Biomarkers, type II collagen, glucosamine and chondroitin sulfate

JOrthopaedTraumatol(2008)9:81–87DOI10.1007/s10195-008-0007-5

ORIGINALARTICLE

Biomarkers,typeIIcollagen,glucosamineandchondroitinsulfateinosteoarthritisfollow-up:the‘‘Magentaosteoarthritisstudy’’

M.ScarpelliniÆA.LuratiÆG.VignatiÆ

M.G.MarrazzaÆF.TeleseÆK.ReÆA.Bellistri

Received:2March2008/Accepted:7April2008/Publishedonline:28May2008ÓSpringer-Verlag2008

Abstract

BackgroundThepurposeofthepresentstudywastodeterminerelationshipbetweendiseaseactivity,systemicmarkersofcartilagedegradation,urinaryC-terminalcross-linkingtelopeptidesoftypeIIcollagen(uCTX-II),andbonedegradation,urinaryC-terminalcross-linkingtelopeptidesoftypeIcollagen(uCTX-I),structuralpro-gressionofosteoarthritis(OA)andpotentialtherapeuticef?cacyoftypeIIcollagen(COLLII)incombinationwithglucosamineandchondroitinsulfate(GC).

MaterialsandmethodsAnobservationalretrospectivestudy,1-yearfollow-up,on104patientswithOA(nodularosteoarthritisofthehand,erosiveosteoarthritisofthehand,EOA,osteoarthritisofthekneeorhip)whoweretreatedwithGCorglucosamine,chondroitinsulfateandcollagentypeII(GCC).Thefollowinginformationwascollectedatentry:demographics,BMI,characteristicsofOA,patientglobalassessment(VAS),C-terminalcross-linkingtelopeptidesofcollagentypesI(uCTX-I)andII(uCTX-II)andradiographs.After6months:VAS,uCTX-IanduCTX-II.After1year:VAS,uCTX-I,uCTX-IIandradiographs.ResultsAfter6monthsand1yearoftreatmentVAS,uCTX-IanduCTX-IImeanvaluesweresigni?cantlylowerthanthebaseline.57weretreatedwithGCCand47withGC.ThegroupthatreceivedGCCshowedasimilarVASmeanvalueafter6monthsand1yearwhencomparedwith

M.Scarpellini(&)ÁA.LuratiÁM.G.MarrazzaÁF.TeleseÁK.ReÁA.Bellistri

RheumatologyUnit,MagentaHospitalItaly,

ViaalDonatorediSangue50,20013Magenta,Milan,Italye-mail:magda.scarpellini@ao-legnano.it

G.Vignati

EndocrineandMetabolicDiseaseCenter,MagentaHospitalItaly,ViaalDonatorediSangue50,20013Magenta,Milan,Italy

thegrouptreatedwithGC.uCTX-IanduCTX-IImeanlevelwaslowerinthegrouptreatedwithGCC(P\0.05).Radiologicalscore(KellgrenandLawrencesummarizedscoreforhands)after1yearshowedareducedprogressioncomparedtothebaselineinthehandosteoarthritisgroup,especiallyafterGCCtreatment(P\0.05).Finally,uCTX-IhasbettercorrelationwithradiologicalscoreandwithGCintheEOAsubgroup(Pearsonindex:R=0.44).

Conclusions(a)uCTX-IanduCTX-IIprovedtobeusefulbiomarkersinOAmonitoring;(b)uCTX-Iisbettercorre-latedwithhandEOAandcouldrepresentapotentialfurthermarkertoassesstheevolutionofEOAbonedam-age;(c)GCslowdownOAprogression;(d)?nallyCOLLIIcouldrepresentafurtherprotectivefactorinOAcartilage.KeywordsOsteoarthritisÁTypeIIcollagenÁCartilageÁuCTX-IÁuCTX-II

Introduction

Articularcartilageisconstructedwithhyalinecartilagetissue.Itiscomposedofchondrocyteslocatedinlacunaeandintheextracellularmatrix.Thechondralmatrixcon-tainswater,collagen,proteoglycans,non-collagenousmatrixproteinsandlipids.Articularcartilageisdividedintofourzones—super?cial,intermediate,deepandcalci-?ed—accordingtomorphology,theorientationofcollagen?ber,andtheproteoglycancontent.ThedominantcollagenofthistissueistypeIIcollagen(COLLII),togetherwithsmallerquantitiesofothercollagens(IX,XII).Numerousstudieshaveshownthatchondrocytesalsohavetissue-speci?cantigens,whichinducetheintroductionofanti-bodiesinpatientswithcartilagegrafts,aswellasthosewithosteoarthritis(OA)[1].

123

82Moreover,ithasbeendemonstratedthatsomechon-drocytescanhavemigratorycapacity,andthemigratoryonescansynthesizeCOLLIIbutnottypeIcollagen(COLLI)[2];thiscanbeinterpretedasfurtherevidencethatjointdamageinvolvesmostlyCOLLII.

Collagensareabigfamilyofproteins,themainoneformingconnectivetissueinallhigheranimals.Connectivetissuescontainamixtureofcells,proteins,complexpoly-saccharidesandinorganicconstituents.COLLII,likeelastineandproteoglycans,islocatedintheextracellularmatrixandisproducedby?broblasts.

ThefunctionalpropertyofCOLLIIistogivestrengthand?exibilitytotheconnectivetissue,resistingtheten-sionssufferedinthedirectionofits?bers.Atthemomentthereare28differentidenti?edtypesofcollagen.Incon-nectivetissue,nativeCOLLIIisarrangedin?brils.Itsfunctionconsistsingivingstrengthand?exibilitytotheconnectivetissue,resistingthetensionssufferedinthedirectionofits?bers.Also,COLLIIispresentinjoints.Ithasaspecialcon?gurationthatgivesparticularelasticpropertiestoprotein:collagen?bersarelocatedintheextracellularmatrixandhavethecapacitytoincreaseorreducetheirvolumeaccordingtothedegreeofcompres-siontowhichtheyaresubjected.Thereforeitprotectstheorgansandtissuesfromruptureandlossofformorstruc-turewhentheyarestressedbysuddenorviolentmovements.Besidesitsstructuralroleintissuesandorgans,collagenisalsoimportantforthedevelopmentoftissues,oneofitsfunctionsistoin?uencethedifferentia-tionandproliferationofnon-specializedcells.OverthelastyearsnumerousmolecularmarkersofcartilagebreakdownhavebeenusedandevaluatedtopredictthestructuralprogressionofOA.Amongtheusedmarkers,oneofthemostreliableresultedurinaryC-terminalcross-linkingte-lopeptidesoftypeIIcollagen(uCTX-II)[3–11].COLLIisametalloproteasecleavageproductofhumanarticularcartilageandisnormallyusedtotestosteoclasticactivityandtoquantifybonereabsorption[10].Thearticleaimsatprovidingacomparisonbetweentwoassociations(gluco-samine–chondroitinsulfateandglucosamine–chondroitinsulfate–nativeCOLLIIpartiallyhydrolyzed)inacohortofItalianpatientswithOA.Furthermore,theroleofurinaryC-terminalcross-linkingtelopeptidesoftypeIcollagen(uCTX-I)anduCTX-IIasbiomarkershasbeenevaluated.

MaterialsandmethodsPatientsandmethods

Thiswasanobservationalretrospective1-yearfollow-upstudy.ThisstudyhasconsideredpatientswhocametoourhospitalbetweenOctober2006andOctober2007andwho123

JOrthopaedTraumatol(2008)9:81–87

wereaffectedbyosteoarthritisofthehand,hiporknee(ful?llingtheAmericanCollegeofRheumatologyCriteria)[12].

Atbaselinethefollowingcharacteristicswerecollected:age,gender,BMI.Eachpatientwasevaluatedatbaseline,after6monthsandafter1yearwithageneralrheumato-logicassessment,apatientglobalassessment(0–100mmVAS),uCTX-IanduCTX-IImeasurement.

FortheurinaryevaluationofdegradationproductsofC-terminaltelopeptidesofTypeIhumancollagentheUrineCrossLapsÒELISAKithasbeenemployedandforproductsofC-terminaltelopeptidesoftypeIItheUrineCartiLapsÒELISAkit(bothfromNordicBioscienceDiagnostic—DK,ItaliansubsidiaryPantec—Turin)hasbeenemployed.Themethodgivesaquantitativedetermi-nationemployingacompetitiveenzymeimmunosorbentassayonmicro-titerwellsbyusingurinefromsecondmorningvoid,theconcentrationisreferredtothecreatinineexcretionwithresultsasnanogram/mmolofcreatinine.Atbaselineandafter1yeararadiologicalevaluationofhands,hiporkneeasneededwasobserved.AllradiographswereevaluatedbytwoexperiencedrheumatologistsandtheKellgrenandLawrencescorewasused[13].Obviously,patientswhoserequiredparameterswere(partially)absentfromourarchiveswerenotconsidered.SecondaryOA,dueforexampletofracture,in?ammatorydiseases(suchasrheumatoidarthritis)orPagetdisease,wasanexclusioncriterion.Patientswithcurrentcorticosteroidtherapyand/oranti-osteoporoticdrugsand/orrenalorhepaticdys-functionintheyearbeforeonsetofthestudywereexcludedtoo.

Hundredandfourpatientswerethusincludedinthestudy(95females,9males),meanagewas61.4±7.2years.ThirtypatientswereaffectedbyhandEOA,54byhandOAand20distributedbetweenkneeorhipOA.

Forty-sevenweretreatedwithglucosamine1,000mg+chondroitinsulfate1,000mg(GC)and57withglucosamine1,000mg+chondroitinsulfate1,000mg+nativeCOLLIIpartiallyhydrolyzed2mg(GCC).

ThestudywasperformedinaccordancetotheDeclarationofHelsinki.Statisticalanalysis

Continuousvariableswereanalyzedintermsofmean±standarddeviation.

StandardStudent’st-testforpairedsamplesorone-wayANOVAwereperformedforcomparingdata,asneeded.AP-value\0.05hasbeenvaluedassigni?cant.

AllanalyseswerecarriedoutwithSPSSsoftwareforwindowsVer13.0.

JOrthopaedTraumatol(2008)9:81–87Fig.1VASinthewholepopulation(GC+

GCC)

Fig.2UrinaryC-terminalcross-linkingtelopeptidesoftypeIcollageninwholepopulation(GCC+

GC)

Results

Analyzingthewholepopulation,VAS(Fig.1)wasreducedsigni?cantlyafter6months(t1:P=0.014),andwasstilllowafter1year(t2:P=0.004)comparedtothebaseline(t0)inbothgroups(GCandGCC).ThesameistrueforuCTX-1(Fig.2),t1:P=0.001andt2:P=0.002.Therewasnosigni?cantdifferencebetweenGCandGCC.uCTX-IinhandOAandhandEOAismorereducedinthegrouptreatedwithGCCatt2(P=0.026)comparedtotheothergroup(Fig.3).

UrinaryC-terminalcross-linkingtelopeptidesoftypeIcollagenintheEOAsubgroupismoresigni?cantlyreduced(P=0.017)att2withGCCcomparedtoGC(Fig.4).

83

Fig.3UrinaryC-terminalcross-linkingtelopeptidesoftypeIcollageninhandOAandhandEOA,GCCvs.

GC

Fig.4UrinaryC-terminalcross-linkingtelopeptidesoftypeIcollageninEOAgroup,GCCvs.GC

UrinaryC-terminalcross-linkingtelopeptidesoftypeIcollagenhasahigherbaselineintheEOAsubgroup.

UrinaryC-terminalcross-linkingtelopeptidesoftypeIIcollagenissigni?cantlyreducedinthewholepopulationbothatt1(P=0.003)andt2(P=0.002)andwithbothtreatments(Fig.5);patientstreatedwithGCCimprovemorequicklyandsteadilyover1year,whereaspatientstreatedwithGCshowsmallerimprovementswhichtendtoregressatt2.

TomakeamoreaccuratecomparisonoftheuCTX-IIimprovements,wechosethegroupaffectedby

hand

123

84Fig.5UrinaryC-terminalcross-linkingtelopeptidesoftypeIIcollageninwholepopulation(GCC+

GC)

Fig.6UrinaryC-terminalcross-linkingtelopeptidesoftypeIIcollageninhandOAandhandEOA,GCCvs.

GC

arthritis(OA+EOA)becauseitisthemorenumerousandhomogeneousfrommanypointsofview(sex,age,BMI).uCTX-IIissigni?cantlyreducedatt1(P=0.01)already,withbothtreatments.However,att2(P=0.017)patientstreatedwithGCCshowfurtherimprovementswhereaspatientstreatedwithGCtendtoregress(Fig.6).

Fromtheradiologicalpointofview,inthehandarthritisgroup(OA+EOA),patientswithGCCshowadecreaseinthebonedecayrateover1year(P=0.009)(Fig.7),startingfromsimilarradiologicscoreatt0.Figure8islikeFig.7,butlimitedtopatientswithhand-EOA(P=0.018),startingwithsimilarradiologicscoreatt0,aswell.

123

JOrthopaedTraumatol(2008)9:81–87

Fig.7EvolutionofradiologicalscoreinhandOAandhandEOA,GCvs.

GCC

Fig.8EvolutionofradiologicalscoreinhandEOA,GCvs.GCC

InEOApatients,analyzingthepossiblecorrelationbetweenradiologicaldata,uCTX-IanduCTX-II,wefoundasigni?cantcorrelationbetweenradiologicalscore(att2)anduCTX-I(bothatt1andt2),withaPearsonindexR=0.59,P=0.0008.

ThereisnocorrelationbetweenradiologicaldataanduCTX-II.

InhandOApatients,thereisaweakcorrelationbetweenradiologicaldataanduCTX-II,withaPearson

index

JOrthopaedTraumatol(2008)9:81–87Fig.9Correlationbetweenradiologicalscore(abscissa)anduCTX-I(ordinate)inhandEOA,GCvs.GCC

R=0.315,P=0.045.Thisdatumseemstocon?rmthattheEOAgroupisbetterdescribedbyuCTX-IthanuCTX-II.

PearsonindexbetweenradiologicalscoreanduCTX-IintheEOAgrouptreatedwithGCisR=0.44.PearsonindexbetweenradiologicalscoreanduCTX-IintheEOAgrouptreatedwithGCCisR=0.12.ThisdatumisconsistentwiththefactthatGCCslowsdowntherateofincreaseofradiologicalscore,thusuCTX-IisabetterindicatorofradiologicalevolutionintheGCgroupcomparedtotheGCCgroup,wheretheeffectofthedrugweakensthecorrelationbetweenthesetwovariables(Fig.9).

Discussion

OAisthemostcommonandfrequentoftherheumaticdiseases.Researchonpathogenesishassuggestednewideasbothinnewdiagnostictestsandintherapy.

Ourstudy,inaccordancewithmanyAA,hascon?rmedthevalidityofuCTX-IIinthediagnosisandquantitativeanalysisofcartilagebreakdown[10,11]alsoduringfollowup,independentlyofarea,localizationandhistologicalgradingofaffectedjoints[14]andatrophicorhypertrophicpatternsofOA[7].

Inourstudy,alsouCTX-Ihasdemonstrateditselfinteresting.Recently,asamatteroffact,thepresenceofboneresorptionhasbeenrecognizedasariskfactorforOAprogression.Arecentstudy[15]demonstratedthatgeneralboneresorption,indicatedbyserumbiomarkermeasure-ments,isincreasedinpatientswithprogressivekneeOA.Theseresultscon?rmedtheroleofbonechangesinthepathogenesisofOA.Itisgenerallybelievedthat

85

degenerationofcartilageinOAischaracterizedbytwophases:abiosyntheticphaseduringwhichthecellsresidentincartilage,thechondrocytes,attempttorepairthedam-agedextracellularmatrix;andadegradationphase,inwhichtheactivityofenzymesproducedbythechondro-cytesdigeststhematrix,matrixsynthesisisinhibited,andtheconsequenterosionofthecartilageisaccelerated.Thecellularreactionpatternduringtheosteoarthriticdiseaseprocessisat?rstglanceratherheterogeneous.However,thereactionpatternscanbasicallybesummarizedin?vecategories:(a)proliferationandcelldeath(apoptosis),(b)changesinsyntheticactivity,(c)changesindegradation,(d)phenotypicmodulationofthearticularchondrocytes,and(e)formationofosteophyte.Severalfactorssuchasretinoicacid,bromodeoxyuridineandIL-1,induceso-called‘‘dedifferentiation’’ormodulationofthechondro-cytesphenotypetoa?broblast-likephenotype.ThechondrocytesstopexpressingaggrecanandcollagentypeII,thoughtheyarestillveryactivecellsandexpresscollagentypesI,IIIandV[16].SomeAA[17]havereviewedtheevidencethatOAabnormalosteoblastsareresponsible:(a)tomaintaintheabnormalmineralization,(b)toreleasefactorsthatcanmodifybothosteoblastsandchondrocytesfunctionsand(c)todegradearticularcarti-lage.OAosteoblastsreleasemoreIL-6,CXCL12,CXCL13andleptinthatcanalsodirectlymodulateCOLLIsynthesis,promotearticularcartilagedegradationandcontributetoin?ammatorystateobservedinOA.Produc-tionofanabnormalcollagenmatrixandasolublefactor(s)byOAosteoblastsleadstoanabnormalosteoidmatrixnotmineralizingnormally.Thisputativefactor(s)contributestocartilagedegradationbutalsotoabnormalosteoblastscellfunction,measurablealsoasuCTX-I.Marcellietal.[18]underlinetherelationshipbetweenhandOAandbonemineraldensityandZolietal.[19]con?rmitalsoinhandEOA,adestructiveformofprimaryOA.Themostimportantproblemfortherapyisthelargepathogeneticprocessnumber.ObviouslytherapeuticinterventioninordertocontrolbothjointcartilagecollagencontentandbonecontenthavetobeconsideredinOAmanagement.Numerouspharmaceuticalandnutriceuticalagents[20]havebeendevelopedandareusedalsoincombination,toimprovetheef?cacyindelayingtheprogressionofstruc-turalchangesinOAcartilage[21,22].ActuallymanyAAhavedescribedtheef?cacyandpotentialdisease-modify-ingeffectofglucosamine,chondroitinsulfate,avocadosoybeanunsaponi?ables,diacerhein,intraarticularhyalu-ronan,ginger,doxycycline,ascorbicacid,manganese,growthfactors[23].

OurgroupwantedtotestthepotentialcapacityofnativeCOLLII,partiallyhydrolyzed,inassociationwithGC,tolessenthedamageandprogressionofOA,basedontheabove.Collagengivesboneits?exibility,helpingit

to

123