Clinical evaluation of the incidence of prosthetic

Clinical evaluation of the incidence of prosthetic

complications in implant crowns constructed with UCLA castable abutments. A cohort follow-up study

Javier Montero

, Guillermo Manzano

, David Beltra´n

, Christopher D. Lynch

, Marı´a-Jesu´s Sua´rez-Garcı´a

, Raquel Castillo-Oyagu¨e

*

aProsthodontics,DepartmentofSurgery,FacultyofMedicine,UniversityofSalamanca(USAL),C/AlfonsoXelSabio,s/n, CampusdeUnamuno,E-37007Salamanca,Spain

bDepartmentofSurgery,FacultyofMedicine,UniversityofSalamanca(USAL),C/AlfonsoXelSabio,s/n, CampusdeUnamuno,E-37007Salamanca,Spain

cTissueEngineering&ReparativeDentistry,SchoolofDentistry,HeathPark,Cardiff,UK

dProsthodontics,DepartmentofBuccofacialProstheses,FacultyofDentistry,ComplutenseUniversityofMadrid(UCM), Pza.Ramo´nyCajal,s/n,E-28040Madrid,Spain

article info

Articlehistory:

Received25May2012 Receivedinrevisedform 31August2012

Accepted3September2012

Keywords:

Dentalimplants

UCLAcastableabutments Cobalt–chromium Implant-supportedcrowns Screwloosening

Ceramicfracture

abstract

Objectives: To evaluate the incidence of prosthetic complications in implant-retained crownsmadewithUCLAcastableabutmentsandtoidentifypossibleriskfactorswitha viewtoestablishingrecommendationstohelppredictthesuccessofsuchrestorations.

Methods:Acohortfollow-upstudywascarriedoutin71partiallydentatepatientsrehabili- tated with93 implant-retainedsingle crowns.Data regardingsocio-demographic back- ground,anatomicalfeatures,implant-,andprosthesis-relatedvariableswererecorded.The incidencerate(%),relativerisk(RR) andoddsratio(OR)wereappliedforpredictiverisk factors.ANOVAandStudentt-testswereusedtocomparequantitativevariables,thechi- squaretestwasusedtocompareproportionsandalsoalogisticregressionanalysiswas performed.Thestatisticalsignificancewassetata=0.05.

Results: Twoimplants(2.2%)werelostduringthefirstyearoffunction.Theincidenceof prostheticcomplications in theobserved mean period (26.215.4 months)was 11.9%, consistingofscrewloosening(10.8%)andceramicfracture(1.1%).Ahighertendencyfor prostheticcomplicationswasnoticedinposterior mandibularcrownsrestoring saddles longerthan10mmwithmesiodistalcantileverslongerthan6mm,havingnaturalantago- nists,afterlong-termuse(>20months),withinitialtorquevaluessuperiorthan30Ncm.

Conclusions: Screwlooseningisthemostfrequentcomplicationinimplant-retainedcrowns fabricatedwithUCLAabutmentscastincobalt–chromium.Nevertheless,theconnection usuallyremainsstableafterretighteningthescrews.Ahighsurvivalratewasrecorded,and theseprosthesesmaybeasuitabletreatmentoption.

Clinical significance:Based onthe studyfindings,the riskofprostheticcomplications is expected toincreasewhenlong-spanposterior edentulousareas arerehabilitated with single implant-supported crowns.The antagonist occlusalplaneshould berestoredto preventtorsionalforcesandoverloading.Implantsystemswithinitialtorquevaluesless than30Ncmshouldbeselected.

#2012ElsevierLtd.Allrightsreserved.

*Correspondingauthor.Tel.:+34607367903;fax:+34913942029.

E-mailaddress:raquel.castillo@odon.ucm.es(R.Castillo-Oyagu¨e).

Availableonlineatwww.sciencedirect.com

journalhomepage:www.intl.elsevierhealth.com/journals/jden

0300-5712/$–seefrontmatter#2012ElsevierLtd.Allrightsreserved.

http://dx.doi.org/10.1016/j.jdent.2012.09.001

1. Introduction

Theindicationsforuseofimplant-supportedprostheseshave expandedinrecentyearsduetotheirrecognizedbiological, functional, and aesthetic properties. However, numerous clinical studies^1,2 have presented the outcome of implant therapy by focusing only on implant survival without providingdetailedinformationontheprosthodonticrehabili- tation.³

Amongthevarietyofabutmentsthatcanbesatisfactorily utilizedforimplant-retainedprostheses,the UCLAcastable typeisoneofthemostpopular.Thisabutmentconsistsofa plasticcylinderthatdirectlyconnectstotheimplantandmay becustomizedbywaxingandcastingusingabasemetalalloy suchascobalt–chromium.Itslow cost,abilitytoovercome problems such as limited inter-occlusal spaces and small interproximaldistancesbetweenimplants,andthepossibility ofimplant angulation errorcorrection areits main advan- tages.^4,5 However,a drawbackofthis abutment isthatthe required laboratory steps could cause implant/abutment misfit,whichmayresultinscrewlooseningand/orfracture alongwithotherbiologicalcomplications.Therefore,further clinicalresearchisnecessary.⁵

Moreover, the biomechanical behaviour of implant- retainedrestorationsmaybecompromisedby,amongstother factors,accumulatederrorsinthecastingofthestructuresor intheveneeringprocedure.Afterstressconcentrationsthat exceed the ultimate strength of the restorative materials, screw loosening, deformation or fractureof the prosthetic framework,andceramicchippingorbreakagemayoccur.^6,7In thisregard,looseningandfracturingofthescrewsthatattach the abutments to the implants are the most prevalent prostheticproblems,rangingfrom4.5to12.7%ofcases.^3,8–13 Preloadisinducedinascrewwhenatorqueisappliedtofix the implant/abutment connection. Such preload keepsthe screwthreadstightlysecuredtothescrew’smatingcounter- partand holdsthe partstogetherbyproducing aclamping force between the screw headand its seat.^14–16 Thescrew elongates, placing the shank and threads in tension. The elasticrecoveryofthescrewcreatestheclampingforcethat pullstogethertheprosthesisandtheimplant.¹⁷Thestrength oftheunionisusuallyproportionaltothetorqueofthescrew whenitistightened,andisexpressedasamomentmagnitude measuredinNcm.^14,15Aspecifictorqueisrecommendedfor each implant system, and even for different diameter abutments within the same system (http://www.genieoss.- com/abutmenttorque.html).

Ifaflexionforceexertedonasingleimplantrestorationis greaterthantheresistanceofthescrew,¹⁴suchscrewsuffersa permanentplasticdeformationandconsequentlossoftensile strengthinitsnucleus.Thisphenomenonelicitsareductionin thecontactforcesandthescrewloosenseasily.Alsoasettling effect (embedment relaxation)¹⁵ is produced because the micro-roughenedsurfacesin contactareprogressively flat- tened under occlusal loading, which results in a partial decreaseoftheinitialpreload.

Afterscrewloosening,micro-movementsoftherestoration under load conditions may irritate the peri-implant tis- sues.^14,16,18 Although the solution usually consists in

tighteningorreplacingthescrew,itisveryinconvenientfor thepatientandthedentist.However,ifthescrewcannotbe retrieved,anextensiverepair,suchasdisuseoftheinvolved implant or remake of the prosthesis becomes critical.¹⁵ Similarly,whentheframeworkorveneeringmaterialbreaks down,therestorationmustbechanged.

In an attempt to prevent prosthetic failures in implant crowns and overcome possible economic constraints of patients,theaimsofthispaperareto:(a)assesstheincidence of prosthetic complications in screw-retainedand cement- retained crowns made with UCLA castable abutments; (b) describe the socio-demographic, anatomical, and implant- and crown-relatedriskfactorsforprostheticcomplications;

and(c)designanapproachtominimizetheunderlyingcauses ofprostheticproblemsinthesetypesofimplantcrowns.

2. Materials and methods

Acohortfollow-upinvestigationwascarriedoutinpartially dentatepatientswithimplant-supportedsinglecrownsover osseointegratedexternalheximplantslocatedineitherthe maxillaorthemandible.Allpatientswererehabilitatedatthe Dental Clinicof the Salamanca University (Spain)between 2005 and 2010. The exclusion criteria were cognitive im- pairment, motility disorders, and serious illness or death.

Additionally,singlecrownssupportedbytwoimplantswere excludedfromthestudybecauseoftheirdifferentbiomechan- ical implications. The study was conducted following the ethicalprinciplesofmedicalinvestigationinvolving human subjectsundertheHelsinkiDeclarationoftheWorldMedical Association(http://www.wma.net)andthe SpanishLaw14/

2007ofJuly3rdforBiomedicalResearch(http://www.boe.es).

All ofthe participants werebriefed aboutthe purpose and process of the study and the patients’ approved written consentwereobtained.Confidentialitywasmaintained.

All of the evaluated crowns were constructed in a standardized manner.Thestructures werevacuum-castin abasemetalalloyofwhite Co–Crforceramics(Heraenium CoCr metal ceramic alloy, Heraeus-Kulzer, Wehrheim, Germany).UCLAcastableabutmentswereusedinall cases to manufacture either screw-retained or cement-retained prostheses. The patterns of the screw-retained structures were custom-shaped by applying modelling wax (Classic modellingwax-blue,RenfertGmbH,Hilzingen,Germany)on the UCLA castable abutments. The patterns were then invested with a commercial phosphate-bonded stone (IPS PressVestSpeed,Ivoclar-VivadentAG,Scha¨an,Liechtenstein) byusingcylinderswithoutametalring.Thevacuumcastingof the Co–Cr specimens was performed with an induction centrifugal machine (MIE-200 C/R, Ordenta, Arganda del Rey,Madrid,Spain)undervacuumpressure(580mmHg)at a temperature of 14658C. Cast frameworks were then retrieved and carefully cleaned using an airborne-particle abrasionwithaluminium-oxidepowders(50mm)for10sata workingdistanceof5mmandapressureof503.5N/cm²to remove the investment residues.¹⁹ In the case ofcement- retained crowns, the abutments and the structures were

waxed-up and cast separately, in two consecutive steps, followingthedescribedprocedureineachphase(i.e.waxing- upthestructuresoverthecastabutments).

Oxidation ofthe crownframeworkswas completedina ceramic oven following the manufacturers’ instructions.

Then, two layers of opaque porcelain were applied that underwent two separate firing cycles of 30min/cycle in a vacuumceramicoven(ProgramatP500/G2,Ivoclar-Vivadent AG, Scha¨an, Liechtenstein). The first layer was heated at 9508C.²⁰ Subsequently, the structures were coated with a compatiblefeldspathicceramic(HeraCeram,HeraeusKulzer, Wehrhein,Germany)throughastratificationtechniqueusing dentineandenamelceramics inthesameoven at8508Cin everycycle.Finally,theglazefiringwasmadeat8108C.

Titanium screws with hexagonal heads were used to connecttheabutmentstothe correspondingimplants. The screwsweretightenedwithatorqueof20–35Ncmaccording tothemanufacturer’sspecificationsforeachimplantsystem.

Cement-retainedprostheseswerelutedwithresin-modified glass-ionomercement(KetacCemPlus,3MESPE,St.Paul,MN, USA).Theexcesscementwasremovedwithaplasticscalerto avoidscratchingorgougingtheabutmentsandrestorations.¹⁹ Allimplantcrownswerereviewedeverysixmonthsduring thefirstyear,andthenannually,usingvisualandradiographic investigation. Data collection was performed by a single operator(DB). Thestudyvariablesweregroupedasfollows:

Group1:socio-demographicbackground(gender,age,smok- inghabits);Group2:anatomicalfeatures(typeofbone,length of the edentulous saddle filled with the single implant restoration, mesiodistal cantilever of the crown, type of antagonist,intensityoftheocclusalcontact);Group3:clinical parametersrelatedtotheimplant(locationinthedentalarch, implantdiameterandlength,durationofthehealingphase, loading protocol, period of function, implant brand); and Group4:clinicalparametersrelatedtotherestoration(typeof retention,torqueofthescrews,numberoftimesthescrews hadtoberetightened,otherprostheticcomplications)(Tables 1and2).ThetorqueofthescrewswasmeasuredinNcmwitha manualtorquedriver(Ref:RTI2035,Biomet3i,UK,Ltd.).Inthe establishedperiodsofrevision, eachpatientwasevaluated twicebythesameobserver(DB).

Themesiodistallengthoftheedentuloussaddlerestored with the implant crown was measured by standardized intraoral digital radiovisiography (Kodak RVG, Kodak,

Germany) using specific software (Kodak Dental Imaging Software6.10.8.3,Kodak,Germany).Eachimagewascalibrated usingthediameteroftheimplantplatformasareference.Then, the saddle length was calculated between the greatest convexities of the adjacent teeth. When the restoration rehabilitated a free-end edentulous area or did not fill an intertooth space completely, the size of the saddle was consideredasthehighestmesiodistaldimensionoftheimplant crown.Fig.1showsanexampleofthemeasuringtechnique.

Forthepurposesofthisstudy,thedifferencebetweenthe highestmesiodistaldimensionofthecrownandthediameter of the implant platform was taken as the ‘‘mesiodistal cantileverofthecrown’’whenmorethana2mm-difference was observed.²¹ The same radiovisiography equipment (Kodak RVG) and system software (Kodak Dental Imaging Software6.10.8.3)wereused(Fig.1).Allofthemeasurements weremadebythesamespecializedoperator.

Thetypeofantagonistand theintensityofthe occlusal contactwiththeimplant crownweredeterminedbydirect inspectionandtheuseofarticulatingpaperof200mm(Bausch Company,Germany).

2.2. Statisticalanalyses

Theintra-examinationerrorwasevaluatedbytheKappatest.

Allofthedatagatheredwerestatisticallyanalyzedaccording to well-established evaluation methods used in related research.^9,10 Descriptivestatistics werecalculated forall of thestudyvariables.

Riskfactorswereexpressedusingtheincidence (%),the relativerisk(RR)andtheoddsratio(OR)coefficientswitha95%

confidence level. Parametric tests were used to compare quantitative variables(Studentt-testforvariableswithtwo categories and ANOVA for variables with three or more categories). The chi-squared test was applied to compare proportions. Astepwise logisticregressionmodel wasesti- matedusingtheoccurrenceofcrownremoval(eithertore- tightenthescrewsorbecauseofotherprostheticcomplica- tions)asthedependentvariable.Thestatisticalanalyseswere completedusingtheStatisticalPackagefortheSocialSciences (softwarev.17.0)(SPSS/PC+,Inc.;Chicago,IL,USA)takingthe cut-offlevelforstatisticalsignificanceata=0.05.^7,19,20

3. Results

Eightpatientswereexcludedfromthereferencepopulation becausetheycouldnotbecontactedduetochangesintheir phone number and/or address details. Nine other patients refused to participate in the study for personal reasons (rejectionrate=10.2%).

Thus, thestudysamplecomprised seventy-onepartially dentate patients (44 males and 27 females) who were rehabilitated, between 2005 and 2010, with a total of 93 individualcrownssupportedbyexternalheximplantsinboth the maxilla (47 implants) and the mandible (46 implants).

None of these patients were lost during the observation period.Themeanfollow-upperiod was26.215.4months.

Kappastatisticsshowedaperfectintra-assessmentcoefficient ofreliability(k=1).

Table1–Patientsampledata(n=71).

Samplesize(N) %

Gender

Male 44 62.0

Female 27 38.0

Age(meanSD=46.912.3)

35years 13 18.3

36–50years 32 45.1

51–64years 21 29.6

65years 5 7.0

Smokinghabits

Activesmoker 26 36.6

Non-smoker 19 26.8

Pastsmoker 26 36.6

Table2–Locationoftherestorations,typeandfeaturesoftheimplants,andcausesofprostheticcomplicationsinthe implant-supportedcrowns(n=93).

Positioninthearch

Incisor,n(%) Canine,n(%) Premolar,n(%) Molar,n(%) Total

Maxilla 10(10.7) 1(1.1) 21(22.6) 15(16.1) 47(50.5)

Mandible 1(1.1) 1(1.1) 2(2.1) 42(45.2) 46(49.5)

Total 11(11.8) 2(2.2) 23(24.7) 57(61.3) 93(100)

N %

Implantbrandandinitialtorque(meanSD=29.55.7Ncm)

Biomet3i^a(20Ncm) 22 23.6

Defcon^b(35Ncm) 9 9.7

Microdent^c(32Ncm) 12 13.0

Mozograu^d(30Ncm) 30 32.3

NobelBiocare^e(35Ncm) 20 21.5

Implantplatform(diameter)

Narrow 9 9.7

Standard(4.1mm) 76 81.7

Wide 8 8.6

Implantlength

<13mm 59 63.4

13mm 34 36.6

Typeofretention

Screw-retained 84 90.3

Cement-retained 9 9.7

Complicationsinimplant-supportedcrowns

Screwloosening 10 10.8

Osseointegrationfailure 2 2.2

Ceramicfracture 1 1.1

Withoutproblems 80 86.0

a Osseotite.Biomet3i,PalmBeachGardens,FL,USA.

b DefconTSA.Impladent,Barcelona,Spain.

c MK.MicrodentSL,Barcelona,Spain.

d MGOsseous.MozograuSL,Valladolid,Spain.

e MKIII.NobelBiocareAB,Go¨teborg,Sweden.

Fig.1–Measurementofthehighestmesiodistaldimensionofthecrownbycalibratingtheimageusingthediameterofthe implantplatformasareference.

Regardingsocio-demographicbackgrounds(Group1),the studysamplewasdrawnmainlyfrommen(62.0%),aged21–80 years(94%),beingactiveorpastsmokers(73.2%)(Table1).

Concerningtheanatomicalfeatures(Group2),themajority ofimplantswereplacedinhealedbone(90.3%),whereasthe remainingimplantswereplacedinfreshextractionsockets (9.7%). The mean length of the saddle restored was 9.72.3mm,and the averagemesiodistalcantileverofthe crown was 5.81.8mm. Of the antagonists, 75.2% were naturalteeth,18.3%wereceramic,5.4%wereacrylic,and1.1%

wasmetallic.Mostrestorationshadproperocclusalcontacts (95.7%).Three implantcrownsshowedpartial contactwith theirantagonists (3.2%) and onecrownwas infra-occluded (1.1%).

Regardingtheimplant-relatedvariables(Group3),80%of the implants werelocated inthe posterior sextantsofthe mouth, mainly in the mandibular molar (45.2%) and the maxillarypremolarregions(22.6%).Thecanineareawasthe leastfrequentlytreated(2.2%).Oftheimplants,81.7%hada standarddiameterand63.4%wereshorterthan13mm(Table 2).96.8%followedasingle-phasehealingprocedure,and98.9%

underwentadelayedloadingprotocol.Onaverage,implants healed during5.52.6monthsbefore loading and werein functionfor20.616.1 months.Two(2.2%)implants failed duetolossofintegrationin thefirst year offunction.The distributionofimplantbrandsisdisplayedinTable2.

With reference to the restoration (Group 4), all of the implants were rehabilitated with single metal–ceramic crowns.Eighty-fourofthecrowns(90.3%)werescrew-retained and9(9.7%)werecement-retained.Thescrewsweretightened with an average torque of 29.55.7Ncm. Eleven screw- retainedcrowns(11.9%)hadprostheticcomplications:screw loosening occurred in ten restorations (10.8%), whereas another screw-retained prosthesis (1.1%) suffered from ceramicfracture(Table2).Theloosescrewswereretightened andremainedstableafterwards,exceptonethathadtobere- tightened twice. Theincidence ofprosthetic problems was significantly different depending on the implant brand (p=0.02), and was higher inDefcon Implants(44.4%) than inMozoGrau(13.3%),Microdent(8.3%),NobelBiocare(5%)and 3IImplants(4.5%),whichshowednosignificantdifferencesin comparisonwithoneother(p>0.05).

The risk of failure was higher in posterior (13.8%) mandibular(15.2%)restorations.Anedentuloussaddlelonger than10mmtoberestoredwithasingleimplantrestoration andacrownwithamesiodistalcantileverlongerthan6mm augmented therisk ofbreakdown. Acleartrendtowards a greaterriskofcomplicationsmaybeobservedinrestorations withnaturalantagonists(14.3%)withrespecttothepresence ofprostheticones(4.3%).Therisk(RR)offailureincreasedafter 20 monthsoffunction(18.2%).Agreater tendency towards prostheticcomplicationswasdetectedinrestorationsscrewed

Table3–Riskfactorsforprostheticcomplicationsinsingleimplant-supportedrestorations(n=93).

Incidence(%) Chi-squaretestp-value RR Confidenceintervalof 95%forRR

Lower Upper

Prosthodonticfactors Location

Posterior 13.8 0.15 1.2 1.1 1.3

Anterior 0.0 NS NS NS

Jaw

Maxilla 8.5 0.31 0.6 0.3 1.3

Mandible 15.2 1.5 1.0 2.3

Lengthoftheedentuloussaddle

>10mm 18.2 0.07 1.7 1.1 2.6

10mm 6.1 0.5 0.2 1.3

Mesiodistalcantileverofthecrown

>6mm 15.9 0.25 1.4 0.9 2.3

6mm 8.2 0.7 0.3 1.5

Typeofantagonist

Natural 14.3 0.20 1.2 1.0 1.6

Prosthetic 4.3 0.3 0.1 2.3

Durationoffunction

>20months 18.2 0.07 1.7 1.1 2.6

20months 6.1 0.5 0.2 1.3

Initialtorque

>30Ncm 14.6 0.46 1.3 0.7 2.3

30Ncm 9.6 0.8 0.4 1.6

Patient-relatedfactors Smokinghabits

Smokers 14.5 0.26 1.3 0.9 1.8

Non-smokers 6.5 0.5 0.1 1.9

Age

>45years 17.1 0.16 1.5 0.9 2.6

45years 7.7 0.6 0.3 1.4

RR,riskratio;NS,nosense.

withinitialtorquevaluesof32–35Ncm(14.6%)incomparison withthoseinsertedat20–30Ncm(9.6%).Activesmokingwas associated with a higher incidence of prosthetic problems thannon-smoking,andpatientsolderthan45yearsshowed higherriskofprostheticfailures(17.1%)(Table3).

In a stepwise logistic regression model in which all potentialpredictorsareincludedtoenvisagetheoccurrence of prosthetic complications and all of the confounding variables are kept under control, a significant effect was foundforthesaddlelength(OR=6.7;IC95%1.5–30.6;p=0.01), thetypeofantagonist(OR=9.4;IC95%1.0–88.8;p=0.05),and theinitialpreloadvalueintroducedasaquantitativevariable (OR=1.2; IC 95% 1.0–1.4; p=0.05). The proportion of the variationofprostheticcomplicationsexplainedbythismodel is23%accordingtotheNagelkerkeR²statistic.

4. Discussion

Awidevarietyoftherapeuticoptionsmaybeconsideredin contemporarypracticeforreplacingteethwithfixedrehabi- litations;rangingfromresin-bondedbridgesorfixedpartial dentures(FPDs),uptotheuse ofimplant-retainedprosthe- ses.^2,22,23 However, for decision making, it is important to knowthesurvivalproportionsandtheincidenceofbiological andtechnicalcomplicationsnotonlyfortheimplants,butalso for the restorations¹⁸. In the present study, the results observed during a mean follow-up period of 26.2 months havebeendocumented.

Thisresearchattemptstoexplorethefactorsassociated with prosthetic complications in implant-retained crowns withaviewtoestablishingclinicalrecommendationstohelp reduce the incidence of such problems in future patients.

Implant crowns have shown high survival rates after the observation period but also an increased prevalence of technicalcomplications,³whichisstillaconcernforclinicians andpatients.

In this study, standardized laboratory procedures for restoration fabrication were used and clinical periodic assessments were made twice by a single operator (DB), obtainingaperfectintra-examinationreliability.Conversely, limitationsofthestudyincludedtheimpossibilityofmeasur- ing the occlusalforces in eachcase, and the fact thatthe patientswererecruitedonlyfromasingleuniversitydental clinic.Furthermore,asthetestedvariableswerenotrandomly assigned,acause–effect relationshipcannotbeestablished.

Therefore,the riskfactors envisagedarejustbased onthe distribution of prosthetic complications concerning such studyvariables.

Themajor finding ofthis investigation isthat during a mean period of 26.215.4 months, the most frequent prosthetic complication was screw loosening showing an incidence of 10.8% (Table 2), which was higher than that previously reported in five toten years offollow-up stud- ies.^10,11,13Thismaybeattributabletodifferencesinthetypeof abutment, the implant system and the initial torque.

Nonetheless,allofthelooseabutmentsremainedstableafter beingretightenedwiththerecommendedtorque,exceptfor onethatwasretightenedtwice.Ourresultsareinaccordance withthesystematicreviewperformedbyJungetal.³inwhich

anaccumulatedincidence of12.7%ofscrewlooseningwas estimatedafterfiveyearsoffollow-up,doublingtheincidence ratedinpreviousstudiesonimplant-supportedFPDs.^11,13

TheuseofUCLAcastableabutmentsmayhavecontributed to the high frequency ofprosthetic complications because theircastingprocedureisverytechnique-sensitiveandmay have somewhat altered the fit at the implant/abutment interfaces. It must also be taken into account that great vertical misfits dot not necessarily imply higher detorque values.²⁴

Less prosthetic failures have been reported for prema- chined²⁵andgold-machinedUCLAabutments.²⁶Tocastthe UCLAabutments,acobalt–chromiumalloywasselectedasits highfracturestrength,elasticmodulus,hardness,corrosion resistanceandlowcost²⁷makeitanalternativeforconstruct- ingscrew-retainedprosthesesandcustomizedabutmentsfor cement-retainedcrowns, offering aneconomicsolutionfor patients.However,furtherresearchonthelong-termsuccess of implant-supported restorations fabricated with the de- scribedmethodisrequired.

Anotherexplanatoryfactorofthehighincidenceofscrew loosening lies behind the use of titanium screws in all restorations. Greaterstabilityhasbeen reportedforgoldor surface-treated titanium screws because of their higher preloadvaluesforanygiventorqueforce.²⁸

Inkeepingwiththeconclusionsofotherauthors,⁹screw looseningmainlyoccurredintheposteriormandibulararea (Table3).Inadditiontothelengthofthemolarrestorations, thechewingforcesexertedontheposteriorsectorsofthearch are three to five times greater than those received by the anterior teeth.²⁹ Hence, that 80% of the restorations were located inthe premolar and molar regions (Table 2) could somewhatjustifythehighincidenceofscrewlooseningfound inthisstudy.

Ourresultscaninferseveralfactorsthatimprovetheriskof screw loosening.Thelength oftheedentulous saddletobe restoredwithasingleimplantrestoration(>10mm),thetype of antagonist, and the initial preload or torque value significantlyincreasedsuchriskaccordingtoamultivariate logisticregressionmodel.Thesefindingsareinkeepingwith the literature.^9,11,30,31 The patient age and the period of function were not significant in the multiple regression model,althoughtheywereintheexpecteddirectionaccording tootherauthors,14,16,32–36sothattheolderthepatientsandthe restorations,thehighertherateofscrewloosening(Table3).

A long saddle to be restored with a single implant restorationand acrown withalong mesiodistalcantilever enhancestheriskofscrewloosening(Table3).Thenegative influenceofcantileversonthelong-termsuccessofprosthetic restorationshasbeendocumentedintheliterature.^30–32When compressiveocclusalloadsareexertednearertothemarginal crestthantotheimplantaxis,atorsionalforceiscreatedthat may increase the risk of screw loosening or fracturing.

Bakaeen etal.³¹concluded thattheriskofscrewloosening coulddecreasebyreducingtheocclusaltableofmolarimplant crownsinboththeirbuccolingualandmesiodistaldimensions tominimizethetorsionalforces.¹⁵Giventhatthelongerthe distance of theload fromthe implant axis, thehigher the torsionmomentand theleverforces,^30–32standard orwide implant platforms would be preferred over narrow ones

wheneverpossibleforreducingcantilevers.However,thiswas notthefocusofthepresentstudy.

A significant relationship was observed between screw looseningandrestorationshavingnaturalantagonists(Table 3).Whenatoothisininfra-occlusion,itofteneruptstoseek occlusalcontact.³⁷Thisphysiologicalprocessisnotpresent whentheantagonistisanimplant-retainedprosthesis,afixed oraremovablepartial denture.^36,37 Moreover,mostnatural teetharenotlocatedintheidealintercuspalrelationshipgiven their tendency for migration and mesial inclination. This phenomenon generates oblique and torsional forces that cause overloading on the restorations.³⁸ On the contrary, prostheticantagoniststendtohavetheocclusalplaneatthe proper level and inclination, which optimizes the stress distribution.Theresultsofthepresentstudyareinagreement with the discussed conclusions on such biomechanical factors.³⁹

Furthermore, the higher the initial torque, the more elevatedtheincidenceofcomplications(Table3).Thisfinding shouldbe viewedwithcaution becauseit is relatedtothe implant brand, as the initial preload is based on the manufacturers’instructions(Table2)andcannotbeconsid- eredasanindependentvariable.Anotherpossibleexplanation for this apparently contradictory observation could be the settlingeffect.¹⁵Ahightorquecanleadtoeffectiveerosionof the micro-roughened surfaces of the screwed joint and, thereby,toapreloaddecrease.Between 2%and 10%ofthe initialpreloadisestimatedtobelostbecauseofthesettling effect.^40,41 Severalauthorssuggest re-tighteningthescrews ten minutesafter tightening them forthe first time.^40,42,43 Moreover,toassurethe longevityoftheimplant–abutment union,theuseofmorse-taperedconnectionsinsteadofhex implantshasbeensuggested.⁴⁴

Ceramic fracture was found only in a screw-retained prosthesis (Table 2). The veneering ceramic used has a stabilized leucite structure (SLS) that makes it particularly resistanttostress andcrack propagation.⁴⁵Moreover, even whenthemetal–ceramicbondispredictableinbothtypesof implant–prosthetic connections, screw-retainedcrowns are morepronetosufferocclusalmicrocracking(neartothescrew access) in comparison with cement-retained prostheses.⁴⁶ Acrylicresinhasmoreresiliencethanceramicasveneering material and,therefore, lower fracturerates,⁴⁷ which is in accordancewiththatrecordedinthecurrentstudy.

Therefore,basedonthefindingsofthisstudy,theriskof prostheticcomplications isexpectedtoincreasewhenlong- span posterioredentulous areas (>10mm) are rehabilitated withsingleimplant-supportedcrowns.Theantagonistocclusal plane should be restored to prevent torsional forces and overloading,and implantsystemswithinitialtorquevalues less than 30Ncm should be selected. Notwithstanding the prostheticcomplicationsdiscussedhere,thehighsurvivalrates obtainedinthisresearch,whicharesimilartothatpresentedin prospectivemulticenterstudiesforsingle-toothrestorations,^8,9 mayleadtotheconclusionthatthistreatmentmodalitywith UCLAcastableabutmentsseemsreliableforroutineuse.

Furthermore,thedurabilityoftheseimplantrestorations maydependonothercriticalfactorsthatmaybeaddressedin futureclinicaltrials,suchasthesurgicaltechnique,implant design,andsurfacecharacteristics,^48,49meanclinicalattach-

mentlevelatimplant,⁵⁰correlationbetweenocclusalscheme and failure mode of the restorative materials,^51,52 and presence bruxism and tooth wear.⁵³ Also the operator experiencemayinfluencethelong-termsuccessofimplant- supportedcrowns.⁵⁴Accordingly,theteachingprogrammesin bothconventionalfixedandimplant-retainedprosthesesare evolving and are sensitive to current clinical trends and evidence-basedpractice.^55,56

5. Conclusions

Withinthelimitationsofthisstudy,thefollowingconclusions maybedrawn:

1. Screwlooseningisthemostfrequentprostheticcomplica- tion inimplantcrowns fabricatedwithUCLAabutments cast in cobalt–chromium. Nevertheless, the connection usuallyremainsstableafterre-tighteningthescrews.

2. A long saddle to be restored with a single implant restoration (>10mm), a crown with a long mesiodistal cantilever(>6mm),anaturalantagonist,andahighinitial torque(Ncm),whichdependsontheimplantbrand,arethe mainriskfactorsfortreatmentfailure.

3. A high survival rate was recorded afterthe observation period(26.215.4months),andtheuseofUCLAcastable abutments may be recommended to fabricate implant- supportedcrownsasasuitabletreatmentoption.

Acknowledgements

Theauthors gratefully acknowledge all thosepatients who participatedinthestudy.Thisstudywaspartlysupportedby the FIS projectEXPTE: PI081020 ofthe Spanish Ministry of Health,bytheresearchprojects:UCM322-2005(sponsoredby NobelBiocareIbe´rica,S.A.),UCM90-2008(sponsoredbyAstra Tech,S.A),UCM41-2009(sponsoredbyAstraTech,A.B.)and the followingprojectsoftheClı´nicoSanCarlos Hospitalof Madrid(CEIC),C.I.10/159–E,C.I.12/240–E,C.I.12/241–E,C.I.12/

242–E,C.I.12/279–EandC.I.12/280–E.

references

1. VigoloP,GivaniA.Clinicalevaluationofsingle-toothmini- implantrestorations:afive-yearretrospectivestudy.Journal ofProstheticDentistry2000;84:50–4.

2. RomeoE,LopsD,MarguttiE,GhisolfiM,ChiapascoM,Vogel G.Long-termsurvivalandsuccessoforalimplantsinthe treatmentoffullandpartialarches:a7-yearprospective studywiththeITIdentalimplantsystem.International JournalofOralandMaxillofacialImplants2004;19:247–59.

3. JungRE,PjeturssonBE,GlauserR,ZembicA,ZwahlenM, LangNP.Asystematicreviewofthe5-yearsurvivaland complicationratesofimplant-supportedsinglecrowns.

ClinicalOralImplantsResearch2008;19:119–30.

4. LewisS,Beumer3rdJ,HornburgW,MoyP.TheUCLA abutment.InternationalJournalofOralandMaxillofacial Implants1988;3:183–9.

5. BarbosaGA,SimamotoJu´niorPC,FernandesNetoAJ,de MattosMdaG,NevesFD.Prostheticlaboratoryinfluenceon theverticalmisfitattheimplant/UCLAabutmentinterface.

BrazilianDentalJournal2007;18:139–43.

6. MischCE.Considerationofbiomechanicalstressin treatmentwithdentalimplants.DentistryToday2006;25.80, 82,84–85;quiz85.

7. Castillo-deOyagu¨eR,OsorioR,LynchC,GilmourA, ToledanoM.Effectofalloytypeandcastingtechniqueon thefracturestrengthofimplant-cementedstructures.

MedicinaOralPatologı´aOralyCirugı´aBucal2011;16:619–25.

8. SchellerH,UrgellJP,KultjeC,KlinebergI,GoldbergPV, Stevenson-MooreP,etal.A5-yearmulticenterstudyon implant-supportedsinglecrownrestorations.International JournalofOralandMaxillofacialImplants1998;13:212–8.

9. BiancoG,DiRaimondoR,LuongoG,PaoleschiC,PiccoliP, PiccoliC,etal.Osseointegratedimplantforsingle-tooth replacement:aretrospectivemulticenterstudyonroutine useinprivatepractice.ClinicalImplantDentistryandRelated Research2000;2:152–8.

10.SimonRL.Singleimplant-supportedmolarandpremolar crowns:aten-yearretrospectiveclinicalreport.Journalof ProstheticDentistry2003;90:517–21.

11.KreisslME,GerdsT,MucheR,HeydeckeG,StrubJR.

Technicalcomplicationsofimplant-supportedfixedpartial denturesinpartiallyedentulouscasesafteranaverage observationperiodof5years.ClinicalOralImplantsResearch 2007;18:720–6.

12.TheoharidouA,PetridisHP,TzannasK,GarefisP.Abutment screwlooseninginsingle-implantrestorations:asystematic review.InternationalJournalofOralandMaxillofacialImplants 2008;23:681–90.

13.KrennmairG,SeemannR,SchmidingerS,EwersR, PiehslingerE.Clinicaloutcomeofroot-shapeddental implantsofvariousdiameters:5-yearresults.International JournalofOralandMaxillofacialImplants2010;25:

357–66.

14.SiamosG,WinklerS,BoberickKG.Relationshipbetween implantpreloadandscrewlooseningonimplant-supported prostheses.JournalofOralImplantology2002;28:67–73.

15.WinklerS,RingK,RingJD,BoberickKG.Implantscrew mechanicsandthesettlingeffect:overview.JournalofOral Implantology2003;29:242–5.

16.ByrneD,JacobsS,O’ConnellB,HoustonF,ClaffeyN.

Preloadsgeneratedwithrepeatedtighteninginthreetypes ofscrewsusedindentalimplantassemblies.Journalof Prosthodontics2006;15:164–71.

17.HaackJE,SakaguchiRL,SunT,CoffeyJP.Elongationand preloadstressindentalimplantabutmentscrews.

InternationalJournalofOralandMaxillofacialImplants 1995;10:529–36.

18.PapaspyridakosP,ChenCJ,ChuangSK,WeberHP,Gallucci GO.Asystematicreviewofbiologicandtechnical

complicationswithfixedimplantrehabilitationsfor edentulouspatients.InternationalJournalofOraland MaxillofacialImplants2012;27:102–10.

19.Oyagu¨eRC,Sa´nchez-Turrio´nA,Lo´pez-LozanoJF,Sua´rez- Garcı´aMJ.Verticaldiscrepancyandmicroleakageoflaser- sinteredandvacuum-castimplant-supportedstructures lutedwithdifferentcementtypes.JournalofDentistry 2012;40:123–30.

20.Castillo-Oyagu¨eR,OsorioR,OsorioE,Sa´nchez-AguileraF, ToledanoM.Theeffectofsurfacetreatmentsonthe microroughnessoflaser-sinteredandvacuum-castbase metalalloysfordentalprostheticframeworks.Microscopy ResearchandTechnique2012;75:1206–12.

21.KimYK,KimSG,YunPY,HwangJW,SonMK.Prognosisof singlemolarimplants:aretrospectivestudy.International JournalofPeriodonticsandRestorativeDentistry2010;30:401–7.

22.PatelPM,LynchCD,SloanAJ,GilmourAS.Treatment planningforreplacingmissingteethinUKgeneraldental practice:currenttrends.JournalofOralRehabilitation 2010;37:509–17.

23.KerschbaumT,HaastertB,MarinelloCP.Riskofdebonding inthree-unitresin-bondedfixedpartialdentures.Journalof ProstheticDentistry1996;75:248–53.

24.BarbosaGA,BernardesSR,dasNevesFD,FernandesNetoAJ, deMattosMdaG,RibeiroRF.Relationbetweenimplant/

abutmentverticalmisfitandtorquelossofabutment screws.BrazilianDentalJournal2008;19:358–63.

25.ByrneD,HoustonF,ClearyR,ClaffeyN.Thefitofcastand premachinedimplantabutments.JournalofProsthetic Dentistry1998;80:184–92.

26.VigoloP,MajzoubZ,CordioliG.Measurementofthe dimensionsandabutmentrotationalfreedomofgold- machined3iUCLA-typeabutmentsintheas-received condition,aftercastingwithanoblemetalalloyand porcelainfiring.JournalofProstheticDentistry2000;84:

548–53.

27.RoachM.Basemetalalloysusedfordentalrestorationsand implants.DentalClinicsofNorthAmerica2007;51:603–27.

28.Stu¨kerRA,TeixeiraER,BeckJC,daCostaNP.Preloadand torqueremovalevaluationofthreedifferentabutment screwsforsinglestandingimplantrestorations.Journalof AppliedOralScience2008;16:55–8.

29.JohnsenSE,SvenssonKG,TrulssonM.Forcesappliedby anteriorandposteriorteethandrolesofperiodontal afferentsduringhold-and-splittasksinhumansubjects.

ExperimentalBrainResearch2007;178:126–34.

30.SatoY,ShindoiN,HosokawaR,TsugaK,AkagawaY.

Biomechanicaleffectsofdoubleorwideimplantsforsingle molarreplacementintheposteriormandibularregion.

JournalofOralRehabilitation2000;27:842–5.

31.BakaeenLG,WinklerS,NeffPA.Theeffectofimplant diameter,restorationdesign,andocclusaltable variationsonscrewlooseningofposteriorsingle-tooth implantrestorations.JournalofOralImplantology2001;27:

63–72.

32.AboyoussefH,WeinerS,EhrenbergD.Effectofan antirotationresistanceformonscrewlooseningforsingle implant-supportedcrowns.JournalofProstheticDentistry 2000;83:450–5.

33.WeissEI,KozakD,GrossMD.Effectofrepeatedclosureson openingtorquevaluesinsevenabutment-implantsystems.

JournalofProstheticDentistry2000;84:194–9.

34.KhraisatA,HashimotoA,NomuraS,MiyakawaO.Effectof lateralcyclicloadingonabutmentscrewlooseningofan externalhexagonimplantsystem.JournalofProsthetic Dentistry2004;91:326–34.

35.TsugeT,HagiwaraY.Influenceoflateral-obliquecyclic loadingonabutmentscrewlooseningofinternaland externalhexagonimplants.DentalMaterialsJournal 2009;28:373–81.

36.PetridisHP,TsiggosN,MichailA,KafantarisSN, HatzikyriakosA,KafantarisNM.Three-dimensional positionalchangesofteethadjacenttoposterioredentulous spacesinrelationtoageattimeoftoothlossandelapsed time.EuropeanJournalofProsthodonticsandRestorative Dentistry2010;18:78–83.

37.CraddockHL,YoungsonCC,ManogueM,BlanceA.Occlusal changesfollowingposteriortoothlossinadults.Part1.A studyofclinicalparametersassociatedwiththeextentand typeofsupraeruptioninunopposedposteriorteeth.Journal ofProsthodontics2007;16:485–94.

38.PellizzerEP,CarliRI,Falco´n-AntenucciRM,VerriFR,Goiato MC,VillaLM.Photoelasticanalysisofstressdistribution withdifferentimplantsystems.JournalofOral

Implantology2011,inpress.

39.SahinS,CehreliMC,Yalc¸inE.Theinfluenceoffunctional forcesonthebiomechanicsofimplant-supported prostheses—areview.JournalofDentistry2002;30:271–82.

40.DixonDL,BreedingLC,SadlerJP,McKayML.Comparisonof screwloosening,rotation,anddeflectionamongthree implantdesigns.JournalofProstheticDentistry1995;74:270–8.

41.GrattonDG,AquilinoSA,StanfordCM.Micromotionand dynamicfatiguepropertiesofthedentalimplant–abutment interface.JournalofProstheticDentistry2001;85:47–52.

42.BreedingLC,DixonDL,NelsonEW,TietgeJD.Torque requiredtoloosensingle-toothimplantabutmentscrews beforeandaftersimulatedfunction.InternationalJournalof Prosthodontics1993;6:435–9.

43.KimKS,LimYJ,KimMJ,KwonHB,YangJH,LeeJB,etal.

Variationinthetotallengthsofabutment/implant assembliesgeneratedwithafunctionofappliedtightening torqueinexternalandinternalimplant–abutment connection.ClinicalOralImplantsResearch2011;22:834–9.

44.KitagawaT,TanimotoY,OdakiM,NemotoK,AidaM.

Influenceofimplant/abutmentjointdesignsonabutment screwlooseninginadentalimplantsystem.Journalof BiomedicalMaterialsResearchPartBAppliedBiomaterials 2005;75:457–63.

45.SeghiRR,DenryIL,RosenstielSF.Relativefracture toughnessandhardnessofnewdentalceramics.Journalof ProstheticDentistry1995;74:145–50.

46.ZaraoneF,SorrentinoR,TrainiT,DilorioD,CaputiS.

Fractureresistanceofimplant-supportedscrew-versus cement-retainedporcelainfusedtometalsinglecrowns:

SEMfractographicanalysis.DentalMaterials2007;23:296–301.

47.BassitR,Lindstro¨mH,RangertB.Invivoregistrationofforce developmentwithceramicandacrylicresinocclusal materialsonimplant-supportedprostheses.International JournalofOralandMaxillofacialImplants2002;17:17–23.

48.JavedF,RomanosGE.Theroleofprimarystabilityfor successfulimmediateloadingofdentalimplants.A literaturereview.JournalofDentistry2010;38:612–20.

49.AnnunziataM,OlivaA,BasileMA,GiordanoM,MazzolaN, RizzoA,etal.Theeffectsoftitaniumnitride-coatingonthe topographicandbiologicalfeaturesofTPSimplantsurfaces.

JournalofDentistry2011;39:720–8.

50.FaggionJrCM,ListlS,TuYK.Assessmentofendpointsin studiesonperi-implantitistreatment—asystematicreview.

JournalofDentistry2010;38:443–50.

51.LorenzoniFC,MartinsLM,SilvaNR,CoelhoPG,GuessPC, BonfanteEA,etal.Fatiguelifeandfailuremodesofcrowns systemswithamodifiedframeworkdesign.Journalof Dentistry2010;38:626–34.

52.BaldassarriM,ZhangY,ThompsonVP,RekowED,Stappert CF.Reliabilityandfailuremodesofimplant-supported zirconium-oxidefixeddentalprosthesesrelatedto veneeringtechniques.JournalofDentistry2011;39:489–98.

53.KatsoulisJ,NikitovicSG,SprengS,NeuhausK,Mericske- SternR.Prostheticrehabilitationandtreatmentoutcomeof partiallyedentulouspatientswithseveretoothwear:3- yearsresults.JournalofDentistry2011;39:662–71.

54.Oyagu¨eRC,Turrio´nAS,ToledanoM,MonticelliF,OsorioR.

Invitroverticalmisfitevaluationofcastframeworksfor cement-retainedimplant-supportedpartialprostheses.

JournalofDentistry2009;37:52–8.

55.AddyLD,LynchCD,LockeM,WattsA,GilmourAS.The teachingofimplantdentistryinundergraduatedental schoolsintheUnitedKingdomandIreland.BritishDental Journal2008;205:609–14.

56.LynchCD,SinghraoH,AddyLD,GilmourAS.Theteaching offixedpartialdenturesinundergraduatedentalschoolsin IrelandandtheUnitedKingdom.JournalofOralRehabilitation 2010;37:908–15.