In amalgam bonding, the mechanism by which the bonding resin attaches to tooth structure is identical to the mechanism by which resin-based composite attaches to dentin and enamel. The attachment of the bonding resin to the amalgam, however, is quite different from that of bonding resin to resin-based composite. The amalgam-to-resin attachment is entirely mechanical rather than chemical. Unset amalgam is condensed into the bonding resin on the tooth surface before it polymerizes, thus incorporating fingers of resin into the amalgam at the interface.¹⁴

The mechanical attachment of amalgam to the bonding resin is enhanced by the incorporation of filler particles, either organic or inorganic, into the resin.^15–18 A probable reason for this improvement in attachment is that the filled bonding resin is more viscous during condensation of the amalgam, and this provides improved penetration of bonding resin into the amalgam for increased mechanical retention.

In vitro studies have demonstrated that the resistance to a shearing load created by amalgam bonding can be equal to or greater than resistance provided by mechanical features such as pins.^15,19,20 Other studies have revealed that Amalgambond Plus (Parkell), with its polymethyl methacrylate filler, provides one of the strongest bonds of amalgam to tooth structure.^{15,17,18,20–22}

In vitro studies have demonstrated that the resistance to a shearing load created by amalgam bonding can be equal to or greater than resistance provided by mechanical features such as pins.

Several clinical studies of bonded amalgam restorations have been reported. Mahler and colleagues^23,24 reported no difference between bonded and nonbonded Class II amalgam restorations in tooth sensitivity two weeks after insertion or in marginal fracture after one year²³ or three years.²⁴ Belcher and Stewart²⁵ compared the clinical success of complex amalgam restorations retained with pins against that of restorations retained with Amalgambond Plus, both with and without filler powder. At two years, all restorations in all three groups were retained with minimal sensitivity, good marginal adaptation and no recurrent caries. Staninec and colleagues²⁶ reported two-year success of bonded amalgam restorations in primary teeth. An in vivo study by Smales and Wetherell²⁷ evaluated 366 bonded amalgam restorations using five bonding materials. Cuspal coverage was accomplished in 178 of the 366 restorations. They reported a 98.6 percent success rate at up to five years.²⁷

We undertook a clinical study to compare the failure rates, marginal adaptation, marginal discoloration, secondary caries rates, sensitivity and tooth vitality of bonded and pin-retained complex amalgam restorations. In this article, we report the results after five years.

	SUBJECTS, MATERIALS AND METHODS

TOP ABSTRACT SUBJECTS, MATERIALS AND METHODS RESULTS DISCUSSION CONCLUSIONS REFERENCES

 
Patients selected for the study had a posterior tooth or teeth requiring restoration that included at least one proximal surface and replacement of one or more cusps. Teeth to be restored had at least one proximal contact and occluded with natural or restored teeth. We took preoperative apical radiographs of all teeth restored in the study. Sixty teeth (57 molars and three premolars) were restored with amalgam in the study, 28 using pin retention and 32 using the amalgam bonding system. For each mouth requiring more than one restoration, a coin toss determined which tooth or teeth received bonded restoration(s) and which received pin-retained restoration(s). For a mouth in which there was only one restoration, we also used the coin toss to determine which type of restoration would be done; then, in the next patient who needed only one restoration, the other type would be done.

Patients read and signed the consent form, which had been approved by the institutional review board of the University of Texas Health Science Center at San Antonio. Patients excluded from participation in the study were those who could not tolerate the procedures, those who had compromised immune systems or compromised salivary flow, and those who stated they would not be available for long-term recall.

Five operators, all faculty members and experienced clinicians (J.B.S., J.O.B., T.G.B., J.W.R., J.W.O.), placed the restorations. One of the authors (J.O.B.) calibrated the operators to guidelines concerning resistance features and the use of the bonding system. The operators confirmed pulpal vitality preoperatively via electric pulp testing and thermal testing. In each case, the clinician took a preoperative radiograph to ensure that there was no radiographic evidence of pulpal pathosis. Operators used rubber dam isolation during tooth preparation, amalgam placement and initial carving of all restorations.

For both groups, any retention form remaining after removal of an old restoration was left intact but was not enhanced. For all restorations, enough occlusal tooth structure was missing or was removed to ensure at least 2 millimeters of amalgam thickness in all occlusal areas of the restorations. A total of 60 teeth were restored in the mouths of 28 patients.

The guideline for pin placement stated that one vertical pin was used for each missing cusp, with a maximum of four vertical pins in any single restoration. Horizontal pins were placed at the operator’s discretion. We used TMS (Thread-Mate System, Coltène-Whaledent) Minim (0.024-inch diameter) vertical pins and Minikin (0.019-in. diameter) horizontal pins. The operators prepared pin channels with depth-limiting TMS pin channel drills to a depth of 2 mm for Minim pins and 1.5 mm for Minikin pins. They applied two coats of Copalite (Cooley & Cooley) varnish after preparing the pin channel, placed the pin, then placed a stainless steel matrix.

The operators used no added mechanical retention form for the resin-bonded restoration group; instead, they used the Amalgambond Plus (Parkell) bonding system. The operator placed a stainless steel matrix before using any portion of the bonding system. The instructions for Amalgambond Plus state that it may be used either with or without incorporation of a polymethyl methacrylate filler powder called high-performance additive, or HPA, powder. In all bonded restorations in this study, operators used the HPA powder, following the manufacturer’s instructions to achieve extra retention. The Amalgambond Plus was refrigerated at a temperature of approximately 40 degrees until immediately before its use. The dentin activator (10 percent citric acid and 3 percent ferric chloride) was applied in accordance with manufacturer’s instructions (approximately 30 seconds to enamel and 10 seconds to dentin). The operator rinsed the prepared tooth surface with air/water spray and briefly dried it with air. The adhesive agent (containing hydroxyethyl methacrylate) was applied to all prepared surfaces, gently thinned with an airstream and left undisturbed for at least 30 seconds. Components of the resin adhesive (three drops of base, one scoop of HPA powder and one drop of catalyst) were mixed together while a spherical amalgam (Tytin, Kerr) was being triturated. The operator coated all walls of the preparation with the resin adhesive using a Kerr Applicator (Kerr) (Figure 1) to provide a thin coat of the material.

View larger version (84K): [in this window] [in a new window]   Figure 1. A bonded restoration in tooth no. 31 that had no mechanical resistance or retention features. A. Preoperative view. B. Preparation. C. Restoration immediately after completion with occlusal contacts marked and all excursive contacts removed. D. Occlusal view after five years.

The operator shaped the occlusal surface of each restoration with burnishers and carvers before removing the matrix. He then removed the matrix and carved the restoration to proper anatomical form. After removing the rubber dam, the operator adjusted the occlusion to provide contact in maximum intercuspation (centric occlusion) and to eliminate contact in eccentric mandibular excursions. He then smoothed the amalgam surface using wet flour pumice in a rubber cup in a low-speed handpiece running at low speed; burnishers also were used. Restorations were not polished.

The operator took preoperative photographs of the preparation and of the restored tooth, both immediately after completion of the restoration and at each recall appointment. Figures 1 through 3 are representative photographs of teeth taken preoperatively of preparations, of restorations immediately after placement and of the same restorations at five years.

View larger version (83K): [in this window] [in a new window]   Figure 2. A pin-retained restoration in tooth no. 15. A. Preoperative view. B. Preparation after pin placement. C. Restoration immediately after completion. D. Restoration after five years.

View larger version (77K): [in this window] [in a new window]   Figure 3. A bonded restoration in tooth no. 14. A. Preoperative view. B. Preparation, showing gingivally deep extension in mesial aspect. C. Restoration immediately after completion. D. Restoration after five years.

	RESULTS

TOP ABSTRACT SUBJECTS, MATERIALS AND METHODS RESULTS DISCUSSION CONCLUSIONS REFERENCES

 
Tables 2 and 3 (page 929 and 930) display the results. Two bonded restorations of the 60 total did not meet selection criteria for the study and were excluded from consideration. Among the 28 total patients, two patients with three restorations between them (two pin-retained and one bonded) did not return for baseline evaluation and therefore were eliminated from the study. One patient with two restorations (one pin-retained and one bonded) was lost from the study after baseline and before the six-month evaluation. One patient, who had one bonded restoration and one pin-retained restoration, was lost to the study after the six-month evaluation owing to a stroke. Another patient, who had a bonded restoration, did not return after the one-year recall. Three patients with a total of three bonded restorations and two pin-retained restorations were unavailable for the five-year recall because they had moved away; at four years, all five of those restorations were performing satisfactorily.

– need for endodontic therapy—two molars restored with pin-retained restorations required endodontic therapy, one after six months but before the one-year evaluation and one at the two-year evaluation; both were successfully restored after endodontic therapy;

– significant tooth fracture adjacent to restoration—three molars restored with pin-retained restorations suffered significant fracture (one each at years one, three and four) but were restorable; one molar restored with a bonded restoration suffered significant fracture, including root fracture, at year five and was extracted;

– secondary caries—one molar restored with a bonded restoration had secondary caries (adjacent to the restoration) at year four, and two molars restored with pin-retained restorations had secondary caries (adjacent to the restorations) at year five; all carious teeth were restorable.

We used the Fisher exact test to compare the failure rate of the two restoration groups, bonded and pin-retained. There was no significant difference in failure rate (P = .071) at five years. In addition, using the Fisher exact test, at five years, we found no significant difference between groups in marginal adaptation (P = .058), marginal discoloration (P = .066), secondary caries (P = .579), sensitivity (P = .500) or tooth vitality (P = .206). During the five years of the study, the only time at which there was a significant difference in tooth sensitivity was at six months (P = .013).

Bonded restorations were performing as well as pin-retained restorations in all categories at five years.

	DISCUSSION

TOP ABSTRACT SUBJECTS, MATERIALS AND METHODS RESULTS DISCUSSION CONCLUSIONS REFERENCES

 
Published results of clinical studies have shown bonded amalgam restorations to function well for periods up to two years.^23–27,29 In these studies, researchers compared bonded amalgam restorations with mechanically retained amalgam restorations. The researchers reported that the bonded restorations were performing as well as the mechanically retained restorations. The restorations in our study yielded similar findings. This study involved replacement of at least one cusp with amalgam, and many of the restorations in the study replaced two or more cusps. These large restorations were exposed to more stress during function than simple Class I or Class II amalgam restorations, and there were more failed restorations in our study than in studies of smaller restorations.

Several clinical investigations have demonstrated no difference in sensitivity between teeth restored with or without bonding.^{23–25,27,29,30} Other studies, however, have shown reduced thermal sensitivity when the amalgam restorations were bonded.^31,32 One of these, reported by Davis and Overton,²⁶ involved teeth that had had symptoms of incomplete tooth fracture before restoration. They found sensitivity to a cold-temperature stimulus (skin refrigerant) to be reduced in the bonded restorations at three and 12 months after placement, compared with baseline. In contrast, restorations that were based and not bonded demonstrated no reduction in thermal sensitivity. In another study, Hadi and colleagues³² compared amalgam restorations bonded with Amalgambond Plus to restorations in which a cavity varnish (Copaliner, Harry J. Bosworth) was used. They surveyed patients the day after restorations were placed and found significantly less sensitivity in the bonded group.

In our study, we found no difference in sensitivity between bonded and nonbonded restorations at any point except at six months. The lack of a difference at baseline is similar to the results of the study by Davis and Overton.³¹ This absence of difference at baseline possibly could be the result of recent mechanical trauma during tooth preparation that opened dentinal tubules, predisposing teeth with both bonded and nonbonded restorations to minor postoperative thermal sensitivity. At six months in our study, and at three and 12 months in the Davis and Overton study, teeth with bonded restorations exhibited less sensitivity than those with nonbonded restorations. This difference might have been caused by improved sealing of tubules provided by the resin adhesive and the elapsed time since restoration.

Marginal discoloration is not a criterion for evaluation of amalgam restorations in most studies. We included it in this study because of the ability, in some areas, to see into the bonding resin. We made no attempt to discriminate staining at the amalgam-resin interface from that at the amalgam-enamel interface.

This study will continue for at least a sixth year to gain additional information. It should be noted that several of the bonded restorations in this study had little or no mechanical retention form (Figure 1), yet none of them had dislodged.

It is interesting that even though the bond strength provided by amalgam bonding agents for amalgam is lower than the bond strength provided by bonding agents for resin-based composite, it seems to be adequate even in very large restorations. This perhaps is because of the fact that amalgam does not exert a significant stress on the bonding mechanism as it sets, whereas resin-based composite—because of polymerization shrinkage—does.

	CONCLUSIONS

TOP ABSTRACT SUBJECTS, MATERIALS AND METHODS RESULTS DISCUSSION CONCLUSIONS REFERENCES

 
Amalgambond Plus, with HPA powder, is retaining complex amalgam restorations well at five years. In view of the results of this and other studies, the bonding of amalgam restorations appears to be a viable alternative to mechanical retention of restorations.

View larger version (125K): [in this window] [in a new window]   Dr. Summitt is a professor, Division of Operative Dentistry, Department of Restorative Dentistry, University of Texas Health Science Center at San Antonio, Dental School, Mail Code 7890, 7703 Floyd Curl Drive, San Antonio, Texas 78229-3900, e-mail "summitt{at}uthscsa.edu". Address reprint requests to Dr. Summitt.

View larger version (112K): [in this window] [in a new window]   Dr. Burgess is a professor and chairman, Department of Operative Dentistry and Biomaterials, Louisiana State University Health Sciences Center, New Orleans.

View larger version (136K): [in this window] [in a new window]   Dr. Berry is a professor, Department of Restorative Dentistry, University of Colorado School of Dentistry, Denver.

View larger version (135K): [in this window] [in a new window]   Dr. Robbins is in private practice in San Antonio and is a clinical professor, Department of General Dentistry, University of Texas Health Science Center at San Antonio, Dental School, San Antonio.

View larger version (130K): [in this window] [in a new window]   Dr. Osborne is a professor and the director of clinical research, Department of Restorative Dentistry, University of Colorado School of Dentistry, Denver.

View larger version (145K): [in this window] [in a new window]   Dr. Haveman is an associate professor, Department of General Dentistry, University of Texas Health Science Center at San Antonio, Dental School, San Antonio.