Background
Although osteoarthritis (OA) of the knee is most often a slowly progressive joint disorder, it is one of the leading causes of disability of the adult population [
1]. Knee OA, a disease of the entire joint, is characterized by joint pain, cartilage degeneration, and an increase in disability [
2]. The progressive nature of OA leads to decreased knee function, affecting an individual’s ability to perform daily activities [
3]. Knee OA also negatively impacts socioeconomic factors, as the associated disability often leads to impaired work performance and early retirement [
4].
Since there is no established disease modifying agent for OA, there are many options for the treatment of knee OA. Among the pharmacologic therapies, non-steroidal anti-inflammatory drugs (NSAIDs) and intra-articular (IA) corticosteroid injections are most commonly prescribed [
5]. These options have inherent limitations, as NSAIDs have potentially serious adverse events associated with their use [
6], and IA corticosteroid injections often provide a relatively short period of effective relief [
7]. Although corticosteroid injection generally has a positive safety profile, it has been shown to cause a transient increase in blood glucose, which may be a concern for diabetic patients [
8]. IA injection of hyaluronic acid (HA) is another treatment option for knee OA pain. HA is nearly ubiquitous in the body, and is a molecule found intrinsically within the knee joint where it provides viscoelastic properties to synovial fluid [
9]. As OA progresses, natural HA concentration and the distribution of HA within the joint shifts towards lower ranges of HA molecular weight, leading to a degradation of the mechanical/viscoelastic properties of the endogenous synovial fluid [
2,
10,
11]. Lower ranges of molecular weight distributions have also been shown to be strongly correlated to pain [
11]. IA-HA administration has aimed to restore this decline in HA concentration and the average molecular weight distribution within the OA knee [
9].
IA-HA has been proposed to have many therapeutic mechanisms of action in the OA knee, including shock absorption, joint lubrication, anti-inflammatory effects, chondroprotection, proteoglycan synthesis, and cartilage matrix alterations [
2]. The correlation between these various effects has created a better understanding of how IA-HA treatment could provide therapeutic effects for patients with knee OA [
12]. There is also evidence suggesting distinct mechanism of action differences between HA products of varying molecular weight. That is, higher molecular weight (HMW) HA has been reported to provide greater anti-inflammatory and proteoglycan synthesis effects, as well as joint lubrication and viscoelasticity maintenance [
9,
13]. There also appears to be evidence of variable safety profiles between HA derived through a biological fermentation process (Bio-HA) and avian-derived HA (AD-HA), as AD-HA has the potential for local IA reactions [
14,
15].
We aim to summarize the mechanisms of action for IA-HA treatment of knee OA described in the current literature in order to determine the validity of the above mechanisms of action. We will systematically assess and outline the defined mechanisms in which HA may provide a therapeutic benefit, while analyzing reported distinction between product characteristic-dependent effects of IA-HA treatment.
Discussion
Our review of the existing literature provides a general consensus that IA-HA for knee OA has beneficial effects through several mechanisms of action; however, the predominant mechanism in which therapeutic effect is provided is not clearly understood [
2]. In perspective, it is not clear which of these mechanisms are clinically relevant, as it is appreciated that beneficial mechanisms of action are not necessarily transferrable to benefit in the clinical setting. At this time, it is presumed that the clinical benefit of IA-HA in knee OA is due to several concurrent mechanisms of action, instead of any one single specific mechanism of action.
The majority of exogenous HA remains in the joint for a few days; however, the clinical therapeutic effects of HA treatment may be seen for up to 6 months, or more. This may suggest that IA-HA contains disease modifying properties, and does not act solely by restoring viscoelastic properties to the synovial fluid [
123]. HA injections may stimulate endogenous production of additional HA by human synoviocytes, aiding in the normalization of HA distribution within the synovial fluid [
124]. A large number of reports describe CD44 binding as a primary mode in which HA provides action against knee OA in non-clinical basic science studies. CD44-mediated effects of IA-HA are shown to contribute to the potential chondroprotection, proteoglycan/glycosaminoglycan synthesis, anti-inflammatory, and subchondral mechanisms. This binding is shown to have a variety of effects on numerous signalling pathways, all of which demonstrate some sort of intervention in the progression of OA [
13,
21,
32,
36,
38,
42,
43,
46,
59,
61,
104,
116]. The suppression of IL-1β and IL-6 and subsequent effects of this suppression have been suggested to be a key factor in the therapeutic mechanism provided by HA-CD44 binding [
123]. It is evident that HA-CD44 binding is a major component in the mechanism in which HA provides therapeutic effect; however, there are additional mechanisms that provide alternate pathways for the effectiveness of HA treatment in OA knees, including ICAM binding, mechanical improvements attributed to shock absorption and lubrication, an increase in cartilage/bone interface type II collagen turnover, as well as analgesic effects through interaction with nerve endings and joint nociceptors [
31,
41,
86,
97,
121,
122]. HA binding to the RHAMM receptor promotes wound repair, activates pro-migration and invasion functions, regulates cellular responses to growth factors, and plays a role in fibroblast migration and motility [
125‐
127]. These results of HA-RHAMM binding are potential factors involved in the disease modification effects of HA treatment for OA.
There is evidence which demonstrates that certain intrinsic properties of particular IA-HA products may provide beneficial results in comparison to other IA-HA products. The most recognized of these intrinsic properties is molecular weight. Contrary to a previous basic science review by Ghosh et al., which suggested a potential benefit of LMW HA in providing rheological property restoration over HMW HAs [
128], the evidence within the current review has demonstrated advantageous results for HMW HA treatments. The current review supports the view that HMW HA provides superior chondroprotective, proteoglycan and glycosaminoglycan synthesis, anti-inflammatory, mechanical, and analgesic mechanisms of action [
61,
89,
97,
111,
122]. A study by Huang et al. demonstrated superior anti-inflammatory effects of HMW HA but superior chondroprotective effects of LMW HA; however, these results regarding chondroprotection are unclear due to lack of additional evidence within the knee OA basic science literature [
72]. An increased production of inflammatory cytokines and chemokines, recruitment of inflammatory mediators, and blood vessel formation have been shown to be a response to LMW HA below 500 kDa. While the average MW of available HA products on the market vary greatly, it should be noted that, to our knowledge, all currently available products worldwide have a molecular weight >500 kDa [
129,
130]. An analysis of HA-CD44 interaction demonstrated that HA size has direct impact on the affinity in which HA binds to the CD44 receptor [
131]. These results demonstrate the capacity of HMW HA to treat the progression of knee OA through CD44 binding of HA. These basic science findings are consistent with systematic reviews of clinical trials and comparative studies which have demonstrated that HMW HA provides greater therapeutic benefit than LMW HA in the treatment of knee OA [
6,
132], although the current literature does not provide consensus regarding the clinical efficacy difference between low and high molecular weight HA [
133].
Traditionally, HA products had been derived from avian sources; however, some available products are produced by biological fermentation. This process avoids the presence of avian-derived molecules which are suggested to be a potential cause of adverse local reactions [
134]. There is a lack of thorough reporting regarding the potential of Bio-HA over AD-HA. One study has suggested AD-HA injection sites may be the cause of synovitis in their patient group, yet the exact pathological agent is unknown [
135]. Results from a second study also outline the potential for hylan AD-HA to cause a foreign body giant cell type granulomatous reaction [
136]. Research has demonstrated that the flare-ups associated with hylan injection may be correlated to the accumulation of hylan or its breakdown products, as injection site flare ups typically do not occur upon first injection [
14]. Avian derived proteins have been shown to be the cause of injection site flare up, as antibodies to chicken serum protein were present in patients who demonstrated injection site adverse reaction after being treated with AD-HA [
15]. There is some high-quality clinical evidence that Bio-HA has a significantly smaller incidence of injection site adverse events than AD-HA [
134]; however, this is not thoroughly investigated within the current literature. More comprehensive investigation of the difference in incidence of injection site adverse events between Bio-HA and AD-HA from both a basic science and clinical perspective is needed.
This review has methodological strength in its systematic and thorough search of available basic science evidence within multiple databases. The current report also demonstrates rigor in its presentation of multiple mechanisms in which HA acts, providing evidence on all mechanisms whether comprehensively or infrequently reported within the current literature. Limitations of the current study arise due to the subjective classification of included article mechanism of action key conclusions. Included articles may briefly mention alternate mechanisms of action, but were not classified into the corresponding category because the mentioned alternate mechanism was not a key result or conclusion of the study. Future research should analyze the relationship between the various mechanisms presented in this report, and clarify the way in which these mechanisms overlap and may work together to alleviate symptoms of knee OA. Future research should also aim to recognize differences between mechanisms exhibited by high and lower molecular weight products, as well as analyze the safety profile differences between Bio-HA and AD-HA.
Conclusions
The non-clinical basic science literature provides evidence for numerous mechanisms in which IA-HA may provide clinical benefit in knee OA. Chondroprotection is the most frequently reported mechanism, with HA-CD44 binding being the most frequently reported source of these effects. IA-HA is also reported to provide proteoglycan and glycosaminoglycan synthesis, anti-inflammatory, mechanical, subchondral, and analgesic effects. There is evidence of favorable results for HMW HA treatments in comparison to LMW HA. Bio HA is also demonstrated to provide an advantageous safety profile over AD-HA, as reports demonstrate the association between injection site flare ups and avian-derived proteins. There are a variety of reported mechanisms in which IA-HA is demonstrated to treat knee OA, as well as numerous product characteristics that impact the results of IA-HA treatment. A thorough understanding of the variety of mechanisms in which IA-HA provides beneficial effects within the OA knee, as well as the characteristic-specific effects of various IA-HA products, is required to recognize the applicability and appropriateness of IA-HA treatment for knee OA. Future research should not only focus on the pain relief provided by IA-HA treatment, but the disease modification properties that this treatment modality may possess as well.
Competing interests
Roy D. Altman: Consultant: Cytori, DuPuy (Bioventus Global), Ferring, Flexion, Iroko, McNeil, Novartis, Oletec, Pfizer, QMed, Rotta, Strategic Science and Technologies, Teva.
Speaker: Ferring, Iroko.
Ajay Manjoo: No conflicts of interest
Anke Fierlinger: Paid employee of Ferring Pharmaceuticals Inc.
Faizan Niazi: Paid employee of Ferring Pharmaceuticals Inc.
Mathew Nicholls: Serves on advisory boards for Ferring.
Authors’ contributions
AF and FN participated in the design and development of the study. RDA provided interpretation of the data, and significant writing contributions to the manuscript. All authors provided critical review and revisions to the manuscript. All authors read and approved the final manuscript.