Article Text


Intra-articular hyaluronan treatment for osteoarthritis
  1. Rheumatology Unit, Arrowe Park Hospital, Arrowe Park Road, Upton, Wirral, Merseyside L49 5PE

    Statistics from

    In 1934 Meyer and Palmer isolated a polysaccharide from bovine vitreous humour and named it hyaluronic acid.1 The name derives from the Greek word “hyalos” (that is, glass like). In 1986 Balazs et al 2 proposed the name hyaluronan as an alternative to hyaluronic acid. “Hyaluronic acid”, “sodium hyaluronate”, and “HA” are terms used in the literature and frequently interchangeably with “hyaluronan”, but the latter has become the preferred name.

    In osteoarthritis (OA) the synovial fluid (SF) is more abundant and less viscous,3 the concentration of hyaluronan is decreased as is its chain length and therefore molecular weight.4 These changes severely decrease the physiological and protective functions of the SF5; shock absorption, traumatic energy dissipation and storage, lubrication and protective coating of the articular cartilage surface and of the inner lining of the synovial membrane, as well as control of the “traffic” between the synovial vessels and cartilage by an exclusion effect on migrating cells and large molecules.

    The concept of viscosupplementation, first proposed by Balazs,6 is based on the hypothesis that intra-articular injection of hyaluronan into OA joints could restore the rheological properties of the SF, promote the endogenous synthesis of a higher molecular weight and possibly more functional hyaluronan, thereby improving mobility, articular function, and decreasing pain. This mechanism is suggested by the finding of hyaluronan with a molecular weight larger than that of the injected solution and within the range of normal SF hyaluronan in humans7 and animals.8

    Hyaluronan used for medical uses has to be extracted from umbilical cord, rooster comb or bacterial cultures then purified and separated from any pyrogenic, inflammatory, immunogenic or chemotactic fractions. The first therapeutic injections of hyaluronan in animal joints were performed on track horses for traumatic arthritis, the treatment proved effective and since then has been widely used in veterinary medicine.9

    Preliminary human clinical studies were performed in the early seventies using the non-inflammatory fraction of hyaluronan (NIF-Na HA). Healon (molecular weight about 2 ×106 Da) became the trade name of the first NIF-Na HA produced by Biotrics Inc (Arlington, MA USA). This trade name for NIF-HA is now used by Pharmacia Upjohn in ophthalmic viscosurgery. Synvisc (Hyalan gel fluid 20, (molecular weight about 7 × 106 Da) Biomatrix, New Jersey, USA) available in Canada and Sweden and recently launched in the UK. Orthovisc (Anika Therapeutics, USA) is marketed in Canada, the Netherlands, and Turkey. Hyalgan (molecular weight about 5 × 105 Da) Fida, Italy) has been launched in Canada, south ease Asia, Italy, France, Germany, and some South American countries and is about to be launched in the UK. Artz (molecular weight about 7 × 105 Da) Seikagaku Kogyo, Japan) has been launched in Japan and Sweden. A number of other products have been or are under development for ophthalmological, veterinary, and musculoskeletal indications.

    Other than the obvious effects of intra-articular hyaluronan by supplementation of the lubricating properties of the SF, evidence for other effects primarily comes from animal studies. In vitro studies have indicated that hyaluronan has a wide range of effects on both chondrocytes and synoviocytes and on the inflammatory cells that invade the synovial cavity during osteoarthritis (OA). In vivo studies in animal models and preliminary studies in humans have indicated that hyaluronan may protect cartilage structure, exert an effect on the inflammatory process, and reduce pain by acting directly on cell receptors.10 It has been reported that hyaluronan may preserve chondrocyte vitality and cartilage structure,11and reduce synovial cell proliferation,12 13 reduce bradykinin induced pain with high molecular weight hyaluronan and kaolin and carrageenan induced arthritis.14 It is speculated that hyaluronan may bind inflammatory mediators and free radicals, removing them from the joint space via lymphatics thereby reducing cartilage damage.15 Figure 1 shows some of the functions of hyaluronan.

    Figure 1

    Physicochemical and cellular functions of hyaluronan.

    The first therapeutic studies in humans of hyaluronan in knee OA were reported by Rydell and Balazs16 and Peyron and Balazs6 and several years later by Weiss et al.17 Since then studies have attempted to determine not only if intra-articular hyaluronan in OA is a therapeutically useful technique but also optimal treatment regimens—that is, frequency and number of injections9 and also if the molecular weight of the injected hyaluronan is an important determinant as to its effectiveness. Studies using the higher molecular weight preparations have shown them to be more effective with a greater duration of action.17 18

    Many cells carry receptors for hyaluronan and this molecule may effect cell-cell, cell-matrix, and cell-receptor-ligand interactions.16 For example, interference with the hyaluronan receptor CD44 abrogated tissue oedema and leucocyte infiltration in a murine arthritis model.17

    The specific receptor for hyaluronan is a glycoprotein belonging to the CD44 adhesion glycoprotein family and its activation allows hyaluronan to modulate cell function directly in a molecular weight dependent fashion21: for example, in vitro studies have shown that high molecular weight hyaluronan preparations are more effective inhibitors of migration and chemotaxis of inflammatory cells.22 Although Ghosh et al 15 23 suggest that higher molecular weight hyaluronan preparations may result in more rapid histological changes in a sheep OA model. The hypothesis that the molecular weight of hyaluronan is a key factor in its clinical efficacy has recently been brought into question by Aviad and Houpt who maintain that the concentration is more important than its molecular weight.24 However, there have been no double blind studies comparing different preparations of hyaluronan that indicate a clinical difference between them.

    The majority of clinical trials over the past 20 years investigating the effectiveness of different preparations of hyaluronan in human knee OA have been performed using Hyalgan and Synvisc. Critical comparisons between these studies have been made difficult because of the use of different hyaluronan preparations, different treatment schedules, number of injections, small numbers of patients, lack of placebo group, etc.

    Another significant difficulty in comparing the various trials using intraarticular hyaluronan in OA has been the well recognised and large placebo response after aspiration of the knee25and also ensuring adequate blinding during the study. The high viscosity of the hyaluronan solutions used tending to unblind the person performing the injection who is usually the person also assessing efficacy and therefore introducing significant bias.

    Despite these difficulties in comparing the results of these studies a significant number do report reduction in pain and improved function (not always significantly) that may persist for up to a year.26 27

    Carraba et al 28 compared one, two or five hyaluronan injections to placebo and arthrocentesis in 100 patients (20 per group) in a double blind, five arm, parallel group “dose effect” trial, with follow up at two months. Assessment criteria were global spontaneous pain on a visual analogue scale, pain on active movement, Lesquesne index for knee OA, and the volume of joint effusions. A superiority of hyaluronan (three or five injections) compared with one single injection, placebo or arthrocentesis on pain and Lesquesne index was found, starting at day 28 and lasting until the end of the study. Addition of dexamethasone to the first of the five hyaluronan injections conferred a slight initial superiority, followed by a significant and lasting superiority of the hyaluronan from day 21 or 28 onwards.

    There have been a considerable number of studies that have failed to show a significant benefit with hyaluronan in knee OA. The first of these,29 recruited 52 patients with constant symptoms—that is, pain—from 150 patients with previously documented abnormalities on knee arthroscopy. Initial physical examination revealed patellar pain in only 29 of the 52 on palpation. Efficacy was assessed by a different person to the one performing the injections. No difference was found in the effect of hyaluronan versus that of placebo when comparing the various efficacy parameters. It was felt that the placebo effect of aspiration and wash out could not be excluded. These findings led to a follow up study30 in which 240 patients with symptomatic, radiological knee OA were randomly assigned to five weekly injection of hyaluronan or vehicle. At 20 weeks (the end of the study) both treatment groups had improved compared with baseline, with no difference between unstratified groups treated with hyaluronan or placebo. Analysis of stratified groups seemed to suggest that a subgroup of older patients with a high Lesquesnes index seemed to benefit more, other studies have failed to confirm this finding.

    Henderson et al,31 in the first double blind placebo controlled single centre study, in 91 patients with moderate to severe OA over one year also failed to show a difference between both arms. Both arms improved but the only difference was a slower return to pre-treatment levels of non-steroidal anti-inflammatory use after hyalgan 20 mg/once weekly for five weeks.

    A meta-analysis of Hyalgan trials was investigated (personal communication) with emphasis on study design, homogeneity, availability of information regarding individual data, and on a similar outcome measure across studies. Of 43 studies (6000 patients) selecting only the ones fulfilling the following criteria: knee joint, randomised, parallel group, placebo controlled with single or double level of blindness were used in the meta-analysis. Only eight (seven European and one US involving 971 patients) studies fulfilled the above criteria.

    Overall effect of Hyalgan compared with placeblo was found to be highly statistically significant (p<0.001) at the end of the treatment cycles (20 mg/2 ml/Hyalgan three to five injections over consecutive weeks) and at month six. The local and systemic rate of adverse events was found to be comparable between Hyalgan and placebo.

    Despite the above findings, improving the rheological properties of SF, by injecting hyaluronan intra-articularly, clearly cannot account for the reported long term beneficial effects of hyaluronan, particularly when one considers that hyaluronan is rapidly eliminated from the joint. In sheep studies the mean metabolic half life of hyaluronan is about 20 hours in healthy knee joints and about 12 hours in inflamed joints.32

    Limited studies in humans have been performed to determine the possible disease modifying effects of hyaluronan, and perhaps explain its possible therapeutic benefits. It has been shown that intra-articular hyaluronan reduces levels of prostaglandin E2 and cyclic AMP in SF of OA patients.33 Creamer et al 34 in a small number of patients found no difference in clinical variables, but reported that SF keratan sulphate a presumed marker of cartilage breakdown was significantly lowered in the treatment group, perhaps indicating a disease modifying effect. To date only one study has examined the possible disease modifying potential of hyaluronan in humans. Listrat et al 35 reported that patients with knee OA who received a series of three intra-articular injections of hyaluronan at three month intervals showed less progression of the disease one year later, as judged by arthroscopy, than controls who received convential treatment, but no hyaluronan injections. The hyaluronan group also scored higher for quality of life and reduced non-steroidal anti-inflammatory drug use during the study period. However, the results have to be viewed with caution despite the idea that hyaluronan may reduce progression of cartilage damage in OA, the patient numbers were small and all had arthroscopic lavage, which may have been responsible for the perceived benefit. Also, the use of arthroscopic readings as an outcome measure in studies of OA progression is uncommon, because they do not provide a measurement of the thickness of intact cartilage, and only estimate the size of large chondral defects. A recent study examined the effect of intra-articular hyaluronan injection on the cartilage changes that occur during the early stages of OA in the canine anterior cruciate ligament transection model.36 They found that the treatment had no effect on morphological changes in the OA joint. However, seven weeks after the last injection of hyaluronan, reduced levels of proteoglycans were seen in the cartilage, perhaps indicating more rapid degradation. However, other investigators12 have reported results that suggest a disease modifying effect of hyaluronan in a similar animal model.

    To date no double blind placebo controlled prospective studies on the use of hyaluronan on joints other than the knee have been reported. Two single, blind studies using intra-articular hyaluronan in the shoulder and hip, have reported similar symptomatic levels and duration of improvement as reported in the knee.37 38

    Five published trials have compared hyaluronan to corticosteroids39-43 and all have been conducted with Hyalgan. Three were open,41-43 one single blind,34 and one double blind.40 Joneset al 40 looked at 63 patients with inflammatory OA—that is, with knee effusions—and reported that intra-articular corticosteroid had a slightly better effect on pain, function, and analgesic consumption than hyaluronan at day seven. Both treatments demonstrated equivalent effects on all parameters up to week five. Hyaluronan was then significantly better on all parameters from week five up to the end of the six month follow up period. Leardini39 reported no significant differences after 12 months between hyaluronan and corticosteroid. The other three studies36-38 were only of two month duration showing no conclusive differences between treatments. Considering that intra-articular corticosteroid injection for OA has an established therapeutic role, it is difficult to make comparisons between the outcome of studies of intra-articular glucocorticoids and that of studies of hyaluronan. Treatment with hyaluronan usually involves a series of injections (depending on which hyaluronan preparation is used) while glucocorticoids are usually given on a single occasion. They certainly exert their effects via different mechanisms increasing the necessity of a placebo arm in these studies.

    Most adverse events seem to be minor and transient local side effects. Painful post-injection reaction occurring in 1–2% of the patients and resolving spontaneously within one to three days. Henderson et al 31 reported patient withdrawal from their study because of increased, albeit transient, pain and swelling at the injection site, or both, occurring more than twice as often with hyalgan than with saline (47% of the 45 patients compared with 22% of the placebo group). It was speculated that Hyalgan or one of its metabolites, may have been acting as either a primary irritant or as an inflammatory mediator in some patients. Other investigators report the incidence of local reactions ranging from 0–18%.44 It is probable that some of these adverse events may be technique related rather than because of the treatment itself—that is, by incorrect placement of the injection.45 No systemic adverse reactions or other significant changes in standard laboratory parameters have been observed after intra-articular hyaluronan in clinical trials of up to one year duration.44 Apparently to date only one case of septic arthritis has been reported in any trial with hyaluronan.46 Despite this, sepsis still remains a possible complication that may occur with any intra-articular treatment, but the risks must surely be increased by frequent puncture of a single or multiple joints.

    To date the evidence that intra-articular hyaluronan is a useful therapeutic measure is still lacking but seems to suggest that it can provide long term symptomatic benefit in some patients. At the present the biochemical properties of hyaluronan that provide the symptomatic and improved function in some patients is not known. Pharmacoeconomic studies are urgently required given the significant placebo effects. Hyaluronan treatment, to date, is limited probably to the knee and involves a course of treatment of a high molecular weight preparation rather than a single injection (and may suit a certain type of clinical practice!).

    The results of further well controlled, large, long term follow up studies are required to confirm the efficacy of hyaluronan and, if so, then to also define the optimal treatment regimens, its long term activity, especially compared with corticosteroids, its use in other joints, and whether it has disease modifying effects.

    In conclusion, the overall effectiveness of intra-articular hyaluronan in OA to date is not yet supported by unequivocal trial based evidence, but probably could be considered as a therapeutic alternative in patients who have failed non-pharmacological and analgesic treatment, particularly if non-steroidal anti-inflammatory drug treatment is contraindicated or surgery is not an option.


    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.