Time required to remove totally bonded tooth-colored posterior restorations and related tooth substance loss

Dent Mater 11:34-40, January, 1995

Time required to remove totally bonded tooth-colored posterior restorations and related tooth substance loss Ivo Krejci, Claudia M. Lieber, Felix Lutz of’Pre~~entive Dentistry, Periodontology and Curiology. Dental Institute, Zurich University, Zurich, SWITZERLAND

Departnzent

ABSTRACT Objecfives. The study was conducted to measure the time required to remove large totally bonded tooth-colored posterior restorations and related tooth substance loss. This information was collected to determine if there were differences between bonded restorations and conventional restorations. Methods. Molars were restored with the following materials: amalgam, composite, glass ionomer cement or glass ceramic cusp coverages. After submitting them to an in vitroaging process, they were attached to a lower jaw model in a phantom head. Six dentists removed the test restorations under standardized, quasi-clinical conditions. Results. While the amalgam was completely removed, some glass ionomer cement vestiges were found. In the glass-ceramic group, the margins were covered with remnants of the composite luting agent in several places. The most restorative material was left in the composite group. The loss of tooth structure after removal of amalgam, glass ionomer cement, composite and glass ceramic was 17.6 mm3, 19.6 mm3. 39.9 mm3 and 41.8 mm3. respectively. When comparing the removal time, the glass ionomer group scored best with an average of 11.9 min followed by the amalgam group with an average of 15.2 min, followed by 24.9 min for the composite group and 30.4 min for the glass ceramic group. Significance. The removal of totally bonded posterior restorations made of composite and ceramic is more technically demanding and more time-consuming than the removal of glass ionomer and amalgam restorations. INTRODUCTION Approximately 70% of operative dentistry is “re-dentistry” (Mjor, 1981). The longevity of low-quality restorations in patients not included in preventive programs is short (Mjor and Toffenetti, 1992). However, apart from isolated cases, restorations that last a lifetime cannot be expected even under optimal prevention with excellent treatment quality (Letzel et al., 1989). Dental restorative materials must, therefore, meet the demand of easy replacement. Amalgam is macro-mechanically retained in the cavity with the aid of 34 Krejci et al./Removal of posterior restorations

undercuts, furrows, pits and pins (O’Brien, 1989). Thus, its removal is relatively simple because after cutting through the restoration, fragments can usually be removed from the cavity in toto. The future of amalgam, however, is uncertain because of controversy surrounding its component, mercury and more posterior tooth-colored restorations, such as composite or ceramic inlays and overlays, are being placed (Krejci et al., 1991; 1994). These restorations are bonded to minimize bacterial penetration (Brannstrom, 19841,to ensure retention, and to prevent marginal discoloration. In order to achieve total bonding, dentin adhesives are used, omitting the placement of a base. It was the purpose of this study to determine the time required to remove large, totally bonded posterior restorations made of composite and ceramic and to compare the results to the removal time of amalgam and glass ionomer cement restorations. In addition, the loss of sound tooth structure after removal was quantitated and the technical problems related to the procedure identified. MATERIALS

AND METHODS

Caries-free extracted human lower molars (n = 24) were collected and stored in 0.5% thymol-solution at 37°C for at least 3 mon. They were randomly assigned to four groups and restored with one of the following materials: amalgam (Valiant, Batch No. 931107, De Trey/Dentsply, Konstanz, Germany), composite (Brilliant Dentin, Batch No. 9301529, Colt&e AG, Alt&itten, Switzerland), glass ionomer cement (Ketac F’il, Batch No. 0020, Espe GmbH, Seefeld, Germany). and glass ceramic (Dicer MGC, Light I 10 L, De Trey/Dentsply!. After cleaning the crowns with pumice, standardized MOD preparations were cut using 80 pm diamond burs (Inlay Set. Intensiv SA, Lugano-Viganello, Switzerland) and finished using 25 pm diamond bum (Inlay Set, Intensiv SA) (Reller et al., 1989). The occlusal box was 3.2 mm deep, 2.8 mm wide in the lateral area and 5.0 mm wide in the central fissure. One proximal box, totally surrounded by enamel, was prepared to

a depth of 5.4 mm and a width of 4.0 mm. The other proximal box, with the cervical margin in dentin, was 6.4 mm deep and 5.0 mm wide. Finally, both buccal cusps were reduced by 2.0 mm, resulting in a MOD cusp coverage preparation. A 3” occlusally divergent preparation was made for each of the four groups so that the silicone impression (Model Silicone, De Trey/Dentsply), needed to determine the cavityvolume, could easily be removed. In the amalgam group, additional retention grooves were prepared with a small bur (diameter 0.7 mm, Les Fils D’Auguste Maillefer SA, Ballaigues, Switzerland): two vertically in each of the two proximal boxes, two horizontally at the transition between the occlusal floor and lingual cusps, and one on the buccally shortened surface. In the glass ceramic group, which was restored using the CEREC method (Siemens AG, Benzheim, Germany) the transition between the occlusal box and the buccal area had to be prepared in a crescent shape to meet CEREC specifications. Various methods to quantitate the loss of sound tooth structure after removal of the restorations were evaluted in pilot studies. Weighing of the teeth proved to be unsuitable, as they were subjected to continuous weight loss because of water evaporation, even under N,-protection or in a closed glass container. In addition, as the preparation of the test teeth for the stress cycles proved impossible without any loss of tooth structure in the root area, all gravimetric methods were excluded. The replica technique, on the other hand, was suitable. This method recorded the dimensions of the cavity prior to restoration with the help of a special silicone (Model Silicone, De Trey/Dentsply). Pilot studies in which six samples were measured in a 3D-scanner (Roulet, 1987; Krejci, 1992) confirmed the stability and durability of the material: a bar of 14.01 pm indented into the material could be measured in all six samples after 3 mon with an accuracy of 1.0 + 1.0 pm. The two remaining lingual cusps were used to achieve exact placement of the silicone impressions onto the sample teeth. Additionally, an occlusally diverging reference furrow of 1.5 mm depth and width was shaped into the outer buccal margin of the cavity with a 25 pm diamond bur, producing a defined and reproducable three-point support. Impressions were then made by placing a steel matrix (Automatrix, De Trey/Dentsply) loosely around the prepared tooth, and the cavity was carefully filled with the impression material, avoiding bubbles. Ten min later, the impression was removed and was reduced with a scalpel under the stereomicroscope at 25x magnification to about 0.1 mm measured from the cavity margin to avoid excess material. Seven impressions were taken per sample. In addition, another impression of the cavity (President Jet light body, Coltene, AG) was cast with epoxy resin (Stycast 1266 Part A & B, Grace N.V., WesterlooOevel, Belgium). This cavity model served as a guide for the dentists during the removal of the restorations. Prior to restoring the cavities, the surfaces of the buccal reference furrows were marked with a red felt tip pen under the stereomicroscope and filled with composite (Brilliant Dentin, Colt&e) to avoid accidental removal of these reference points when taking out the restoration. In the amalgam group, cavities encircled with a metal matrix (Matrice en acier inoxidable KFG, Fltickiger & Huguenin, La Chaux-De-Fonds, Switzerland) and wedges (Ahorn Interdentalkeile, Hawe-Neos Dental AG, Gentilino,

Switzerland) were restored with the amalgam. The amalgam cusp coverages were pm-burnished, cut, post-burnished and stored in water. After 24 h, the occlusal surface was finished and polished with rubber cups (Shofu Dental Corporation, Menlo Park, CA, USA), and flexible discs (Soflex, 3M Dental Products, St. Paul, MN, USA) (Geiger et al., 1989) were used approximally and occlusally under the stereomicroscope. In the glass ionomer cement group, a similar metal matrix was fixed with two interdental wedges (Ahorn Interdentalkeile, Hawe-Neos Dental AG). After the cavity was filled with the glass ionomer cement, it was immediately covered with a protective varnish (Ketac-Glaze, Espe GmbH), which was light-cured (Coltolux II, Colt&e AG) for 10 s. The varnish prevented the material fi-om dehydrating during the initial setting phase. After the restoration had been stored in water for 24 h, it was finished and polished under water-cooling. Occlusally, 15 um diamond bum (Composhape Set, Intensiv SA) were used, and proximally and occlusally, flexible discs (Soflex, 3M Dental Products) were applied. For the indirect composite group, the cavity was encircled with a transparent matrix (Hawe-Molar-bands Transparent, Hawe-Neos Dental AG) and two light-transmitting wedges (Ahorn Interdentalkeile, Hawe-Neos Dental AG). The cavity was coated with a separating gel (Haereus Kulzer GmbH, Wehrheim/TS, Germany), then filled with a fine-hybrid composite (Brilliant Dentin, Colt&e AG). The composite was placed in two layers that were light-cured (Coltolux II, Colt&ne AG) for 60 s each. A&r removal from the cavity, the overlay was post-cured using light and heat (D.I., 500~Ofen, Colt&e AG). The cavity was cleaned with water spray, and the enamel margins were etched for 30 s (Etchant Gel Kit, Colt?me AG). The 1:l mixture of Sealant Base and Catalyst (A.R.T. Bond, Colti?ne AG) was rubbed into the dentin surface for 60 s with a brush. The cavity was air-dried for 60 s, and then covered with Adhesive (A.R.T. Bond, Colt&e AG), that was pre-cured (Coltolux II, Colt&e AG) for 60 s. The inner surface of the overlay was slightly roughened using a 25 pm diamond bur Kiejci, 1992) before applying the adhesive (Duo Bond Kit, Coltone AG) and the composite luting agent (Duo Cement, Coltone AG). A&er inserting the overlay into the cavity, the composite luting agent was light-cured (Coltolux II, Coltene AG) from the cervical area through the light wedges and along the proximal, oral, buccal and occlusal surfaces each for 60 s. Finishing and polishing were done as in the glass ionomer cement group. The glass ceramic overlays were made using the CEREC method. The cavity was prepared for the optical impression using Dicer MGC Imaging Liquid (De Trey/Dentsply) and a thin, homogeneous layer of opaque powder (CEREC Powder, VITA, H. Rauber GmbH & Co. KG, Bad Sackingen, Germany). The CEREC intraoral camera was used to take an optical impression, and after correcting the data manually, the overlay was milled out of a Dicer MGC glass ceramic block (Light shade, I lOL, De Trey/Dentsply) (E-Version, Model No. 2987519 D 3255, Series No. 01718 S 02, Software COS 2.1, Siemens AG, Benzheim, Germany). Milling was carried out with a D 64 diamond disk. After the milling procedure, the overlay was manually adjusted. The buccal and proximc-axial enamel margins were bevelled in a 0.5 mm crescent shape with 40 and 15 pm diamond bum (Fett et al., 1991; Htirzeler et al., 1990). The enamel margins were then etched for 30 s

Dental Materials/January

7995

35

Probe, Perio-Probe, Mouth mirror 14.0

Amal - Comp / Cer - GI

2.5

Amal - Comp i Cer - GI

2.5

Amal - Comp i Cer GI

1.5

Comp

7.5

Comp - Amal / GI - Cer

2.5

Comp - Amal / GI - Cer

Amal / GI - Cer

Hand excavator Red contra-angle & bur ejector (Micro-Mega SA, Besanqon Cedex. France) Turbtne (S 30. Bren Air SA, Biel, Switzerland) Metal matrix band (Matrice en acrer rnoxrdable KFG, Fluckiger & Huguenrn. La Chaux-De-Fonds,

Swttzerland)

Crown and collar scissors (Aesculap AG, Tuttlingen. Germany) Wooden wedges (Ahorn Interdentalkeile, Hawe-Neos Dental AG. Gentilino, Switzerland)

(De Trey Etch, De Trey/Dentsply), rinsed for 30 s with water spray and air-dried. The primer (Prisma Universal Bond 3, De Trey/Dentsply) was applied to the entire dentin cavity and air-dried after 30 s. A thin layer of adhesive (Prisma Universal Bond 3) was applied to the enamel and dentin and subsequently light-cured for 60 s (Coltolux II, Colt&e AG). A transparent matrix (Hawe-Molarbands Transparent, Hawe-Neos DentslAG) and two light wedges (Hawe luciwedge, Hawe-Neos Dental AG) were placed around the cavity The inner surfaces of the overlays were etched for 30 s with ammonium-bifluoride (Dicer Etching gel, De Trey/Dentsply), rinsed with water spray and air-dried. The Silane Activator and Silane Primer (De Trey/Dentsply) were mixed for 5 min and the mixture was applied to the etched inner side of the overlay After 30 s, the overlay was air-dried, and composite luting agent (Dicer MGC Dualhartend, De Trey/Dentsply) was applied to both the restoration and the surface of the cavity. After insertion of the overlay into the cavity, the composite luting agent was light-cured iColtolux II, Coltene AG) for 60 s each from the cervical through the lighttransmitting wedges, approximal, oral, buccal and occlusal. The luted restoration was finished according to the technique described for the glass ionomer cement group. The restorations were stored in water at 37°C for 1 wk. then they were loaded in an in vitro chewing apparatus to simulate aging (Krejci, 1992) by exposing them to the following mechanical, thermal and chemical loads: 1,200,OOO chewingcycles, max. 49 N; 3,000 thermal cycles, 5”-55°C dwell time 2 min in each bath; 300 min. 75%~aqueous ethanol solution. This cycle corresponded to about five years of clinical function (Krejci, 1992). Removal of the restorations was carried out in a phantom head (Teaching Simulator, J. Morita Europe GmbH, Frank&&Main, Germany), which had been fixed to a patient chair (chair EMDAquasar, EMDA, Frankfurt/M ain,Germany; Unit: EMDA varidul with EMDA Fontana FHSV, EMDA, Operation light: Starlite D44, Bitter, Karlsruhe, Germany). The restored tooth was fixed into the lower jaw model at the site of tooth 36 so that the cervical margin of the proximal box was 1.5 mm coronal to the alveolar bone margin, and the restored cusp was placed buccally The proximal box with the cervical margin in dentin was placed randomly either mesially or distally A radiograph (Heliodent 70,70 kV, 7 mA, Siemens AG, Benzheim, Germany) was taken of this situation.

36 KrejY et aLi Removal of posterior restorations

Narrow and wide suction tips (Dijrr Dental GmbH, Bietigherm-Bissrgen. Germany) X-ray film (Kodak Ultra-speed, Eastman Kodak Company. New York. USA) Epoxy model of the unfilled cavity Rotating Instruments: Cylindrical preparation and finishing diamonds (Inlay Set. lntensrv SA, Lugano-Viganello,

Switzerland)

- Large and small tungsten carbide bur (Komet, H 33 L. FG 021 und H 7, FG 006, Gebruder Brasseler GmbH. Lemgo. Germany)

- Flame-shaped finishing dramond (Composhape. FG 4205 L lntensiv SA, Lugano-Viganello,

Switzerland)

Each of the six dentists with varying clinical experience (Table 1) removed one restoration per group in two sessions. The dentists were informed that the aim of the studv was the total removal of the restorations as well as finishing to prepare the cavities for n new restoration made out of the same material as fast as possible. They were allowed to examine the tooth clinically for as long as necessary and the) used the radiographs and the epoxy models of the cavities for reference. To become familiar with the phantom situation. each clinician removed a composite MOD restoration prior to the removal of thr test restorations in both sessions. Each dentist worked with a trained chairside assistant and had ;t set of standardized instruments (Table 2 1. A new set of burs was used for each restoration. The instruments used and thts time required for each stage were recorded. After removing the restorations according to the guidelines. the test tooth and the adjacent artificial teeth were collected. The prepared cavities were examined under a stereomicroscope (M5, Wild-Leitz AG, Heerbrugg, Switzerland) and were described qualitatively with respect to preparation form and remaining restorative material. These remnants were then removed by the test coordinator under the stereomicroscope using 15 ~_undiamond burs (Inlay Set, Intensiv SA, LuganoViganello, Switzerland). The adjacent teeth were scored with regard to damage caused by drilling using a reticle-equipped stereomicroscope according to the following criteria: Score 0. Tooth not touched. Score 1: Defect less than 0.5 mm deep, can

A

25.6

25.6

43.4

48.8

B

17.8

30.2

39.5

65.0

C

12.4

16.3

31.7

21.7

D

13.2

7.7

60.4

36.4 44.9 22.5

E

20.9

20.9

42.6

F

15.5

17.0

33.3

17.5f 5.0

19.6f 7.8

41.8f 10.3

39.9f 16.6

112k38

254~104

227r80

Mean+S.D.

Relative Percentage ofRestaratiie Rf3moval 100

A

06:56

08:16

12%

16:07

B

07%

II:09

11:53

19:14

C

07:49

1155

41:33

4o:oo

0

18:18

05:45

25:38

3259

To better illustrate the values, the volumes of a sound and a prepared crown were additionally determined to relate the results to a non-prepared and a prepared tooth. For this part of the study, a similar silicone impression (President Jet light body, Colt&e) was taken of three sound lower molars from the crown to the enamel-cement margin. These impressions were weighed, filled with the same silicone material to the level of the enamel-cement margin, reweighed and the averages were calculated. The same procedure was carried out with two prepared teeth: the first had a conventional overlay preparation and the second a CEREC specific overlay preparation. Statistical analysis was done by using a statistical analysis program (Stat View IV Abacus Concepts, Inc., Berkeley, CA, USA) on a work station (Quadra 800, Apple Computer Inc. Cupertino, CA, USA). The volume loss data was statistically analyzed using Kruskal-Wallis test, followed by Mann-Whitney U-tests to highlight significant differences between specific groups. The same statistical procedure was applied for the removal time data.

RESULTS The amounts of dental tissue lost during the

removal of the restorations are summarized in Table 3. The times required for the removal of the restorations are listed in 15.2.t9.3min 11.9*4.7min 24.9*11.4min30.4+11.5min Table 4. Figs. l- 4 illustrate the steps chosen MeankSD by each dentist and the time required for each step in the four test groups. The semiquantitative scores of the amount of the remaining restorative material in the cavities are listed in Tables 5a-c. Undercuts were be polished and fluoridated. Score 2: Defect deeper than marked with an asterisk, as they did not conform to a cavity 0.5 mm, would need an adhesive restoration. prepared for an overlay. With the amalgam cusp coverages, The initial impressions were then weighed using precision no remaining material was found. Only a few remnants were scales (AT 261 Delta Range, Mettler-Instruments, N&&onleft in the cavities of the glass ionomer group. In contrast, the Uster, Switzerland). The teeth were air-dried, the buccal composite and glass ceramic groups had several remnants of composite stop removed with a probe, and the preparation various dimensions, particularly in the proximal boxes and surface was covered with a thin layer of low-viscosity silicone along the cavity margins. In one of the glass ceramic samples, impression material President Jet light body, Colt&e). The the pulp chamber was accidentally opened. buccal reference point and the cusp tops were not covered by Table 6 shows the scores given to each of the adjacent teeth. the material. The silicone impressions were then placed on All defects in these teeth were within enamel only. their respective teeth. During setting, the impressions were fixed in place with standardized pressure. After setting, the DISCUSSION excess was carefully removed under the stereomicroscope using a scalpel. These impressions were again weighed, and The optical contrast to the tooth substance and the purely mechanical retention of amalgam restorations allow their the difference from the initial weight recorded. The volumetric difference was calculated using the specific weight of the relatively fast and careful removal. These characteristics were confirmed in the present study. No remaining amalgam was light body silicone f?om the formula: found in any of the specimens in the amalgam group. There was slight tooth structure loss though the restorations in this volume = mass / 1.28 g/mm” (11 study had no secondary caries which had to be removed. Therefore, the enlargement of the cavity after amalgam The maximal deviation of the measurements in a group was removal, as found in clinical studies (Barbakow et al., 1988), 1.29 mm?. E

21:03

18:39

27152

F

29%

15:30

30:22

28:31

45:29

Dental Materials/January

1995

37

Dentist

0

5

10

15

20

25

30

35

40

45

Dentist

l"l"l"l"l"l"l"I"I"Ic

0

5

10

15

20

25

30

35

40

[min]

brnl

Mean

Mean

Fig. 7. Time required to remove amalgam

Dentist

45

1”1”1”1”1”1”1”J~‘I”l~

0

5

10

15

20

restorations.

25

Fig. 2. Time required to remove glass ionomer

30

35

40

45

Dentist

5

10

15

20

25

30

~~~I”I”I”I”I”I’~~‘~I”~

l”l”l’~l”l”l”l”I”I”lc [min]

A B C D E F

Fig. 3. Time required to remove composite

0

cement restorations.

restorations.

to remove glass ceramic

40

45

* [mrnl

A B C D E F

Fig. 4. Time required

35

restorations.

the molar areas with no or insufficient optical magnification, the lack of selection of rotating instruments and the need to mechanically finish the cavity. The probable reason why only a few rem Proximal Box nants of glass ionomer cement were found along Cervical in Enamel the cavity margins was because of the porous _ _ low Short Axial Wall nature of the material and its opaque surface. low Cervical low This enabled the glass ionomer cement to he _ _ low Long Axial Wall easily recognized a&r drying the restored tooth. _ low * Pulpo-Axial Adhesion to the tooth surface seemed to have straight straight straight straight Margin jagged jagged decreased after longer functional stresses, so that most parts of the filing could be removed using ckolusatBox hand instruments of the material in toto. Thew BuccalWal13 factors and the low hardness of the material IOW &80X low allowed a fast removal of the glass ionomer Od Wall restoration with an amount of tooth substance loss comparable to that of amalgam, despite Proximal Box having a tooth-like color and primary adhesion Cervical in Dentin to the dental tissues. i _ _ Short Axial Wall low The removal of the composite as well as the _ high low low Cervical glass ceramic overlays caused a loss oftooth ma_ low Long Axial Wall terial, which was about twice as high compared low Pulpo-Axial low to amalgam and glass ionomer cement straight straight straight Margin jagged jagged jagged restorations. Where remnants of filling material were found after restoration removal, they were carefully removed under the stereomicroscope by the test supervisor before measuring cannot be explained by the removal of carious tooth substance the amount of tooth material lost. Thus, the tooth material only, but it obviously reflects technical problems, too. lost was in fact smaller than that which would have resulted Clinically, these problems may be due to the limited view in with the removal of all filling remnants.

38

Krejci et al.1 Removal of posterior restorations

Proximal Box Cervical in Enamel Short Axial Wall Cervical Long Axial Wall Pulpo-Axial Margin

low low -

low low

straight

jagged

straight

low low

low low

-

low high

high

jagged

high high

-

* *

-

low low high

medium low

low

-

straight

straight

lOW

low

low

high -

high

-

medium jagged

-

jagged

0cGlusalBox Buccal Wall Box Oral Wall Proximal Box Cervical in Dentin Short Axial Wall Cervical Long Axial Wall Pulpo-Axial Margin

* *

straight

-

*

jagged

Special Findings

Proximal Box Cervical in Enamel Short AxialWall Cervical Long Axial Wall Pulpo-Axial Margin

Occlusal Box Buccal Wall

straight

jagged

low

medium medium

straight

straight

*

-

VBrY invasive WED.

medium medium low

-

-

-

-

low low

jagged

jagged

jagged

low* straight

??

-

-

-

BOX

low

-

-

Oral Wall

low

-

low

Proximal Box Cervical in Dentin Short AxialWall Cervical Long Axial Wall Pulpo-Axial Margin

medium

*

-

high*

-

-

high

low high medium

low

jagged

jagged

*

-

low -

straight

jagged

straight

Special Findings pulpal

-

-

medium medium medium jagged

The slightly lower substance loss in the glass ceramic group compared to the composite group maybe explained by two factors: the harder glass ceramic material was more readily distmguished from the tooth than the composite, which subjectively was perceived as “tooth-like” by the test dentists. Furthermore, the glass ceramic restorations with a limited choice of shades was easier to discern from the tooth material. More tilling material remnants were recorded in the composite group than in the glass ceramic group. On the other hand, the material removal was more aggressive in the glass ceramic group, and in one specimen, the pulp was exposed at the transition of the pulpal wall to the cervical floor. Comparisons of the present results to other studies are difficult. Other results on the subject WIlar et al., 1992; Robinson and Milk, 1989; Davies et al., 19901, quantitated only the surfaces ofthe cavities, but not their volume. In addition, the restored teeth were not placed in a phantom head. Thus, the clinical situation was not simulated, and the teeth were accessible from all sides. Furthermore, only small mesio-distal composite restorations and composite inlays without total bonding were removed. Ceramic, amalgam and glass ionomer restorations were not considered. Finally, the removed restorations were not submitted to an aging process. Nevertheless, these authors also concluded that the substitution of toothcolored posterior restorations would cause problems. In summary, the removal of tooth-colored adhesive posterior composite and ceramic restorations was considerably more difficult compared to the removal of amalgam and glass ionomer cement. The long time required for the removal and the inconsistent, extensive loss of tooth material were the major problems. These problems were probably caused by the poor visual and tactile differences between the restorative materials and the tooth, and by total bonding to enamel and dentin. A further problem was restorative material remnants, especially at the cavity margins. From the perspective of this study, the following recommendations can be made: 1) clinically acceptable, adhesive ceramic and composite restorations should be repaired, rather than removed in toto; 2) total bonding should not be applied in posterior restorations. Instead, the placement of a base as demarcation between restoration and cavity floor makes any later replacement easier; 3) if replacing adhesive tooth-colored restorations, after etching enamel margins for the new restoration, the etching pattern must be carefully inspected to identify any remaining restorative material; and 4) dental fees should include the increased time needed to remove composite or ceramic toothcolored posterior restorations. Dental Materials/January 7995 39

Amalgam Premolar Molar

1 1

0 1

1

1

1

1

1

1

1

1

Premolar Molar

0 1

1 1

1 I

0 0

1 1

0 1

Composite Premolar Molar

1

1

1

0

2

2

1

1

1 1

2 2

Glass Ceramic Premolar Molar

1 0

1

1

2

1

1 0

0 1

0 2

GIC

In the future, color indicators, which allow a reversible contrast coloring ofthe composite luting agent and composite restorative materials, should be developed. An alternative could be the use of selective removal techniques, such as lasers (Liesenhoff Pf nl., 1990), in place of non-selective. rotating instruments. Received April 2 1, 1994 / Acwpted June 29. 1994 Address correspondence and reprint reqursts to: Iv0 Krejci Department of I’revontivt~ Ikntistq Periodontology and Cariolog~ Dental Institute. %urich University Plattenstrasse 11 CH-8024 Zurich Switzerland

REFERENCES Barbakow E GaberthiielT, Lutz F, Schiipbach P (1988). Maintenance ofamalgam restorations. Quintessence Int 19:861870. Brk-mstrijm M (1984). Communication between the oral cavity and the dental pulp associated with restorative treatment. Oper Dent 9:57-68.

40 Krejci et a/./ Removal of posterior restorat/ons

Davies BR, Milk BJ, Robinson PB ( 1990). The effect of the removal of indirect composite resin restorations on cavitysize. J Dent Res 69:976, Abstr. No. 175. Fett HP, MBrmann WH, Krejci I, Lutz F (1991). The effects of’ short bevels and silanization on marginal adaptation of computer machined MOD inlays. Quintessence Int 22:823829. Geiger F: Reller U. Lutz F ( 1989). Burnishing, finishing, and polishing amalgam restorations: A quantitative scanning electron microscope study. Quintessence Int 20:461-468. Hikzeler M, Zimmermann E, Mijnnann WH (1990 ).Marginale adaptation von maschinell hergestellten onlays ill r&o, Schuviz Monatschr Zahnrned 100: 715-720. Krejci I ( 1992 1. Zahnfarbene Restaurationen. Miinchen: Hanser: Krejci I, Giintert A. Lutz F i 1994 1. Clinical and SEM evaluation of a fine hybrid composite inlay/overlay over 12 months. Quintessence Int 25:403-409. Krejci I. Lutz F, Fiillemann J ( 1991). Zahnfarbene Inlays/Overlays. Schweiz Monatsschr Zuhnmed 102:73-80. Letzel H, van’t Hof MA, Vrij hoef M&IA, Marshall GW Marshall SJ ( 1989). A controlled clinical study of amalgam restorations: survival. failure. and causes of failure. kzt Mafvr 5:115-121. Liesenhoff T Lenz H, Seiler T ( 1990). Root canal preparation with 308 nm excimer laser under simultaneous control b> sp&ral tissue identifkation. In: Innovation andTechnologie en Biologie et Mkdecine. Proceedings of the 2nd World Congress on the Impact of Lasers on Dental Science. May 2% 31. Paris, 38-39. Milk BJ, Robinson PB, Davies BR c1992 ). Effects of thr IXmoval of composite resin restorations on Class I1 cavities. I+ Dent -J 173:210-212. Mjiir lA ( 1981). Placement and replacement of restorations. Oper Dent 6149-54. Mjijr IA. Toff‘enetti F (19%). Placement and replacement 01 amalgam restorations in Italy. Oper Dent 17:70-73. O’Brien WcJ (1989). Dental Materials: Properties and Selection. Chicago: Quintessence Publishing Co.. 245-283. Reller (7, (Niger I? Lutz F ( 1989). Quantitative investigation of different finishing methods in conventional cavity preparations. Quintessence Inf 20:453-460. Robinson PB, Millar BJ ( 1989). The effects of the removal ot composite resin restorations on cavity size. .I Denf Res 68:604, Abstr. No. 366. Roulet J-F ( 1987 1. Degradation of Dental Polymers. Basel: Karger_ 92-113.