Viscosupplementation in the Therapy for Osteoarthritic Knee

: Osteoarthritis (OA) is a degenerative disease that is spreading worldwide due to an aging population. This is not simply a disease caused by worn out joints, but a complex disease accompanied by various mechanisms such as inﬂammatory reactions. Among various joints, knee joints show degenerative changes earlier than other joints because they carry most of the weight load, causing social-economic problems. In the case of OA of the knee that does not respond to relatively simple conservative treatments such as physical therapy or medication, intra-articular injection is preferred. However, intra-articular injection drugs have limited effectiveness and uncertainties. There are several intra-articular viscous supplement drugs such as hyaluronic acid. Tissue regeneration active materials such as polydeoxyribonucleotide and polynucleotide are also newly used. The objective of this paper was to compare effects of intra-articular supplementation drugs used for degenerative arthritis of the knee. label study, and one retrospective observational study. Most of the papers showed similar or statistically signiﬁcant improvement in pain control or functional improvement of knee arthritis compared to IA HA. Papers related to PDRN compared the IA HA alone group with the PDRN and HA combination group, and report that pain and function improved in the combined administration group.


Introduction
Osteoarthritis (OA) is a disease in which inflammation of the cartilage and synovial membrane causes exudate in the joint cavity, stiffness, swelling, and joint pain [1]. According to data analysis in the United States, 54.4 million adults were diagnosed with OA between 2013 and 2015. If this trend continues, it is estimated that 78.4 million adults will be diagnosed with OA by 2040 [2]. The social cost of OA is at a level that cannot be ignored.
OA is a disease that cannot be explained by simple degenerative changes. It is a disease caused by several complex factors, such as genetic changes and lack of sex hormones [3]. Recent studies have focused on molecular markers caused by chondrocyte aging [4]. Although various treatment methods for OA have been suggested, there is a limit to clearly selecting one treatment method due to its complex pathophysiology. Non-pharmacologic modalities such as physical exercise and rehabilitation have been recommended for the treatment of knee OA. Oral drugs are also used when effects of non-pharmacologic modalities are limited. Various drugs ranging from non-steroidal anti-inflammatory drugs (NSAIDs) to opioids are used for OA treatment [5].
Since these oral medications can cause systemic side effects, intra-articular injection has been continuously increased to minimize systemic adverse effects and enhance local effects. Injecting drugs into the joint cavity is increasingly used recently. These drugs include steroids, hyaluronic acid (HA), and cross-linked hyaluronic acid (CL HA), polydeoxyribonucleotide (PDRN), and polynucleotide (PN). This review article aims to summarize effects and evidence of intra-articular injections, especially viscosupplements, in patients with knee OA. Steroids, analgesics, and platelet rich plasma (PRP) were excluded.
We searched the databases PubMed, PubMed Central, Medline, Google Scholar, and Google using the key words "viscosupplement", "intra-articular supplement", "hyaluronic In the 1980s, a substance called polyvinylpyrrolidone (PVP) hyaluronate was developed in Russia that could replace synovial fluid. This was administered to OA patients. It was reported in 1984 that PVP hyaluronate injected into the joint space had anti-inflammatory, prolonging, and anti-commissural effects as well as lubricating effects without major side effects. In 1997, the US Food and Drug Administration (FDA) approved viscosupplementation as a conservative technique for the management of OA [6]. The structure of HA is shown in Figure 1 [7]. Since then, until the early 2000s, articles have been published, showing that hyaluronic acid (HA) is more effective in OA compared to the control group (saline group) or that it is non-inferior to previously used corticosteroid (CS) intra-articular (IA) injection. After 2010, papers comparing new HA with existing drugs have been published. Since late 2010, articles have been published, claiming that new substances such as PRP are non-inferior to HA. Most articles have proved that HA is effective for OA, although some papers have reported that there is no difference in efficacy between HA and the control group. Research papers comparing the efficacy of HA with other drugs or treatments from 1984 until recently are summarized in Table 1.   Table 1 summarizes 25 RCTs and 1 clinical study that showed meaningful results among a total of 250 IA HA papers searched. In 3 studies, it was revealed that there was no difference in efficacy compared with the saline control group, and 23 papers reported that it was superior to the saline control group or there was no difference in efficacy compared to the IA steroid group or the NSAID oral administration group.
Since the 1990s, many review articles on effects of HA have been published. Some review articles have reported that HA is ineffective in improving knee pain and function in patients with OA of the knee. However, there are more papers reporting that it is effective in improving knee pain and function with OA of the knee.
When looking at meta-analysis of the efficacy of some control groups (saline or arthrocentesis), the control group is also reported to have efficacy. Thus, it is necessary to consider the placebo effect [34]. As mentioned above, studies on knee OA, knee intraarticular supplement, hyaluronic acid review, or systemic review were searched in several search databases such as PubMed, PubMed Central, and Medline. Among 147 articles searched for, articles on effects of other drugs such as PRP and ozone were excluded under the premise that HA was effective. Furthermore, papers on OA in areas other than the knee were excluded. Among papers published from 1993 to 2021, 12 papers reported a negative effect of HA on knee arthritis, 66 papers reported a positive effect of HA, and 12 papers reported some positive or partially negative effects of HA. Major review articles about effects of HA on knee OA are summarized in Table 2. Table 2 summarizes wellorganized 12 meta-analysis and 11 systematic reviews among the searched review articles for a total of 98 IA HA. Four papers revealed that IA HA had no statistical significance in the improvement of knee arthritis compared to the control group. In 19 papers, it was reported that it was effective in improving knee arthritis compared to the control group, and there was no difference in efficacy compared to the IA steroid or NSAID oral administration group. 2015 [40] Meta-analysis pro IA HA vs. NSAIDs, IA CSs, IA PRP, or IA placebo; The current highest level of evidence suggests that IA HA is a viable option for knee OA.
2015 [41] Meta-analysis pro IA HA vs. IA placebo; IA HA is safe and efficacious through 26 weeks in patients with symptomatic knee OA.
2017 [42] Systemic review pro statistically significant improvements in pain, function and stiffness up to 26 weeks were found with IA HA therapy compared with IA placebo or controls 2018 [43] Meta-analysis pro IA-HA vs. IA CS Both HA and CS injections were effective therapies for patients with knee OA.
2019 [44] Meta-analysis pro IA HA provides a moderate symptomatic benefit to knee OA patients and without major safety concerns 2019 [45] Systemic review pro OARSI guideline (2019) In the absence of comorbidity, the recommendation level is 2, and in the presence of cardiovascular and gastrointestinal comorbidity, the recommendation level is 1B.

[46]
Meta-analysis pro Efficacy of IA HA: early-moderate OA vs. end-stage OA Regardless of OA stage, IA HA provided significant pain relief compared to the saline group, but the level of pain relief was lower in end-stage OA patients.
2019 [47] Systemic review pro ESCEO 2019 Weak recommendation to the use of IAHA in patients who have contraindications to NSAIDs, or if the patient is still symptomatic despite the use of NSAIDs.
2020 [48] Meta-analysis pro Six months of treatment IA HA and the combination of IA HA and triamcinolone improve pain and/or physical function in patients suffering from knee OA.

Guidelines for the Use of HA
Osteoarthritis Research Society International (OARSI) revised in 2019 reported that the use of IA HAs needed caution in patients with OA of the knee at the recommended level. In the absence of comorbidity, the recommendation level was level 2. In the presence of cardiovascular and gastrointestinal comorbidity, the recommendation level was level 1B. Level 1B is a level approved by more than 75% of experts and conditionally recommended by more than 50% of experts. Level 2 is a level approved by 60 to 74% of experts. Although this guideline mentioned that IA HA could bring pain relief for more than 12 weeks after treatment, it recommended only when all core treatment failed [48]. The European Society for Clinical and Economic Aspects of Osteoporosis, Osteoarthritis, and Musculoskeletal Diseases (ESCEO) also published a revised guideline in 2019 and recommended that IA HA could be used for NSAIDs contraindication or for those who had failure with the use of NSAIDs (weak recommendation) [51]. In addition, the American College of Rheumatology (ACR) recommended the use of IA HA when CS injections failed to reduce symptoms such as joint pain [57].
On the other hand, some guidelines do not recommend the use of IA HA in patients with OA. First, the American Academy of Orthopedic Surgeons (AAOS) published the Clinical Practice Guideline on Treatment of Osteoarthritis of the Knee (2nd edition) in 2013 strongly discouraged the use of IA HA [58]. In the New England Journal of Medicine (NEJM) published in 2021, there is insufficient evidence to support a meaningful effect of IA HA with many side effects such as injection site reaction and joint swelling [59]. However, the evidence was based on one meta-analysis paper published by Rutjes et al. [38].

Hyaluronic Acid Physiology
The clinical benefit of IA HA for knee OA is from two perspectives: mechanical viscous replenishment in the joint cavity and re-establishment of intra-articular HA homeostasis by inducing the production of endogenous HA [60,61]. As mentioned above, hyaluronic acid is a substance present in cartilage and synovial fluid with molecular weights ranging from 6500 to 10,900 kDa [62]. During arthritis, endogenous HA present in cartilage and synovial fluid is broken down into substances of high molecular weight and low molecular weight (depolymerization). Consequently, viscoelastic properties of the synovial fluid are reduced. Therefore, when HA is injected from the outside, it acts as a lubricant instead of an endogenous HA with a lower molecular weight. In addition, it induces the relief of pain caused by degenerative arthritis through various mechanisms [63]. There is also a report showing that exogenous HA can affect joints by slowing the degradation rate of type II collagen in the joint cavity [64]. After administering HA and methylprednisolone acetate (MPA) three times, the cartilage was biopsied 6 months later and examined for articular surface morphology, territorial matrix, chondrocyte number, and ultrastructure. It has been reported that the HA group has a higher chondrocyte density than the MPA group with improved appearance of the regional matrix, reconstitution of the superficial layer, and catabolism of chondrocytes [65].
Mechanistically, HA is known to attenuate phagocytosis and decrease cyclic adenosine monophosphate, fibronectin, and prostaglandin associated with inflammatory responses [66]. GoldBerg et al. have reported that HA has disease-modifying activity in OA patients through these actions [67]. There is also a paper showing that IA administration in patients with knee OA can significantly reduce hydrogen peroxide and peroxide in synovial fluid and protect cartilage by inhibiting apoptosis caused by hydrogen peroxide in chondrocytes [68].

Hyaluronic Acid Contraindication and Adverse Effects
There is no absolute contraindication of intra-articular injection of HA other than acute inflammation in the joint cavity, although drug effect may be reduced in the following cases. It is prohibited for use in diseases such as extensive bone edema, bone fissure or stress necrosis on MRI, acute diseases such as gout [69], and scleroderma [70]. In addition, in unstable meniscal tear, ligament laxity, and limb malalignment, not only viscosupplement, but also treatment of the underlying disease should be performed. Radiologic chondrocalcinosis is not a contraindication to HA [69]. For severe OA, HA is not contraindication, but it is known to be ineffective [50]. Eymard et al. have reported that IA HA should be considered only in patients with low BMI and low radiologic severity because high BMI and radiologic severity significantly increase the risk of IA HA failure [71].
Many papers have mentioned adverse effects of HA. Some papers have reported serious complications such as septic arthritis after the use of HA [72]. However, most research articles report only minor side effects associated with the use of HA. The degree of side effects is not particularly high compared to CS [73].

Comparison of HA Products/HA Products Mixed with Other Substances
HA is produced as various drugs due to differences in source (animals using modified organisms, bacterial fermentation), average molecular weight (MW) (low MW HA~high MW HA), MW distribution, molecular structure, crosslinking method, concentration, and complex formulations such as mannitol and CS. Initially, rooster crests were used as a source of HA. Recently, modified bacteria have been used as the source [24]. Several research papers have been published on whether there is a difference in the effect depending on the characteristics of HA. HA can also be classified as follows according to its MW: (1) low MW (500-730 kDa); (2) intermediate MW (800-2000 kDa); and (3) high MW (2000-6000 kDa), including cross-linked formulations of HA. One paper has compared Sinovial ® , an intermediate MW HA, and Synvisc ® , a high MW HA, and reported that they show similar efficacy and safety despite their difference in MW [74]. One paper comparing a cross-linked HA, Hylan, a medium MW HA (avian origin), and a low MW HA (bacterial origin) has reported that effects of the three groups are similar except that Hylan has a slightly higher adverse effect [75]. Altman et al. have compared products with average MW above 3000 kDa and below and found that HA with high molecular weight can provide more favorable efficacy and a lower rate of discontinuation due to treatment-related side effects [76]. There is no significant difference in the effect between HA alone and HAnOX-M ® , a combination of mannitol and HA. Both agents have good effects [24,77,78]. Oxygen free radical neutralization by mannitol can retard degradation of both linear and cross-linked HA in several in vitro models of oxidative stress [79]. Several studies have published the efficacy of CS and HA complex formulations. Zhang et al. have observed radiological and pathological differences by administering control, HA, dexamethasone, and HA to an OA rat model. The knee joint space gap was significantly wider in the HA group. Less osteophytosis was observed in the dexamethasone group. In addition, it was observed that cartilage lesion formation was attenuated in the group using HA. Proteoglycan staining was stronger in the group using dexamethasone [80]. Clinically, in a paper comparing HA, the same amount of HA + triamcinolone (TA) 18 mg (Cingal ® ), and a control group, the drug group showed a significant effect on pain and function up to 26 weeks compared to the control group. In addition, the Cingal ® group was significantly superior to the HA alone group at weeks 1 and 3, although they showed a similar pattern from weeks 6 to 26 [81].

HA Administration Method (Frequency, Dose)
Initial HA is mostly administered based on five times (once per week). However, as more studies comparing effects according to the HA usage method are published, there are now more cases where three times (once per week) are the default. According to Stitik et al., in 24 studies of 2168 people, when hyalgan was injected once a week and when a 3-week course and a 5-week course were compared, there was no statistical difference in pain reduction between the two [82]. Another paper reported that a single injection of HA did not have a statistically significant effect compared to the saline group and that there was no significant difference in the case of 5 or more injections compared to the group injected less than 5 times. Therefore, they reported that 2 to 4 administrations of HA were the most effective [83].
Since 2004, high-dose single injection therapy has been proposed as an alternative to dispensing equal amounts of HA. It is economical by reducing the number of hospital visits and procedures with a single injection. It has the advantage of reducing risks such as side effects of the procedure. Suppan et al. have reported that when 5 mL of 1% sodium hyaluronate is administered in a single dose, there is no difference in terms of efficacy or side effects between the two groups (single dose vs. administration of 2.5 mL of the same product three times) [84]. As mentioned above, although there is still controversy about the use of HA for OA of the knee, since there are no major side effects, it is recommended to consider the use of IA HA in patients who are ineffective in palliative treatment or find it difficult to use oral drugs.

Cross-Linked Hyaluronic Acid
Cross-linked hyaluronic acid (CL HA) is a high MW HA made by chemically crosslinking natural HA. Not all high MW HAs are cross-linked, but most are cross-linked HAs. The structure of CL HA is as follows (Figure 2) [7]. Many papers have reported that high MW CL HA is more effective than low MW HA. It has been reported that one administration of high MW HA has a similar effect to three administrations of low MW [85,86]. However, due to the specificity of crosslinking, pathological development of granulomatous inflammation, a pseudoseptic response, is also known [75]. Although the number of cases reported for adverse events is higher with CL HA, such adverse events are not severe enough to require medical intervention [86]. Many studies have reported that CL HA is good for pain and knee function compared to the control group (saline). There are also papers on the effect and frequency of administration with general HA. Papers on how CL HA differs from regular HA are listed in Table 3. Table 3 summarizes 21 well-designed RCT papers among a total of 25 papers searched for IA CL HA. Seven papers have shown that it is effective for pain caused by OA of the knee compared to the saline group. There were two papers each comparing IA Steroid and oral administration of NSAIDs, and there was no significant difference in efficacy or better results. Another paper compared non-crosslinked HA and reported that a single administration of IA CL HA produced similar results to multiple administrations of IA HA.

Cross Linked Hyaluronic Acid Physiology
The background for the development of CL HA is that the half-life of HA injected to compensate for the decreased synovial HA concentration and chain length in OA patients is too short. It has been reported that unmodified low MW HA has a half-life of about 10 h, whereas CL HA can persist for an average of 8.8 ± 0.9 days [107]. The durability of the injection site is extended because the decomposition rate is slow while the elasticity and viscosity are increased [108].

Cross Linked HA Administration Method (Frequency, Dose)
In the initial paper announcing the efficacy of CL HA, CL HA was administered once a week for 3 weeks, just like natural HA [74,96,109]. However, in the case of CL HA, most papers have reported a similar effect to natural HA administered multiple times with one use [110,111]. In a paper, Hylan G-F 20, a CL HA, was administered at various doses and frequencies. It was found that a single administration of 6 mL of the drug was as effective as administration of 2 mL three times [112]. Although CL HA has been used clinically for a shorter period than HA, the basic structure of the two drugs is the same, and research results are continuously reported that even one administration of CL HA shows similar effects to multiple administrations of HA. Therefore, it is recommended to consider the use of CL HA as well as HA in patients with OA of the knee.

Polynucleotide/Polydeoxyribonucleotide
Polynucleotide (PN) and polydeoxyribonucleotide (PDRN) are DNA fragments extracted from testis and semen of salmon or trout, respectively. As they are DNA polymers of testis and semen, they are substances widely used in regenerative medicine. PN and PDRN have been used for skin regeneration since the 1990s. They have been used for the treatment of osteoarthritis since the 2010s. Structure of PN and PDRN are shown as follows ( Figure 3). PDRN is known to exert anti-inflammatory action without metabolic side effects, unlike NSAIDs or steroids. Because of this, it has been used to treat plantar fasciitis and supraspinatus tendinopathy patients in the pain area. It has been reported to be effective [113,114]. It is also effective for rheumatoid arthritis in animal models [115]. Yoon et al. have reported that, in patients with knee OA, compared to HA alone injection, combined administration of HA with PDRN has a statistically significant effect in VAS and WOMAC except for stiffness [116].
Like PDRN, PN is known to induce anti-inflammatory responses when administered in vivo [117]. Unlike PDRN, it is a substance with a high molecular weight that can bind to a large amount of water. Thus, it is a substance that can increase viscoelasticity. Several studies have demonstrated that these drugs are effective for OA of the knee. Related research articles are summarized in Table 4. Table 4 summarizes 6 papers on PN and 1 paper on PDRN in patients with osteoarthritis of the knee. As for the papers on PN, there were four RCTs, one open label study, and one retrospective observational study. Most of the papers showed similar or statistically significant improvement in pain control or functional improvement of knee arthritis compared to IA HA. Papers related to PDRN compared the IA HA alone group with the PDRN and HA combination group, and report that pain and function improved in the combined administration group.

PN/PDRN Physiology
The intra-articular use of PN serves to provide nutrients to restore the homeostasis of cartilage and environment within the joint. In addition, PN is a polymer material that combines with a large amount of water to form a three-dimensional gel. Because of these characteristics, when it is administered intra-articularly, pain caused by OA can be significantly reduced [122]. As is known, PN is physiologically present in the extracellular environment. It consists of simple nucleotides, nucleosides, and nitrogen bases that constitute the basic matrix of cells [126]. It also has the advantage of supporting the physiological repair process of cartilage by supplying these substrates to chondrocytes [122]. In addition, PN is a high-purity DNA polymer present in the human body. Thus, it has few side effects such as foreign body reaction [118]. It has been reported that PN is not toxic in chronic or acute toxicity studies [119]. As mentioned earlier, PDRN has anti-inflammatory properties. However, there are few clinical studies on it. Bitto et al. have reported that PDRN can affect levels of cytokines, particularly high mobility group box-1 (HMGB-1), tumor necrosis factor-α (TNF-α), interleukin (IL)-6, and IL-10 in human chondrocytes. In particular, upregulation of IL-10 suggests an anti-inflammatory action [115].
According to Baek et al., when a pro-inflammatory mediator was put into human chondrosarcoma cells to create a stress situation followed by treatment with PN and PDRN, they had a significant anti-inflammatory effect compared to the control group [117].
Although they are substances originated from regenerative medicine, there are no papers yet on the effect of PDRN or PN to regenerate chondrocytes in OA in vivo. It is known that they have a protective effect on chondrocytes in vitro [127]. However, in an experiment in which PDRN was administered to a collagen-induced arthritis mice model, it was confirmed that the onset of arthritis was significantly attenuated compared to that in the control group [115].

PDRN/PN Administration Method (Frequency, Dose)
It has been reported that PN can be used in a 40 mg/2 mL prefilled syringe and PDRN can be used at 5.625 mg/3 mL (Placentex Inj ® ; Mastelli srl, Sanremo, Italy) [116,122]. The number of uses of PN or PDRN can be once a week to about 3-5 times, like HA [119,124,128]. PDRN/PN has recently been used in knee osteoarthritis compared to HA or CL HA. However, it has been used in the human body for a long time, has no major side effects, and has a large viscous replenishment effect. Therefore, it can be used in patients with knee osteoarthritis that is not suitable or effective for using HA or CL HA.

Conclusions
We summarized viscoelastic supplements used to treat knee OA. HA has been used in clinical practice for decades. However, even authoritative societies for arthritis report that its use is controversial. Although many papers have reported the effectiveness of hyaluronic acid in the management of osteoarthritis of the knee for pain reduction, the quality of the evidence is still insufficient. However, many studies have reported positive results regarding the efficacy of high (molecular weight) MW or (cross-linked hyaluronic acid) CL HA. In the case of PN and PDRN, there are not many published research papers since they are recently used for the treatment of osteoarthritis, although they are already widely used and clinically recognized. Thus, more studies with higher quality of evidence are needed in the future. However, as drugs that have been used for a long time without major side effects, HA and CL HA can be used in patients with osteoarthritis of the knee where there is no improvement in conservative management. The use of PDRN/PN may be a new option in patients who do not improve even after using HA and CL HA.