Prosthodontics

Comparison of luting cements for minimaiiy retentivecrown preparations

William D. Browning, DDS, MSVStephen K. Nelson, DMD /Roman Cibirka, DDS, MSVMichael L. Myers,

Objective: The purpose ol this study was to compare the retentive strengths of resin, giass-ionomer, andzinc phosphate cements under adverse conditions. Method and materials: Thirty extracted teeth weremounted and prepared in their long axis. The axial wail height was 3 mm and the convergence angie was28 degrees. These conditions increased the role ot the cement and decreased the role of the preparationin providing retention of the casting. The axiai surtace area was determined. Copings were tabricated witha ring aligned in the long axis to facilitate removal of the crown. They were cemented with a resin cement,a glass-ionomer cement, or a zinc phosphate oement. A block randomization scheme was used to assigncements so that the mean surface areas of the teeth were equivalent in all groups. The copings wereloaded in tension, and the amount of force required to remove the coping was recorded. The stressrequired to remove the coping was oaioulated. Results: The mean stress required to remove the copingswas 9.4, 5.0, and 3.1 MPa for the resin, glass-ionomer, and zinc phosphate cements, respectively.Conclusion: The resin cement group was signifioantly stronger than both the glass-ionomer cement andthe zinc phosphate cement groups. The glass-ionomer oement was significantly stronger than the zincphosphate cement, (Quintessence int 2002:33:95-100)

Key words: glass-ionomer cement minirrîally retentive crown, resin cement, retention, zinc phosphate

cement

CLINICAL RELEVANCE: When the axia! wall height orthe convergence angle is unfavorable, it is likely that aresin cement will resist dislodgment better than will amore traditional cement.

The preparation of a posterior tooth to receive acrown has been shown to expose 1 to 2 million denti-nal tubules,' leaving the tubules patent and facihtattngmovement of fluid within the tubules. Investigationhas shown that provisional restorations often allowiiiicroleakage.2 TÍIUS, when chemical or temperature

'Associate Professor, DeparlmenI of Oral Rehabilitation, Medical College

of Georgia, Sctiool ot Denlistry, Augusta, Georgia.

^Associate ProOssor, Department of Oral Rehabilitation. Medical Cdleg.

of Georgia, School ot Dentistry, Augusta. Georgia.

^Professor, Departrnert of Orai Rehab.litatior,, Medical College ot Georgia,

School of Dentistry, Augusta, Georgia

Dentistry, t120 15th Street, Augusta. Georgia 30912-1260.

wbrowni n©mail.mcg.edu

gradients are created between the prepared tooth andthe oral environment, fluid movement in the tubules isstimulated. The result is postoperative pain. In addi-tion, bacteria and their toxins, as well as chemicalsrelated to both provisional and permanent cements,gain access to these open tubules.'

According to the theori' of pulpal hydrodynamics,the use of materials and techniques capable of sealingthe dentina! tubules after tooth preparation shouldreduce postoperative pain. The use of resin liners inconjunction with amalgam restorations has beenshown to reduce postoperative sensitivity clinically*and have the potential to be successful in conjunctionwith crown preparations. If anecdotal reports basedon tbe observations of clinicians using this techniqueare correct, resin liner placement following crownpreparation does reduce postoperative pain.

Initially, the effects of using a combination of a resinliner and traditional cement for the retention of crownswere unknown. Subsequent research found vei ; definiteinteractions between resin liners and some of the mate-rials used for crown cementation.^* Some material com-binations resulted in no loss of retention,« while othercombinations exhibited dramatic, negative effects."

95

Browning et

In light of these unpredictable interactions, the useof resin cements tnanufactured to be compatible withspecific adhesive liners became attractive. Subsequentresearch indicated that this type of cement has theclinical potential to provide more retention and conse-quently less postoperative sensitivity caused bymicroleakage.^ Questions remain about the perfor-mance of resin cements, however. Comparison of theretcnfion provided by resin cements to that providedby more traditional products has yielded inconsistentresults."'" Studies investigating the same productshave yielded different results. Not only the valuesachieved but also the relative rankings of identicalproducts have varied from study to study.''''"-'^

Differences in study design make it difficult todirectly compare the results of different studies andthe different products examined. Some research hasnot taken the surface area of the prepared tooth intoconsideration,''- although other studies have."'"Measuring stress rather than the force required toremove the casting controls for differences in toothsize. This also facilitates the comparison of dataamong studies. Failure to consider surface area notonly makes such comparisons difficult, but also intro-duces a potential for important bias. Without effectiverandomization or a design that assures that the meansurface area per group is equivalent, the resultsobserved may be the result of differences in the aver-age size of the teeth assigned to each group ratherthan the cement used.

The height of the walls and the convergence anglealso play a significant role in the retention of crowns.Shorter walis and walls with greater taper increase therole of the cement in retaining the coping and aremore likely to demonstrate significant differencesamong cements. In contrast, longer walis with lesstaper raise the mean stress required to remove castingsfrom a group of samples.'^ At present, there is no stan-dardization of wall height or convergence angle fromstudy to study.

Finally, the length of time specimens are stored,after cementation and before removal, has differedfrom study to study as well, from a low of 24hours'"'^ to a high of 7 days."-'"

The purpose of this study was to compare the per-formance of a resin cement, a glass-ionomer cement,and a zinc phosphate cement under adverse condi-fions. The specimens used in this study were preparedwith both excessive axial taper and short axial walls.These conditions increased the role of the cement anddecreased the roie of the preparation in providingretenfion of the casting against a tensile force.

The hypothesis that cement type is significantlyassociated with the mean amount of stress required toremove cemented copings was tested against the

hypothesis that there is no difference among '' '^cements. The significance level used was 5°ifi. in addi-tion, the mode of failure for each cemented castingwas recorded.

METHOD AND MATERIALS

Preparation of specimens

Thirty freshly extracted teeth were cleaned, inspectedfor fracture or other damage, and mounted in clear0.875-inch-diameter polyvinyl chloride tooth sleeveswith an autopolymerizing acrylic resin (OrthodonticResin, LD Caulk). The roots of the teeth were notchedto resist dislodgment from the sleeve when the speci-men was tested in tension. A brass alignment blockwas used in conjunction with a polishing device(Ecomet Polisher/Grinder, Beuhler) to assure thatboth ends of the polyvinyl chloride tooth sleeve wereat right angles to the length of the sleeve. Theextracted teeth were mounted in their long axis, andthen the combination of brass alignment block andgrinder was used to flatten the occlusal surface by wetgrinding with 240-grit silicon carbide paper. This pro-cedure ensured that when the sleeves were placed ona fiat surface the teeth would be vertical, the occlusalsurfaces would be borizontal, and cast rings placed ata right angle to the occlusal surface would parallel thelong axis of the teeth.

The specimens were placed on the flat surface of asurveyor table and prepared with a high-speed hand-piece. The surveyor was modified to hold the hand-piece in a fixed position so that the diamond bur usedfor preparation was parallel to the long axis of eachtooth. The teeth were prepared for complete crowns.Water irrigafion and a medium-grit diamond (852-037,Brasseler) were used to produce preparations withaxial walls that had 14 degrees of taper per wall, pro-viding a convergence angle of 28 degrees. Specimenswere positioned so that the axial wall height was stan-dardized at 3 mm. Each bur was used to cut only twopreparations.

Fabrication of copings

Impressions of the prepared teeth were made with reg-ular and light-bodied polyvinyl siloxane; a 1-oz plasficmedicine cup was utilized as the impression tray. Theimpressions were poured in type V dental stone to cre-ate working dies. Three coats of die spacer and a thinlayer of cyanoacrylate were applied to each die beforewaxing. The copings were waxed with a flat occlusalsurface that paralleled the occlusal plane of the pre-pared tooth.

96 Volume 33, Number 2, 2002

Browning et ai

The waxed copings were invested witfi the hygro-scopic technique and cast in type 111 gold (HarmonyHard, WilHams), The castings were divested and theinternal surface was air ahraded with 50-iim alu-minum oxide particles used under a pressure of 60 psi.Finally, a ring, cast in the same alloy, was soldered at aright angle to the occlusal plane of the copings-

Determination of axial surface area

A thin replica of each occlusal surface was formed byplacing an autopolymerizing resin (Castolite, Beuhler)in the occlusal portion of the impression of eachpreparation. Once set, the replicas were removed fromthe impressions, A permanent-marking pen was usedto highlight the perimeters with a thin black line,images of these occlusal replicas were digitized, alongwith a circle of known circumference, on a flathedscanner (Arcus II 36 hit, AGFA), If the convergenceangle, the axial wall height, and the circumference ofspecimens are known, the axial surface areas can becalculated-'

The scanned images of each occlusal surface wereidentified hy specimen number and imported into soft-ware designed to organize and store them (ScionImage PC. Scion Image Software). From the scanned¡mages of the control and of each occlusal surface, thelength of the circumference of each specimen wasdetermined with specialized software (NIH ImageSoftware V1.6, NIH), The computer utilized was aGateway 2000 laptop interfaced with the scannerthrough an external adapter (SCSI Adapter. Adaptec),These data were used to calculate and record the sur-face areas in square milhmeters.

Luting of copings

Castings were checked for seating discrepancies with adisclosing medium (Fit Checker. GC} and adjusted,where needed, to allow complete seating. Finally, theinterna! surfaces of the copings were cleaned with asteam cleaner set at 70 psi.

Specimens were arranged horn smallest to largest,according to axial surface area. Randomization wasaccompiished by first dividing the specimens into 10blocks of three specimens each. Next, each of thethree cements was assigned at random to one of thethree teeth in each block. The cements used were aresin cement (Permalute. Ultradent), a glass-ionomercement (Ketac-Cem Maxicap, ESPE), and a zincphosphate cement (Fleck's, MizQ')-

To avoid the introduction of bias, these same 10bJocks of three specimens were used to assure thatcementation of the casting was also performed ran-domly. This was done because of concerns that one

material might suffer if all specimens were cementedfirst, another gain fi-om being in the middle, and thethird suffer hecause of operator fatigue. This distribu-fion created three groups of 10 specimens each.

One operator cemented all of the copings.Materials were mixed in strict accordance with theprotocol set forth by each manufacturer. The glass-ionomer cement was mixed with an amalgamator(CAPMIX, ESPE) following the manufacturer's rec-ommended trituration time.

Copings were seated with a specialized device thathad been machined at tbe Medical College of Georgiafor that purpose. This device allowed a known load tobe applied along the long axis of tbe teeth. Specimenswere loaded at 5 kg for a period of 10 minutes. Excessluting agent was removed from the margins before itbecame fully set. All specimens were then stored indisfilled water for 7 days at 37°C.

Testing of tensile bond strength

The same device that was used to apply the load dur-ing cementation of the copings was used to provideproper alignment of the force used to dislodge thecopings, directed in tension without a rotational com-ponent. Testing was done with a materials-testinginstrument (Instron) unit used at a crosshead speed of1,27 mm/min. The copings were loaded in tensionuntil tbey were dislodged or suffered a fracture. Thepeak load, in newtons, was recorded.

In addition, the copings and axial walls of thepreparations were examined under x3,5 magnificationto determine the mode of failure. Four failure modeswere defined:

1. Root fracture was defined as cohesive failure of thedentin while the coping remained luted in place.

2. Failure at the metal-cement interface w'as recordedwhen a clear preponderance, greater than 75%, ofcement remained on the axial wail of the tooth.

3. Failure at the dentin-cement interface was recordedwhen a clear preponderance, greater than 75^0, ofcement remained on the internal metal surface ofthe casting,

4. If neither the dentin nor the casting had a clearpreponderance of retained cement, ie. less tban75''/o of cement was retained, the failure was classi-fied as mixed.

Statisticai analysis

A one-way analysis of variance (ANOVA) procedurewas used to determine ii tbe type of cement used wassignificantiy associated witb tbe stress required toremove the copings, A multiple comparison test,

97

Browning et ai

il Pa)

Stre

ss (

10

8

6

4

2

0

A

Permaiute1

Ketac-Cem

C•Fi eck's

FIg 1 Mean stress values required to dislodge dental castingstested with each luiing cement. A, B, and C, indicate groups thaiditfer from one another signiticantly (analysis of variance andSfudent-Newman-Keuis all-pairmise multipie comparisons proce-dure). The error bars indicate slandard deviations. The samplesize was 10 per grcup.

TABLE 1specimens

Cement

PermaiuteKetac-CemFleck's

Modes of cement failure in coping

RootIracture

700

Modes of tailure

Metal-cement Dentin-ce mentintertace interface

3 D0 S0 9

Mixed

021

Student-Newman-Keuls all-pairwise test, was used todetermine if there were significant differences amongthe three cements. The significance level was set at 5%.

RESULTS

The block randomization scheme used to assign thespecimens to the three groups produced three groupsthat were very similar in mean axial surface area (one-way ANOVA; P = ,99), The mean surface areas for theresin, glass-ionomer, and zinc phosphate groups were91,1, 91.8, and 92,1 mm , respectively.

For each specimen, the area of the axial surface andthe force required to dislodge the casting were used tocalculate the stress required to remove the casting.The means (± standard deviations} were 9.4 (± 2,4),5,0 (± 1,5), and 3.1 (+ 0.9) MPa for the resin, glass-ionomer, and zinc phosphate groups, respectively (Fig1), There was a statistically significant associationbetween the type of cement used and the stressrequired to remove the coping (one-way ANOVA;P<,001),

The data were also analyzed to determine if therewere significant differences among the three cements.The mean tensile bond strength of the resin cementwas significantly higher than that of both the glass-ionomer cement and the zinc phosphate cement [P <.001; one-way ANOVA and Student-Newman-Keulsall-pairwise multiple comparisons test). The mean ten-sile bond strength of the glass-ionomer cement wassigniflcantly higher than that of the zinc phosphatecement (one-way ANOVA and Student-Newman-Keuls all-pairwise multiple comparisons test; P < ,02),

Seven of the 10 specimens luted with the resincement fractured before the coping was dislodged.Accordingly, the true mean for this group of speci-mens is higher than the mean calculated but in actual-ity is unknown. Because the three groups were equiva-lent in mean surface area, a comparison of the meanforce required to produce failure (either failure of thecoping to resist being dislodged or a fracture of theroot) can be compared. The mean force was 843.1,448.5, and 286,6 N for the resin, glass-ionomer, andzinc phosphate groups, respectively. There was a sta-tistically significant association between the type ofcement used and the mean force applied at failure(one-way ANOVA; P < .001), ITie force at failure wassignificantly higher for the resin group than either ofthe other two cements (one-way ANOVA andStudent-Newman-Keuls all-pairwise multipie compar-isons test; P < ,001), The glass-ionomer group had asignificantly higher mean force at failure than did thezinc phosphate group (one-way ANOVA and Student-Newman-Keuls all-pairwise multiple comparisons test;P<,02).

The modes of failure by cement type are reported inTable 1. Failure arising from root fracture and failureat the metal-cement interface were found only in theresin cement group. The glass-ionomer and zinc phos-phate groups failed predominately at the dentin-cement interface.

DISCUSSION

Observation of the data suggests that the mode of fail-ure for the resin cement differed from that of the glass-ionomer and zinc phosphate cements. A chi-squareanalysis demonstrated a significant associationbetween the type of cement used and the mode of fail-ure (Pearson chi-square analysis; P < ,001},

Differences among studies in terms of the outcomemeasured (force or stress), the convergence angleused, the axial wall height, and the treatment of speci-mens after preparation makes comparison of resultsdifficuh. However, studies such as this one are usefulto the profession only to the extent that the results can

Volume 33. Number 2. 20D2

Browning et al •

be applied to clinical practice. Likewise, relating tbeperformance of one product to tbat of anotber isimportant- Tbe present study is similar to two previousstudies.^"* in that all three studies recorded stress astbe outcome of interest and all three inciuded Ketac-Cem as one of the cements studied. These two factorsprovide a means to relate the results of the presentstudy to the other two. Nevertheless, comparisonsmust be made with caution because, althougb the gen-eral effect of design factors sucb as wall bcight andconvergence angle are known, the exact extent towhich they effect tbe outcome measured is not.

In the study by Emst et al,'« tbe wall height andstorage time were the same as were used in the pres-ent study. The convergence angle, however, was 10degrees rather than 28 degrees. The greater conver-gence angle used in the present study should be asso-ciated with lower values rather than higher values'^for the same material. The mean for the Ketac-Cemgroup in that study was 2.36 MPa; the value was 5.0MPa in the present study. The present results appearto differ substantially from those of the previous study.The resin cement investigated in the study by Emst etal,'" F21, demonstrated only 25% of the strength ofKetac-Cem; in tbe present study, Permalute resincement demonstrated lSS' 'b of the strengtb of Ketac-Cem. In both studies, however, the mode of failure forKetac-Cem was predominately at the cement-dentininterface,

Tbe study by Jobnson et al tested both Ketac-Cemand Fleck's. In tbat study, the wail height was 4 mmratber tban tbe present 3 mm, and the convergenceangle was 20 degrees rather than 28 degrees. Therewere other differences as well. Specimens were ther-mocycled before mechanical testing, and a desensitiz-ing agent was used on one balf of tbe specimens testedand not on tbe otber half. The use of the desensitizerproved not to be a significant factor. The longer waifsand reduced taper should provide higher retentivestrengths for the same material,'^

In both the present study and the study by Johnsonet al,« the resin cement had the highest mean values,Ketac-Cem had the second highest, and Fleck's hadthe lowest. In botb tbe present study and the studyby Jobnson et al,« the differences among the threecements were significant. The resin cement tested mthe earlier study, Resinomer (Bisco Dental), demon-strated 128% of tbe strength of Ketac-Cem and 195%of tfie strength of Fleck's; in the present study, Perma-lute demonstrated 188% of the strengtb of Ketac-Cemand 303% of the strength of Fleck's.

In botb studies, tbe mode of failure for tbe resincements was predominately root fracture. For Heck s,failure was predominately at tbe dentin-cement inter-face In the study by Jobnson et al, failure for Ketac-

Cem witb dentinal desensitizer was at tbe metal-cement interface; tbe mode of failure in specimenswithout desensitizer was mixed. This result differsfrom that of the present study, in which Ketac-Cemfailure occurred predominately at the dentin-cementinterface.

In botb tbe study by jobnson et al' and tbe presentstudy, tbe increase in tensile bond strength acbievedwitb the use of a resin cement, beyond that availablewitb more traditional cements, is impressive. To theextent that inferences can be made about clinical per-formance from a laboratory study, these results indi-cated that the use of a resin cement can help to pro-vide adequate retention when chnical circumstancesmake it impossible to acbieve tbe desired wall beightand convergence angle. However, tbe proportion ofspecimens cemented with resin cement tbat failedbecause of root fracture introduces new questions;Will sucb high bond strengths lead to the fracture ofabutment teetb? Sbould the use of these cements berestricted to adverse clinical situations? Long-termclinical testing is needed.

CONCLUSION

Witbin tbe limitations of tbis experiment it can beconcluded tbat;

1. Tbe resin cement provided significantly betterretention tban did both tbe glass-ionomer and zincphosphate cements,

2. Tbe glass-ionomer cement provided significantlybetter retention than did tbe zinc pbospbatecement.

3. Failure resuhing from fracture of tbe root wasobserved only in tbe resin cement group.

REFERENCES

1, Richardson D, Pashley DH, Dentin permeability; eftect oícrown preparation, Int J Prosthodont 1991 ;4;219-225.

2, Langland H, Langland LK_ Pulp reaction to cro»Ti prepara-tion, impression, temporarj' croun fixation and permanentcementation, J Prosthet Dent 1965:15; 129-143.

3, Brännström M. Communication between the oral cavity andthe dental pulp associated with restorative treatment, OperDent 19a4;9;57-68.

4 Pashley DH, Smear layer-Physiological considerations,Oper Dent 1984;9(suppl J|;13-29.

5. BrowTiing WD, Johnson WW. Gregory PN. Reduction ofpostoperative pain: A douhle-hlinded, randomized clinicaltrial, J Am Dent Assoc 1997;128:1661-1667,

6, Johnson GH, Hazelton LR, Bales DJ. Crown retention withuse of a resin sealer on prepared dentin [abstract 1918].JDent Res 1996;75;257.

99

3rowning et ai

7. Mausner IK. Goldstein GR, Georgescu M, Effect of twodentinal desensitizing agents on retention of complete castcoping using four cements, J Prosthet Dent 1996;75:129-134.

8 Johnson GH, Lepe X, Bales DJ. Crown retention with use ofa 5% glutaraldehyde sealer on prepared dentin. J ProsthetDent 1998;79:671-676.

9. Gorodovsky S, Zidan O. Retentive strength, disintegrationand marginal quality of luting cements. J Prosthet Dent1992;68:269-274.

10. Ernst CP, Wenzel N, Stender E, Willershausen B. Retentivestrengths of cast gold crowns using glass-ionomer, com-pomer or resin cement, f Prosthet Dent 1998;79:472-476.

11. Swift EJ, Lloyd AH, Felton DA. The effect of resin desensi-tizing agents on crown retention, J Am Dent Assoc 1997;128:195-200.

12. EI-Mowafy OM, Fenton AH, Forrester N, Milenkovic M.Retention of metal ceramic crowns cemented with resincements: Effects of preparation taper and height J ProsthetDent 1996;76:524-529.

THE SINGAPORE INTERNATIONAL DENTALEXHIBmON & CONFERENCE ¡SIDEC 2002)

11-14 July. 2002 SingaporeThe Mandarin Singapore

Thfime Dentistry Ohâilen^BS in The New Millennium

SCIENTIFIC PROGRAMME

ParticipatiDji OoursesFuli Day PragrammasLimitad Atlenûance ClinicsTable ClinicsPoster Presentalla n

LECTURES & HONDS-ON/UMITEa ATTENDANCE

fteslhetic Dentistry Orthodnnlic!

Oental Oeramics Oral nehabilitalionDental Implants Oral Surgery

EndQdonlics PeriodûKEicsPraclica Management ProslhadanlicsRestoraliue Denlistry

THE EXHIBITION

[dealevent for exiijbi tors to show and damonstrale theijätest technalogical advances in material scierca antequipnient supplies.

ORAL PRESENTATION SESSION

Vou are invited to corïlribute papers. The papers sough

for should be based on research or practicûl/diinjca

nlact

Ths SecretariatSIOEC 2002Orchard Dentai Centra Pta LtdE63 Orchard Road «05-07. Singapore 235Tel: (65) 734 3t6Z Fax: (651 732 t97S

Homepage;http://www.stng net.com. s 9/-qualiser/siile

100 Volume 33, Number 2, 2002