| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
|
J Am Dent Assoc, Vol 131, No 1, 43-49.
© 2000 American Dental Association |  |
COVER STORY |
 |
ABSTRACT |
|---|
 TOP ABSTRACT MATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
Methods. The authors placed traditional Class I and Class II bonded and unbonded amalgam restorations in 76 patients. Panavia 21 (J. Morita USA Inc.) was the bonding agent selected, and Aristaloy CR (Englehard Dental) and Tytin (Kerr Corp.) were the amalgam alloys used. Postoperative sensitivity and marginal fracture were evaluated at yearly intervals, for up to three years of clinical service.
Results. At the patients’ appointment for polishing one to two weeks after restoration placement, and at each yearly recall appointment, the authors found no significant difference in postoperative sensitivity between bonded and unbonded restorations for either amalgam alloy (
2 analysis, 
= .05). In addition, there was no significant difference between bonded and unbonded restorations for either amalgam alloy with respect to marginal fracture (analysis of variance and Tukey’s contrasts at = .05). Moreover, no cusp fractures were observed for either bonded or unbonded restorations.
Conclusions. After three years of clinical service, amalgam bonding for traditional Class I and Class II restorations had no effect on postoperative sensitivity or marginal integrity.
Clinical Implications. The merit of using adhesive bonding agents for traditional Class I and Class II amalgam restorations was not demonstrated in this three-year clinical study.
Many dental practitioners are using adhesive agents to bond amalgam to tooth structure.1 The potential advantages of this procedure, as suggested by in vitro test results, are reduced microleakage, which could lead to a reduced incidence of postoperative sensitivity2; a reduced degree of marginal fracture3; increased strength of the prepared tooth4; and retention of restorations in less retentive preparations, with the potential for conserving tooth structure.5 Although these advantages would be of obvious merit, their validity requires confirmation in the clinical environment.
If bonding agents do not exhibit benefits in clinical service, it would be difficult to justify a procedure that is more time-consuming and costly for the dentist and patient than placing unbonded amalgam restorations. Furthermore, the procedure involving the bonding agent used in this study as well as other bonding agents is subject to technical problems, and the potential exists for producing substandard restorations. Placing a uniform layer of the adhesive on cavity walls is difficult, and pooling can readily occur in line angles and at proximal margins. In addition, the amalgam may not be as well-condensed as it should be because of displacement of the adhesive during condensation. Furthermore, the adhesive agent adheres to the condensing instrument, and periodically must be removed during condensation, thereby increasing the time for placement. In addition, the adhesive can become embedded in the amalgam, which can weaken the restoration.6,7 This may be particularly important in highly stressed regions of the restoration. These problems also have been associated with other bonding agents, particularly those that have a high initial viscosity.8
A few clinical studies have examined the potential advantages of bonding agents. In a one-year study, our research group compared bonded with unbonded Class I and Class II restorations. The results showed that bonding did not reduce the incidence of postoperative sensitivity or result in a reduced degree of marginal fracture.9 However, another study demonstrated early success when bonding was used for large, complex restorations, which primarily involves bonding to a single surface.10 In a single-surface application, polymerization shrinkage forces that are perpendicular to the bonding interface between the amalgam and the tooth pull the amalgam toward the tooth surface, which may not affect the bond. However, for traditional Class I and Class II restorations, which involve bonding to two fixed opposing surfaces, polymerization forces can disrupt the bond.
We conducted a study to examine the effectiveness of bonding traditional Class I and Class II amalgam restorations. Specifically, we extended our one-year study9 to three years of clinical service and added another amalgam alloy to the experimental design.
The high-copper–content amalgam Aristaloy CR (Englehard Dental) was selected because previous studies have shown that it exhibits inferior marginal fracture performance to that of most present-day high-copper–content alloys12; therefore, it might lead to more discriminating results. We added another high-copper–content alloy (Tytin, Kerr Corp.) to the study because it has exhibited superior marginal fracture performance compared with that of Aristaloy CR.12 Furthermore, it also has shown a greater propensity for postoperative sensitivity,13 which could lead to better discrimination in evaluating sensitivity.
Patient population.
The Institutional Review Board for Human Research of the Oregon Health Sciences University, Portland, reviewed and approved this investigation. We selected 76 patients from surrounding schools and the community who required two or more amalgam restorations. This resulted in 140 restorations in 38 patients for the Panavia-Aristaloy CR project and 93 restorations in 38 patients for the Panavia-Tytin project. For both projects, 56 patients (74 percent) were in the 14- through 21-year-old age group and 20 patients (26 percent) were in the 22- through 40-year-old age group; the mean age of patients was 19.2 years and the median age was 15.5 years.
Because of the effect of tooth type and restoration size, as shown by Osborne and Gale,14 we attempted to place equal numbers of bonded and unbonded restorations into each category. The table
MATERIALS AND METHODS
TOP
ABSTRACT
MATERIALS AND METHODS
RESULTS
DISCUSSION
CONCLUSIONS
REFERENCES
We selected Panavia 21 (J. Morita USA Inc.) as the bonding agent for evaluation because in vitro tests showed that this material exhibited the greatest bond strength and lowest level of microleakage of four commonly used amalgam bonding agents.11 Recent tests have continued to show the superiority of Panavia 21.8 Therefore, we considered this adhesive to be an excellent example of a potentially successful amalgam bonding agent. 
presents characteristics of the teeth in the study sample at the two-year recall appointment for Tytin and at the three-year recall appointment for Aristaloy CR. As in most clinical studies, we faced the problem of maintaining the original sample throughout the duration of the study. Therefore, the number of patients evaluated for postoperative sensitivity soon after restoration placement was considerably higher than the number of patients evaluated for sensitivity and marginal fracture at one, two and three years.
|
To maintain consistency, only one experienced practitioner placed all of the restorations, and to reduce the influence of patient variation, each patient received both bonded and unbonded restorations.
Postoperative sensitivity and marginal fracture were the two clinical characteristics evaluated. We chose postoperative sensitivity because it has been reported to occur with a small percentage of newly placed amalgam restorations. Marginal fracture was selected because it can be a precursor to secondary caries and, even in the absence of secondary caries, it has been known to lead to premature replacement of the restoration.15 Furthermore, if the bond is substantial, a reduction in the characteristic fracture of amalgam at the restoration margin might be expected.
Postoperative sensitivity. Postoperative sensitivity was recorded at the polish and recall appointments. Teeth that were sensitive before treatment were excluded from the sensitivity data. To approximate complaints of sensitivity that are typically heard in dental practices, we recorded sensitivity as a dichotomous response (that is, yes or no) based on patients’ responses to the following questions.
We used 2 analysis to compare the numbers of teeth evoking a response of postoperative sensitivity for both bonded and unbonded restorations.
Marginal fracture.
We measured the extent of marginal fracture at yearly recall appointments by comparing occlusal photographs of the test restorations with a scale of numbered photographs of restored teeth that illustrates increasing amounts of marginal fracture (Figure 1).16 Each restoration was assigned the number corresponding to the restoration in the scale that most closely approximated it in terms of marginal fracture. Two evaluators who were experienced in using this photographic scale assessed the test photographs independently, and their assessments were averaged. A previous study has shown a linear relationship between the scale numbers and the mean width (in micrometers) of the marginal fracture exhibited by the restorations in the photographs.17 Therefore, the scale numbers recorded for all restorations in the study were treated as interval data that allowed the use of a parametric statistical method to test for differences between bonded and unbonded restorations with respect to marginal fracture. We assessed the degree of interevaluator agreement by using the 
statistic, as described by Cohen.18 This calculation produced a of 0.81. According to Landis and Koch,19 values greater than 0.75 represent excellent agreement beyond chance.
|
|
RESULTS |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
presents the results of the postoperative sensitivity evaluation at the polish recall appointment. 2 analysis showed no difference in sensitivity between bonded and unbonded restorations made from either amalgam alloy. Furthermore, no patients reported postoperative sensitivity for either bonded or unbonded restorations made of either alloy at the one-, two- or three-year recall visits.
|
and 4 show the results of the marginal fracture evaluation for the Aristaloy CR and Tytin amalgam alloys, respectively. Using analysis of variance, or ANOVA, and Tukey’s contrasts (at P = .05) for each recall visit, we found no difference between bonded and unbonded restorations in regard to marginal fracture. Because the null hypothesis of no difference in either postoperative sensitivity or marginal fracture for bonded vs. unbonded amalgam was not rejected, the power of the statistical tests was analyzed. For the 2 test, the chance of rejecting the null hypothesis at P = .05 was only 27 percent, while for the ANOVA tests, the chance of rejecting the null hypothesis at P = .05 was only 9 percent.
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
Spherical particle alloys. When first introduced, spherical particle alloys were considered to be better adapted to cavity walls than were lathe-cut particle alloys but, to some extent, the opposite has been the case.13 The condensation technique appears to be more critical with spherical alloys than with lathe-cut alloys. Using larger-faced condensers to condense, rather than penetrate, the amalgam mix and directing condensation forces against cavity walls are important techniques for achieving good adaptation of amalgam restorations made from spherical particle alloys.
Furthermore, microleakage, which is considered to be a significant cause of postoperative sensitivity owing to the hydrodynamic effect,22 has been shown to be related to setting dimensional change, surface texture of the amalgam at the interface between the amalgam and the tooth, and amalgam mix plasticity (that is, microleakage increases with more setting contraction, a rougher surface texture and lower mix plasticity [or drier mix]).23 In general, spherical particle alloys exhibit negative characteristics in regard to each of these three factors.24 Because practitioners prefer drier mixes, manufacturers usually proportion the alloy and mercury in encapsulated products to produce these drier mixes. However, the result is a tendency for greater microleakage and the potential for more postoperative sensitivity.
Alternatives to adhesive bonding. In individual practice settings in which spherical particle alloys are used, amalgam bonding may tend to reduce the incidence of postoperative sensitivity. However, if cost and the procedural difficulties associated with amalgam bonding are of concern, two alternatives may be considered. One involves changing to admix particle alloys (combinations of spherical and lathe-cut particles), which, as a group, have demonstrated less microleakage than spherical particle alloys in laboratory tests.13 A second alternative is to seal the dentin with a resin-based dentin bonding agent, followed by the usual placement of amalgam. These alternatives can avoid the technical problems often associated with the adhesive agent used for amalgam bonding.
Marginal fracture. We found no difference between bonded and unbonded restorations in regard to marginal fracture. However, Tarim and colleagues3 conducted an in vitro test, the results of which showed reduced marginal fracture with bonded restorations. However, their results were based on a scanning electron microscopic examination of margins of restored extracted teeth subjected to a simulated clinical loading pattern. This disparity in results points to the need for caution in comparing laboratory test results with those of clinical applications.
Although the degree of marginal fracture is an important clinical criterion for success, its reduction by the use of adhesive techniques may also imply a relationship between bonding and strengthening of the restored tooth. The most direct proof of tooth strengthening would be a reduced incidence of cusp fracture, and this was not found in this study. After three years of clinical service, we found no cusp fractures in any of the 125 restored teeth examined. However, because of the relatively short period of this study, this result does not unequivocally show that bonding would not reduce the incidence of cusp fracture.
Evaluating cusp fracture in a scientifically controlled clinical study is a tenuous proposition. Cusp fractures of teeth restored with amalgam can be caused by an excessive buccolingual width of the restoration, a traumatic loading incident or, in the case of a fatigue mechanism, an extended time in clinical service. Moreover, the relatively small number of cusp fractures that would be recorded in a carefully controlled clinical study tend to preclude meaningful statistical analyses of the data. Therefore, it may be difficult to determine whether amalgam bonding will strengthen a tooth enough to prevent cusp fracture in the clinical environment, particularly in light of the relatively low strength of the bond between the amalgam and the adhesive.25
We found no difference between bonded and unbonded restorations in regard to marginal fracture.
Nature of bonding. An-other factor to consider is the basic nature of bonding amalgam to tooth structure, which involves three elements:
If the bond between the amalgam and the adhesive fails and a gap is formed, but the bond between the adhesive and the tooth remains intact, corrosion products can seal the gap, as they do when no bonding agent has been used. On the other hand, if the bond between the adhesive and the tooth fails and a gap is formed, an environment is established that can promote the development of secondary caries. This is similar to the scenario of a direct-filling composite that fails to bond to tooth structure, which is considered by some to increase the probability of secondary caries formation. Therefore, an intact bond between the adhesive and the tooth is essential for the success of a bonded amalgam restoration. This requires the careful application of acid etching, priming agent and adhesive, all of which work best with rubber dam isolation.
Other applications of amalgam bonding, such as for extensive restorations, may have merit, particularly if they involve bonding adhesive to a single surface; in such cases, polymerization contraction forces do not stress the bond. However, several in vitro studies have shown a reduction in bond strength26,27 and debonding28 when bonded restorations are subject to cyclical loading (as in mastication). Therefore, dental practitioners should exercise caution when deciding whether to use amalgam bonding agents in nonretentive applications. Perhaps a prudent approach would be to use bonding as an adjunct to established methods for retention of large restorations that involve a minimal amount of remaining tooth structure.
|
CONCLUSIONS |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
|
FOOTNOTES |
|---|
|
REFERENCES |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
