Restorative Dentistry

Three-year evaluation of computer-machinedceramic inlays: Influence of luting agentRita Zueliig-Singei, DDS*/Roland W. Bryant, BDS, MDS. PhD^ *

Objective: Changes in the occlusal marginal adaptation of computer-machined inlay restorations wei-eassessed over .Í vears.Method and materials: Each of 21 patients received two Cerec {Vtia Mark II) two- or three-suifacerestorations, with maigins entirely in enamel, placed using tivo of four luting agents: a homogeneousmicrofilled ivsin composite, afine hybrid composite, a relatively coarse hybrid composite, and an encap-sulated glass-ionomer cement. Replicas of the restorations were made at baseline 6 months orI year (93% of original restorations), 2 years [79%). attd i years (8S%). The percentage of continuousmargin on the occlusal margin was evaiuatedfor the enamel-luting agent and luting agent-porcelain in-terfaces using a scanning electron microscope.Results: Mean continuous margin values were high for all groups at the enamel-luting agent interface;there were no significant differences among the luting agents. For each luting agent, the percentage ofcontinuous margin decreased significantly ai the luting agent-porcelain interface over 3 years; differencesamong the luting agents at 3 years were not significant (Kruskal-Wallis test). After 3 years, the wear of theluting agents was .úgnificantly different. The least wear occurred with the microfilled resin composite; thecoarse hybrid composite showed the most wearConclusion: The occlusal margins of Cerec inlays exhibited consistently high-qualiry adaptation at 3years. The microfilled resin composite luting agent was more wear resistant than the hybrid compositesand the glass-ionomer cement. (Quintessence Int 1998;29:573-5íi2)

Key words: Cerec inlays, continuous margin, luting agent, marginal adaptation, wear

Clinical relevance

Cerec restorations exhibited consistently high-quality occlusal margins after 3 years of ciinicaluse, regardless of the luting agent. A microfilledcomposite luting agent exhibited less evidence ofwear and appears to be the preferred choice for lut-ing porcelain inlays.

Patient demand for tooth-colored posterior restorationsis increasing. For larger restorations, inlays are supe-

rior alternatives to direct resin composite fillings.'- Cer-amics are among the most biocompatible materials devel-oped for dental restorations'-'' and are more esthetic anddurable in clinical conditions tban composite materials.

'Private Practice, Zurich. Switzerland.

•'Professor. Deparlment of Conservative Dcnlistry. University of Syd-ney, Sydney. Australia; Ciinical Dentistry, Wcstmcad Hospitat. West-mead, New Soutti Wales. Australia.

Reprint requests: Dr Rita Zuellig-Singer. Schnetkenmannslrasse 'J.8044 Zuricti, Switzertand. Fax: 4t-l-252-0574.

The Cerec metbod (Siemens), a computer-aided de-sign-computer-aided macbining restorative techniquethat enables the direct chairside placement of ceramicrestorations without auxiliary laboratory support, hasbeen used clinically since 1986.'' Tbe Cerec tecbniqueconsists of three-dimensional scanning of the cavity,immediate data transformation, and three-axial milling,which IS integrated into the mobile unit.'

As with other indirecdy fabricated restorations, theweakest part of tbe Cerec restoration is the lutingagent layer that is exposed at the margin. The long-term success of these restorations depends mainly onthe success of the luting agent.' which requires a lowshrinkage rate, high wear resistance, bonding to botbtootb structure and ceramic, polishability. and colorstability. Previous results have generally indicated thatdual-cured resin composites are preferred for the lut-ing of ceramic inlays because of their ability to setcompletely and tbeir greater resistance to occlusalloading compared with glass-ionomer luting agents.''"The advantage of glass-ionomer cement is the reliablechemical bond to tooth structure and its fluoride re-

573

Zueliig-Singer/Bryant

TABLE 1

Groupcode

Dt

0^Ct

K

Luting agents investigated

ProductDual CemenlDuo CementCR inlay cementKetac-Cem(incapsulated)

" Information provided by the manulactur< Dual cur d.

ManufacturerVivadentVita ZahnfabrikKurarayESPE

Fillersize'

0.04 |jmMean 0.5 p m0.1-20 MmGlass-ionomer

Filiercontent'(wt%)

60-6268

76

This article reports on changes in occiusal marginaladaptation of Cerec restorations, analyzed under a scan-ning electron microscope (SEM), after a 3-year periodof clinical use. One microfilied resin composite, twohybrid composites with different particle sizes, and oneglass-ionomer cement luting agent were tested.

Method and materials

Twenty one patients were selected randomly from pa-tients registered at the United Dental Ho.spital in Sydney.Each padent required two Class II restorations in perma-nent premolars or molars. Written patient consent wasohtained at the start of the project, and the protocol wasapproved hy the Human Ethical Research Committee ofthe hospital.

Oral hygiene instructions were given, and sealingwas carried out prior to placement of restorations. Eachpatient received two Class II (mesio-occlusal. disto-occlusal, or mesio-occlusodistal) Cerec restorations.One operator treated all patients.

Tooth preparation

After the injection of local anesthetic solution, cavitieswere prepared with cylindrical diamond burs of lOO-pmgrit and finished with cylindrical and inverse conical di-amond burs of 25-tim grit and margin trimmers. Featuresof the cavity design were slightly divergent walls (di-verging toward the occiusal snrface) in the proximal por-tion, parallel or slightly divergent walls in the occiusalportion, and enamel margins that were not beveled. Ifthere were undercitts in the occiusal portion, they wereallowed to remain, to be filled out with the luting agent.

Rubber dam was placed, and an encapsulated glass-ionomer lining cement (Ketac-Bond, ESPE) was ap-plied to the air-water-s prayed and dried dentin surfacesof the prepared cavity. In very deep cavities, a spot ap-

plication of calcium hydroxide (Dycai, Dentsply) wasused prior to the glass-ionomer lining cement, Cerec liq-uid (Vita Zahnfabrik) was brushed onto the cavity wallsand margins and air blown to a thin layer with pres.sur-ized air. Cerec powder (Vita Zahnfabrik) was thensprayed evenly onto the cavity,

inlay fabrication

The camera head was positioned over the tooth to takean optical impression of the cavity, after which the cav-ity was protected from desiccation by a wet cotton pel-let, A static pseudoplastic image of the tooth appearedon the screen. Individual marker points were set to de-fine the cavity floor and the approximal contours. Therestoration walls were determined automatically,' andthe electronically designed inlay was milled, undercomputer control, out of a ceramic block (Mark II, VitaZahnfabrik) by a diamond-coated disk. The material tagwas ground away with diamond burs, and the approxi-mal surfaces were polished with flexible disks (Pop-onflexible disks, 3M Dental). The inlays were tried in and,if necessary, adju.sted to achieve a good fit.

Luting procedures

The luting agents were systematically allocated to spe-ciftc restorations in a way that minimized the influenceof tooth type, position, and restoration size. Based onlhe four luting agents investigated, four experimentalgroups were formed: Dual Cement (group D) Duo Ce-ment (group 0), CR inlay cement (group C), and Ketac-Cem (group K) (Table I ). The distrihution of molars andpremolars m each group is shown in Table 2.

After milling of the inlay was complete, the operatorwas advised (by reference to an allocation form) whichone of the four luting agents had been designated for theparticular restoration. Until this stage, the operator who

574 Volume 29, Number 9

Zueiiíg-Singet/Bryanl

TABLE 2 Nutnber of inlays assessed

Assessment

Baseline

Recall 1 (6 mo - 1

Recall 2 (2 y)

Recail 3 (3 y]

Group D

PM M

5 3

/) 4 3

4 2

4 2

Group 0

PM

7

6

5

6

M

4

4

3

4

Group C

PM

8

8

6

6

PM = numbei of premoiars; U = nurrber ol rroiars.• One miay tracturad at £ years and was repiaced by direct resin composite

M

5

5

4

4"

Group K

PU

5

5

4

4

M

5

4

5

5

had not been involved in allocating the luting tech-niques, was unaware as to which luting agent was to beused.

Group D. The iuting technique for group D consistedof the following steps. Silane coupling agent (Silanit,Vivadent) was mixed, according to the manufacturer'sinstructions, 15 minutes prior to use. The inlay's fittingsurface was etched for 60 seconds with Cerec etch (VitaZahnfabrik}, containing 4.9% hydrotluoric acid, thor-oughly rinsed and dried, and then placed into the silanecoupling agent liquid for 3 minutes and dried with theair .syringe.

Strips of transparent matrix hands (Ash alpine matrixstrips. De Trey) were placed interproximally and held inplace with wooden wedges. The enamel walls of thecavity were etched for 15 to 30 .seconds with 37% phos-phoric aeid (etching gel. Vita Zahnfabrik), after wbichthe tooth was rinsed for 20 seconds and air dried. Nodentin primer was used.

On both the fitting surface of the inlay and the cavitywalis, a low-viscosity bonding agent (Heliobond, Vi-vadent) was applied, blown out to a thin layer, and lightcured for 10 seconds. After it was mixed according tothe manufacturer's instructions, luting agent D (DualCement) was applied to the cavity walls. The inlay wasinsened up to three quarters of the way into the eavity,and gross excess of the luting agent was removed. Aftercomplete seating of the inlay, the small amount of re-maining excess was left for curing using a light frombuccal, lingnal, and ocelusal directions for 40 secondseach.

After removal of the rubber dam, excess luting agentwas removed from the approximal part with oscillatingfiles (Eva profm lamineer, Dentatus) and from the oc-elusal part with fine diamond burs. The restoration wasthen eontoured, occlusally adju.sted, and prepolishedwith 30-|im-grit diamond burs. A topical fluoride solu-tion was applied to all margins of the restoration.

Group O. A similar luting proeedure was carried outfor gronp O restorations. The inlay was etched in thesame way as for group D. Components A and B of thesilane coupling agent (Silicoup, Heraeus Kulzer) weremixed according to instructions, applied to the fittingsurface of the inlay with a brush, and the surface wasgently dried with air. The enamel of the cavity wasetched, washed, and dried as described previously. Nodentin primer was applied.

A bonding agent (Duo bond. Vita Zahnfabrik) wasmixed and brushed on the cavity and the fitting surfaceof the inlay. Luting agent O (Duo Cement) was mixedaccording lo the manufacturer's instructions and theinlay was seated, cured, contoured, and treated with flu-oride as described for group D restorations.

Group C. For group C restorations, etching of theinlay was carried out as previously described. Themixed porcelain bonding agent (Clearfil porcelain bond,Kuraray) was applied to the inlay with a brush, andwithin 30 .second.s of the application, the inlay was airdried for 2 to 3 seconds. An eichant gel (K-etehant gel,Kuraray) was used to etch the whole cavity (enamel,dentin, and glass-ionomer lining) for 15 to 30 seconds.After the cavity was rinsed with water and dried, ahonding agent (Clearfil porcelain bond, Kuraray) wasmixed and brushed on the cavity.

The powder and liquid components of luting agent C(Clearfil CR inlay cement) were mixed according to themanufacturer's instructions and applied to the cavity.The previously descrihed techniques for two-stepseating, curing, contouring, and fluoride treatment werefollowed.

Group K. For group K restorations, the inlay wasetched as described for group D. The encapsulatedglass-ionomer luting agent (Ketae-Cem) was mixed for10 seconds and apphed to the cavity. A two-step seatingprocedure, described for group D, was followed. Imme-diately after placement, the luting agent was protected

575

Zuellig-Singet/Bryant

Figs la tole Examples ot the SEM evaluation ctiteria. (E) enamel; (P] porcelain; (L) luting agent.

Fig la Group K at baseline, showing a oontinuous margin. Fig l b Group K at 3 years, witti evidence of gap (marginalopening) (arrow) and evidence ol wear.

Fig 1c Gtoup D at 2 years, willn evidence ol Hash (marginalexoess) (arrow).

from exposure to air with a light-cured bonding agent(Silux bonding agent. 3M Dental). Ten minutes after theluting agent was mixed, cotitonring was carried out asdescribed for group D.

Finishing procedures

One inlay was placed and contoured in each of two pa-tient appointments. During a third appointment, diskswere used to finish the restorations. The final smooth-ness was achieved using a polishing paste {PrismaGloss. Caulk) applied with a rubber cup. after which theteeth were treated with a topical fluoride.

Evaluation procedures

Polyivinyl siloxane) impressions of each restorationwere obtained at baseline (President. Coltene) and at

each recall assessment {Extrude, Kerr) at 6 months to 1year, 2 years, and 3 years.

Epoxy replicas (Stycast 1292. Emerson & Cuming}were prepared for examination with a scanning electronmicroscope (Amray 1810, Amray), and the occlusal sur-faces were analyzed quantitatively according to themethod described by Lutz.'' Eor technical reasons, onlythe occlusal margins were evaluated. The interfaces be-tween the enamel and luting agent and between the lut-ing agent and porcelain were scored separately.

The evaluation criteria used were continuous marginand gap. Continuous margin de.scribes a perfect adapta-tion of the luting agent to enamel or porcelain, and gapis defined as a marginal opening between the lutingagent and enamel or porcelain. An example of the crite-rion continuous margin is .shown in Eig la and one of thecriterion gap is shown in Eig lb. Additionally, ,/7CÎ,Î/Ï{marginal excess of luting agent), wear (cementationspace is not filled up to the vertical level of the enameland the restoration), enamel fraciure (microscopic frac-ture of enamel prisms adjacent to the margin), and inlayfracture (microscopic fracture of porcelain adjacent tothe margin) were recorded. Figure lb shows an exampleof wear and Eig 1 c shows one of flash.

The entire occlusal margin was followed with theSEM, and each section appearing on the screen wasevaluated at x2üO magnification. The distribution of theabove-defined criteria was estimated in percentages foreach image. Eor each restoration, the distribution wascalculated as the mean of the values determined foreach screen. The mean and standard deviation were thencalculated for the restorations of each experimentalgroup. Statistical analyses were canied out with Kruskal-Wallis and Wilcoxon tests.

576 Volume 29, Numbet

Zuellig-Singer/Bryant

cu BaselineESI 6 mo-1 y

Fig 2 Mean percentage ol continuous margin (CM) at theenamel-luting agent interface for ihe four luting agent groups ateach assessment. (The value for one standard déviation is shownon each bar.) The line above the bars connects the baseline re-sults that are significantly different among the four groups. (D)Dual Cement; (0) Duo Gement: (G) GR inlay cement; (K) Ketao-Cem.

Luting agent

Fig 3 Mean peroentage of oontinuous margin ¡CM) at the lutingagem-porcelain interface lor the lour luting agent groups at eachassessment. (The value lor one standard deviation is shown oneach bar.) The lines below some bars Indicate a signilicant differ-ence between ihe baseline assessmenl and the 3-year reoall. (D)Dual Cement; (0) Duo Cement; (C) GR inlay cement; (K) Ketac-Cem.

Results

The number of inlays evaluated and the distribution ofpremolars and molars at baseline and at each recall isshown in Table 2. Fotiy-two inlays in 21 patients wereevaluated at baseline, and 33 inlays in 18 patients wereavailable for evaluation at the 3-year recall. At ?, years,tbree of the original 21 patients could not be contacted.In addition, one restoration in a molar of group C, in apatient who did retum, had been replaced with compos-ite after being assessed at the 2-year recall as havingsuffered bulk fracture.

The results of the SEM analysis at the enamel-lutingagent interface are shown in Fig 2. For example, at base-line, the restorations of group D showed a continuousmargin on 90.5% of the entire occiusal margin at theenamel-luting agent interface, meaning that 9.5% of theocciusal margin presented a gap. At baseline, there was astatistically significant difference (Kruskal-Wallis: P <.01) among the luting agents. Group C restorations ex-hibited the highest percentage of continuous margin, andgroup D restorations exhibited the lowest. After 3 years,the differences among the groups were not statisticallysignificant, nor were the differences between the baselineand 3-year results for each group. Tbe mean percentageof continuous margin for all luting agents combined was95.8 ± 6.2 at baseline, 97.7 ± 3.8 after 6 months to 1 year,97.0 ± 3.7 after 2 years, and 94.6 ± 6.0 after 3 years.

The results for the luting agent-porcelain interfaceare presented in Fig 3. Differences among tbe lutingagents at baseline and at 3 years were not statisticallysignificant. Compared with the adaptation at baseline,the loss of adaptation at the luting agent-porcelaininterface after 3 years of function was significant for

Fig 4 Mean percentage of flash for the four luting agent groupsal each assessment. (The value for one standard deviation isshown on each bar.) (D) Dual Cement; (O) Duo Cement: (C) CRinlay cement, (K) Ketac-Cem.

each group (Wiicoxon, F < .05, for group D restorationsand Wiicoxon, P < .01, for group O, C, and K restora-tions}. The mean percentage of continuous margin forail luting agents was 98.7 ± 2.5 at baseline, 98.6 ± 2.0after 6 months to I year. 96.4 ± 3.1 after 2 years, and94.1 ± 4.4 after 3 years.

In Fig 4 the mean percentages of flash (marginal ex-cess of luting agent) are shown. For example, at base-line, tbe restorations of group D exbibited flash on23.0% of the assessable margins. At baseline, group Dbad the highest percentage, followed by groups O(12.8%). C (11.3%), and K (8.0%). After 3 years, groupD still showed the most flash (22.9%), followed bygroups O (10.8%), C (6.5%), and K (2.8%). The differ-ences among the luting agents were not statistically sig-nificant, nor were the results between baseline and 3years for each group.

577

Zueilig-Singer/Bryant

Luting agent

Fig 5 Mean percentage of weai (or the four tuting agent groupsat each assessment. (Tlie value for one standatd deviation isshown on each bar] Lines above some bais connect baseiine or3-year recali results that are significantiy différent among the tourgroups. The iines below some bars indicate a significant differencebetween the baseline assessment and tiie 3-year recaii. (D) DuaiCement: (O) Duo Cement: (C) CR iniay cement: (K) Ketao-Cem.

Figs 6a to 6c Exampiesof tiie SEM findings of grcup D restorations. (E) enamei; (P) pofceiain; (L) iuting agent

Fig 6a At baseiine, showing a continuous margin Fig 6b At 3 years, with evidence of wear facet (arrow).

Fig 6c At 3 years, showing a continuous margin aid mini-mai wear in an area without occiusai contact.

The mean percentages of wear are shown in Fig 5.Wear was rare at baseline, except for group K with13.7%. After 3 years, the wear of tbe four luting agentswas significantly different (Kruskal-Wallis: P < .001).Tbe least wear occurred with group D inlays, for whichthe luting agent was a microfilled eomposite. Group Cmlays exhibited most wear of the luting agent, followedby group K mlays. The increase of wear after 3 years offunction was signiHcant for the restorations in group OfWilcoxon: P < .01). group C (Wiicoxon: P < 01} andgroup K (Wiicoxon: P < .05}.

Bnamel and inlay fractures occurred in fewer thati1% of the assessable margins and appeared to occurrandomly across all four groups. Tberefore furtheranalysis ot these two criteria was omitted.

Some examples of the SEM findings are shown inFigs 6 to 9.

578Voiume 29, Numbe 9, 1938

Zueiiig-Smger/Bryant

Figs 7a and 7b Examples of the SEM findings oí group O restorations. (E) enamei, (P) porceiain; (L) iuting agent.

Fig 7a At baseiine. siiowing a continuous margin Fig 7b At 3 years, same aiea as in Pig 7a, showing a con-tinuous margin.

Figs 6a and 6b Exampie' of the SEtvl findings of group C restorations (E) enamel; (P) porcelain; (L) iuting agent.

Fig 8a At baseline, showing a continuous margin Fig 8b At 3 years, same area as in Pig 8a. showing a con-tinuous margin and evidence ol wear.

Discussion

The long-term success of a Cerec restoration, as formost indirectly fabricated restorations, depends mainlyon its most vulnerable component, the luting agentlayer, and its adaptation to enamel and porcelain.^ Toevaluate the quality of this layer, several criteria wereselected. First, as a measure of the integrity of the mar-ginal adaptation, the criterion continuous margin wasevaluated at both the enamel-luting agent interface andat the luting agent-porcelain interface. A continuousmargin provides the best opportunity for optitnal bond-lno to enamel to minimize microleakage. Because of the , ,^ r.r-„M- .^•' " e ^" ciiaiin-i •>-• . ", L • Fig 9 Examples of the SEM lindings of group K restora-elastic deformation of the tooth and the restorative ^^^^^ ^ ^ enamei, (P) porceiain; (L) iuling agent. At 3 years,material during function, even a minute gap at the same area as in Fig la, showing a continuous maigin andenamel-luting agem interface can widen to a critical evidence of wearsize, allowing bacteria to enter the gap.'-* This results in

579

Zueilig-Singer/Bryant

microleakage and can eventually progress to caries.Leakage at the luting agent-ptncelain interface can re-sult in marginal discoloration, but damage to tooth sub-stance will not necessarily follow. Therefore, it is postu-lated that, in particular, the enamel-luting agentinterface should have a high percentage of continuou,smargin.

Second, the criterion wear was assessed to determinethe wear of the luting agent. Ideally, the luting agentshould form a stable continuity between enamel andporcelain, filling the cementation space up to the clini-cal surface of the enamel and the porcelain. If there isverticai loss of the luting agent because of wear, bacte-ria can easily accumulate in the cementation space, pro-tected from oral hygiene. It was anticipated that thecomposite luting agents would show les.s wear than theglass-ionomer cetnent."

Finally, the percentage of marginal excess, or Hash,was measured. The complete removal of flash reducesirregularities and roughness and thereby bacterial accu-mulation at the margins.

Initially, all luting agents provided very good mar-ginal adaptation, with 90% to 99% continuous margin atthe enamel-luting agent interface and 98% to 100% atthe luting agent-porcelain interface. Etching of theenamel and conditioning of the porcelain resulted in anexcellent bond by the composite luting agents, as de-scribed by Buonocore'" and Fett et al.' Glass-ionomercement appeared to adhere satisfactorily to enamel andporcelain without etching of the enamel or silanizationof the porcelain. Initially, the relatively coarse CR inlaycement showed the best marginal adaptation at theenamel-luting agent interface.

After 3 years, all luting agents exhihited reliable ad-hesion to the enamel margin with 93% to 96% continu-ous margin. There was no significant difference amongthe luting agents.

At the luting agent-porcelain interface, the loss ofadaptation for all luting agents was significant at 3years and there was 92% to 96% continuous margin.The finding that the composite luting agents (as recom-mended by Thurmond et al'* and Fett et al"), used incombination with their eorresponding silane couplingagents, did not produce a more durable adaptation toporcelain than the glass-ionomer cement differed fromthe observation made by Âberg et al.''' They comparedfired porcelain inlays luted with composite or glass-ionomer cement. After 3 years, the clinical durability ofthe composite luting agent was significantly superior tothat of the glass-ionomer cement. Van Dijken andHörstedt'-' used an SEM evaluation technique and alsofound a significantly superior marginal quality for theinlays luted with composite after I year. In both of these

clinical studies, Fuji T glass-ionomer cement (GCl wasu.sed. In the present study, the restorations were lutedwilh Ketac-Cem Maxicaps (ESPE). The improvedbonding might be a result of the consistent optimummix achieved by encapsulation of the glass-ionomer lut-ing agent.

The marginal adaptation found in the present study iscomparable with other findings in the literature for resinluting agents. In a study by H(Jrzeler et al,-" the mar-ginal adaptation of Ceree inlays cemented with compos-ite luting agent was examined after 3.5 years of func-tion. At the enamel-luting agent interface and at theluting agent-porcelain interface, 97% and 85% of themargins, respectively, were continuous. Hofmann et al-'obtained somewhat comparable results after 5 years.They reported 69% and 74.5% continuous margins,respectively, for the above-mentioned interfaces. Simi-lar results were found for composite inlays in an in vivostudy by Krejci et al.-- After I year, the SEM evaluationrevealed 96.7% continut)us margin at the enamel-lutingagent interface and 94,9% at the luting agent-compositeinterface. Krejci et al-* evaluated Empress inlays(Ivoclar) cemented with composite luting agent after 1.5years of clinical function. Only 66.8% of the enamel-luting agent interface and even less of the other inter-face were identified as continuous.

In the present study, despite meticulous finishing,some excess luting agent remained. Luting agents Dand O had good color match and fine texture and, as aconsequence, marginal excess was barely visible duringfinishing. Luting agents C and K were more opaque andrelatively coarse, and remaining excess was observedless frequently in SEM evaluation. The high percentageof flash in group D and group 0 restorations may havecontributed to the relatively low percentage of con-tinuous margin at baseline. Marginal excess might ad-here less to the surface enamel adjacent to the eavitymargin because of incomplete etching and applicationof enamel bonding. The small decrease in marginal ex-cess after 3 years of function did not approach statisticalsignificance.

Some restorations, particularly those with the glass-ionomer luting agent, exhibited wear at baseline.Polishing of the restorations resulted in more abra.sionof the glass-ionomer cement than of enamel or porce-lain, despiie the fact that polishing was carried out 1week after cementation to allow complete setting of thecement.

The microfilled composite luting agent exhibited nosignifieant increase in wear alter 3 years of function. Incontrast, the two hybrid composite luting agents and theglass-ionomer cement revealed a significant loss (wear)of the layer uf the luting agent after 3 years. It is possi-

580 Voiume 29, Number 9,

Zuellig-Singet/Bryant

ble that the relatively high frequency of marginal excessfor the microfilled composite might have provided pro-tection for the layer of luting agent.

The fitidings for wear in the present study corre-spond with the findings in the literature. Krejci et al--"tested porcelain inlays in an in vitro study and measuredthe wear of a microfilled and a fine hybrid compositethai served as luting agents. The wear of the luting com-posites at the end of the loading test, which corre-sponded to approximately 5 years of clinical use, rangedfrom 4.9 un for the microfilled composite to 12.3 ¡amfor the fine hybrid composite. Kawai et al- concluded,from an iu vitro study, that a microfllled composite wasthree to four times more wear resistant than a hybridluting agent. Isenberg et al- evaluated the depth and thewidth of the interfacial gap for two hybrid and one mi-crofilled composite luting agents in Cerec restorations.After 3 years, the microfilled composite showed lesswear than the two hybrid composites. This conclusion isin agreement with other in vivo and in vitro studies.-'-"Suzuki et al- tested various luting agents for differentinlay materials. The glass-ionomer cement exhibitedmore wear than the composite luting agents after an invitro wear test.

The use of quantitative marginal analysis is a widelyaccepted method of assessment for describing the mar-ginal performance of dental restorations,'-'--'"The mar-gin of a restoration shows a wide diversity of morphol-ogy. The behavior of a restoration at its margin istherefore an important parameter in predicting thelongevity of the restoration.^" In this study, only the oc-clusal margins could be analyzed because of technicallimitations associated with taking impressions of the ap-proximal margins. A further limitation in this type of as-sessment can be the cleaning of the surface prior to tak-ing the impres.sions. Marginal gaps may be fllled withdebris and improperly rated as continuous margin,--'Margin analysis using an SEM does not allow a deter-mination of the amount of wear of the luting agent.Only a qualitative observation was possible: assessingwhether the cementation space was filled up to the levelof the enamel and the porcelain or whether the marginshowed wear or marginal excess.

Generally, the statistical analysis did not revealstrong signiflcance among the different luting agentsbecause the results were widely dispersed, resulting inhigh standard deviations. In vivo studie.s often showsuch an outcome.

Conclusion

The SEM findings of this study indicated that CerecTestorarions exhibit consistently high-quality occlusal

margins after 3 years, regardless of the luting agent.Thus, Cerec restorations can be recommended to pa-tients. The marginai adaptation is more durable at theenamel-luting agent interface than at the luting agent-porcelain interfaee.

Although there was no difference in marginaladaptation among the luting agents used in this study, amicrofilled composite luting agent exhibited less evi-dence of wear. Based on these 3-year findings, a micro-filled composite luting agent appears to be the pre-ferred choice for luting the restoration. Eurtherdevelopments may optimize the wear characteristics ofthe composite luting agents and may strengthen thebond between the porcelain and the composite by amore stable silanization.

Acknowledgments

The authors would like to acknowledge the stipport received from ad-ministration and clinical staff at Ihe United Dental Hospital, Sydney,and the School of Dental Medicine. Zurich. The assistance of DrMengini provided with the statistics is gratefully appreciated.

References

1. Sheth PJ, Jensen ME, Sheth JJ, Cotnparative evaltiation of threeresin inlay techniques: Microleakage studies. Quintessence Int1989:20:831-836.

2. Shortall AC, Baylis RL, Baylis MA, Grundy JR, Marginal sealcomparisons between resin-bonded Class 11 porcelain ¡niays,posterior composite restorations and direct composite inlays. IntJ Prosthodont 1989:2:217-223.

3. Bergmann M. Side-effecls of amalgatn and its alternatives:Local, systetnic and environtnental. Int DentJ 1990:40:4—10.

4. Batimann MA, Heidemann D. Biocompatibility of dental inlayceratnics. In: Mörmann WH led|. State of the Art of the Ccrec-Method. International Symposiutit on Cottiputer Restorations.Berlin: Quintessence, 1991:373-37(5,

5. Craig RG. Restorative Dental Materials, ed 8. St Louis: Mosby,1989:4^1-498.

6. Mörmann WH, Brandestini M. Lutz F. Barbakow F, Götsch T.CAD-CAM eeratnic inlays and onlays: A case report after 3years in place, JAm Dent Assoc 1990:120:517-520.

7. Mörmann WH, Brande.stini M, Lutz F, Barbakow F. Chairsidecompuler-aided direct ceramic inlays. Qtiintessence Int1989:20:329-339.

8. Mörmann W. Krejci I. Clinical and SEM evaluation of Cerec in-lays after 5 years in situ. In: Mörmann WH led]. State of the Artof the Ccrec-Melhod. Intemational Symposium un ComputerRestorations. Berlin: Quinlessence, 1991:25-32.

9. Geppert W, Roulet J-F. In vivo marginal integrity of MOD Dicorinlays luted with adhesive techniqties labstract 48|. J Dent Res1986:65:731,

10. Lambrechts P, Inokoshi S, Van Meerbeek B, Willetns G. BraemM, Vanherle G. Classification and potential of ctimposite lutingmaterials. In: Mörmann WH (ed). State of the Art of tbe Cerec-Melhod. Inlernationai Symposiutn on Computer Restorations.Berlin: Qiimtesscnce, 1991:61-90.

581

Zuellig-Singer/Bryant

11. Muzynski BL, Greener E, Jameson L, Maione WE Fluoride re-lease troni glass ionomers used as luting agents. J Pronlhet Denl

12. Melnnes-Ledoux PM, Ledous WR, Weinberg R. A bond strengihstudy of luted cüstabIe ceramic restorations. J Dent Res I989;ö8:823-825.

13. Lutz F. BeitrSge zur Entwicklung von Seitenzahn-Komposits.Zürich: KAR PAR P2M Verlag, 1980:')/6.

14. Jensen ME, Chan CDN: Polymerization shrinkage and micro-leakage. [n: Vanherle G, Smith DC {cd]. International Sympo-sium on Posterior Composite Re.sin Dentul Restorative Mate-rials. Utrecht. The Netherlands: Peter Szulc, 1985:243-262.

15. Van Dijken JWV, Hörstedt P. Marginal breakdown of Hred cera-mic inlays cemented with polyalkeroate (ionomer] cement orresin composite. J Dent 1994;22:265-272.

16. Buonocore MG. A simple method of increasing Ihe adhesion ofacrylic filling materials to enamel surfac:e. J Dent Res 1955;34:849-853.

17. Fett HP, Mörmann WH. Krejci I, LML F, The effects of shortbevels and siianization on marginal adaptation of computer-tna-chined mesio-occlusodistal inlay.s. Quintessence Int 1991;22:823-829.

18. Thurmond JW, Barkmeier WW. Wilwerding TM. Effect of por-celain surface treatments on bond strengths of composite tesinbonded to porcelain. J Prosthet Dent 1994;72:355-359,

19. Aberg CH. Van Dijken JWV, Olofsson A-L. Three-year compari-son of fired ceramic inlays eemenled with composite resin orglass ionomer cement. Acta Odontol Scand 1994;52:140-149.

20. Hiirzeler MB, Fell H, Mörmann W. Marginal adaptation ofCerec-CAD-CAM inlays after 3.5 years. In; Mörmann WH (ed).State of Ihe Art of iJie Cerec-Method. International Symposiumon CotnpuKr Restorations. Bedin: Quintessence, 1991:417-^23.

21. Hpfmann N. Popp M. Klaiber B. Klinische und rasterelektro-mikro.îkopische Nachumersuchung von Ccrec-lnlays nach fünfJahren Liegedauer Disch Zahnärztl Z I995;50;835-839,

22. Krejel I, Güntert A, Lutz F. Scanning electron microscopic andclinical examination of composite resin inlays/onlays up to 12months in .situ. Quintessence Int i 994;25:403^09.

23. Krejci I, Krejci D, Luiz F. Clinical evaluation of a new pressedglass-ceramic inlay material over 1.5 years. Quintessence Inlt992;23:l8l-l8ö.

24. Krejci I, Lutz P, Reimer M, Hein:!mann JL. Wear of ceramic in-lays, their enamel antagoni.sts, and luting cements. J ProsthetDem I993;69:425^3O.

25. Kawai K, Isenberg BP. Lemfeider KF. Effect of gap dimensionon composite resin wear. Quinlessence Int 1994; 1:53-58.

26. Isenherg BP, Essig ME, Leinfclder KR Three-year clinical eval-uation of CAD/CAM restorations. J Esthel Dent Í992;4173-176.

27. O'Neal SJ. Miracle RL, Leinfelder KE. Evaluating interfacialgaps for esthetic inlays. J Am Dent As,soc 1993; 124:48-54.

28. Shinkai K. Suzuki S. Leinfelder KF, Katoh Y. Effect of gap di-mension on wear resistance of luting agents. Am J Dent 1995;8:149-151.

39. Suzuki S. Lcinfelder KF. Shinkai K. Wear resistance of resin ce-ments. Am J Dent I995;8;83-87.

30 Roulet J-F. Reich T, Biunck Ü. Quantitative margin analysis inthe scanning electron micro,scope Scanning Microsc 1989;3:147-158.

; Í ^ ^

Answers to QI 6/98 Questions

^*=^==;^ 1.2.3.4.

DCDA

5.6.7.8.

BDBA

9.10.11.12.

CDDB

13.14.15.16.

DDBC

582Volume 29. Nürnberg, 1993