs used in RA have had no significant OA benefit. Furthermore, OA is very slowly progressive and most studies of OA disease mechanisms and therapy have been in advanced cases. The search for a disease-modifying osteoarthritis drug (DMOAD) may require beginning therapy early in the disease and following subsequent changes for years.

The medical community now recognizes that pain reduction in OA may not correlate with structural improvement and may, in fact, be associated with increased joint damage. Relieving nociceptive pain with NSAIDs or anti-nerve growth factor (NGF) therapy may promote more joint load and promote ambulation that damages cartilage, which has no pain receptors. This therapy is in contrast to new therapies in RA where there is good concordance of pain and functional improvement with disease modification. Therefore, it should not be surprising that drugs or techniques that are primarily analgesic may not achieve status as disease-modifying OA therapies.

The medical community now recognizes that pain reduction in OA may not correlate with structural improvement and may, in fact, be associated with increased joint damage. (Source: 123RF)

Current Treatment Recommendations

Early intervention has been shown to be key to optimal OA management but has often been neglected.6 Primary care providers are often on the frontline of evaluating and treating OA, and can encourage patients to exercise, lose weight (if applicable), and utilize physical therapy as soon as pain is reported.

Patient Education, Exercise, and Diet

The American College of Rheumatology (ACR) has published pharmacologic and non-pharmacologic treatment recommendations for both OA in the knee and hip (see Table I).6 Patient education, exercise, and dietary weight management are strongly recommended. All patients with knee or hip OA should be in an exercise program that focuses on strengthening, range of motion, and joint protection. Water-based exercises are often more successful and better tolerated than land-based programs. In moderate to severe medial tibiofemoral OA, a valgus, unloaded knee brace is recommended. Patellar taping and walking aids may reduce pain in select patients.

An algorithm for the treatment of knee OA from the European Society for Clinical and Economic Aspects of Osteoporosis, Osteoarthritis, and Musculoskeletal Diseases (ESCEO) provides similar recommendations for education, exercise, and weight reduction.7

Acetaminophen, NSAIDs, Supplements, and Injections

ESCEO gave a weak recommendation for acetaminophen/paracetamol to be used for long-term analgesic therapy, but considered prescription glucosamine sulfate and/or chondroitin sulfate or topical NSAIDs for background treatment (see Table II). If patients are still symptomatic, intermittent or longer cycles of NSAIDs or COX-2 inhibitors should be used. If symptoms still persist, intra-articular (IA) hyaluronate or IA corticosteroids may be considered.

Topical NSAIDs, acetaminophen, and oral NSAIDs continue to be the most widely used medications for OA of the knee and hip. These include topical diclofenac and capsaicin. Diclofenac has similar efficacy and better tolerability than the oral diclofenac,6,7 while capsaicin, derived from hot chili peppers, has shown to be more effective than placebo in OA of the knee.7

In the past, acetaminophen, also marketed as paracetamol, was considered as a first-line analgesic for OA in the knee and hip. However, recent systematic reviews question this recommendation due to its minimal efficacy.8 Its more potent analgesic effect, at 3 g to 4 g daily, comes with a significant risk of liver damage and should be avoided for long-term use. NSAIDs may be more effective than placebo or acetaminophen, but cardiovascular and gastrointestinal side effects can be significant, particularly in older patients.

A systematic analysis concluded that there was limited evidence that the COX-2 inhibitor celecoxib was more effective than placebo and some NSAIDs for reducing OA pain.9 NSAIDs produced significant pain relief at 2 weeks, but relief decreased over time and declined by 8 weeks. The risk/benefit ratio for NSAIDs and COX-2 inhibitors in OA should be determined on an individual patient basis, always attempting to find the minimal effective dose. Duloxetine may be considered in patients with contraindications or lack of response to oral NSAIDs. Both the ACR and ESCEO guidelines strongly recommend against using long-term strong opioids; these medications have been found to not be better than NSAIDs or non-opioid analgesics in improving OA pain and function.10

Although glucosamine and chondroitin sulfate supplements have been widely used in knee and hip OA for the past 20 years, their efficacy remains controversial. Most clinical trials with glucosamine and chondroitin have had strong placebo effects. Some reviews have concluded that glucosamine and chondroitin do not provide clinically relevant benefits to patients with the disorder in the knee or hip.11 However, other reviews, particularly those conducted outside of the US, recommend their early use in OA of the knee or hip.12,13 According to ESCEO guidelines, only prescription-grade glucosamine sulfate has been shown to have consistent bioavailability and clinical efficacy.7 These guidelines strongly recommend prescription glucosamine sulfate or chondroitin sulfate for long-term background therapy in OA of the knee (see Table II). There is no good evidence that other nutritional supplements have clinically meaningful effect in treating OA in the knee or hip. However, an observational study found that vitamin D supplementation was associated with less cartilage loss in knee OA over 4 years.14

The ACR knee and hip guidelines conditionally recommend duloxetine as well as intra-articular (IA) hyaluronic acid injections.6 Duloxetine and tramadol, however, are considered to be the last pharmacological option in ESCEO guidelines.7 One study showed that duloxetine 30 mg/day and gabapentin 300 mg/day were both more effective than 1 g/day of acetaminophen for pain reduction in OA of the knee.7 Tramadol, recommended in some guidelines, has been associated with higher mortality rate when compared to NSAIDs, but not when compared with codeine.15

Intra-articular steroid injections may result in short-term pain improvement, but relief typically lasts less than a few months and serial injections may have potential negative effects, so they are discouraged.6,7 Patients with a significant knee effusion may have more benefit from joint drainage and steroid injections. There are five FDA-approved injectable corticosteroids but no good evidence of each’s superiority.

Intra-articular hyaluronic acid injections have been advocated by many orthopedic specialists, although most studies have failed to reveal clinically significant pain relief. The ACR does not recommend IA hyaluronic acid injections because the largest and best clinical trials demonstrated minimal benefit over placebo injections with high costs and occasional toxicity, including pain flare-ups.6 However, the ESCEO guidelines gave only a weak recommendation for IA injections of either corticosteroids or hyaluronic acid in patients who have failed background therapy and oral NSAIDs (see Table II).7 A recent review concluded that IA hyaluronan was comparable to IA corticosteroids for long-term pain relief, but slightly inferior for short-term pain relief.16

Emerging Non-Pharmacologic Therapies


A systematic review found that acupuncture and electroacupuncture may have some benefit in OA of the knee,17 while another analysis found that pulsed electromagnetic field therapy relieved pain and improved function in knee OA specifically.18 There were no significant adverse events; therapy of less than 30 minutes was more effective than longer durations. There is inconsistent evidence that transcutaneous electrical nerve stimulation also relieves knee or hip pain.7


Peripheral nerve block, either by radiofrequency ablation or cryoneurolysis, has been shown to decrease OA knee pain in comparison to sham procedures.19 At the knee, the peri-articular or IA sensory branches of the genicular nerve are targeted. Pain relief lasted for up to 6 months, but joint functional improvement only lasted 3 months. A local anesthetic nerve block may predict nerve ablation efficacy. This procedure was more cost-effective than IA steroid injections. There are a number of ongoing trials using either freezing the nerve (cryoneurolysis) or heat denervation (radiofrequency ablation) to treat refractory OA knee pain.

Exercise and Meditation

There have been a number of new studies demonstrating efficacy for yoga, Tai-Chi, and mindfulness training to benefit the condition as well. A systematic analysis concluded that yoga may be effective for improving pain and function in knee and hip OA compared to exercise and non-exercise control groups.20 Tai-Chi and physical therapy are likely to result in similar benefits in knee OA but Tai-Chi may be more cost-effective.21

Emerging Pharmacologic Therapies


Extended-release triamcinolone acetonide has been approved in the US for IA injection in OA of the knee. The corticosteroid is formulated in microspheres that release the drug slowly, which reduces systemic absorption. As a single injection, it lasts longer with less adverse effects than shorter-acting agents.22 Trials using extended-release triamcinolone compared to placebo or immediate-release triamcinolone have been generally favorable. Intramuscular corticosteroid injections may be better than placebo for pain reduction in hip OA but concerns regarding potential steroid side effects have tempered its use.

Intra-articular opioid injections have not been effective in knee OA, but a buprenorphine transdermal patch applied to knee OA was more effective in relieving pain with less adverse effects and better compliance when compared to chest application of the opioid patch.23

Stem-cell or platelet-rich plasma (PRP) injections into knees and hips with OA have been increasingly popular despite few studies demonstrating their efficacy. Platelet-rich plasma is loosely defined as a volume of plasma with a greater platelet concentration than normal. The mechanism of action of PRP is thought to be related to increased concentration of growth factors which may increase collagen and matrix synthesis. There have been a number of randomized clinical trials, usually comparing PRP injections to hyaluronan injections with systematic analysis demonstrating equivalent efficacy.24 In small, well-controlled studies, injection of bone marrow derived mesenchymal stem cells for OA pain improved and cartilage growth was noted.25 However, no long-term, well-controlled studies using PRP or stem cells in knee or hip OA are available.

Because of the absence of evidence for safety or positive results, FDA has not approved these procedures for OA treatment. Stem cell clinics may circumvent regulations since treatment using a patient’s own cells does not require approval, but there is a lack of product standardization.26

Biologics and Disease-Modifying Osteoarthritis Drugs (DMOADs)

Anti-NGF Therapies

In contrast to rheumatoid arthritis, where more than 20 DMARDs have revolutionized therapy over the past 30 years, no DMOAD for osteoarthritis has been approved. However, new biologic drugs that target OA pain, rather than structure, have been developed, and the most promising are antibodies against NGF.

Tanezumab, an anti-NGF monoclonal antibody in development by Pfizer and Eli Lilly and Company, was found to be effective in one large randomized double-blind controlled study.27 A rapidly progressive osteoarthritis, referred to as RPOA, and possible osteonecrosis was later found to be associated with the drug which caused FDA to halt trials with this or similar agents until 2015. Since that time period, evaluation of the adverse event profile found that tanezumab was not associated with osteonecrosis but was associated with RPOA. Risk factors for RPOA included higher doses of tanezumab, pre-existing subchondral insufficiency fractures, and co-administration with NSAIDs.28

The most recent study evaluated 2.5 mg and 5 mg tanezumab in 700 patients with moderate to severe OA pain.29 The drug was given subcutaneously every 8 weeks for a total of 24 weeks, and the 5 mg dose resulted in improvement vs placebo in pain and function. The improvement was modest but significantly better than studies with NSAIDs or COX-2 medications. RPOA occurred in 2% of the 2.5 mg dose and 0.4% in the 5 mg dose compared to none in the placebo-treated patients.

A pooled analysis reported that tanezumab significantly improved pain and function in patients with OA of the knee or hip, including those with severe symptoms, those older than 65, and those with diabetes.30 The Western Ontario and McMaster Universities Osteoarthritis (WOMAC) pain index demonstrated that all 3 doses of tanezumab improved pain significantly better than placebo or 500 mgs of naproxen twice-daily and there was no difference in efficacy for the 5 mg or 10 mg dose of tanezumab (see Figure 1).

Figure 1. Change in Western Ontario and McMaster Universities Osteoarthritis Index Pain Scores from baseline. Modified from Reference 29.

Serological biomarkers were predictive of which tanezumab-treated patients developed RPOA, especially in patients who were chronic NSAID users, and may prove useful in the future.31 Unfortunately, there is no evidence that these agents are truly disease-modifying and long-range studies are needed to determine whether their significant short-term efficacy in pain reduction and improved function outweigh their potential for accelerated joint damage (see more detail in “Inside the Potential on Nerve Growth Factor Antagonists” from PPM’s June 2019 issue).

Fasinumab, another anti-NGF monoclonal antibody in development by Teva Pharmaceuticals, was evaluated in 342 patients with hip or knee OA who had inadequate response to analgesics.32 The drug was given in 4 doses every 4 weeks over 16 weeks and patients were followed to Week 36. There were significant and clinically important reductions in pain and physical function with all 4 doses of fasinumab compared to placebo. A greater than 30% improvement in WOMAC pain scores were noted in 60 to 70% of fasinumab vs 45% of placebo-treated patients. Adverse side effects occurred in 17% on active drug vs 10% on placebo. As seen with tanezumab, a dose-dependent RPOA was noted in a small number of patients. One out of 337 treated patients developed a destructive arthropathy. Extensive radiologic evaluation both at baseline and throughout the study was done to look for RPOA, osteonecrosis, or subchondral insufficiency fractures.

Historically, animal model testing of OA has consistently demonstrated significant pain reduction from anti-NGF therapy as well but have also caused a synovitis with more rapid cartilage degeneration, especially when started early in the OA disease model. There have been more than 15 clinical trials and two systematic reviews concluding that anti-NGF agents relieve pain and improve function in OA of the knee and hip.30

Overall, anti-NGF therapy was found to be more effective than analgesics, NSAIDs, and selective COX-2 inhibitors in relieving pain and improving function in patients with moderate to severe OA of the knee or hip (see Table III). It is likely that they will be available in the near future, although there may be a small risk of OA rapidly progressing. Careful radiologic monitoring will be necessary if they are approved.

Other New Analgesics for OA

Potential new approaches to analgesia in OA may also include medications that selectively target delta and kappa opioid receptors with some positive results in OA of the hip.33 Drugs targeting ion channels, including a synthetic trans-capsaicin that blocks vanilloid receptors, have shown promise in OA preliminary trials, for example.33 A single IA injection of a synthetic trans-capsaicin provided significant pain reduction in subjects with moderate OA in the knee.33 Novel drugs that target cannabinoid receptors are also being evaluated.

Modifying Structural OA Changes

Despite ongoing research from academic centers around the world, there are no DMOADs with proven efficacy. With a better understanding of the complicated interplay of inflammatory cytokines, bone marrow lesions, and cartilage degradation, animal models for knee and hip OA have evaluated a number of pathways and a few human trials of a potential DMOAD look promising.

The search for a DMOAD has been linked to the identification of OA effector genes, including transforming growth factor beta 1 (TGFB1), fibroblast growth factor 18 (FGF18), cathepsin K (CTSK), and anti-interleukin inhibitors (IL-1, IL-6 and IL-11).25 Trials with anakinra, an IL-1 receptor antagonist, tocilizumab, which targets IL-6, and the anti-tumor necrosis factor (TNF) inhibitor adalimumab are underway in knee OA, although targeting proinflammatory cytokines has not been helpful in preliminary OA trials. A gene therapy that produces therapeutic growth factor is also being trialed in early human studies.25

Bone morphogenetic protein, a member of the transforming growth factor family, has been approved for spinal fusion therapy and appeared beneficial after IA injections in Phase I OA trials and is being evaluated in Phase II OA trials.33 A recombinant fibroblast growth factor, sprifermin, was found to increase cartilage thickness of the femoral-tibial joint in a Phase II OA trial;34 a 5-year dose-ranging study is in progress. Human serum albumin has anti-inflammatory effects and was found to have efficacy in three clinical trials in more than 400 patients with severe knee OA.35 Wnt signaling, a group of signal transduction pathways that pass signals into a cell through cell surface receptors, is important in cartilage and bone homeostasis. In an animal model of OA, a small molecule Wnt pathway inhibitor inhibited a number of pathways involved in inflammation and cartilage breakdown. In early human studies, a single IA injection of the Wnt signaling inhibitor improved symptoms.36

Bisphosphonates used to treat osteoporosis, including alendronate and zoledronate, reduced OA hip and knee pain (zoledronate reduced subchondral bone lesions in a 2-year study in Australia).37 IA injections of atelocollagen, a type of collagen with reduced antigenicity, has been evaluated for potential disease-modifying and analgesic effect with positive early results.38 Strontium, also used in osteoporosis, improved pain and reduced cartilage loss in knee OA.39 Cathepsin K, a lysosomal cysteine protease expressed in osteoclasts, is upregulated in joints in early OA.40 A cathepsin k inhibitor and a tropomyosin receptor kinase inhibitor both demonstrated OA structural modification in animal models of OA and are being tested for evidence of reduction in bone marrow and cartilage loss in human knee osteoarthritis.40,41

Discussion and Summary

In the future, identifying pain phenotypes may be helpful in predicting OA management. For example, a trial with IA botulinum toxin was effective only in a subset of OA knee patients with nociceptive pain.25 Individuals with chronic, widespread pain indicative of a strong component of central sensitization would likely benefit more from medications like duloxetine or that focus on biopsychological therapy. For those with OA in the knees, synovitis and/or knee effusion may occur, needing IA drainage and either corticosteroids or hyaluronic acid injections. Lack of long-term efficacy and toxicity has led to more judicious use of analgesics, including acetaminophen and opioids in OA, as well as more topical NSAIDs.

Overall, there is a great need for newer analgesics in the treatment of knee and hip OA and many recommendations may be emerging (see Table IV). Biologics that target OA pain and function appear to be ready for primetime, notably anti-NGF agents. It has become increasingly recognized that the treatment of pain in OA may not improve OA pathology, but rather increase structural damage. This structural damage is a major concern with approval of anti-NGF agents and may determine whether they truly will revolutionize OA pain management. In the future, the search for disease-modifying osteoarthritis drugs will need to reconcile the poor correlation of pain reduction with structural improvement in recent drug trials.