Quality assessment of marginal sealingusing 7 dentin adhesive systemsSlavoljub ZivkovJc, MSc, PhD'

Objective; The goal of this investigation was to assess tn vitro the quality oi marginal sealing of compos-ite-dentin adhesive systems and human dentin. Method atid materials: Forty intact human premoiars andthird moiars were extracted from sub|ects ot both sexes and of different ages. After the enamel iayer wasremoved, a Ciass V cavity was formed on the buocai surtace, and a wedge cavity was formed on the iir-gual surtace. These were restored with resin oomposite materials and their corresponding dentin adhesivesystems. The quaiity ol marginal sealing was evaluated by assessing the iinear penetration of silver nitratedye. Results: The best marginai sealing between composite materials and the cavity wails, in both wedgeerosions and Class V cavities, was provided by Scotchbond Muiti-Purpose/Vaiux and Syntac/Helioprogtessystems. Dye penetration was slightiy greater with the XR-Bond/Heroulite, Giuma/Pekatill, and Superiuxtjniversal Bond 2/Supetiux Soiar systems The greatest microleakage was observed in Triptcn/Opalux andDenthesive/Chafisma specimens. Conclusion; The use of an adhesive system and the correspondingresin composite does not eliminate miorcleakage completeiy when the cavity margins are in dentin.(Quintessence int 2000:31:423-429)

Key words: adhesive systems, marginai seaiing, microleakage, resin composite

CUNICAL RELEVANCE; Because composite adhesivesystems achieve a high degree of physical and ohemicaibonding of dentai tissues and restorative materiais, theymay significantly improve marginai sealing in dentin.

Rapid technological development dnring the secondhalf of this century has left an impact on restora-

tive dentistry. Currently, scientific and technologicalachievements have improved existing dentai materialsand offered new ones with extretnely good estheticand biophysical characteristics, in particular resincomposite materials and glass-ionomer cements.'-*

However, despite the unquestionable advantages ofresin composite materials (such as their hardness andresistance to pressure, wear, and stretching), contrac-tion during honding may cause formation of marginalmicrogaps between the restoration and dental tis-sues.'-'^ Dimensional changes in composite restora-tions, in addition to contraction resulting from poly-merization, may also lead to high values in coefficientsof thermal expansion of the organic component of the

'Clmic (or Conservative Dentisiry and Endodontics. Faculty ofStomatology. University of Belgrade, Belgrade, Yugoslavia.

Reprint requests: Dr Slavoljub Zivkovic, Clinic for Conservative Dentistryand Endodontics, Faculty of Stomatology, tJniversity of Belgrade, DrSubotica 8, Betgrade, Yugoslavia. Fa*;: 3B1 -t t-6a5-361.

material. This can lead to the development of micro-gaps, which further enhance the penetration of oralfluids and microorganisrns.'"•''•'•' According to manyauthors, the microgap is the most common cause ofpathologic changes in dental pulp in teeth that havebeen restored with composite fillings.''*'"'

The introduction of enamel and dentin adhesivesystems (DASs), which promote a high degree of phys-ical and chemical bonding between dental tissues andrestorative materials, has significantly decreased thecontraction of materials and improved tbe retentionand marginal scaling of restorations.'"'^" Because thebonding of composite materials to enamel has beenmore or less solved successfully, the focus has shiftedto the bond between restorative materials and thedentin surface.-' The main reason for the interest isbecause, in Class II marginal sealing. 1 or more sidewalls or margins along the cavity may end in dentin orcement.*"' '"^^ The goal of this investigation was toassess, semiquantitatively in vitro, the quality of mar-ginal seai hetween specific resin composite adhesivesystems and human dentin.

METHOD AND MATERIALS

Forty intact human premolars and third molars,extracted from subjects of both sexes and oí differentages, were used. Teeth were kept in distilled water at4°C until the beginning of the experiment.

Quintessenoe International 423

• Zivkovic

Fig 1 Exposed dentin on the buccal sur-face of an intact human premclar. Theenamel layer has been completely removedwith a high-speed handpiece and a cylindn-cal diamond bur

Fig 2 Cross section through the middle of the tooth afier prepa-ration (A] Wedge erosion cavify on the lingual side: (B) Class Vcavity on the buccal side.

TABLE 1 Materials used in the study

Restorative materials

Denthesive/CharismaTripton/OpaluxSyntac/HelioprogresGluma/PekafillScotchbond Multi-Purpose/ValuxXR-Bond/HerculiteSuperlux Universal Bond 2/Superlui: Solar

Manufacturer

KuizerICI DentalVivadentBayer3M DentalKerrDMG

Fig 3 Cross section ot the tooth affer restoration. (A) Restorationof the wedge erosion cavity on the lingual surface; (B) restorationof the Class V cavity on the buccal surface.

On both the buccal and lingual surfaces, the enamellayer was completely removed with a high-speed hand-piece with a cylindrical diamond bur, used with watercooling. Tlie enamel layer was removed until a smoothsurface of exposed dentin was obtained (Fig 1). A ClassV caiity was prepared on the buccal side: on the lin-guai side, a wedge erosion cavity was formed (Fig 2).

The Class V cavities and wedge erosions wererestored with 1 of 7 DASs and the corresponding resincomposite material (Table I). Restorations were pol-ished and teeth were immersed in physiologic salineand kept at 37°C for 7 days. During that time, theteeth were periodically exposed to thermocycling, 300times, for 30 to 40 seconds. Specimens ahernatedamong 3 water haths with temperatures from i^C to7''C, 35°C to 37°C, and 55''C to 57''C.

The tooth surface, except the restoration site and 1mm surrounding it, was covered with 2 layers of nailpolish and was immersed in a 50% solution of silvernitrate for 6,5 hours. Following rinsing with distilledwater for 1 mintite, the specimens were immersed in a10% solution of universal film developer and keptthere for 120 minutes; afterward they were rinsed anddried.

After the nail polish was removed with a diamonddisk, the teeth were cut through the center of therestoration, parallel to the buecolingual surface (Fig3). A binocular magnifying glass with an insertedmicrometer ruler was used at x25 enlargement tomeasure the penetration of dye at each cross sectionon both the gingival and the occlusal walls, along thetooth-restoration interface.

The results, which represented the depth of pene-tration of dye expressed in millimeters, were statisti-cally evaluated with analysis of variance (ANOVA)and Í test. The means and standard de\'iations of spec-imens were analyzed with ANOVA, while the t testwas used to analyze differences between the resultsobtained on the occlusal and gingival walls in theClass V cavities and wedge erosions.

RESULTS

The mean penetration of silver nitrate is shown inTables 2 to 4 and Figs 4 and 5. The greatest microleak-age in Class V cavities was observed with the Tripton/Opalux and Denthesive/Charisma systems. Specimens

424 Volume 31, Numbers

• Zivkovic

TABLE 2 Mean depth of penetration (mm) of silver nitrate solution along the tooth-restoration interface inClass V cavities

Denthesive/Charisma

Tnpton/Opalux

Scotchbond Supetlux UniversalSyntac' Gluma/ Multi-Purpose/ XR-Bond/ Bond 2/

Helioprogres Pekatit! Valux Herculile Superlux Solar

Mean values Ocd^sal Gingival Occlnsal Gingival Oecljsal Gingival Occlusal Gingival Occljsal Gingival Dec al Gingival Occlusal GingivalMeanSD95%confidenceintervalfor rneanMinimumMaximum

2.020,51

2,45

1.35

2.050.31

1,702,40

2,180,35

2.47

1.70

2 320.40

1.702.70

0.460.38

0.14

0.7B

0,00

0,97

0.830.76

0.19

1.46

0.00

1.90

0,810,77

0 850.78

0.430.38

0,600,49

0.490.39

0.930 69

0.730.62

1,45

0,00

2,15

1.50

0.00

1.95

0.75

0.001,00

1.01

0 001 40

0,82

0,00

1,05

1.050 75

0,16 0.19 0.11 0,18 0,16 0 35 0.21 0 42

1.51 1.26 1.67

0.00 0.00 0.001.90 2.00 2.02

Analysis ot variarP<0.0001)andi

:e testing revealed a Wgh slalislically significant difference among groups in OlassV cavities or the occlusal walls (F = 17 419-n the gingival walls |F = 9.178; P < 0.0001),

TABLE 3 Mean depth of penetration (mm) of silver nitrate solution along the toot h-restoration interface inwedge erosions

Mean values

Mean

SD

confidenceintervalfor mean

Uppe

Minimum

Maximum

Analysis of variancana or Ihe gingiva

Denthesive/Charisma

Occlusal

2.250.14

2.13

2.37

1.952.40

e testingwalls IF

Gingiva

2.250.14

2.13

2,37

1,972,40

14.697

Tripton/Opalux

Occlusa

2.190,07

2.13

2.25

2.102,30

a high stallPe 0.0001

Gingival

2.240.90

1.48

2.99

0.002,62

tiqally sig

Syntac/Helioprogres

Occljsal Gingival

0.53 0 630.47 0 70

0.13 0.04

0.93 1.22

0.00 0.001.45 2,02

ificant difference a

Gluma/Pekafill

Occlusal Gingiva

0.45 0.570.37 0,53

0 14 0.12

0 76 1.02

0 00 0.001.05 1.55

Hong groups in mec

ScotchbondMulti-Purpose/

Valux

Occlusal Gingiva

0.43 0 620.32 0.49

0.21 0.20

0,75 1.04

0 00 0.000,87 1,37

ge erosions or tlie

XR-Bond/Herculite

Occlusa

0.570.55

0.11

1,03

0,001,40

occ(usa[

Gingiva

0.440,36

0 14

0.74

0.000,96

«alls (F =

Superlux UniversalBond 2/

Superlux Solar

Occlusal Gingival

0,54 0.700,73 0,68

0.00 0,13

1.15 1.27

0.00 0.002,22 2.20

29.362; P< 0.0001)

TABLE 4 Statistically significant differences for depth of penetration of silver nitrate solution

Scotchbond Superlux UniversalDenthesive/ Tripton/ Syntac/ Gluma/ Mulli-Purpose/ XR-Bond/ Bond 2/Charisma Opalux Helioprogres Pekafill Valux Herculite Superlux Solar

Mean values Occlusal Gingival Ocdusal Gmgii/al Occlusal Gingival Occlusal Gingivai Occiussl Gingival Occlusal Gingival Occlusal Gingival

Leve ne'stest forequality oívariances/ test forequality ofmeansSignificance 0,241 0,112 0,923 0,825 0.750 0.598 0.258 0.415 0,788 0.921 0.746 0 096 0.584 0.344

19.215 15.969 17,335 0,722 0 075 0 174 6.265 2.961 0,182 0.022 2,237 4.354 0.054 0.951

0.001 0.001 0.001 0.410 0.788 0,683 0.025 0 107 0.676 0.885 0.157 0.056 0.820 0.346

-1.224 -1.696 0.099 0.225 -0.325 0.540 1 179 0.839 -0.274 -0.101 -0,330 1.785 0.560 0.979

Mann-Whitney U 26,5 22,5 31,0 — — — 25 5 _ _ _ _ _ _ _

Quintessence international 425

• Zivkovic

Superlux Universal Bond 2/Superlux Solar

XR-Bond/Heroulite

Scctchbond Multi-Purpose/Valux

Gluma/Pekafill

Syntac/Helioprogres

Tripton/Opalux

Denthesi ie/Charisma

1

JHHIP0.00 0.50 1,00 1,50 2,00 2,50

Microleakage (mm)1 Occiusal 1 Gingival

Fig 4 Mean microleakage (mm) in Class V cavities

Superlux Universal Bond 2/Superlux Solar

Scotchbond Multi-Purpose/Valu«

Microleakage (mm)i Occiusal • Gingival

Fig 5 Mean microleakage (mm) in wedge erosions

426 VolumeSI, Numbers, 2000

• Zivkovic

in these 2 groups exliibited a significantly higher levelof penetration fhati did teeth restored with Syntac/Helioprogres, Gluma/Pekafill, Scotchbond Multi-Purpose/Valux, XR-Bond/Herculife, and SnperluxUniversal Bond 2/Superlux Solar,

The AiNOVA test showed high statisticaliy signifi-cant differences in penetration of silver nitrate atnongrestorative systems on the ücclusal [F = 17,419; P <0,0001) and gingival walls {F - 9.178; P < 0.0001) inClass V cavifics. Wedge erosions restored wilhTripton/Opalux and Denthesive/Charisma also exhib-ited greater penetration of marker than other restora-tive systems. The ANOVA showed statistically signifi-cant differences of penetrafion of dye amongcomposite adhesive systems on the occlusal (F =29,362; P < 0,0001) and gingivai walls (F = 14,897; P< 0,0001). No statistically significant differences werefound hetween levels of occlusal and gingival dye pen-etration in the Class V cavities and wedge erosions.

Generally speaking, the semiquantitative experi-ments indicated that the hest marginal sealing wasachieved by the Scotchhond JVfulti-Purpose/Valux,Syntac/Helioprogrcs, XR-Bond/Herculite, Gluma/Pekafill, and Superlux Universal Bond 2/SuperluxSolar systems; the greatest microleakage was notedwith the Tripton/Opalux and Denthesive/Charismasystems.

DISCUSSION

Several methods are availahle for evaluating the qual-ity of marginal seaiing of restorative materials for den-tal tissues, such as scanning electron microscopic evai-uation, semiquantitative methods with dye solution,quantitative methods with isotopes, and autoradio-graphic methods. Therefore, resuhs obtained throughdifferent laboratory methods very often differ, depend-ing on the author and method, thus preventing directcomparisons among studies.

In this study, the dye penetration method was usedbecause of its precision in evaluation of marginal seal-ing and its ability to reveal an existing microgap. Inaddition to its ability to give exact data on linear pene-tration and the possibility of direct reading of the pen-etrated marker by microscope, the main advantage ofthis method is its simplicity: It can he performed evenin small laboratories without any special equipment.

One disadvantage of this method is the size of theparticles {0.059 mm for silver nitrate); they are, in fact,much smaller than the molecules of the endotoxinsand hacteria that penetrate through the microfissure.Therefore, they are clinically unable to imitate themcompletely. Because of their size and ability to pene-trate even intact dentin, penetration of silver ions may

also have an effect on the poor interpretation ofmicroleakage studies.

The results ohtained in this study showed that, onaverage, the linear penetration of dye was the greatestwith the Tripton/Opalux and Denthesive/Charismasystems (more than 2,0 mm). The next greatest was inGluma/Pekafill, Superlux Universal Bond 2/SuperluxSolar systems, and XR-Bond/Herculite groups (lessthan 1,0 mm). The least amount of dye penetrationwas noted in Syntac/Helioprogres and ScotchhondMulti-Purpose/Valux specimens (less than 0.5 mm).

Similar results have heen ohtained hy other authors.For instanee, Barkmeier and Cooley' reported similarresults for Dentin Adhesive, Gluma, and Scotchhond2, In comparison to the present results, Fitchie et al"obtained greater penetration of dye with theGluma/Pekafill system. Also, McConnell et a l"ohtained similar values for linear penetration of dye inan experimental study performed on unprepared ero-sions and abrasions. There were no greater differencesin dye penetration at erosion lesions and Class Vpreparations, Crim"'' tested the marginal sealing ofGluma, Tenure, and XR-Bond systems on Class V cavi-ties. On average, penetration of dye was least forGluma (0,29 mm), greater for XR-Bond (0,33 mm),and the greatest for Tenure (0.45 mm), Kanca^*observed less microleakage with the Seotchhond 2 sys-tem than with the Gluma and Tenure adhesive systems.

The results of the present study revealed no statisti-cally significant differences between Class V cavitiesand wedge erosion lesions in the depth of dye markerpenetration along the interface of the restoration andthe wall of the eavities. The fact that there was slightlyless linear penetration of dye marker in wedge erosionlesions than in Class V cavities can be explained, first,hy the shape of the cavity, whieh is smoother; hence,for the same volume, the composite material is spreadover a greater surfaee, which helps to reduce its thick-ness and polymerization contraction."'^-^^ Anotherreason could he the depth of the cavity. In shallowercavities, there is synergistn between adhesion and con-traction forces in dentin; in addition, contractionstress during polymerization of composites is weakerin shallow cavities than in conventional Class V cavi-ties, and this has a direct effect on the efficacy of mar-ginal sealing,'*-" The ahsence of significant differencesbefween occlusal and gingival cavity walls in linearpenetration of dye can he easily explained by the factthat the whole cavity was placed within dentin.

Based on current knowledge, the majority of inves-tigators share the opinion that polymerization con-traction is solely responsible for the development ofmicrogaps; the extent of the gaps depends on the poly-merization contraction of resin composite materials,its quantity, and the surface that it covers."'^'^^ In

Ouintessenoe Inteinational 427

addition, the results may be influenced by a variety offactors, which can be divided into 2 groups; factorsdetermined by the tooth itself (type, age, dentin sur-face, the distance from the enamel-dentin border, andthe presence of caries) and factors related to method-ology (preservation of tooth until the experiment, thetype of solution, tooth preparation, tooth handling,and thermocycling),''^'""''

However, because all these factors were identicalfor all tested specimens and under maximal controlduring this whole experiment, the microgap develop-ment in the present study can be attributed to thepolymerization contraction of composite materials; ie,the strength of the adhesive bond between the restora-tion and the dentin surface. Because better marginalsealing was obtained with newer adbesive systems, aquestion arises about the influence that the smearlayer might have on the quality of that bond. Onegroup of investigators has concluded that it is neces-sary to preserve the smear layer as a natural barrierand, hence, obtain more effective bonding,''* whilethe other believes that, because it is a weak point anda barrier to bonding between dentin and organicresins, the smear layer should be completelyremoved,'"•"

Experiments by Yu et al'* assessed the possiblepathways of micropenetration, which could be withinor through the smear layer, between tbe smear layerand adhesive material, or between the adhesive systemand the restorative materials. One disadvantage of themethod used in the present experiment is that only thedepth, and not the pathways, of dye penetration couldbe observed under the microscope.'^

The most extensive penetration of dye in tfiepresent experiments was noticed in specimens treatedwith the Denthesive and Tripton adhesive systems,which preserve the smear layer completely. The forma-tion of microgaps can be explained by the contractionof the material before bonding between dentin andadhesive resin has been achieved, and not by thepreservation of the smear layer, because older genera-tion adhesive systems were also examined. In Scotch-bond Multi-Purpose, Syntac, and SuperliLx UniversalBond 2 systems, the smear layer is modified with con-ditioners and prepared for bonding with resin, whilein the Gluma system the smear layer is completelyremoved. In all 4 of these adhesive systems, the bondwith dentin was achieved by impregnation and inter-penetration of resin into the demineralized subsurfacedentin layer, which confirms the signiflcance that con-ditioning might have in achieving the adhesive bondbetween the dentin and the composite material,*'"- ' *

Although the bond between these adhesive systems(except Gluma) is first established on the smear layer,which is conditioned and modified, and not on the

dentin surface layer, a strong adhesive connectionwith tooth collagen was achieved through polymeriza-tion of resin. The bonding of these adhesive systems isbased on tbe reaction of hydroxyetbyl methacryiatewith collagen, while the monomer, which contains analdehyde group, plays the role of an adhesive. The effi-cacy of adhesion of methacryiate, which containshydrophobic and hydrophilic groups, to dentin isachieved by interpénétration of resin into the previ-ously conditioned part of dentin,"*"

CONCLUSION

1. The marginal sealing properties of all the testedadhesive systems and their corresponding resincomposites were not able to eliminate microleak-age completely in dentin cavities,

2, The adhesive systems Scotchbond Multi-Purposeand Syntac exhibited the best marginal sealing, XR-Bond, Gluma, and Superiux Universal Bond 2allowed slightly greater dye penetration. The great-est microleakage was noted with the Tripton andDenthesive systems,

3. The marginal leakage at the occlusal and gingivalwalls in Class V cavities and wedge erosion lesionsin dentin was not signiflcantly different.

4, The adhesive-composite systems of newer genera-tions (Scotchhond Multi-Purpose/Valux, Syntac/Helioprogres, XR-Bond/Herculite, SuperiuxUniversal Bond 2/Superlux Solar, and Gluma/Pekafill) significantly improved marginai sealing,

REFERENCES

1. Asmussen A, Utio S. Adhesion of restorative resins todentin: Chemical and physiochemical aspects, Oper Dent1992;17(suppl 5);62-67.

2. Fortin D. Swift EJ, Denehy GE, Reinhardt IW. Bondstrength and microleakage of current dentin adhesives.Dent Mater 1994:10:253-258.

3. Goracci G, Mori G, Bazzucchi ML. Marginal and biocom-patibility of a fourth-generation bonding agent. Dent Mater1995;ll:343-347.

4. Holtan ]R, Nystrom GP, Rensch SE, Phelps RA, DouglasWH. Microleakage of five dentinal adhesives. Oper Dent1994;19:189-193.

5. Prati C, Toehi E, Hamingni CA, Selghini M. One year clini-cal study of Seotehbond MP and Clearfil liner bond in classV [abstract]. J Dent Res 1994;73:129.

6. Swift E|, Perdigao [, Heymann HO, Bonding to enamel anddentin: A brief history and state of the art. Quintessence Int1995;26:96-110.

7. Barkmeier WW, Cooiey RL, Laboratory evaluation of adhe-' sive systems, Oper Dent 1992;17{suppl 5):50-61.

8. Bertolotti R. Conditioning of the dentin substrate OperDent 1992;17(suppl 5):131-136.

428 Volume 31, Numbers, 2000

• Zivkovic

9. Crim GA. Assessment of microleakage of three dentinalbonding systems: A six-month evaluation. Quintessence lntt991;22:387-389.

10, Erickson RE, Surface intractions of dentin adhesive materi-als, Oper Dent 1992;17(suppl 5);81-94.

11. Van Meerbeek B, Lambrechts P, Inokoshi S, Braem M,Vanherle G. Factors affecting adhesion to mineralized tis-sues. Oper Dent 1992;17(suppl 5]:111-124.

12, Barkmeicr WW, Erickson RL. Sheer bond strength of com-posite to enamel and dentin using Scoichbond Multi-Purpose, Am J Dent 1994:7:175-179.

15. Crim GA. Assessment of microleakage of three dentinalbonding systems. Quintessence lnt 1990;21:295-297.

14. Fitchie ]G, Pucket AD, Hembree JH, Williams M. Evalua-tion of a new dentinal bonding system. Quintessence lnt1993;24:65-70.

15. Ben-Amar A, Liberman R, Serbro L, Moses P. The effect ofdental pretreatment on microleakage in class V compositeresin restoration with two dentinal adhesive systems. Quin-tessence lnt 198^:20.905-905

16. Pintado M, Douglas WH. The comparison of microleakagebetween two different dentin bonding resin systems. Quin-tessence lnt 1988;19:905-907.

17 Wieckowski G, Yu HY, Davis E, |oint RB. Microleakage invarious dentin bonding agent composite resin systems. OperDent 1992:17(suppl 5|:62-67.

18. Yu XY. Davis EL, Joint RB. Wieczkowski G. Originationand progression of microleakage in a restoration with asmear layer-mediated dentinal bonding agent Quintessencelnt 1992;23:551-555,

19. Watanabe I, Nakabayashi N. Bonding durability uf photo-cured phenyl-p in TEGDMA to smear layer-retainedbovine dentin. Quintessence lnt 1993:24:335-342.

20. Yocbimura H, Maseki T, Nara Y, Katsuyama S. Dogon ¡L.Bond strength and microleakage of latest adhesive compos-ite systems [abstract 1057]. J Dent Res 1991:70:395

21. CroU TP, Donly KJ. Dentin and enamel bonded class Vcomposite restoration. Quintessence lnt 1992;23:463-465.

22. Reid ]S, Saunders WP, Chen YY. The effects of bondingagent and fissure sealant on microleakage of compositeresin restorations. Quintessence lnt 1991;22:295-298.

25. Swift EJ, Le Valley BD. Microleakage of etched-dcntin com-posite resin restorations. Quintessence lnt 1992;25:505-508.

24. Tjan AH, Tan DE, Microleakage at gingival margins of ClassV composite resin restorations re bonded with various low-viscosity resin systems. Quintessence lnt i991;22:565-573,

25. Yoshioka H, Katoh Y. Marginal closure of visible light curedcomposites in cervical cavities [abstract 984]. J Dent Res1991:70:389,

26. Ziemecki TU Dennison JB, Charbeneau GT, Clinical evalu-ation of cervical composite resin restoration placed withoutretention, Oper Dent 1987;12:27-33,

27 McConnell TA, Richards ND, Mixon |M, Mitchell RJ.Microleakage of unprepared abrasion lesions vs preparedcervical lesions ¡abstract 985J. J Dent Res 1991;70:589.

28. Kanca J. Tlie effect of microleakage of four dentin-enameibonding systems. Quintessence lnt 1989,20:559-561.

29. Dijken JW. A three years evaluation of Gluma and Gluma/Scotchbcnd for restoration of cervical erosions, Scand JDent Res 1990;98:341-344.

3Ü. Ferrari M, Yamamoto K, Vichi A, Finger WJ. Clinical andlaboratory evaluation of adhesive restorative systems. Am JDent 1994;7:217-219.

31, De Lange C, Kortland PJ, Luxwolda RJ, Swijnenburg JT.Restoration of noncarious cervical erosion lesions: A twoyear clinical report, Fenestra Res Forum 1996:2:1-4.

52. Titley K, Chemecky R, Marie B, Smith D, Penetration of adentin bonding agent inlo dentin. Am J Dent 1994;7:190-194.

55. Calamia |, Kohli SS, Shuiman A, Kaim J. Microleakage ofclass V composite restorations using the state of the artdentin adhesive system (abstract 982]. ( Dent Res 1991;70:388

54. Nonaka T. Baez RS Dentin conditioning and microleakage[abstract 978] ] Dent Res 1991:70:588.

55. Retief DH, O'Brien JA, Smith LA. In vitro investigation andevaluation of dentin bonding agents. Am J Dent 1988,1:178-183.

56. Rigsby DF, Retief DH, Bidez MW, Russe CM. Microleakageof class V restorations subjected to temperature and loadcycling [abstract 981]. J Dent Res 1991;70:388,

57. Torstenson B, Brännström M. Contraction gap under com-posite resin restorations: Effects of hydroscopic expansionsand thermal stress. Oper Dent J 1988:13:24-31,

38. Inokoshi S, Iwaku M, Fusayama T. Pulpal response to anew adhesive restorat ive resin, ] Dent Res 1982;61:1014-1019.

Quintessence International 429