ABSTRACT

Introduction: Viscosupplementation by intraarticular injection of hyaluronic acid (HA) is awidely used treatment for lower limbosteoarthritis. However, the injected HA israpidly degraded by reactive oxygen species(ROS), limiting its time of intra-articularresidence. Optimizing clinical effectiveness ofviscosupplementation by reducing HAdegradation in situ, and therefore increasingthe time of contact with the diseased tissue, is achallenging research approach. Mannitol, apowerful ROS scavenger, is a good candidate for

this. The aim of this study was to compare in vitrothe resistance to ROS-mediated degradation oftwo marketed viscosupplements (one linear andone cross-linked) to that of two novelviscosupplements combining HA and mannitol.Methods: A HA viscosupplement at aconcentration of 10 gIL (HA 1%), wascompared to a HANOX-M, a novelviscosupplement made of a mixture of HA andmannitol. In a second experiment, Hylan G-F20, a partially cross-linked viscosupplement,was compared to a HANOX-M-XL, a novelcross-linked viscosupplement made of a HAand mannitol (35 g/L). The four HAviscosupplements were subjected to oxidativestress generated by the addition of hydrogenperoxide (H202) and their rheological behavior(elastic moduli [G’], viscous moduli [G”], andcomplex viscosity [j11*~]) were compared beforeand after the oxidative stress exposure.Resuits: The two viscosupplements notcontaining mannitol HA were rapidlydegraded by H202, as demonstrated by thedramatic decrease of IrI*~. On the other hand,the rheological properties of HA containingmannitol were not substantially modified inthe presence of H202.

Rheumatol TherDOl 10. 1007/s40744-014-0001-8

ORTG!NAT~RBSEARdH

Mannitol Preserves the Viscoelastic Propertiesof Hyaluronic Acid in an In Vitro Model of OxidativeStress

Thierry Conrozier Pierre Mathieu Marguerite Rinaudo

Received: April 22, 2014© The Author(s) 2014. This article is published with open access at Springerlink.com

Electronic supplementary material The onlineversion of this article (doi:1O. 1007/s40744-014-0001-8)contains supplementary material, which is available toauthorized users.

T. Conrozier (~)Service de Rhumatologie, Centre Hospitalier deBelfort-Montbéliard, Belfort, Francee-mail: thierry_conrozier@hotmail.fr;tconrozier@chbm.fr

P. MathieuRhumatologue, Chazelles-sur-Lyon, France

M. RinaudoBiomaterials Applications, Grenoble, France

Published online: 25 July 2014L\Adis

Rheumatol Ther

Conclusion: This in vitro study demonstratesthat mixing mannitol with HA protects theviscosupplement from ROS-mediateddegradation and might therefore increase itsintra-articular residence time withoutsubstantially modifying its rheologicalbehavior. This in vitro study has to befollowed by clinical trials designed to assesswhether the addition of mannitol to HA mightimprove the efficiency and/or the duration ofaction of viscosupplementation.

Keywords: Antioxidant; Hyaluronic acid;Hydrogen peroxide; Mannitol; Osteoarthritis;Reactive oxygen species; Rheology; Viscoelasticproperties; Viscosupplementation

INTRODUCTION

The hallmark of osteoarthritis (OA) is anincreased degradation of the articular cartilageextracellular matrix (ECM) molecules, leading toprogressive cartilage destruction, combined withseveral other joint tissue changes including boneremodeling (osteophyte formation, subchondralscierosis, cyst formation), synovium 10w-gradeinflammation, and frequently synovial fluid (SF)effusion [1].

Viscosupplementation by intra-articularinjections of high molecular weight (HMW)hyaluronic acid (HA) is a widely used treatmentfor knee OA, aimed to restore the rheologicalhomeostasis of the SF and to precipitate arestoration of normal HA metabolism [2, 3]. HAbiologic functions depend on interactions withspecific binding proteins called hyaladherinsthat modify the conformation of HA such asthe celi surface hyaluronan receptor (CD44), HAmediated motility (RHAMM), and hyaluronanbinding proteins of the ECM. Indeed, SF is aviscoelastic system consisting in a complex of

HA and proteins constituting a physical 3D-network [4—7]. HA is an anionic polyelectrolyte,constituted of alternating units of Nacetylglucosamine and D-glucuronic acidmolecules. In healthy SF the polysaccharidechain can be up to 12,000 disaccharide unitsand has an average MW of 4—6 million Da. HA isproduced in synovial joints by hyalocyteslocated in the synovial membrane and isreleased into the SF [3]. It plays a major role inlubrication, shock absorption, and viscoelasticbehavior of the SF [3] involving entanglementsin HA and proteins/HA association based mainlyon electrostatic interactions. This viscoelasticbehavior, which is directly related to both MWand concentration of HA [8], allows the HA 3Dnetwork to adapt to the mechanical stressapplied. At low shear, such as occurs during aslow movement, the linear chains of HA alignslowly in the direction of flow and behave like aviscous fluid. When the joint is subjected to fastimpact (i.e., running or jumping) HA moleculesdo not have time enough to realign and exhibitelastic behavior, allowing shock absorption. Thischange in behavior is called viscoelastic behaviorbased on the temporary network formed bychain entanglements (as observed in dynamicexperiments) and usually corresponds to a nonNewtonian behavior in flow experiments.

In OA joints, SF viscoelasticity, andconsequently ability to protect cartilage, aredramatically lowered when compared withhealthy SF as a result of a decrease of both HAMW and concentration [2].

Viscosupplementation effect is not fullyclarified and is probably due to severalmechanisms such as promotion of endogenousHMW HA production [9], interaction with painreceptors, and various anti-inflammatory effects[10—16].

There are now more than 25 commercialviscosupplement formulations available

L\ Adis

Rheumatol Ther

worldwide from different manufacturers. Theseproducts widely vary in their MW,concentration, volume, indication, residencetime into the joint [17—19], and recommendeddosing regimens, ranging from one to fiveinjections at weekly intervals. Despite the longhistory of this therapy and its inclusion inseveral guidelines for the management of OAsubject to certain conditions [20—26], the use ofviscosupplementation in OA remains a topic fordebate regarding treatment efficacy and safety,ideal dosing regimen, cost-efficacy, differencesbetween commercial preparations, andpredictors of clinical response [27—30]. Thesediscrepancies may originate from differences inefficacy between the studied products, varyingin concentration, MW, and molecular crosslinkage [20]. Indeed, the HA injected into thejoint is rapidly degraded, limiting the time ofintra-articular residence from a few days forlinear molecules [17] to a few weeks for thesolutions of cross-linked HA [18, 19]. Amongthe many pathogenic mechanisms contributingto HA degradation, reactive oxygen species(ROS) derivatives play a major role [31, 32].Then OA is a degenerative joint disease ofmultifactorial origin in the pathogenesis ofwhich ROS play a deleterious effect [33]. It hasbeen shown that interleukin-1 f3 (IL-1 f3) activatesthe production by the chondrocytes of largeamounts of ROS which are themselves involvedin the production and/or the activation ofcollagenase matrix metalloproteinase-1 and inthe phenomena of chondrocyte apoptosis [34].In addition to their effect on the degradation ofthe ECM, ROS are directly involved in themechanisms of degradation of HA. Optimizingclinical effectiveness of viscosupplementationby reducing HA degradation in situ, andtherefore increasing the time of contact withthe diseased tissue [35], is a challenging researchapproach. We [36] and others [32] have

previously shown that the addition ofmannitol (a polyol well known for its ROSscavenging properties) with HA, protects thelatter to ROS-mediated degradation andtherefore may increase the injected HAresidence time into the joint.

The aim of the present study was to comparein vitro the resistance to ROS-mediateddegradation of two marketed viscosupplements(one linear and one cross-linked) to that of twonovel viscosupplemerits combining HA andmannitol.

METHODS

This article does not contain any studies withhuman or animal subjects performed by any ofthe authors.

Viscosupplements to be Assessed

In the first experiment, a HA viscosupplementwith an average MW of 800 kDa at aconcentration of 10 gIL [HA 1%] (Go-on®;Rottapharm SpA, Monza, Italy), was comparedto a HANOX-M (HAppyVisc®; LABRHA, Lyon,France), a novel viscosupplement made of amixture of HA (15.5 gIL, average MW 1 MDa)and mannitol (35 gIL). Both were ofbiofermentation origin and were flot cross-linked.

In the second experiment, Hylan G-F 20, apartially cross-linked viscosupplement,obtained from cockscomb extraction, at a HAconcentration of 8 gIL with a stated MW of 6MDa (Synvisc®; Sanofi-Aventis, France) wascompared to a HANOX-M-XL (HAppyCross®;LABRHA), a novel cross-linked viscosupplementalso made of a mixture of a non-animal originHA (16 gIL) and mannitol (35 g!L).

The average MW of each linear HA wasdetermined using steric exclusion

L\ Adis

Rheumatol Ther

chromatography using a chromatograph(Waters® Alliance® GPC/V 2000; WatersCorporation, Milford, MA, USA) equipped withthree detectors in une (26). The injectedconcentration was 2 g/L, with an injectionvolume of 100 iL using two columns in series(OHpak® 805 and 806; Shodex, Munich,Germany). Ail samples were filtered through amembrane with pores of 0.2 jim (CelluloseAcetate Membrane Filter; Sartorius, Surrey, UK)prior to injection in order to retain anyaggregates. The eluent used was 0.1 M ofNaNO3 at an elution temperature of 30 °C anda flow rate of 0.5 mL/min.

Induction of Oxidative Stress

Hydrogen peroxide (H202) application is a wellknown method to degrade polysaccharidesthrough a radical mechanism [37]. The 4 HAviscosuppiements mentioned above weresubjected to oxidative stress generated by theaddition of H202 (30%; Cari Roth GmbH & Co.KG, Lauterbourg, Germany) at a finalconcentration of 2.7% and/or 5.4% (v/v). For thatpurpose, the commercial samples were added ofH202 in a ratio 10:1 and stirred before being placedrapidly on the rheometer (delay time betweenaddition of H202 and first measurement was2min).

Rheological Assessment

The rheoiogical properties were measured using acone-plate rheometer (ARES-G2 Rheometer; TAInstruments, New Castle, DE, USA) at 25 °C on thedifferent solutions as a function of the oxidativestress time at room temperature. The dynamicexperiments were carried out in the region oflinear viscoelasticity, where the G’ elastic and G”viscous moduli are independent of the appliedfrequency. The dynamic moduli expressed in Pa

(G’, the elastic contribution, and G”, the storagecontribution) and the complex viscosity (jr~*j)were taken at the angular frequency (w) of 1 Hz.

RESU LTS

The two viscosupplements flot containingmannitol HA were rapidly degraded byhydrogen peroxide, as demonstrated by thedramatic decrease of the complex viscosity(~1*i). On the other hand, the rheologicalproperties of the HA solutions containing35 g/L of mannitol were not substantiallymodified in the presence of H202 over aperiod of 15 or 30 min. The resistance tooxidative stress was much better for the crosslinked (HANOX-M-XL) than for the non-crosslinked (HANOX-M) HA containing mannitol,the cross-linking being the only differencebetween them (same origin, and same HA andmannitol concentrations). Likewise, at baseline,HANOX-M was much more viscous than HA1%, because of its higher MW (average MW 1.2MDa versus 789 kDa) and concentration(15.5 gIL versus 10 gIL).

Figure 1 shows the rheological behaviors ofthe two linear viscosupplements. It clearlydemonstrates that the presence of mannitol inHANOX-M stabilizes HA against oxidativedegradation. Conversely, it shows that theelastic modulus (G’) of HA 1% solution wasparticulariy decreased. The values of rheologicalparameters (elastic and viscous moduli,complex viscosity) taken after 15 min arerepresented in Fig. 2. This histogram is veryindicative of the benefit of mannitol on HAstability after H202 exposure.

Under the same experimental conditions,the kinetics of degradation of the two crosslinked viscosupplements was compared (seeFigs. 3, 4). The influence of mannitol is again

L\ Adis

Rheumatol Ther

G)

C

- n n n = n n n o n n = n an n o n n = n u n n n n n na n n n o — — —

1.000-O time global (mm)

Fig. 1 Elastic (G’) and storage (G”) moduli at 1 Hz ofhyaluronic acid (HA) solutions as a function of oxidativedegradation time on linear hyaluronans. Filled triangle G’

(Pa) open triangle G” (Pa) on HANOX-M;filled square G’(Pa) open square G” (Pa) on HA 1%

obvious. Nevertheless, the level of the elasticmodulus (G’) of HANOX-M-XL, which is muchhigher than that of Hylan G-F 20, indicates thatits degree of cross-linkage is higher than that ofHylan G-F 20. The better resistance to

degradation of HANOX-M-XL is likely due toboth mannitol and its high level of crosslinkage. Figure 4 represents the rheologicalparameters after 15 min of oxidativedegradation.

200.0-

D-

Q

[200.0

ô ____ ô -- ~ r

2~5 5.0 7.5 id.o 12.5‘- -1.00015.0

0

F5Ha nox- M

rL4~HA 1%

L~Ji -10.8 -271-14.9 -41.6

G -7.4 -25.5

Fig. 2 Decrease (%) of elastic (G’) and viscous (G”)moduli and complex viscosity (Ir~j) after 15 min ofoxidative stress by hydrogen peroxide. Comparison of two

linear non cross-linked viscosupplements: HANOX-M andhyaluronic acid (HA) 1%

L\ Adis

Rheumatol Ther

Fig. 3 F.lastic (G’) and storage (G”) moduli at 1 Hz ofhyaluronic acid (HA) solutions as a function of oxidativedegradation time on crosslinked hyaluronans. Filled triangle

-1G)Ç)G)

4-o-50-60

Hanox-M-XL Hylan G F-20

U I n’~I -3.3 -42.2

5G -3 -48.8

~G” -14.3 -11

Fig. 4 Decrease (%) ofelastic (G’) and viscous (G”) moduli and complex viscosity (~r~j) after 15 min ofoxidative stress byhydrogen peroxide. Comparison of two cross-linkcd viscosupplements: HANOX-M-XL and Hylan G-F 20

DISCUSSION

This in vitro study, using a powerful model ofoxidative stress, shows that the addition of highconcentration of mannitol (35 gIL) to HAprotects HA from degradation. Mannitol(C6H1406) 15 a polyol well known for itsproperties as a ROS scavenger. The beneficial

effect related to its antioxidant power towardsthe rich reactive hydroxyl function has beenshown in various diseases [38—41].

In viscosupplementation, the rationale of itscombination with HA is legitimate: Whenadministered intra-articularly, HAmacromolecules of the viscosupplement,containing many OH groups, react with ROS,

time global (mm)

G’ (Pa) open triangle G’ (Pa) on HANOX-M-XL; filledsquare G’ (Pa) open square G” (Pa) on Hylan G-F 20

L’. Adis

Rheumatol Ther

resulting in the rupture of the macromolecularchains and accelerated degradation by radicalmechanism [31]. This rapid depolymerization ofHA is the main cause for the short intra-articularhalf-llfe of viscosupplements made of non-crosslinked HA, cross-linking being another way toprotect HA from degradation by ROS [18, 19]. Thechemical characterjstjcs of mannitol make it anantioxidant of choice in combinatjon with HA.As with HA, mannitol has a very good safetyprofile, as demonstrated by animal tests showingthat it was non-cytotoxic, non-genotoxic, noncarcinogenic, and non-mutagenic, even at highdoses [42]. In humans, mannitol is widely usedper os and by intravenous injection at very highconcentrations, including hyperosmolarconcentrations [43]. Its resistance to heat alsopermits sterilization by autoclaving, unlike otherantioxidants that are thermolabile (e.g.,polyphenols, vitamin C). Furthermore, mannitoldoes not increase the ionic strength of themedium and thus does flot significantly alterthe rheological behavior of the HA, and ishydrosoluble, unlike antioxidants such asvitamin E or betacarotene [44]. Two open-labelclinical trials assessing viscosupplementscontaining antioxidants (mannitol 5 g/L [45] orits isomer sorbitol 40 gIL [46]) in patientssuffering from knee OA have been recentlypublished. In both, safety was similar to that ofconventional viscosupplements not containingantioxidant, and none of them has revealed anyserious or unexpected adverse effects. In theabsence of a control group, the efficacy cannotbe conclusively assessed; however, the publishedresuits strongly suggested the beneficial effects ofthe treatments on pain and function.

Our data also suggested that HA crosslinking, high MW and high concentrationmay play a role in the improved resistance tooxidative stress and are other effective ways toprotect HA from degradation.

CONCLUSION

In summary, this in vitro study demonstratesthat mixing high concentrations of mannitolwith HA, protects the viscosupplement fromROS-mediated degradation and might thereforeincrease the intra-artjcular residence time of thegel without substantially modifying itsrheological behavior. Thereby, thecombination of HA and mannitol might be asimple mean to improve the efficiency and/orthe duration of action of jointviscosupplementation. These in vitro resultsfully justify the in vivo studies already inprogress with both HANOX-M and HANOX-MXL whose aims are to confirm or refute thisexciting hypothesis and to assess theeffectiveness and safety of this new promisingassociation on large samples of patients.

ACKNOWLEDGMENTS

The study has been sponsored by LABRHA(Laboratoire de Rhumatologie Appliquée, 19place Tolozan, 69001, Lyon, France). The authorsthank the Laboratory Rhéologie et Procédés ofGrenoble University for their technicalcontribution. Ail named authors meet the ICMJEcriteria for authorship for this manuscript, takeresponsibility for the integrity of the work as awhole, and have given final approval for theversion to be published. Ah authors had fullaccess to ail of the data in this study and takecomplete responsibility for the integrity of the dataand accuracy of the data analysis. The study hasbeen sponsored by LABRHA (Laboratoire deRhumatologie Appliquée, 19 place Tolozan,69001, Lyon, France).

Conflict of interest. Thierry Conrozier hasreceived fees of consultant from LABRHA of

L\ Adis

Rheumatol Ther

which he is a regular consultant and member of

the scientific board. Marguerite Rinaudio hasreceived honoraria from LABRHA for

conducting the study. Pierre Mathieu does flot

declare any conflict of interest.

Compliance with ethics guidelines. This

article does not contain any studies with

human or animal subjects performed by any of

the authors.

Open Access. This article is distributedunder the terms of the Creative CommonsAttribution Noncommercial License which

permits any noncommercial use, distribution,

and reproduction in any medium, provided the

original author(s) and the source are credited.

REFERENCES

1. Pelletier JP, Martel-Pelletier J. The pathophysiologyof osteoarthritis and the implication of the use ofhyaluronan and hylan as therapeutic agents inviscosupplementation. J Rheumatol. 1993;20(suppl39): 19—23.

2. Balazs EA, Denhinger JL. Viscosupplementation: anew concept in the treatment of osteoarthritis.J Rheumatol Suppi. 1993;9:3—9.

3. Balazs EA. Viscosupplementation for treatment ofosteoarthritis: from initial discovery to currentstatus and results. Surg Technol Int.2004;12:278—89.

4. Fam H, Bryant JT, Kontopoulou M. Rheologicalproperties of synovial fluids. Biorheology.2007;44:59—74.

5. Mathieu P, Conrozier T, Vignon E, Rozand Y,Rinaudo M. Rheologic behavior of osteoarthriticsynovial fluid after addition of hyaluronic acid: apilot study. Clin Orthop Relat Res. 2009;467:3002—9.

6. Rinaudo M, Rozand Y, Mathieu P, Conrozier T. Roleof different pre-treatments on composition andrheology of synovial fluids. Polymers.2009;1:16—34.

7. Conrozier T, Mathieu P, Vignon E, Piperno M,Rinaudo M. Differences in the osteoarthriticsynovial fluid composition and rheology between

patients with or without tiare: a pilot study. ClinExp Rheumatol. 2012;30:729—34.

8. Muas M, Rinaudo M. Characterization andproperties of hyaluronic acid (hyaluronan). In:Dimitriu S, editor. Polysaccharides: structuraldiversity and functional versatility. New York, NY:Marcel Dekker; 2004. p. 535—49.

9. Bagga H, Burkhardt D, Sambrook P, March L. Longterni effects of intra-articular hyaluronan onsynovial fluid in osteoarthritis of the knee.J Rheumatol. 2006;33:946—50.

10. Greenberg DD, Stoker A, Kane S, Cockrell M, CookJL. Biochemical effects of two different hyaluronicacid products in a co-culture model ofosteoarthritis. Osteoarthr Cartil. 1993;1:97—103.

11. Wang CT, Lin YT, Chiang BL, Lin YH, Hou SM.High molecular weight hyaluronic acid downregulates the gene expression of osteoarthritisassociated cytokines and enzymes in fibroblast-likesynoviocytes from patients with earlyosteoarthritis. Osteoarthr Cartil. 2006;14:1237—47.

12. Li P, Raitcheva D, Hawes M, et al. Hylan G-F 20maintains cartilage integrity and decreasesosteophyte formation in osteoarthritis throughboth anabolic and anti-catabolic mechanisms.Osteoarthr Cartil. 2012;20:1336—46.

13. Li J, Gorski DJ, Anemaet W, et al. Hyaluronaninjection in murine osteoarthritis prevents TGFbeta1-induced synovial neovascularization and tibrosisand maintains articular cartilage integrity by aCD44-dependent mechanism. Arthritis Res Ther.2012; 14:R151.

14. Waddell DD, Kolomytkin 0V, Dunn S, Marino AA.Hyaluronan suppresses IL-lbeta-inducedmetalloproteinase activity from synovial tissue.Clin Orthop Relat Res. 2007;465:241—8.

15. Yasuda T. Hyaluronan inhibits prostaglandin E2production via CD44 in U937 human macrophages.Tohoku J Exp Mcd. 2010;220:229—35.

16. Boettger MX, KUmmel D, Harrison A, Schaible HG.Evaluation of long-term antinociceptive propertiesof stabilized hyaluronic acid preparation (NASHA)in an animal model of repetitive joint pain.Arthritis Res Ther. 2011; 13:R1 10.

17. Lindenhayn K, Heilmann HH, Niederhausen T,Walther HU, Pohlenz K. Elimination of tritiumlabelled hyaluronic acid from normal andosteoarthritic rabbit knee joints. Eur J Clin ChemClin Biochem. 1997;35:355—63.

18. Lindqvist U, Tolmachev V, Kairemo K, Astrôm G,Jonsson E, Lundqvist H. Elimination of stabilised

L\ Adis

Rheumatol Ther

hyaluronan from the knee joint in healthy men.Clin Pharmacokinet. 2002;4:603—13.

19. Larsen NE, Dursema HD, Pollak CT, Skrabut EM.Clearance kinetics of a hylan-basedviscosupplement after intra-articular andintravenous administration in animal models.J Biomed Mater Res Part B. 2012;100B:457-.62.

20. Bellamy N, Campbell J, Robinson V, et al.Viscosupplementation for the treatment ofosteoarthritis of the knee. Cochrane Database SystRev. 2006;2:CD005321.

21. Hochberg MC, Altman RD, April Xl’, et al. AmericanCollege of Rheumatology 2012 recommendationsfor the use of nonpharmacologic andpharmacologic therapies in osteoarthritis of thehand, bip, and knee. Arthritis Care Res.2012;64:465-.74.

22. Zhang W, Doherty M, Arden N, et al. EULAR evidencebased recommendations for the management of hiposteoarthritis: report of a task force of the EULARStanding Committee for International ClinicalStudies Including Therapeutics (ESCISIT). AnnRheum Dis. 2005;64:669—81.

23. Jordan KM, Arden NK, Doherty M, et al. EULARRecommendations 2003: an evidence basedapproach to the management of kneeosteoarthritis: report of a Task Force of theStanding Committee for International ClinicalStudies Including Therapeutic Trials (FSCISIT).Ann Rheum Dis. 2003;62:1145—55.

24. Zhang W, Moskowitz RW, Nuki G, et al. OARSIrecommendations for the management of bip andknee osteoarthritis, part II: OARSI evidence based,expert consensus guidelines. Osteoarthr Cartil.2008;16:137—62.

25. Zhang W, Nuki G, Moskowitz RW, et al. QARSIrecommendations for the management of hip andknee osteoarthritis: part III: changes in evidencefollowing systematic cumulative update of researchpublished through January 2009. Osteoarthr Cartil.2010; 18:476—99.

26. Pendleton A, Arden N, Dougados M, et al. EULARrecommendations for the management of kneeosteoarthritis: report of a task force of thestanding committee for international clinicalstudies including therapeutic trials (ESCISIT). AnnRheum Dis. 2000;59:936—44.

27. Miller LE, Block JE. US-approved intra-articularhyaluronic acid injections are safe and effective inpatients with knee osteoarthritis: systematic reviewand meta-analysis of randomized, saline-controlledtrials. Clin Med Insights Arthritis MusculoskeletDisord. 2013;6:57—63.

28. Bannuru RR, Vaysbrot EE, Sullivan MC, McAlindonTE. Relative efficacy of hyaluronic acid incomparison with NSAIDs for knee osteoarthritis: asystematic review and meta-analysis. SeminArthritis Rheum. 2014;43:593—9.

29. Rutjes AW, Jûni P, da Costa BR, Trelle S, Ntiesch E,Reichenbach S. Viscosupplementation forosteoarthritis of the knee: a systematic review andmeta-analysis. Ann Intern Mcd. 2012;157:180—91.

30. Colen S, van den Bekerom MP, Mulier M,Haverkamp D. Hyaluronic acid in the treatment ofknee osteoarthritis: a systematic review and metaanalysis with emphasis on the efficacy of differentproducts. Bio Drugs. 2012;26:257—68.

31. Stem R, Kogan G, Jedrzejas MJ, Soltés L. The manyways to cleave hyaluronan. Biotechnol Adv.2007;25:537—57.

32. Mendoza G, Alvarez AI, Pulido MM, et al. Inhibitoryeffects of different antioxidants on hyaluronandepolymerization. Carbohydr Res.2007;342:96—102.

33. Grange L, Nguyen MV, Lardy B, et al. NAD(P)Hoxidase activity of Nox4 in chondrocytes is bothinducible and involved in collagenase expression.Antioxid Redox Signal. 2006;8:1485—96.

34. Rousset F, Nguyen MV, Grange L, Mord F, Lardy B.l-Terne oxygenase- 1 regulates matrixmetalloproteinase MMP-1 secretion andchondrocyte ccli death via Nox4 NADPH oxidaseactivity in chondrocytes. PLoS One.2013;20(8):e66478.

35. Chevalier X. Intraarticular treatments forosteoarthritis: new perspectives. Curr DrugTargets. 2010;1 1:546—60.

36. Conrozier T, Lardy B, Rinaudo M. Effect of highconcentrations of mannitol on the stability ofhyaluronan in an oxidative stress model inducedby xanthine/xanthine oxidase. Osteoarthr Cartil.2014;22:5478.

37. Soltés L, Kogan G, Stankovska M, et al. Degradationof high-molar-mass hyaluronan andcharacterization of fragments. Biomacromolecules.2007;8:2697—705.

38. Belda JI, Artola A, Garcfa-Manzanares MD, et al.Hyaluronic acid combined with mannitol toimprove protection against free-radical endothelialdamage: experimental model. J Cataract RefractSurg. 2005;31:1213—8.

39. Magovern GJ Jr, Bolling SF, Casale AS, Bulkley BH,Gardner TJ. The mechanism of mannitol inreducing ischemic in jury: hyperosmolarity or

L\ Adis

Rheumatol Ther

hydroxyl scavenger? Circulation. 1984;70(suppll):91—5.

40. Suzuki J, Imaizumi S, Kayama T, Yoshimoto T.Chemiluminescence in hypoxic brain—the secondreport: cerebral protective effect of mannitol,vitamin E and glucocorticoid. Stroke.1985; 16:695—700.

41. Fu W, Jiao X. The effect of mannitol and anisodimon the prevention of free radical injury to postischemia flaps: an experimental study. Br J PlastSurg. 1995;48:218—21.

42. Mannitol: INCHEM International program forChemical safety (IPCS): http://www.inchem.org/documents/jecfa!jecrnono/v21 jelO.htm (2014).Accessed June 6 2014.

43. VIDAL: http://www.vidal.fr/recherche/index/q:mannitol/sorbitol/classe_therapeutique (2014).Accessed June 6 2014.

44. Sies H. Relationship between free radicals andvitamins: an overview. Int J Vitam Nutr Res Suppl.1989;30:215—23.

45. Borras-Verdera A, Calcedo-Bernal V, OjedaLevenfeld J, ClaveI-Sainz C. Efficacy and safety ofa single intra-articular injection of 2% hyaluronicacid plus mannitol in knee osteoarthritis over a6-month period. Rev Esp Cir Ortop Traumatol.2012;56:274—80.

46. Heisel J, Kipshoven C. Safety and efficacy findingsfrom a non-interventional study of a newhyaluronic acid/sorbitol formulation (GO-ONmatrix) for intra-articular injection to relieve painand disability in osteoarthritis patients. Drug Res.2013; 63: 445—9

L\ Adis