Article Text


Intra-articular therapy in osteoarthritis
  1. I Uthman1,
  2. J-P Raynauld2,
  3. B Haraoui2
  1. 1Department of Internal Medicine, Faculty of Medicine, American University of Beirut, Beirut, Lebanon
  2. 2Osteoarthritis Research Unit, Department of Rheumatology, Hôf.pital Notre-Dame, Centre Hospitalier de l’Université de Montréal, Montréal, Québec, Canada
  1. Correspondence and reprint requests to:
 Dr Imad W Uthman, American University of Beirut Medical Centre, PO Box: 113–6044, Hamra, 1103 2090, Beirut, Lebanon; 


The medical literature was reviewed from 1968–2002 using Medline and the key words “intra-articular” and “osteoarthritis” to determine the various intra-articular therapies used in the treatment of osteoarthritis. Corticosteroids and hyaluronic acid are the most frequently used intra-articular therapies in osteoarthritis. Other intra-articular substances such as orgotein, radiation synovectomy, dextrose prolotherapy, silicone, saline lavage, saline injection without lavage, analgesic agents, non-steroidal anti-inflammatory drugs, glucosamine, somatostatin, sodium pentosan polysulfate, chloroquine, mucopolysaccharide polysulfuric acid ester, lactic acid solution, and thiotepa cytostatica have been investigated as potentially therapeutic in the treatment of arthritic joints. Despite the lack of strong, convincing, and reproducible evidence that any of the intra-articular therapies significantly alters the progression of osteoarthritis, corticosteroids and hyaluronic acid are widely used in patients who have failed other therapeutic modalities for lack of efficacy or toxicity. As a practical approach for a knee with effusion, steroid injections should be considered while the presence of symptomatic “dry” knees may favour the hyaluronic acid approach. The virtual absence of serious side effects, coupled with the perceived benefits, make these approaches attractive.

  • osteoarthritis
  • intra-articular therapy
  • corticosteroids
  • hyaluronic acid
  • kD, kilodalton
  • MPA, methylprednisolone acetate
  • NaPP, sodium pentosan polysulfate
  • NSAIDs, non-steroidal anti-inflammatory drugs
  • WOMAC, Western Ontario and McMaster Universities Osteoarthritis Index

Statistics from

Osteoarthritis is a disease characterised by a mixture of degradative and reparative processes in the articular cartilage, subchondral bone associated with marginal osteophyte formation, and low grade inflammation. Osteoarthritis involves mainly the hips, knees, spine, and the interphalangeal joints. Its clinical presentation is usually monoarticular or oligoarticular with fluctuations in intensity and localisation over time. It is therefore logical to consider local therapeutic modalities in order to avoid untoward systemic effects. Several compounds have been used intra-articularly in open label and in double blind randomised clinical trials. In this article we review the various intra-articular therapeutic modalities used in the treatment of osteoarthritis.


Corticosteroids are very potent anti-inflammatory agents. Intra-articular injections have been used for nearly five decades for the symptomatic treatment of osteoarthritis.1 A survey of rheumatologists in the US suggested that more than 95% use this treatment sometimes and more than half of them, frequently.2 Intra-articular corticosteroids are also recommended in the recent American College of Rheumatology guidelines for the management of acute knee pain in patients with osteoarthritis.3

Corticosteroids exert their anti-inflammatory action by interrupting the inflammatory and immune cascade at several levels including: impairment of antigen opsonisation, interference with inflammatory cell adhesion and migration through vascular endothelium, interruption of cell-cell communication by alteration of release or antagonism of cytokines (interleukin-1), impairment of leukotriene and prostaglandin synthesis, inhibition of production of neutrophil superoxide, metalloprotease and metalloprotease activator (plasminogen activator), and decreased immunoglobulin synthesis.4

Earlier studies have reported that corticosteroids injections may suppress cartilage proteoglycan synthesis, worsen cartilage lesion, or even cause degenerative lesions in normal cartilage.5 In more recent reports, using animal model (rabbits and guinea pigs) of osteoarthritis, several investigators have reported that low dose intra-articular corticosteroids (sufficient to suppress catabolism) normalised cartilage proteoglycan synthesis and significantly reduced the incidence and severity of cartilage erosions and osteophyte formation.6

In a study using the Pond-Nuki dog model of osteoarthritis, an animal model which closely resembles the changes observed in human osteoarthritis, intra-articular injection of low dose triamcinolone acetonide (5 mg), a long acting corticosteroid preparation, reduced osteophyte formation and cartilage lesions and exerted a protective effect on cartilage histological changes (confined the spread of cartilage erosions, reduced the formation of fissures and cell clones, and preserved matrix proteoglycans).6 In a more recent study, also using a dog experimental model of osteoarthritis, intra-articular injections with methylprednisolone acetate (MPA) significantly reduced the incidence and size of osteophytes and the severity of cartilage lesions; this was associated with a decrease in chondrocyte stromelysin synthesis.7 The beneficial effect of corticosteroids on cartilage can be due to several possible mechanisms: in vivo suppression of metalloproteases synthesis (probably by suppression of gene expression), decreasing the synthesis of possible physiological activators such as plasminogen activator/plasmin or by increasing the level of tissue inhibitors of metalloproteases, or by suppression of the synthesis of synovial factors such as interleukin-1.6–8

In humans, repeated corticosteroid injections in knees of patients with chronic arthritis presented no evidence of destruction or accelerated deterioration.9 In a recent study, a single intra-articular injection of triamcinolone acetonide to patients with knee osteoarthritis provided short term pain relief.10 Increased benefit was associated with both clinical evidence of joint effusion and successful aspiration of synovial fluid at the time of injection.10

To demonstrate the difference in outcome and rate of cartilage degradation among patients with knee osteoarthritis treated with intra-articular corticosteroid injections, we have studied 70 patients fulfilling the American College of Rheumatology criteria for primary knee osteoarthritis who were randomly chosen to receive intra-articular injections of a corticosteroid (40 mg triamcinolone acetonide) or a vehicle, at three month intervals, for a prospective period of two years. Patients with a standing radiograph showing a joint space width of the knee internal compartment of less than 2 mm were excluded. Conventional treatment (analgesics and non-steroidal anti-inflammatory drugs (NSAIDs) except indomethacin) was prescribed according to the practitioner’s usage and was systemically recorded. The radiological progression of joint space narrowing of the injected knee over two years was our primary outcome variable. The progression was assessed blindly at one and two years of treatment with the validated Buckland-Wright protocol. Clinical evaluation of the patients using validated measures Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC), global assessments with visual analogue scales, 50 foot walking time) was also conducted every three months by blinded investigators; 68 patients completed the study. Baseline characteristics between the two groups regarding clinical and anatomical measurements were similar.11,12 No difference was noted between the two treatment groups at one and two years with respect to loss of joint space over time (p = 0.57). At the one and two year follow up evaluations, the patients injected with triamcinolone acetonide showed a trend toward greater symptom improvement, especially at the first year follow up, for the WOMAC pain subscale, night pain assessment, and the range of movement (p = 0.05) compared with patients injected with saline. Moreover, using area under the curve analyses, knee pain and stiffness were significantly improved throughout the two year study, by repeated injections of triamcinolone acetonide compared with saline injections (p<0.05).12

Different corticosteroid formulations have been used over the years, with similar general efficacy: triamcinolone acetonide, MPA, and prednisolone acetate were administered in single or multiple repetitive injection regimens.10–20 In one trial, triamcinolone acetonide had a longstanding effect superior to betamethasone16; this observation, however, was not shared by other investigators who demonstrated a short lived beneficial effect not lasting beyond one week.15,18,20 Despite several observed flaws in the design of these trials, valuable information has been gathered, especially regarding the short term safety of this approach. Moreover, the aspiration of the effusion, when present, seems to bring additional short term clinical benefit to triamcinolone acetonide injections.10 Despite their wide use in clinical practice, no predictors of response could be identified in two recent randomised double blind trials.10,20

The perceived efficacy and lack of major toxicity have made intra-articular corticosteroid injections one of the mainstays of the management of osteoarthritis, in particular, osteoarthritis of the knee.


Viscosupplementation is a novel, safe, and possibly effective form of local treatment for osteoarthritis.21 It aims at supplying replacement hyaluronic acid into the joint space to return the elasticity and viscosity of the synovial fluid to normal.22 Hyaluronic acid, a polysaccharide consisting of a long chain of disaccharides (β-D-glucuronyl-β-D-N-acethylglucosamine), is a natural component of cartilage and plays an essential part in the viscoelastic properties of the synovial fluid. Because of its hyaluronic acid content, joint fluid acts as a viscous lubricant during slow movement of the joint, as in walking, and as an elastic shock absorber during rapid movement, as in running. It is considered not only a joint lubricant, but also a physiological factor in the trophic status of cartilage. Hyaluronic acid has a very high water binding capacity: when 1 g of hyaluronic acid is dissolved in physiological saline, it occupies three litres of solution.23 The estimated total hyaluronic acid in a human knee joint is from 4–8 mg.23

Intra-articular injections of hyaluronic acid with a molecular weight between 500 and 750 kilodaltons (kD) have been studied using corticosteroid injections24–28 or NSAIDs29 as control treatment, and by conducting placebo controlled clinical trials.30–37 These studies suggested that intra-articular injection of hyaluronic acid may improve the clinical condition and have a long term beneficial effect in knee osteoarthritis patients especially if osteoarthritis was less than moderate in grade.30–37 A review by Maheu looked at five different clinical trials, comparing different regimens of Hyalgan (a hyaluronan preparation; Sanofi-Synthelabo) versus corticosteroid injections in the osteoarthritis knee, with follow ups from 2–12 months.38 One study showed initial superiority of Hyalgan over corticosteroid injections and in three studies, equal efficacy over time. The fifth study used a combination of Hyalgan and corticosteroid injections initially. The corticosteroid injections seemed to increase the long term efficacy of the hyaluronic acid suggesting that the combination of these two local treatments would be promising. In a recent randomised clinical trial of 495 patients with osteoarthritis of the knee, Altman et al demonstrated that one intra-articular injection of Hyalgan each week for a period of five weeks is at least as effective as continuous treatment with naproxen for 26 weeks, with fewer adverse reactions.39 The long term structure modifying properties of Hyalgan were investigated through a double blind randomised placebo controlled trial using a standardised arthroscopic score after one year of follow up. Hyaluronic acid proved superior to placebo in two out of three parameters used to quantify the severity of osteoarthritis lesions.40

On the other hand, however, a double blind placebo controlled study involving 91 patients with radiologically confirmed knee osteoarthritis concluded that intra-articular administration of 750 kD hyaluronic acid offered no significant benefit over placebo during a five week treatment period, but incurred a significantly higher morbidity.41 Furthermore, in a study that evaluated the effect of joint lavage with lactated Ringer’s solution in 23 patients, the secondary goal was to determine if any additional benefit could be obtained by injecting the knee with hyaluronic acid after washout treatment.42 Improvement was noted at one and two year follow ups; however, there were no statistically significant differences in outcome for the hyaluronic acid and placebo groups. A one year double blind control study involving 52 patients compared the effect of intra-articular injections of hyaluronic acid and placebo, both administrated weekly for five weeks.43 Though both groups improved from baseline, there was no statistically significant difference in any of the relevant variables at any time point. Another large randomised double blind placebo controlled trial also found no significant difference at 20 weeks between the two groups when compared to their baseline evaluation. However, once stratified according to age and disease severity, hyaluronic acid proved more efficacious than placebo for patients over 60 years of age who had the most severe knee osteoarthritis.44

Intra-articular injection with hyaluronic acid is a relatively safe approach; however, some adverse events have been reported. In a retrospective study of 336 patients treated by five Canadian rheumatologists over 2.5 years (1537 injections involving 458 knees) it was suggested that the incidence of local adverse events after injection of hylan G-F 20 was relatively low and was influenced by the injection technique,45 and with a medial approach and a partially flexed knee, the incidence was 5.2%; with a straight medial approach, 2.4%; and with a straight or lateral approach, 1.5%. In another study of 22 patients who received a total of 88 injections in 28 knees, post-injection “flares” occurred in 27% of patients and after 11% of injections.46 In some cases, acute synovitis was associated with joint swelling lasting up to three weeks.47 Other rare side effects included haemarthrosis,40 muscle pain,48 and pseudogout.49,50

In conclusion, in the majority of studies, a clinical benefit of treatment was reported compared with the injected control group. Compared with treatment with local corticosteroids, hyaluronic acid showed equal pain relief in the first few weeks after therapy, with hyaluronic acid demonstrating more sustained benefit up to 60 days.51 In all trials reporting adverse effects, the primary adverse effect with hyaluronic acid was pain at the injection site.51 Although considerable evidence supports the positive effects of hyaluronic acid on joint cellular and immunological function there is still insufficient information to permit a conclusion concerning the effect of this treatment, if any, on the progression of osteoarthritis in humans.52


Other substances such as orgotein, radiation synovectomy (dysprosium-165hydroxide macroaggregate, yttrium-90 silicate), dextrose prolotherapy, silicone, saline lavage, saline injection without lavage, analgesic agents (bupivicaine, morphine), NSAIDs (tenoxicam, indoprofen, phenylbutazone), glucosamine,53 somatostatin,54 sodium pentosan polysulfate (NaPP),55 chloroquin,56 mucopolysaccharide polysulfuric acid ester,57 lactic acid solution,58 thiotepa,59 cytostatica (thiotepa, azetepa, and osmium acid),60 have been investigated as potentially therapeutic in the treatment of arthritic joints. Orgotein is the pharmaceutical form of the bovine enzyme Cu-Zn superoxide dismutase. The anti-inflammatory properties of orgotein were discovered in 1965.61

A randomised double blind study comparing orgotein injections with intra-articular MPA injections found that orgotein could be used safely and effectively without serious adverse reactions.62 The efficacy of orgotein was compared to that of betamethasone over a one year period in 419 patients with knee osteoarthritis.63 Though betamethasone acted more quickly, orgotein at low doses (4 or 8 mg) was comparable to the corticosteroid from week 4, and up to a year of follow up. The main adverse reactions to orgotein were pain, swelling, stiffness, prickling or burning sensations, skin rashes, or a feeling of heaviness at the injection site.62,63

An observational prospective study evaluating the effects of radiation synovectomy was conducted in 40 patients with knee osteoarthritis over a one year period.64 A marked improvement in pain and evaluation scores occurred at three months, but had disappeared by six months after treatment. The safety and efficacy of dysprosium-165 hydroxide macroaggregate was compared to yttrium-90 silicate for radiation synovectomy of the knee in a multicentre double blind clinical trial, with no significant difference in clinical response between the two treatment groups.65 No clinically significant side effects were observed.65 Reeves and Hassanein in a randomised prospective double blind placebo controlled study of dextrose prolotherapy for knee osteoarthritis concluded that injection with 10% dextrose resulted in a statistically significant clinical improvement.66

Intra-articular silicone injections have been used in few studies.67–69 Wright et al conducted a pilot study in five patients, with a control group of 25 outpatients, to evaluate intra-articular silicone as an artificial lubricant for osteoarthritis joints.68 Sequential analyses showed a significant benefit from saline compared with silicone at one week follow up, and no significant difference at one month.

A randomised placebo controlled study with 20 patients was carried out to assess saline lavage of knee osteoarthritis versus intra-articular saline injection without lavage.70 Though both groups showed improvement, knee washout conferred no further benefit. However, in a recent randomised single blind study, tidal knee irrigation with saline in 77 patients with knee osteoarthritis showed a greater reduction in pain than did conservative medical management.71 In a 24 week placebo controlled study in 98 patients, Ravaud et al compared joint lavage to a single corticosteroid injection and the combination of both interventions.72 Intra-articular corticosteroids had a beneficial effect on pain as early as the first week, but it was lost by week 12. On the other hand, joint lavage had a delayed onset of action (week 4) but lasted up to the final 24 week evaluation.

Since pain is the main symptom in knee osteoarthritis, intra-articular analgesic agents were tested in three studies. In a single blind trial, 20 patients were randomised to receive either intra-articular bupivicaine or placebo73; the local anaesthetic agent had a short lived significant effect (less than 24 hours). In a crossover placebo controlled design, intra-articular morphine (100 mg) was shown to be superior to placebo and had a long lasting effect, up to nine days, which was the last evaluation period for this study.74 Stein et al examined intra-articular morphine in chronic arthritis in a randomised double blind comparison between intra-articular morphine (3 mg), dexamethasone (4 mg), and saline (3 ml) in 44 patients with chronic inflammatory arthritis or osteoarthritis of the knee.75 They concluded that intra-articular morphine produces analgesia of similar magnitude to dexamethasone and it may have anti-inflammatory actions in chronic arthritis.

Trials on the intra-articular injection of NSAIDs have been reported by a number of investigators. A single tenoxicam intra-articular injection was also superior to placebo with no local side effects.76 Egsmose et al in a triple blind controlled study comparing the effect of IA indoprofen with placebo in hip joint osteoarthritis, did not find any superior effect of intra-articular indoprofen compared to placebo as regards pain relief or increased joint mobility.77 Moens and Moens tried intra-articular injection of phenylbutazone in a patient with gonarthrosis.78 Glucosamine was shown to be safe and provided a greater benefit than placebo.53 Somatostatin led to pain reduction and increased joint mobility with no reported adverse reactions.54 Finally, NaPP isolated from beechwood hemicellulose has anticatabolic effects in osteoarthritis by direct enzyme inhibition and gene expression alteration of metalloproteinases. One injection of NaPP each week for a period of four weeks was administered intra-articularly in 15 patients with knee osteoarthritis and they showed a greater improvement in pain and mobility, as well as synovial fluid viscosity, at two months compared with a control group 16 patients injected with saline55; however, data on long term efficacy and safety are not available.


Despite the lack of strong, convincing, and reproducible evidence that any of these intra-articular therapies significantly alters the short term outcome and even less so the progression of osteoarthritis, corticosteroids and hyaluronic acid are widely used in patients who have failed other therapeutic modalities for lack of efficacy or toxicity. As a practical approach for a knee with effusion, steroid injections should be considered, while the presence of symptomatic “dry” knees may favour the hyaluronic acid approach. The virtual absence of serious side effects, coupled with the perceived benefits, make these approaches attractive.


View Abstract

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.