Management of Osteoarthritis: Pharmacotherapy


Osteoarthritis (OA) is the most common chronic joint disorder and is characterized by a degradative and repair process of cartilage and subchondral bone associated with synovial inflammation that may be triggered by a variety of biochemical as well as mechanical insults. OA should not be considered as a degenerative cartilage disease anymore but as a dynamic process including destruction and repair, as well as inflammation.

Pharmacologic therapies include oral (systemic) agents, topical agents, dietary supplements, and intra-articular (IA) therapies. All of these therapies have, to some degree, a symptomatic effect, and belong to the symptom-modifying osteoarthritis drug (SMOAD) family. Some of them may also have a beneficial effect on the structural damage and could be considered disease-modifying osteoarthritis drug (DMOAD). A better understanding of OA pathogenesis is bringing new opportunities for novel targeted therapies, some of which are already under clinical investigation.

The primary outcome measures recommended by Outcome Measures in Rheumatology Clinical Trials (OMERACT) for clinical trials of OA comprise four core domains: pain, physical function, patient global assessment, physician global assessment and, for studies of at least 1 year in duration, joint imaging. For the study of DMOADs, primary endpoints should include joint space narrowing, pain, and physical function ( Table 23-1 ).

Table 23-1

Clinicians Guide to Anti-Inflammatory Therapy

Type of Risk No or Low GI Risk NSAID GI Risk
No CV risk (without aspirin) NS-NSAID (cost consideration) Coxibor NS-NSAID + PPIor Coxib + PPI for patients with previous GI bleeding
CV risk (with aspirin) Naproxen * Addition of PPI if GI risk warrants gastroprotection Add PPI irrespective of NSAIDCoxib + PPI for those with previous gastrointestinal bleeding

CV, cardiovascular; GI, gastrointestinal; (NS) NSAID, (nonselective) nonsteroidal anti-inflammatory drug; PPI, proton pump inhibitor.

From Lin J, Zhang W, Jones A, Doherty M. Efficacy of topical non-steroidal anti-inflammatory drugs in the treatment of osteoarthritis: meta-analysis of randomised controlled trials. BMJ 2004;329:324.

* NS or selective (low-dose) NSAID without established aspirin interaction if naproxen is ineffective.


Acetaminophen (Paracetamol)

Acetaminophen (also known as paracetamol) is a simple analgesic that has both analgesic and antipyretic actions.

As a treatment for symptoms of OA, acetaminophen at up to 4 g/day is widely recommended as first-line therapy, for example, by the European League Against Rheumatism (EULAR), the American College of Rheumatology (ACR), and the Osteoarthritis Research Society International (OARSI).

Symptomatic Efficacy

The description of efficacy and safety of acetaminophen for the treatment of OA has been well updated in a recent meta-analysis. Acetaminophen was given at a dosage of 2600 to 4000 mg/day in the patient’s knee or hip OA. Most of the studies lasted around 6 weeks and reported a superiority of acetaminophen compared with placebo. A pooled analysis of five trials demonstrated a statistically significant reduction in pain compared with placebo (pain decreased by 4 points on a scale of 0–100 in the acetaminophen group compared with placebo). The clinical significance of such a low effect remains debatable. No beneficial effect was observed on physical function.

Comparing acetaminophen with nonselective nonsteroidal anti-inflammatory drugs (NSAIDs) (ibuprofen, diclofenac, naproxen, and others) or selective cyclooxygenase-2 (COX-2) inhibitors (celecoxib, rofecoxib, and others), acetaminophen was less effective than NSAIDs in terms of pain reduction (pain decreased by 6 points on a scale of 0 to 100 in people who took NSAIDs compared with acetaminophen), improvement in global assessments (patient and investigator), stiffness and improvement in some parameters of functional status such as the Western Ontario and McMaster Universities (WOMAC) function scale, but not in the 50-foot walk time, nor in the Health Assessment Questionnaire (HAQ) or the Lequesne’s algofunctional index.


In people taking NSAIDs, gastrointestinal (GI) discomfort occurred more frequently than in those taking acetaminophen (relative risk [RR], 1.47; 95% confidence interval [CI], 1.08–2.00) or a COX-2-selective NSAID.

In this meta-analysis, no conclusion could be drawn concerning the serious adverse events, including serious GI, renal, and cardiovascular safety, because of the small number of participants and the relatively short period of time in most studies (6 days to 2 years).

García Rodríguez and associates reviewed the data on the risk of upper GI complications associated with NSAIDs or acetaminophen, and provided evidence that high-dose acetaminophen may not be as safe as previously thought. Surprisingly, in two long-term studies, acetaminophen was associated with an increased risk for serious upper GI complications, with a clear dose-effect relationship. There was also evidence of a very strong interaction between the use of acetaminophen at doses of 2000 mg/day or more and NSAIDs (RR, 16.6; 95% CI, 11-24.9). However, a bias due to channeling cannot be ruled out in such an epidemiologic study. Hepatic toxicity was also reported in patients taking high doses (4 g/day) in particular when associated with alcohol consumption. Finally, recent data suggest that high-dose acetaminophen could induce hypertension, although the explanation remains unclear.


Acetaminophen is considered a safe treatment of OA, but the effect-size (improvement expressed as the fraction of the standard deviation) is modest (0.21; 95% CI, 0.02–0.41). NSAIDs have a higher effect size than acetaminophen for pain relief (effect size 0.32; 95% CI, 0.24–0.39) but a less favorable safety profile. The choice of treatment for each patient ultimately depends on both effectiveness and safety profile in addition to availability, cost, and patient acceptance. According to the recommendation by EULAR, ACR, and OARSI, acetaminophen remains the first line oral analgesic in the management of OA. Additional randomized controlled trials (RCTs) are necessary to better identify the patients with OA who are more likely to benefit from acetaminophen.


Opioids are potent analgesics targeting the opioid receptors. Biologic studies have shown that opioid receptors are present in inflamed OA synovial tissue.

Symptomatic Efficacy

In a systematic literature analysis of RCTs, Avouac and colleagues found that opioids significantly decrease pain intensity but have small benefits on function compared with placebo in patients with OA. In two meta-analyses, Cepeda and coworkers sought to determine the effect of oral tramadol (mean daily dose: 200 mg, during 1 week to 3 months) in patients with symptomatic hip or knee OA. The placebo-controlled studies indicated that participants who received tramadol had less pain (−8.5 units on a 0 to 100 scale; 95% CI, −12.0 to −5.0) than patients who received placebo. The reduction in the WOMAC index was 8.5% larger in the tramadol group than the placebo group. In the three placebo-controlled studies that followed participants for more than eight weeks, tramadol was more effective than placebo concerning pain intensity and patient global assessment. Thus, pooled results suggest that in people with OA, tramadol may decrease pain, improve overall well-being, slightly decrease stiffness and slightly improve function more than placebo.

In knee OA, no difference has been shown between tramadol (200 mg/day) and placebo among nonresponders to naproxen; in contrast, patients responding to naproxen had a significant improvement with tramadol and were able to reduce their dosage of the NSAID.


According to Cepeda and associates, participants who received tramadol had 2.27 times the risk of developing minor adverse events and 2.6 times the risk of developing major adverse events, compared with participants receiving placebo (minor adverse events: 39% versus 18% and major adverse events: 21% versus 8% respectively). The most common adverse events were nausea, vomiting, dizziness, constipation, somnolence, tiredness, and headache. There was no report of any life-threatening event. No comparison in terms of adverse events with other opioid drugs used in OA has been performed to date. Adverse events, although reversible and not life-threatening, may lead to discontinuation of the drug and could limit the use of tramadol.


Therefore, for patients suffering from severe pain or for which other analgesics/NSAIDs are contraindicated, opioids may be a reasonable alternative but should be monitored closely.


NSAIDs decrease inflammation and pain by inhibiting the enzyme cyclooxygenase (COX) and thereby reducing prostaglandin production, COX-1, found in most normal tissues including the GI tract, kidneys, and platelets, is considered a housekeeping enzyme that regulates the production of prostaglandins and thromboxane A 2 for several physiologic functions such as protection of gut mucosal integrity, vascular vasoconstriction, and platelet activation; COX-2, found particularly in the kidney, brain, bone and reproductive organs, is upregulated substantially in any tissue with inflammation or injury and is a key mediator of inflammation, explaining its role in arthritis pain. NSAIDs differ in their ability to inhibit COX-1 and COX-2: nonselective NSAIDs inhibit COX-1 and COX-2, whereas coxibs inhibit selectively COX-2.

Nonselective Nonsteroidal Anti-inflammatory Drugs

Symptomatic Efficacy

Several trials have compared nonselective NSAIDs with acetaminophen for hip and knee OA. All studies except one have found nonselective NSAIDs to be superior to acetaminophen.

In the Ibuprofen, Paracetamol Study in Osteoarthritis (IPSO) study, the analgesic efficacy of ibuprofen 400 mg as a single dose and as multiple doses (1200 mg/day) was compared with that of acetaminophen either as a single dose of 1000 mg or as multiple doses (3000 mg/day) in patients with knee or hip OA. Over 2 weeks, pain intensity, stiffness, and functional disability decreased significantly more in the ibuprofen than in the acetaminophen group.

Pincus and colleagues found significantly higher levels of improvement for diclofenac plus misoprostol than for acetaminophen over 6 weeks. This superiority of diclofenac plus misoprostol was observed especially in patients with more severe OA, whereas patients with mild OA had similar improvements with both therapeutic options. Melo Gomes and coworkers found that the efficacy of diclofenac/misoprostol in treating the signs and symptoms of OA was at least comparable to that of piroxicam or naproxen.


Gastrointestinal Disorders

In a meta-analysis of dyspepsia and NSAIDs, Ofman and associates found that dyspeptic symptoms occur in 4.8% of patients receiving NSAIDs compared with 2.3% receiving placebo and that they are the most common reason for cessation of therapy. Dyspeptic symptoms are dose-dependent but are a poor predictor of peptic ulcers: of those investigated for dyspepsia, 50% have a normal endoscopy, 15% gastro-esophageal reflux disease, 25% peptic ulcers, and 2% malignancies. Endoscopic abnormalities are more likely in patients older than 45 years of age. Serious events such as perforation or bleeding from upper GI ulcers occur in 2% of NSAID users, but are usually not preceded by symptoms. The link between ulcers on endoscopy and these complications needs to be investigated further. It has been shown that prostaglandin analogues (misoprostol 800 µg/day), proton pump inhibitors (PPIs), and double-dose H2-receptor antagonists are effective at preventing chronic NSAID-related endoscopic gastric and duodenal ulcers (misoprostol: RR, 0.17; 95% CI, 0.11–0.24, PPIs: RR, 0.40; 95% CI, 0.32–0.51, H2-receptor antagonists: RR, 0.44; 95% CI, 0.26–0.74).

NSAIDs may also cause lower intestinal side effects, such as ulcers, bleeding, inflammation and scarring in the small intestine and colon: lower intestinal events account for nearly 40% of the serious GI events in a population of patients with rheumatoid arthritis (RA) treated with naproxen.

Predictors of serious toxicity are

  • Age 65 years or older: the risk rises for each decade after age 50, with an RR for those between age 50 and 60 years of 1.8 (compared with those younger than 50), and an RR of 9.2 for those older than the age of 80 years.

  • People with a past history of peptic ulcer disease: for example, the RR for serious toxicity using naproxen in such patients is 13.5.

  • People also taking anticoagulants: RRs exceed 6.0.

  • People also using corticosteroids: RRs vary between 2 and 6.

  • People using NSAIDs for prolonged periods.

Cardiovascular Effects

Singh and coworkers reviewed the published evidence and assessed the risk of acute myocardial infarction with nonselective NSAIDs. Nonselective NSAIDs as a class were associated with an increased risk for acute myocardial infarction (RR, 1.19; 95% CI, 1.08–1.31). Similar results were found for diclofenac (RR, 1.38; 95% CI, 1.22–1.57) and ibuprofen (RR, 1.11; 95% CI, 1.06–1.17) but not for naproxen (RR, 0.99; 95% CI, 0.88–1.11). Increases in blood pressure, which can be seen with any NSAID, may be one explanation for these increases in cardiovascular risk.

Renal Effects

Prostaglandins control renal blood flow, glomerular filtration rate, and salt and water excretion by the kidney. NSAIDs may cause sodium retention, hyporeninemic hypoaldosteronism, prerenal azotemia, acute interstitial nephritis, and nephritic syndrome. Acute renal dysfunction has been reported with NSAIDs. Moreover, many of the nonselective NSAIDs have been implicated as causing chronic renal failure with RRs ranging from 2 to 8.

Thus, special vigilance is needed in patients treated with antihypertensive drugs, those at risk for renal diseases such as diabetes, and in the elderly.

Other Adverse Effects

Other adverse effects including skin rashes, allergic reactions, mouth ulcers, headaches, tinnitus, exacerbation of asthma, and aggravation of inflammatory bowel disease have been described.

Selective Cyclooxygenase-2 Inhibitors (Coxibs)

Symptomatic Efficacy

In order to assess the effectiveness of COX-2–selective NSAIDs for the management of OA and RA, Chen and colleagues undertook a meta-analysis of RCTs for each COX-2 selective NSAID compared with placebo and nonselective NSAIDs. The coxibs were found to be similar to nonselective NSAIDs for the symptomatic relief of OA.


Gastrointestinal Disorders

Compared with non-selective NSAIDs, COX-2 selective NSAIDs (celecoxib, rofecoxib, and lumiracoxib) were found to be associated with significantly fewer clinical upper GI events, the reduction being around 50%. Although coxibs offer protection against serious upper GI events, the amount of evidence for this protective effect may vary across individual drugs. Moreover, no difference has been demonstrated when a coxib is compared with a classic NSAID plus PPI. Importantly, in patients at high GI risk, such as patients with previous ulcer bleeding induced by nonselective NSAIDs, neither celecoxib nor diclofenac plus omeprazole adequately prevents ulcer recurrence : in this population, celecoxib plus a PPI is more effective than celecoxib alone for the prevention of recurrent ulcer bleeding.

Recent studies have shown that coxibs might have the additional advantage of a lower risk for GI complications, which cannot be achieved by other strategies. (Currently available gastroprotective therapies such as PPI do not protect the lower GI tract.) A post hoc analysis of serious lower GI clinical events in a prospective, double-blind trial in RA patients randomly assigned to naproxen 500 mg twice daily or rofecoxib 50 mg daily, found a rate of serious lower GI events (serious bleeding, perforation, obstruction, ulceration, or diverticulitis) per 100 patient-years of 0.41 for rofecoxib and 0.89 for naproxen (RR, 0.46; 95% CI, 0.22–0.93; P = 0.032). Serious lower GI events were 54% lower with the use of the selective COX-2 inhibitor rofecoxib. A more recent systematic review identified studies of NSAID or coxibs that reported on lower GI integrity (e.g., permeability), visualization (e.g., erosions, ulcers), and clinical events. Coxibs had significantly fewer harmful effects than nonselective NSAIDs in three of four integrity studies, one endoscopic study (RR for mucosal breaks: 0.3), and two randomized studies (RR for lower GI clinical events: 0.5; hematochezia: 0.4). Nevertheless, these data have not been confirmed in a recent prospective trial assessing lower GI clinical events in patients with OA or RA randomly treated with the COX-2–selective inhibitor etoricoxib (60 or 90 mg daily) versus the nonselective NSAID diclofenac (150 mg daily): no statistically significant decrease in lower GI clinical events was seen in this study. The risk of a lower GI clinical event with NSAID use seemed to be constant over time, and the major risk factors were a prior lower GI event and older age. More trials are warranted to more precisely estimate the effects of nonselective NSAIDs and coxibs on the lower GI tract.

The concept of CSULGIEs (Clinically Significant Upper and/or Lower GI Events) as the first and only available composite endpoint designed to evaluate the entire GI tract, may help physicians to consider the entire GI tract when choosing NSAIDs and protective therapies.

Two trials comparing celecoxib with a non-selective NSAID using CSULGIEs as a primary endpoint are in progress. The study outcomes will provide important information on the entire GI safety of these two commonly used therapies.

Cardiovascular Effects

There is still a debate on a specific cardiovascular risk of coxibs compared with classic NSAIDs. A coxib-induced imbalance between PGI 2 secretion in endothelial cells and thromboxane A 2 (TXA 2 ) platelet synthesis could explain why COX-2 inhibitors may increase the risk of CV events. In a meta-analysis of the incidence of serious vascular events in RCTs, coxibs were associated with a 42% relative increase in the incidence of serious vascular events compared with placebo, corresponding mainly to an excess of myocardial infarctions. There was no significant heterogeneity among the different coxibs. However, the incidence of serious vascular events was similar between a coxib and nonselective NSAID except naproxen.

Renal Effects

Concerning renal events, significant heterogeneity of renal effects across agents is observed suggesting that there is no class effect. Celecoxib is associated with a lower risk of both renal dysfunction (RR, 0.61; 95% CI, 0.40–0.94) and hypertension (RR, 0.83; 95% CI, 0.71–0.97) compared with NSAID controls.

Other Adverse Effects

COX-2 inhibitors share most of the side effects seen with the classical NSAIDs. Recently, lumiracoxib has been withdrawn because of its hepatotoxicity.


The choice between a classic NSAID and a coxib depends on the patient’s history and comorbidities. In any case, COX-2–selective and classic NSAIDs should be used at the lowest effective dose for the shortest possible duration of treatment. However, it is not rare that chronic use of the highest dose is necessary to achieve the clinical goals. Further trials assessing the relative efficacy of COX-2–selective NSAIDs versus combination of nonselective NSAIDs and gastroprotective agents in people at standard risk and those at a higher risk are needed.

Topical Nonsteroidal Anti-inflammatory Drugs and Capsaicin

In a meta-analysis of RCTs of short duration (less than 4 weeks), topical NSAIDs were found to be superior to placebo in relieving OA pain and stiffness and improving function in the first 2 weeks of treatment only. Effect sizes for pain at weeks 1 and 2 were modest at 0.41 (95% CI, 0.16–0.66) and 0.40 (95% CI, 0.15–0.65), respectively. However, topical NSAIDs were inferior to oral NSAIDs in the first week of treatment and associated with local side effects such as rash, itch, or burning. These results were in contradiction with those of a meta-analysis of long-term studies of topical NSAIDs in knee OA that found a pooled effect of topical NSAIDs at 4 weeks or beyond superior to placebo/vehicle in pain relief (mean effect size −0.28; 95% CI, −0.42 to −0.14), suggesting that topical NSAIDs are effective for pain relief in knee OA for a longer duration.

Topical capsaicin cream contains a lipophilic alkaloid extracted from chili peppers. Its efficacy in knee or hand OA is supported by a meta-analysis of RCTs: capsaicin cream was better than placebo in providing pain relief in OA (odds ratio [OR], 4.36; 95% CI, 2.77–6.88). However, complete blinding was impossible because of the initial discomfort associated with topical capsaicin (local burning, stinging or erythema).



Glucosamine is a natural precursor of the cartilage extracellular matrix component glycosaminoglycan. It has been widely promulgated as a remedy for OA on the basis that it might provide a substrate for matrix synthesis and repair. There are no dietary sources of glucosamine, and commercially available glucosamine is derived from shellfish. As a nutritional supplement, it is available in three forms: glucosamine hydrochloride (GH), glucosamine sulfate (GS), and N-acetyl-glucosamine.

The mechanisms of action for glucosamine in OA are poorly understood. In vitro studies have shown that adding GS to human chondrocytes results in increased proteoglycan synthesis. Antiarthritic effects of glucosamine are presumed to result from the provision of glucosamine as a substrate for articular cartilage glycosaminoglycan synthesis, stimulating the production of cartilaginous matrix. In addition, anti-inflammatory properties have been proposed.

The recommended dose of glucosamine in OA is 1250 mg/day.

Symptomatic Efficacy

The efficacy of glucosamine has been described in a meta-analysis according to the Cochrane guidelines. Comparing GS or GH versus placebo, glucosamine was significantly superior to placebo in decreasing pain (on a scale of 0 to 100, glucosamine achieved a 13-point greater improvement than placebo) and in improving the Lequesne’s Index (with a difference in the change from baseline between glucosamine and placebo of 2.3 units on the Lequesne scale). In contrast, there was no statistical difference between glucosamine and placebo for the WOMAC total score nor for its different subscales (pain, stiffness, function).

In a randomized trial, glucosamine (1500 mg/day), chondroitin sulfate (1200 mg/day), and the combination of both all failed to show a benefit on pain in a group of patients with OA of the knee, but post-hoc analyses suggested that the combination of glucosamine and chondroitin sulfate may be effective in the subgroup of patients with moderate-to-severe knee pain.

Structural Efficacy

Three meta-analyses suggested the ability of GS to delay structural progression in knee OA. In Poolsup and associates, the risk of disease progression was reduced by 54% (pooled RR, 0.46; 95% CI, 0.28–0.73; P = 0.0011). The number needed to treat was nine (95% CI, 6–20). Similar results were described by Richy and colleagues, who showed significant efficacy of glucosamine on joint space narrowing with an effect size of 0.41 (95% CI, 0.21–0.60; P < 0.001), which corresponds to a low to medium effect. These results were corroborated by those reported by the Cochrane Collaboration. It is noteworthy that these results are based on the assessment of joint space narrowing on x-ray studies. Because several biases are known to interfere with this measure (level of pain, positioning, and others), the clinical consequence of the reported differences remains questionable. Moreover, data on the long-term use of glucosamine in knee OA are sparse. One trial in radiographic knee OA (Kellgren-Lawrence grade 2 or grade 3 changes and joint space width of at least 2 mm at baseline) compared the structural efficacy of glucosamine 500 mg three times daily, chondroitin sulfate 400 mg three times daily, the combination of glucosamine and chondroitin sulfate, celecoxib 200 mg daily, or placebo over 24 months. At 2 years, no treatment achieved a predefined threshold of clinically important difference in the loss of joint space width as compared with placebo. Further long-term trials and trials evaluating different forms of glucosamine are warranted before the widespread use of this agent for the prevention of structural progression could be recommended.


Long-term treatment with glucosamine is well tolerated. The number of reported adverse effects was not significantly different between glucosamine and placebo. The most common problems reported to be associated with glucosamine were generally transient and considered mild to moderate, including abdominal pain, dyspepsia, diarrhea, increased blood pressure, fatigue, and rash.

Chondroitin sulfate

Chondroitin is a highly hydrophilic polysaccharide macromolecule. Its hydrocolloid properties confer much of the compressive resistance of cartilage. Proposed mechanisms of action of chondroitin sulfate include restoration of the extracellular matrix of the cartilage, prevention of further cartilage degradation, and/or a role in overcoming a dietary deficiency of sulfur-containing amino acids that are essential building blocks for cartilage extracellular matrix molecules. The recommended dose of chondroitin in OA is 1000 to 1200 mg/day.

Symptomatic Efficacy

A recent meta-analysis examined data on symptomatic and structural efficacy and safety of chondroitin in knee or hip OA, versus placebo or no treatment. The analysis revealed a high degree of heterogeneity among the trials. Large-scale, methodologically sound trials indicated that the symptomatic benefit is minimal, with an effect size of −0.03 (95% CI, −0.13 to −0.07) corresponding to a difference of 0.6 mm on a 100-mm VAS. These results differ from a previous meta-analysis in knee or hip OA, in which the aggregate effect size was 0.78 (95% CI, 0.60–0.95). When only high-quality or large trials were considered, this effect size was substantially lower, suggesting that quality issues and publication biases may have played a critical role. The effect sizes were relatively consistent for pain and functional outcomes.

In the recent RCT referred to earlier, which compared glucosamine (1500 mg daily), chondroitin sulfate (1200 mg daily), glucosamine plus chondroitin sulfate, celecoxib (200 mg daily), and placebo for 24 weeks in patients with knee OA, it was shown that the response to chondroitin sulfate, either alone or in combination with glucosamine, was not significantly higher than the response to placebo.

Structural Efficacy

Differences in changes between chondroitin and placebo groups revealed a small effect in favor of chondroitin: 0.16 mm on minimum joint space width (95% CI, 0.08–0.24) and 0.23 mm on mean joint space width (95% CI, 0.09–0.37), that corresponds to small effect sizes (e.g., 0.12 and 0.18, respectively). These effects are small, and their clinical significance is uncertain. Moreover, there was no evidence for structural efficacy of chondroitin versus placebo in radiographic knee OA in a 24-months study.


The RR for any adverse event with chondroitin sulfate does not differ from that of placebo.


Several studies (both in vitro and in animal models) suggest that the active diacetyl derivative of diacerein, rhein, may inhibit interleukin-1 (IL-1) production and the secretion of metalloproteinases without affecting the synthesis of prostaglandins. Diacerein has also been implicated in the regulation of transforming growth factor (TGF) beta-1 and beta-2 in articular chondrocytes. The recommended dose of diacerein in OA is 100 mg/day.

Symptomatic Efficacy

Rintelen and coworkers performed a systematic meta-analysis of RCTs with diacerein for knee and/or hip OA to provide an evidence-based assessment of its symptomatic efficacy. Diacerein was significantly superior to placebo during the active treatment phase (reducing pain, changes in functional impairment, global efficacy rating by patients), and comparable to standard treatments (mostly NSAIDs). However, diacerein, but not NSAIDs, showed a carryover effect, persisting up to 3 months after treatment, with a significant analgesic-sparing effect during the follow-up period. A meta-analysis performed by Fidelix and colleagues reviewed RCTs in order to confirm the effectiveness and safety of diacerein. When compared with placebo, pain on a VAS (0–100 mm) showed a statistically significant difference in favor of diacerein. However, a subgroup analysis according to the localization (knee or hip OA) showed an absence of efficacy. There was no improvement in the Lequesne’s index, either in the group as a whole or in the subgroup analyses. Concerning total hip replacement, no evidence of a benefit was found. When comparing diacerein with NSAIDs, no significant difference was found in terms of pain on VAS, WOMAC function, or analgesic intake. Nevertheless, the studies were heterogeneous, evaluated different joints (knee or hip) for a short period (mean of 2.5 months), and the results were expressed with very large confident intervals. One 8-week study in hip OA, comparing diacerein plus NSAID with placebo showed better results in the former group in reducing pain and improving function (Lequesne’s index). Diacerein was compared with other symptomatic slow-acting drugs in osteoarthritis such as NRD 101 (a new hyaluronic acid [HA] high-molecular-weight polysaccharide) in knee OA for 1 year or harpadol in knee or hip OA for 4 months. In separate subgroup analyses, there was no difference between the groups in the following outcomes: pain (VAS 0–100 mm), Lequesne’s index, global efficacy patient assessment, painful days in the previous month and radiographic progression.

Structural Efficacy

In the 3-year controlled ECHODIAH trial including 507 patients suffering from painful OA of the hip, diacerein was compared with placebo. Structural progression was evaluated radiographically, and defined as the time to a 0.5-mm loss of joint space width. Using a life table approach, progression was significantly less frequent, and average time-to-progression significantly longer, in the diacerein group.


Tolerability assessments revealed a statistically significant inferiority of diacerein versus placebo. Diacerein is an anthraquinonic derivative; therefore, diarrhea was the most frequent adverse event (from 39% to 42% of patients treated by diacerein). The severity of diarrhea was mild to moderate, occurred within the first 2 weeks of the treatment, and resolved on continuing treatment. In most patients, this did not result in treatment interruption. The second most prevalent side effect is discoloration of the urine, which has no clinical consequences but may interfere with the blinding design in clinical trials. Finally, allergic events affecting the skin (pruritus, rash) are not rare.

Avocado-Soybean Unsaponifiables

It has recently been shown that avocado-soybean unsaponifiables (ASUs) can counteract stress-activated signaling pathways in chondrocytes. The recommended dose in OA is 300 mg/day (100 mg avocado oil and 200 mg soybean oil).

Symptomatic Efficacy

The symptomatic efficacy of ASU in hip or knee OA has been evaluated in a meta-analysis of RCTs. The average trial duration was 6 months. Both pain reduction and the Lequesne index favored ASU (effect size 0.39; 95% CI, 0.01–0.76; P = 0.04, and 0.45; 95% CI, 0.21–0.70; P = 0.0003, respectively), but results were heterogeneous between trials. The number of responders following ASU compared with placebo (OR, 2.19; P = 0.007) corresponded to a number needed to treat 6 (4–21) patients. Meta-analysis data support better chances of success in patients with knee OA than in those with hip OA.

Structural Efficacy

One randomized, double-blind, placebo-controlled pilot trial failed to demonstrate a structural effect of ASU in 108 patients suffering from hip OA (Kellgren-Lawrence grade 1 to 3). However, in a post hoc analysis, ASU significantly reduced the progression of joint space loss as compared with placebo in the subgroup of patients with advanced joint space narrowing, suggesting that ASU could have a structural effect. Confirmation in a larger placebo-controlled study is required.


ASU is generally well tolerated. Occasional regurgitations with unpleasant taste can be avoided by taking the pill in the middle of a meal. Allergic reactions are rare. Hepatic disorders can occur but remain exceptional.


S-adenosylmethionine (SAMe) is a dietary supplement now available in the United States. SAMe is proposed as an antidepressant, a medication for cholestasis and liver disorders, a treatment for migraines, and a therapy for fibromyalgia or OA. In OA, its mechanism of action remains controversial.

The efficacy of SAMe has been assessed in a meta-analysis of RCTs versus placebo or NSAIDs. When compared with placebo, SAMe was more effective in reducing functional limitation (effect size 0.31; 95% CI, 0.099–0.520), but not in reducing pain, but data were based on only two studies. SAMe appeared to be as effective as NSAIDs in reducing pain (effect size 0.12; 95% CI, −0.029 to 0.273) and in improving functional limitation (effect size 0.025; 95% CI, −0.127 to 0.176) without the adverse effects often associated with NSAID therapies.

Only gold members can continue reading. Log In or Register to continue

May 19, 2019 | Posted by in RHEUMATOLOGY | Comments Off on Management of Osteoarthritis: Pharmacotherapy

Full access? Get Clinical Tree

Get Clinical Tree app for offline access