Microleakage of glass ionomerlcomposite resin restorations: A laboratory study. 1. The influence of glass ionomer cement E. Dell Kingsford Smith, MDS(Syd), FRACDS* F. Elizabeth Martin, MDS(Syd), FRACDS*
Key words: Composite resins, dye penetration, glass ionomer cement, microleakage. Abstract Acid-etching of enamel margins, often combined with bevelling and use of a low viscosity resin, has resulted in improved sealing of composite restorations. When enamel is absent and margins involve dentine or cementum, microleakage is more often observed. The use of glass ionomer cement, as both a restorative material and as a base combined with a veneer of composite resin, has been advocated as a means of minimizing microleakage where margins are placed in dentine or cementum. Four restorative techniques incorporating glass ionomer cement were used, in vitro, to restore cervical cavities. After thermal cycling, dye penetration was scored along occlusal and gingival margins. In all instances the gingival margin exhibited a less reliable seal than the occlusal margin. The sandwich restorations produced a superior seal at both occlusal and gingival margins when compared with glass ionomer restorative cement. (Received for publication November 1990. Revised April 199 1. Accepted April 199 1.)
Introduction The absence of a seal at restoration margins may contribute to tooth discoloration, adverse pulpal response, postoperative sensitivity and recurrent Microleakage, defined as 'the passage
*Department of Operative Dentistry, The University of Sydney. Australian Dental Journal 1992;37(1):23-30
of bacteria, fluids, chemical substances, molecules and ions between the tooth and its restoration',' may be used in an attempt to predict marginal sealing ability. Cervical lesions occurring in response to caries, erosion and abrasion often have both enamel and dentine or cementum margins, and are frequently encountered in clinical practi~e.~.' Cervical restorations are, therefore, often used to investigate marginal sealing ability with respect to both gingival dentine or cementum margins and occlusal enamel margins. Acid-etching of enamel margins, often combined with bevelling and use of a low viscosity resin, has resulted in improved sealing of composite resin restorations both in vitro and in vivo.6-'1Where enamel is absent and margins involve dentine or cementum, microleakage is more often observed.Z Several restorative techniques have been introduced which reportedly bond successfully to enamel, dentine and cementum, and result in an improved marginal seal. These include the glass ionomer restorative cements,'* composite resins in combination with dentine bonding agent^'^ and, more recently, a glass ionomer cement base combined with a veneer of composite resin,I4 often called the sandwich restoration. The purpose of this investigation was to compare the abilities of four restorative techniques incorporating glass ionomer cement to achieve a seal, in vitro, where the cavosurface margins of the cavities were placed in both enamel and on the root surface. Part 2 of this investigation will examine the abilities of four bonding systems to obtain a seal under the same conditions. 23
Dentino-enamel junction L
v/k, Ketac -Fil/ ChemFiI
vk
Ketac-Bond/ Vitrabond
............................ .......................... Silux ........... P
Fig. 1. -Diagram of a buccolingual, longitudinal section showing the cavity prior to modification of the occlusal margin.
Materials and methods Sound, extracted premolars were obtained and stored under humid conditions at 4 "C.Storage time between extraction and use in the study was at least 24 hours and no more than four months. Forty teeth which showed minimal extraction damage and had been exposed to long-term systemic fluoride were chosen, periodontal ligament remnants were mechanically removed? and the teeth cleansed with pumice and water. Uniform, 3 mm diameter cavities were prepared, as illustrated in Fig. 1, on the buccal and lingual surfaces of the teeth using a water-cooled tungsten carbide bur$ at high speed. T h e cavities were bisected by the cemento-enamel junction so that the occlusal margin was bounded by enamel and the gingival margin by cementum and dentine. A retentive groove was placed$ in dentine along the gingivo-axial line angle.
tMorse Scaler (1). Neos Dental, Switzerland. tKornet H245 008 Gebr. Rrasseler GrnbH & Co. KG, Lerngo, Germany. @Ash Rd. PC%. Ash Co., Wevbridge, ITK.
24
Fig. 2. -Diagram of buccolingual, longitudinal sections of the finished restorations.
T h e eighty prepared cavities were restored using one of four techniques, namely: Group K with Ketac-Fil, [( a glass ionomer restorative cement. Group KS using Ketac-Bond 1) as a base to restore lost dentine, with a surface veneer of Silux.7 Group VS using Vitrabondl as a base to restore lost dentine, with a surface veneer of Silux. Group C restored using Chemfil,** a glass ionomer restorative cement. A representative buccolingual, longitudinal section of the resulting restorations is illustrated in Fig. 2. The manipulation of all materials conformed to the manufacturers' directions unless otherwise stated. Group K Cavity surfaces were cleansed with 10 per cent poly acrylic acid?? for 20 seconds, washed for 20
(IEspe GrnbH, Seefeld/Oberbay, Gerrnanv. 73M Dental Products Division, St Paul, Minn, USA. "De Trey Division, Dentsply Ltd, Weybridge, Surrey, U K . t%-C Dentin Conditioner. G-C Dental Industrial Carp., Tokyo, Japan.
Australian Dental Journal 1992;37:1.
the distal. Two minutes after curing the matrix was removed.
No leakage =N
%
I,---=P I I -
-wT
Fig. 3.-The system used to score dye penetration, where Grade
N - no dye penetration. Grnde P - partial dye penetration along the occlusal or gingival wall. Grade W - dye penetration along the cavity wall, but not including the axial wall. Grnde T - total dye penetration along the cavity walls extending onto the axial wall.
seconds, then dried. Ketac-Fil was placed using a syringe and a matrix$$ was adapted immediately, removed after five minutes, then varnish 1 applied to the restoration surface. Group KS The enamel cavosurface margin was modified to form a 1 mm wide, 45 " bevel using the high speed bur. The cavity was also cleansed with polyacrylic acid then Ketac-Bond was placed on all dentine surfaces, including the gingival cavosurface margin. Six minutes from starting to mix the Ketac-Bond, the base was contoured and excess removed from the enamel margin. The enamel bevel and KetacBond surfaces were then etched for 60 seconds, rinsed and dried. A layer of Deltongs light-cured resin was applied to all etched surfaces, thinned with compressed air and cured. Silux was then adapted to the cured resin surface, a matrix applied and held in place while the composite was cured for a total of 40 seconds-20 seconds from the mesial aspect and 20 seconds from
$$Hawe Neos Dental, Switzerland. #Johnson &JohnsonDental Products Company, E. Windsor, NJ, USA. Australian Dental Journal 1992:37:1.
Group vs The enamel cavosurface margin was modified as described for Group KS and the cavity cleansed with polyacrylic acid. Vitrabond was then placed on all dentine surfaces, including the gingival cavosurface margin. After curing, the enamel bevel and Vitrabond surfaces were etched for 60 seconds, rinsed and dried. Scotchbond was applied to all etched surfaces and cured, then Silux was placed and cured as described for group KS. Group c Cavity surfaces were cleansed with polyacrylic acid, washed, then dried. Chemfil was placed and a matrix adapted immediately, removed after five minutes, then varnish** applied to the restoration surface. All specimens were stored for 24 hours in 37 "C deionized water then retrieved and prepared for thermal cycling.IS The specimens were subjected to thermal cycling in two stages - prior to, and then after finishing of the restorations. The first stage consisted of 100 cycles, at 24 hours after restoration, between baths held at 8"+2"C and 48 O f 2 "C. The cold bath contained 0.25 per cent basic fuchsin,ll the warm bath deionized water only. A cycle consisted of 25 seconds in each bath and 5 seconds bath change - a total of 60 seconds per cycle. After this stage of cycling, the specimens were stored for a further six days, then the restorations were finished using Sof-Lex discs. One day after finishing, the principal stage of thermal cycling was initiated. The specimens were cycled 200 times per day for 10 days and between cycling episodes were kept in 37 "C deionized water. On completion of the cycling the specimens were mounted and sectioned to obtain between three and five buccolingual, longitudinal sections. Three surfaces from each specimen were then randomly chosen and examined for dye penetration at x 16 magnification. The extent of dye penetration was scored for both occlusal and gingival margins using the system illustrated in Fig. 3, where: Grude N corresponds to no dye penetration occurring at the tooth-restoration interface. Grude P indicates partial dye penetration along the occlusal wall contained by the dentino-enamel junction and along the gingival wall by the start of the gingival retentive groove. 11BDH Chemicals Ltd., UK. 25
s
loo
c
n
rc
0
-
NPWT
NPWT
KS
K
NPWT
vs
Fig. 4. -Microleakage results for the occlusal margins.
v)
cd
NPWT
NPWT
KS
K
NPWT
vs
Fig. 5. -Microleakage results for the gingival margins.
Table 1. Microleakage results (percentage of surfaces per grade) Depth Group K Group KS Group VS of leakage Occlusal Gingival Occlusal Gingival Occlusal Gingival
N P W T
19 37
19 25
0 6 65 29
63 25 12 0
4 42 23 31
54 17 25 4
4 34 4 58
Grade Wcorresponds to dye penetration past the dentino-enamel junction along the occlusal wall, and past the start of the retentive groove along the gingival wall, but not including the axial wall. 26
Grade T refers to total dye penetration along the occlusal or gingival walls and extending onto the axial wall. Microleakage patterns were compared using two by four Chi-square analysis. Results Accurate assessment of the depth of dye penetration around the Chemfil restorations was made impossible by dye absorption into the bulk of the restorative material, and as a consequence the results from group C were discarded. Microleakage results for the remaining groups K, KS and VS are shown in Table 1 and Fig. 4, 5. Statistical analysis of the data is presented in Table 2. Australian Dental Journal 1992;37:1.
Fig. 6.-A sectioned surface from the Ketac-Bond and Silux restoration (group KS).
Table 2. Comparison of microleakage results at occlusal and gingival margins Comparison of groups Occlusal margin K:KS K:VS KS:VS Gingival margin K:KS K:VS KS:VS
A difference found at 0.001 level
Leakage observed
Yes Yes No
K>KS K>VS Similar
Yes Yes No
K>KS K>VS Similar
General observations No restorative technique completely resisted microleakage at either the occlusal or gingival margin under these experimental conditions. At the occlusal margin, no technique totally confined Australian Dental Journal 1992:37:1.
leakage to the enamel cavity wall, although all did so in more than 50 per cent of cases. The gingival margin of all groups exhibited more severe microleakage than the occlusal margin, and this observed difference was found to be statistically- significant at the 0.001 level in all cases. When leakage occurred in restorations using glass ionomer cement as a base material in combination with composite resin, the dye penetration was invariably at the tooth-restoration interface. A sectioned surface from group KS showing the composite and glass ionomer bond remaining impervious to dye can be seen in Fig. 6 .
The occlusal margin The Ketac-Fil restorations (group K) appeared to provide the least reliable seal at the occlusal margin. The ability of Ketac-Fil to prevent leakage in less than 20 per cent of cases is in marked 27
contrast to the results achieved using the other two restorative techniques, where leakage was prevented in more than 50 per cent of cases. When the results for groups KS and VS were compared, no statistical difference could be found.
The gingival margin Cavities restored with Ketac-Fil showed a significant difference in the pattern of gingival leakage when compared with groups KS and VS. The two most severe grades of leakage, W and T, were scored for more than 90 per cent of surfaces examined in group K, compared with 54 per cent and 62 per cent for groups KS and VS, respectively. When the microleakage results for groups KS and VS were compared, no statistical difference could be found. From observation of the leakage patterns in Fig. 5, although the two groups showed the same ability to completely prevent leakage, group KS showed a much lower incidence of the worst grade of leakage, grade T, than did group VS. Discussion General observations No restorative technique tested in this investigation completely resisted microleakage at either the occlusal or the gingival margin, and in all techniques the gingival margin exhibited significantly greater leakage than the occlusal margin. These two findings are consistent w h those of the vast majority of other microleakage studies. The significance of the difference found in the behaviour of occlusal and gingival margins within each restorative technique also confirms the suitability of cervical restorations in testing materials for use at different sites on the tooth surface. A hrther finding of this investigation, that no restorative technique tested was capable of providing a predictable seal at the gingival margin is also consistent with other similar s t u d i e ~ . ' ~ - ' ~ This consensus emerged despite vast differences in materials, cavity design and some areas of the experimental regimen, for example thermal cycling. Although the temperatures used for thermal cycling were very similar, the amount of thermal cycling varied from nonezo to some studies with several h ~ n d r e d ~ and ~ - ~another ~ . ~ several ~ * ~ ~thousand," as used in this study. That the interface between composite resin and glass ionomer cement remained uniformly impervious to dye in the sandwich restorations is in agreement with the observation of other^.^',^^ One studya7did report leakage between composite resin and glass ionomer cement, although only when extensive dye penetration had occurred at the 28
dentine wall and combined with cracks and fissures in the glass ionomer cement. These observations of the composite resin and glass ionomer cement bond remaining generally intact appear consistent with the reports of values for bond strengths between composite resin and etched glass ionomer in vitro which exceed the cohesive strength of the ~ e m e n t . ' ~It*is~also ~ * interesting ~~ to note that etch times for the glass ionomer cement bases varied considerably, from 20 ~ e c o n d s ' to ~ *60 ~ ~seconds2' as used in this study. This variation did not, however, appear to have any correlation with reported dye penetration at the composite resin and glass ionomer cement interface. The absorption of basic hchsin dye into the bulk of the Chemfil restorative material in this study contrasted with the behaviour of Ketac-Fil. This variation in behaviour occurred despite both being water-hardening cements. It can only be speculated, therefore, that some other difference in the two materials, such as the type of polyacid used or packaging and mixing characteristics, was the cause. Although no other reported microleakage study has included Chemfil restorations to enable comparison, a comprehensive clinical study of Chemfil reported no cases of gross surface staining over two years.26 More than 20 per cent of the Class V restorations observed did, however, show slight or obvious surface staining over the same time period.
The occlusal margin In this investigation, the glass ionomer restorative cement, Ketac-Fil, provided the least reliable seal occlusally when compared with the two sandwich restorations in which the enamel margin was restored with composite resin. This finding is in agreement with the few other microleakage studies which have cited data for the occlusal margin16~z2 and appears to confirm the superiority of the acidetched composite resin bond over the glass ionomer bond with enamel. When the two sandwich restorations were compared, microleakage results at the occlusal aspect were not significantly different, a finding again in agreement with other similar investig a t i o n ~ . It' ~appears, ~ ~ ~ ~therefore, ~~ that within the limits of this study differences in bonding resins and glass ionomer base materials do not greatly influence the reliability of the etched enamel bond with composite resin. The gingival margin As with the occlusal margin, cavities restored with Ketac-Fil showed significantly greater leakage at the gingival aspect than those restored using a glass ionomer base combined with a veneer of composite Australian Dental Journal 1992;371.
resin. This finding concurs with that of another similar study.I6 When the two sandwich restorations investigated in the present study were compared, no significant difference could be found in their leakage patterns. These results are in agreement with some studies20but conflict with others where differences were f o ~ n d . ' ~In. 'the ~ only other study'8 where both Vitrabond and Ketac-Bond were compared in the sandwich restoration, Ketac-Bond exhibited significantly greater gingival microleakage than Vitrabond. The significance level used was 0.05, a substantially lower level than the 0.001 used in the present study, which may explain the conflicting result. Despite some differing findings, most microleakage studies investigating the sandwich restoration concluded that although leakage at the gingival margin was not completely eliminated, it could be significantly decreased by this restorative t e ~ h n i q u e . ' ~ , ' ~ . ' ~ . ~ ~ It should be noted that these conclusions are based solely on laboratory testing and substantiation by clinical trials would be desirable. Although results of clinical microleakage studies are rarely reported, it has been suggested that marginal staining occurring with time around restorations may indicate the presence of microleakage clinially.^^ Observations of marginal staining at one year showed 8 per cent of Ketac-Fil cervical restorations with marginal staining, compared with two per cent of Silux restorations bonded to etched enamel and a glass ionomer cement base.27,28No distinction was made, however, between staining occurring at occlusal or gingival margins, so although the sandwich restoration performed slightly better than Ketac-Fil alone, any comparison with the present study is difficult.
Conclusions Within the limits of this investigation, none of the restorative techniques studied completely sealed either the occlusal enamel margins or the gingival margins positioned on the root surface. In all instances, however, the gingival margin exhibited a less reliable seal than the occlusal margin. The sandwich restoration using either Vitrabond or Ketac-Bond produced a superior seal at both occlusal and gingival margins when compared with the glass ionomer restorative cement, Ketac-Fil. Acknowledgements Assistance with statistical analysis by Associate Professor R. W. Bryant, preparation of histograms by Dr D. M. Roessler, and with specimen preparation by Dr N. Andrews is gratehlly acknowledged. Australian Dental Journal 1992;371
This project was partly supported by a grant from the Australian Dental Research Fund Trust.
References 1. Going RE. Microleakage around dental restorations: a
summarizing review. J Am Dent Assoc 1972;84:1349-57. 2. Bauer JF, Henson JL. Microleakage: a measure of the performance of direct filling materials. Oper Dent 1984;9:2-9. 3. Crim GA, Garcia-Godoy F. Microleakage: the effect of storage and cycling duration. J Prosthet Dent 1987;57:574-6. 4. Sognnaes RF, Wolcott RB, Xhonga FA. Dental erosion. Erosion-like patterns occurring in association with other dental conditions. J Am Dent Assoc 1972;84:571-6. 5 American Dental Association Council on Dental Research and Council on Dental Therapeutics. Report of symposium: Root surface caries. J Am Dent Assoc 1983;106:496. 6 Hembree JH, Andrews JT. In siru evaluation of marginal leakage using an ultraviolet-light-activated resin system. J Am Dent Assoc 1976;92:414-8. 7 Hembree JH, Andrews JT. Microleakage of several acidetch composite resin systems: a laboratory study. Oper Dent 1976;1:91-7. 8. Ortiz RF, Phillips RW, Swartz ML, Osborne JW. Effect of composite resin bond agent on microleakage and bond strength. J Prosthet Dent 1979;41:51-7. 9. Crim GA, Swartz ML, Phillips RW. An evaluation of cavosurface design and microleakage. Gen Dent 1984;Jan-Feb:56-8. 10. Qvist V. Marginal adaptation of composite restorations performed in vivo with different acid-etch restorative procedures. Scand J Dental Res 1985;93:6&75. 11. Torstenson B, Brannstrihn M. Composite resin contraction gaps measured with a fluorescent resin technique. Dent Mater 1988;4:238-42. 12. American Dental Association Council on Dental Materials and Devices. Status report on the glass-ionomer cements. J Am Dent Assoc 1979;99:221-6. 13. American Dental Association Council on Dental Materials, Instruments and Equipment. Dentin bonding systems: an update. J Am Dent Assoc 1987;114:91-5. 14. McLean JW, Powis GR, Prosser HJ, Wilson AD. The use of glass-ionomer cements in bonding composite resins to dentine. Br Dent J 1985;158:410-4. 15. Wenner KK, Hawkins IK, Smith CD, Fairhurst CW. Microleakage of root restorations. J Dent Res 1985;64:Abstr 71:182. 16. Srisawasdi S, Boyer DB, Reinhardt JW. The effect of the removal of the smear layer on microleakage of class V restorations in vim. Dent Mater 1988;4:384-9. 17. Prati C, Montanari G. Comparative microleakage study between the sandwich and conventional three-increment techniques. Quintessence Int 1989;20:587-94. 18. Prati C, Nucci C, Montanari G. Microleakage of mixed class V restorations placed with GIC bases or dentin bonding agents. Schweiz Monatsschr Zahnmed 1989;99:1117-21. 19. Schwartz JL, Anderson MH, Pelleu GB. Reducing microleakage with the glass-ionomer/resin sandwich technique. Oper Dent 1990;15:186-92. 20. Phair CB, Zidan 0, Gomez-Marin 0, Han S. Marginal leakage in the composite resinlglass-ionomer cement class V restoration. J Dent Res 1987;66:Abstr 45112. 21. Mathis RS, DeWald JP, Moody CR, Ferracane JL. Marginal leakage in class V composite resin restorations with glassionomer liners in wirro. J Prosthet Dent 1990;63:522-5. 29
22. Gordon M, Plasschaert AJM, Stark MM. Microleakage of several tooth-coloured restorative materials in cervical cavities. A comparative study in vim. Dent Mater 1986;2:228-31. 23. Gordon M, Plasschaert AJM, Soelberg KB, Bogdan MS. Microleakage of four composite resins over a glass-ionomer cement base in class V restorations. Quintessence Int 1985; 16:817-20. 24. Sneed WD, Looper SW. Shear bond strength of a composite resin to an etched glass ionomer. Dent Mater 1985;1:127-8. 25. Hinoura K, Moore BK, Phillips RW. Tensile bond strength between glass ionomer cements and composite resins. J Am Dent Assoc 1987;114:167-72. 26. Knibbs PJ, Plant CG, Pearson GJ. A clinical assessment of an anhydrous glass-ionomer cement. Br Dent J 1986; 161:99- 103.
30
27. Tyas MJ, Twhey A, Clark J. Clinical evaluation ofthe bond between composite resin and etched glass ionomer cement. Aust Dent J 1989;34:1-4. 28. Tyas MJ, Burns GA, Byrne PF, Cunningham PJ, Dobson BC, Widdop FT. Clinical evaluation of Scotchbond: one year results. Aust Dent J 1986;31:159-64.
Address for correspondenceheprints: Department of Operative Dentistry, Faculty of Dentistry, The University of Sydney, United Dental Hospital, 2 Chalmers Street, Surry Hills, New South Wales, 2010.
Australian Dental Journal 1992;37:1.
Key words: Composite resins, dye penetration, glass ionomer cement, microleakage. Abstract Acid-etching of enamel margins, often combined with bevelling and use of a low viscosity resin, has resulted in improved sealing of composite restorations. When enamel is absent and margins involve dentine or cementum, microleakage is more often observed. The use of glass ionomer cement, as both a restorative material and as a base combined with a veneer of composite resin, has been advocated as a means of minimizing microleakage where margins are placed in dentine or cementum. Four restorative techniques incorporating glass ionomer cement were used, in vitro, to restore cervical cavities. After thermal cycling, dye penetration was scored along occlusal and gingival margins. In all instances the gingival margin exhibited a less reliable seal than the occlusal margin. The sandwich restorations produced a superior seal at both occlusal and gingival margins when compared with glass ionomer restorative cement. (Received for publication November 1990. Revised April 199 1. Accepted April 199 1.)
Introduction The absence of a seal at restoration margins may contribute to tooth discoloration, adverse pulpal response, postoperative sensitivity and recurrent Microleakage, defined as 'the passage
*Department of Operative Dentistry, The University of Sydney. Australian Dental Journal 1992;37(1):23-30
of bacteria, fluids, chemical substances, molecules and ions between the tooth and its restoration',' may be used in an attempt to predict marginal sealing ability. Cervical lesions occurring in response to caries, erosion and abrasion often have both enamel and dentine or cementum margins, and are frequently encountered in clinical practi~e.~.' Cervical restorations are, therefore, often used to investigate marginal sealing ability with respect to both gingival dentine or cementum margins and occlusal enamel margins. Acid-etching of enamel margins, often combined with bevelling and use of a low viscosity resin, has resulted in improved sealing of composite resin restorations both in vitro and in vivo.6-'1Where enamel is absent and margins involve dentine or cementum, microleakage is more often observed.Z Several restorative techniques have been introduced which reportedly bond successfully to enamel, dentine and cementum, and result in an improved marginal seal. These include the glass ionomer restorative cements,'* composite resins in combination with dentine bonding agent^'^ and, more recently, a glass ionomer cement base combined with a veneer of composite resin,I4 often called the sandwich restoration. The purpose of this investigation was to compare the abilities of four restorative techniques incorporating glass ionomer cement to achieve a seal, in vitro, where the cavosurface margins of the cavities were placed in both enamel and on the root surface. Part 2 of this investigation will examine the abilities of four bonding systems to obtain a seal under the same conditions. 23
Dentino-enamel junction L
v/k, Ketac -Fil/ ChemFiI
vk
Ketac-Bond/ Vitrabond
............................ .......................... Silux ........... P
Fig. 1. -Diagram of a buccolingual, longitudinal section showing the cavity prior to modification of the occlusal margin.
Materials and methods Sound, extracted premolars were obtained and stored under humid conditions at 4 "C.Storage time between extraction and use in the study was at least 24 hours and no more than four months. Forty teeth which showed minimal extraction damage and had been exposed to long-term systemic fluoride were chosen, periodontal ligament remnants were mechanically removed? and the teeth cleansed with pumice and water. Uniform, 3 mm diameter cavities were prepared, as illustrated in Fig. 1, on the buccal and lingual surfaces of the teeth using a water-cooled tungsten carbide bur$ at high speed. T h e cavities were bisected by the cemento-enamel junction so that the occlusal margin was bounded by enamel and the gingival margin by cementum and dentine. A retentive groove was placed$ in dentine along the gingivo-axial line angle.
tMorse Scaler (1). Neos Dental, Switzerland. tKornet H245 008 Gebr. Rrasseler GrnbH & Co. KG, Lerngo, Germany. @Ash Rd. PC%. Ash Co., Wevbridge, ITK.
24
Fig. 2. -Diagram of buccolingual, longitudinal sections of the finished restorations.
T h e eighty prepared cavities were restored using one of four techniques, namely: Group K with Ketac-Fil, [( a glass ionomer restorative cement. Group KS using Ketac-Bond 1) as a base to restore lost dentine, with a surface veneer of Silux.7 Group VS using Vitrabondl as a base to restore lost dentine, with a surface veneer of Silux. Group C restored using Chemfil,** a glass ionomer restorative cement. A representative buccolingual, longitudinal section of the resulting restorations is illustrated in Fig. 2. The manipulation of all materials conformed to the manufacturers' directions unless otherwise stated. Group K Cavity surfaces were cleansed with 10 per cent poly acrylic acid?? for 20 seconds, washed for 20
(IEspe GrnbH, Seefeld/Oberbay, Gerrnanv. 73M Dental Products Division, St Paul, Minn, USA. "De Trey Division, Dentsply Ltd, Weybridge, Surrey, U K . t%-C Dentin Conditioner. G-C Dental Industrial Carp., Tokyo, Japan.
Australian Dental Journal 1992;37:1.
the distal. Two minutes after curing the matrix was removed.
No leakage =N
%
I,---=P I I -
-wT
Fig. 3.-The system used to score dye penetration, where Grade
N - no dye penetration. Grnde P - partial dye penetration along the occlusal or gingival wall. Grade W - dye penetration along the cavity wall, but not including the axial wall. Grnde T - total dye penetration along the cavity walls extending onto the axial wall.
seconds, then dried. Ketac-Fil was placed using a syringe and a matrix$$ was adapted immediately, removed after five minutes, then varnish 1 applied to the restoration surface. Group KS The enamel cavosurface margin was modified to form a 1 mm wide, 45 " bevel using the high speed bur. The cavity was also cleansed with polyacrylic acid then Ketac-Bond was placed on all dentine surfaces, including the gingival cavosurface margin. Six minutes from starting to mix the Ketac-Bond, the base was contoured and excess removed from the enamel margin. The enamel bevel and KetacBond surfaces were then etched for 60 seconds, rinsed and dried. A layer of Deltongs light-cured resin was applied to all etched surfaces, thinned with compressed air and cured. Silux was then adapted to the cured resin surface, a matrix applied and held in place while the composite was cured for a total of 40 seconds-20 seconds from the mesial aspect and 20 seconds from
$$Hawe Neos Dental, Switzerland. #Johnson &JohnsonDental Products Company, E. Windsor, NJ, USA. Australian Dental Journal 1992:37:1.
Group vs The enamel cavosurface margin was modified as described for Group KS and the cavity cleansed with polyacrylic acid. Vitrabond was then placed on all dentine surfaces, including the gingival cavosurface margin. After curing, the enamel bevel and Vitrabond surfaces were etched for 60 seconds, rinsed and dried. Scotchbond was applied to all etched surfaces and cured, then Silux was placed and cured as described for group KS. Group c Cavity surfaces were cleansed with polyacrylic acid, washed, then dried. Chemfil was placed and a matrix adapted immediately, removed after five minutes, then varnish** applied to the restoration surface. All specimens were stored for 24 hours in 37 "C deionized water then retrieved and prepared for thermal cycling.IS The specimens were subjected to thermal cycling in two stages - prior to, and then after finishing of the restorations. The first stage consisted of 100 cycles, at 24 hours after restoration, between baths held at 8"+2"C and 48 O f 2 "C. The cold bath contained 0.25 per cent basic fuchsin,ll the warm bath deionized water only. A cycle consisted of 25 seconds in each bath and 5 seconds bath change - a total of 60 seconds per cycle. After this stage of cycling, the specimens were stored for a further six days, then the restorations were finished using Sof-Lex discs. One day after finishing, the principal stage of thermal cycling was initiated. The specimens were cycled 200 times per day for 10 days and between cycling episodes were kept in 37 "C deionized water. On completion of the cycling the specimens were mounted and sectioned to obtain between three and five buccolingual, longitudinal sections. Three surfaces from each specimen were then randomly chosen and examined for dye penetration at x 16 magnification. The extent of dye penetration was scored for both occlusal and gingival margins using the system illustrated in Fig. 3, where: Grude N corresponds to no dye penetration occurring at the tooth-restoration interface. Grude P indicates partial dye penetration along the occlusal wall contained by the dentino-enamel junction and along the gingival wall by the start of the gingival retentive groove. 11BDH Chemicals Ltd., UK. 25
s
loo
c
n
rc
0
-
NPWT
NPWT
KS
K
NPWT
vs
Fig. 4. -Microleakage results for the occlusal margins.
v)
cd
NPWT
NPWT
KS
K
NPWT
vs
Fig. 5. -Microleakage results for the gingival margins.
Table 1. Microleakage results (percentage of surfaces per grade) Depth Group K Group KS Group VS of leakage Occlusal Gingival Occlusal Gingival Occlusal Gingival
N P W T
19 37
19 25
0 6 65 29
63 25 12 0
4 42 23 31
54 17 25 4
4 34 4 58
Grade Wcorresponds to dye penetration past the dentino-enamel junction along the occlusal wall, and past the start of the retentive groove along the gingival wall, but not including the axial wall. 26
Grade T refers to total dye penetration along the occlusal or gingival walls and extending onto the axial wall. Microleakage patterns were compared using two by four Chi-square analysis. Results Accurate assessment of the depth of dye penetration around the Chemfil restorations was made impossible by dye absorption into the bulk of the restorative material, and as a consequence the results from group C were discarded. Microleakage results for the remaining groups K, KS and VS are shown in Table 1 and Fig. 4, 5. Statistical analysis of the data is presented in Table 2. Australian Dental Journal 1992;37:1.
Fig. 6.-A sectioned surface from the Ketac-Bond and Silux restoration (group KS).
Table 2. Comparison of microleakage results at occlusal and gingival margins Comparison of groups Occlusal margin K:KS K:VS KS:VS Gingival margin K:KS K:VS KS:VS
A difference found at 0.001 level
Leakage observed
Yes Yes No
K>KS K>VS Similar
Yes Yes No
K>KS K>VS Similar
General observations No restorative technique completely resisted microleakage at either the occlusal or gingival margin under these experimental conditions. At the occlusal margin, no technique totally confined Australian Dental Journal 1992:37:1.
leakage to the enamel cavity wall, although all did so in more than 50 per cent of cases. The gingival margin of all groups exhibited more severe microleakage than the occlusal margin, and this observed difference was found to be statistically- significant at the 0.001 level in all cases. When leakage occurred in restorations using glass ionomer cement as a base material in combination with composite resin, the dye penetration was invariably at the tooth-restoration interface. A sectioned surface from group KS showing the composite and glass ionomer bond remaining impervious to dye can be seen in Fig. 6 .
The occlusal margin The Ketac-Fil restorations (group K) appeared to provide the least reliable seal at the occlusal margin. The ability of Ketac-Fil to prevent leakage in less than 20 per cent of cases is in marked 27
contrast to the results achieved using the other two restorative techniques, where leakage was prevented in more than 50 per cent of cases. When the results for groups KS and VS were compared, no statistical difference could be found.
The gingival margin Cavities restored with Ketac-Fil showed a significant difference in the pattern of gingival leakage when compared with groups KS and VS. The two most severe grades of leakage, W and T, were scored for more than 90 per cent of surfaces examined in group K, compared with 54 per cent and 62 per cent for groups KS and VS, respectively. When the microleakage results for groups KS and VS were compared, no statistical difference could be found. From observation of the leakage patterns in Fig. 5, although the two groups showed the same ability to completely prevent leakage, group KS showed a much lower incidence of the worst grade of leakage, grade T, than did group VS. Discussion General observations No restorative technique tested in this investigation completely resisted microleakage at either the occlusal or the gingival margin, and in all techniques the gingival margin exhibited significantly greater leakage than the occlusal margin. These two findings are consistent w h those of the vast majority of other microleakage studies. The significance of the difference found in the behaviour of occlusal and gingival margins within each restorative technique also confirms the suitability of cervical restorations in testing materials for use at different sites on the tooth surface. A hrther finding of this investigation, that no restorative technique tested was capable of providing a predictable seal at the gingival margin is also consistent with other similar s t u d i e ~ . ' ~ - ' ~ This consensus emerged despite vast differences in materials, cavity design and some areas of the experimental regimen, for example thermal cycling. Although the temperatures used for thermal cycling were very similar, the amount of thermal cycling varied from nonezo to some studies with several h ~ n d r e d ~ and ~ - ~another ~ . ~ several ~ * ~ ~thousand," as used in this study. That the interface between composite resin and glass ionomer cement remained uniformly impervious to dye in the sandwich restorations is in agreement with the observation of other^.^',^^ One studya7did report leakage between composite resin and glass ionomer cement, although only when extensive dye penetration had occurred at the 28
dentine wall and combined with cracks and fissures in the glass ionomer cement. These observations of the composite resin and glass ionomer cement bond remaining generally intact appear consistent with the reports of values for bond strengths between composite resin and etched glass ionomer in vitro which exceed the cohesive strength of the ~ e m e n t . ' ~It*is~also ~ * interesting ~~ to note that etch times for the glass ionomer cement bases varied considerably, from 20 ~ e c o n d s ' to ~ *60 ~ ~seconds2' as used in this study. This variation did not, however, appear to have any correlation with reported dye penetration at the composite resin and glass ionomer cement interface. The absorption of basic hchsin dye into the bulk of the Chemfil restorative material in this study contrasted with the behaviour of Ketac-Fil. This variation in behaviour occurred despite both being water-hardening cements. It can only be speculated, therefore, that some other difference in the two materials, such as the type of polyacid used or packaging and mixing characteristics, was the cause. Although no other reported microleakage study has included Chemfil restorations to enable comparison, a comprehensive clinical study of Chemfil reported no cases of gross surface staining over two years.26 More than 20 per cent of the Class V restorations observed did, however, show slight or obvious surface staining over the same time period.
The occlusal margin In this investigation, the glass ionomer restorative cement, Ketac-Fil, provided the least reliable seal occlusally when compared with the two sandwich restorations in which the enamel margin was restored with composite resin. This finding is in agreement with the few other microleakage studies which have cited data for the occlusal margin16~z2 and appears to confirm the superiority of the acidetched composite resin bond over the glass ionomer bond with enamel. When the two sandwich restorations were compared, microleakage results at the occlusal aspect were not significantly different, a finding again in agreement with other similar investig a t i o n ~ . It' ~appears, ~ ~ ~ ~therefore, ~~ that within the limits of this study differences in bonding resins and glass ionomer base materials do not greatly influence the reliability of the etched enamel bond with composite resin. The gingival margin As with the occlusal margin, cavities restored with Ketac-Fil showed significantly greater leakage at the gingival aspect than those restored using a glass ionomer base combined with a veneer of composite Australian Dental Journal 1992;371.
resin. This finding concurs with that of another similar study.I6 When the two sandwich restorations investigated in the present study were compared, no significant difference could be found in their leakage patterns. These results are in agreement with some studies20but conflict with others where differences were f o ~ n d . ' ~In. 'the ~ only other study'8 where both Vitrabond and Ketac-Bond were compared in the sandwich restoration, Ketac-Bond exhibited significantly greater gingival microleakage than Vitrabond. The significance level used was 0.05, a substantially lower level than the 0.001 used in the present study, which may explain the conflicting result. Despite some differing findings, most microleakage studies investigating the sandwich restoration concluded that although leakage at the gingival margin was not completely eliminated, it could be significantly decreased by this restorative t e ~ h n i q u e . ' ~ , ' ~ . ' ~ . ~ ~ It should be noted that these conclusions are based solely on laboratory testing and substantiation by clinical trials would be desirable. Although results of clinical microleakage studies are rarely reported, it has been suggested that marginal staining occurring with time around restorations may indicate the presence of microleakage clinially.^^ Observations of marginal staining at one year showed 8 per cent of Ketac-Fil cervical restorations with marginal staining, compared with two per cent of Silux restorations bonded to etched enamel and a glass ionomer cement base.27,28No distinction was made, however, between staining occurring at occlusal or gingival margins, so although the sandwich restoration performed slightly better than Ketac-Fil alone, any comparison with the present study is difficult.
Conclusions Within the limits of this investigation, none of the restorative techniques studied completely sealed either the occlusal enamel margins or the gingival margins positioned on the root surface. In all instances, however, the gingival margin exhibited a less reliable seal than the occlusal margin. The sandwich restoration using either Vitrabond or Ketac-Bond produced a superior seal at both occlusal and gingival margins when compared with the glass ionomer restorative cement, Ketac-Fil. Acknowledgements Assistance with statistical analysis by Associate Professor R. W. Bryant, preparation of histograms by Dr D. M. Roessler, and with specimen preparation by Dr N. Andrews is gratehlly acknowledged. Australian Dental Journal 1992;371
This project was partly supported by a grant from the Australian Dental Research Fund Trust.
References 1. Going RE. Microleakage around dental restorations: a
summarizing review. J Am Dent Assoc 1972;84:1349-57. 2. Bauer JF, Henson JL. Microleakage: a measure of the performance of direct filling materials. Oper Dent 1984;9:2-9. 3. Crim GA, Garcia-Godoy F. Microleakage: the effect of storage and cycling duration. J Prosthet Dent 1987;57:574-6. 4. Sognnaes RF, Wolcott RB, Xhonga FA. Dental erosion. Erosion-like patterns occurring in association with other dental conditions. J Am Dent Assoc 1972;84:571-6. 5 American Dental Association Council on Dental Research and Council on Dental Therapeutics. Report of symposium: Root surface caries. J Am Dent Assoc 1983;106:496. 6 Hembree JH, Andrews JT. In siru evaluation of marginal leakage using an ultraviolet-light-activated resin system. J Am Dent Assoc 1976;92:414-8. 7 Hembree JH, Andrews JT. Microleakage of several acidetch composite resin systems: a laboratory study. Oper Dent 1976;1:91-7. 8. Ortiz RF, Phillips RW, Swartz ML, Osborne JW. Effect of composite resin bond agent on microleakage and bond strength. J Prosthet Dent 1979;41:51-7. 9. Crim GA, Swartz ML, Phillips RW. An evaluation of cavosurface design and microleakage. Gen Dent 1984;Jan-Feb:56-8. 10. Qvist V. Marginal adaptation of composite restorations performed in vivo with different acid-etch restorative procedures. Scand J Dental Res 1985;93:6&75. 11. Torstenson B, Brannstrihn M. Composite resin contraction gaps measured with a fluorescent resin technique. Dent Mater 1988;4:238-42. 12. American Dental Association Council on Dental Materials and Devices. Status report on the glass-ionomer cements. J Am Dent Assoc 1979;99:221-6. 13. American Dental Association Council on Dental Materials, Instruments and Equipment. Dentin bonding systems: an update. J Am Dent Assoc 1987;114:91-5. 14. McLean JW, Powis GR, Prosser HJ, Wilson AD. The use of glass-ionomer cements in bonding composite resins to dentine. Br Dent J 1985;158:410-4. 15. Wenner KK, Hawkins IK, Smith CD, Fairhurst CW. Microleakage of root restorations. J Dent Res 1985;64:Abstr 71:182. 16. Srisawasdi S, Boyer DB, Reinhardt JW. The effect of the removal of the smear layer on microleakage of class V restorations in vim. Dent Mater 1988;4:384-9. 17. Prati C, Montanari G. Comparative microleakage study between the sandwich and conventional three-increment techniques. Quintessence Int 1989;20:587-94. 18. Prati C, Nucci C, Montanari G. Microleakage of mixed class V restorations placed with GIC bases or dentin bonding agents. Schweiz Monatsschr Zahnmed 1989;99:1117-21. 19. Schwartz JL, Anderson MH, Pelleu GB. Reducing microleakage with the glass-ionomer/resin sandwich technique. Oper Dent 1990;15:186-92. 20. Phair CB, Zidan 0, Gomez-Marin 0, Han S. Marginal leakage in the composite resinlglass-ionomer cement class V restoration. J Dent Res 1987;66:Abstr 45112. 21. Mathis RS, DeWald JP, Moody CR, Ferracane JL. Marginal leakage in class V composite resin restorations with glassionomer liners in wirro. J Prosthet Dent 1990;63:522-5. 29
22. Gordon M, Plasschaert AJM, Stark MM. Microleakage of several tooth-coloured restorative materials in cervical cavities. A comparative study in vim. Dent Mater 1986;2:228-31. 23. Gordon M, Plasschaert AJM, Soelberg KB, Bogdan MS. Microleakage of four composite resins over a glass-ionomer cement base in class V restorations. Quintessence Int 1985; 16:817-20. 24. Sneed WD, Looper SW. Shear bond strength of a composite resin to an etched glass ionomer. Dent Mater 1985;1:127-8. 25. Hinoura K, Moore BK, Phillips RW. Tensile bond strength between glass ionomer cements and composite resins. J Am Dent Assoc 1987;114:167-72. 26. Knibbs PJ, Plant CG, Pearson GJ. A clinical assessment of an anhydrous glass-ionomer cement. Br Dent J 1986; 161:99- 103.
30
27. Tyas MJ, Twhey A, Clark J. Clinical evaluation ofthe bond between composite resin and etched glass ionomer cement. Aust Dent J 1989;34:1-4. 28. Tyas MJ, Burns GA, Byrne PF, Cunningham PJ, Dobson BC, Widdop FT. Clinical evaluation of Scotchbond: one year results. Aust Dent J 1986;31:159-64.
Address for correspondenceheprints: Department of Operative Dentistry, Faculty of Dentistry, The University of Sydney, United Dental Hospital, 2 Chalmers Street, Surry Hills, New South Wales, 2010.
Australian Dental Journal 1992;37:1.
