DETERIORATION OF RESTORATIVE MATERIALS AND THE RISK FOR SECONDARY CARIES AJ.
GOLDBERG
Department of Prosthodontics, School of Dental Medicine, University of Connecticut Health Center, Farmington, Connecticut 06032 Adv Dent Res 4:14-18, June, 1990
ABSTRACT his article reviews the literature related to marginal deterioration of amalgam restorations and the risk for secondary caries. Background information on this characteristic material failure is provided, inT cluding how it is measured clinically, its rate of progress with time, and the responsible microstructural mechanisms. The association between degree of marginal fracture and recurrent or secondary caries is reviewed in clinical and in vitro studies. Earlier clinical investigations and a more recent laboratory study demonstrated a strong correlation between gap size and recurrent caries for a caries-active environment, supporting replacement of defective restorations for preventive reasons. However, more recent clinical data show the complex and equivocal nature of this association. Poor specificity and validity for detection of recurrent caries with an explorer and mirror further complicate interpretation of some trials and decision-making in clinical practice. Improved methods of diagnosing recurrent caries and further studies of the factors influencing its initiation and progression are needed.
INTRODUCTION This paper examines available data on the relationship between the deterioration of amalgam restoration margins and the risk for secondary caries. Restorative resin composites are also discussed, but related more to mode of failure than risk of secondary caries. Relatively few data are available related to secondary, marginal, or recurrent caries of prosthodontic procedures, so this area is not reviewed here. The paper covers several aspects of the problem. First, there is a brief discussion of how restorative material deterioration is measured clinically. Second, data are presented on the characteristics of this deterioration and its rate. Third, microstructural mechanisms responsible for clinically observed amalgam failures are discussed. Finally, specific evidence concerning the relationship between marginal deterioration and recurrent caries is reviewed. Because all restorative materials deteriorate with time, another way of looking at this question is to ask "at what Presented at a symposium "Re-restoration of Teeth: Materials Failure, Secondary Caries, and Socio-economic Factors'', March 9,1988, during the 66th General Session of the International Association for Dental Research, Montreal, PQ, Canada 14
point does the deterioration justify replacement?" Other papers in the symposium address other factors which affect the decision to replace a restoration. MARGINAL DETERIORATION OF AMALGAM RESTORATIONS For amalgam, the common mode of chronic clinical deterioration is marginal ditching or dyscrasia, most commonly observed at the cavo-surface margins (Leinfelder and Lemons, 1988). Clinicians are familiar with the wide range of marginal gaps routinely observed. In practice, this amalgam-enamel interface is typically evaluated with an explorer and mirror. Margins into which the explorer can penetrate are sometimes referred to as "open" or possibly "defective". Traditional teaching and practice associate such imperfections, especially when also accompanied by "sticking" of the explorer, with marginal or recurrent caries. However, this relationship is equivocal (Merrett and Elderton, 1984; Soderholm et al., 1989), as discussed below and at a previous symposium (Anusavice, 1989). Although the marginal areas of amalgam restorations do deteriorate with time, there is little overall
Vol. 4
DETERIORATED MARGINS AND SECONDARY CARIES
loss of material from the occlusal surface or from surfaces distant from margins. In contrast, resin composite restorations usually exhibit general loss of material across the entire occlusal surface (Kusy and Leinfelder, 1977). This results in exposed enamel walls at the cavo-surface margin, but the geometry is such that it may not be scored as "defective marginal adaptation7', according to some clinical evaluation criteria (Phillips et al, 1972; Ryge and Snyder, 1973). Some composites, particularly in posterior application, develop excessive localized wear which has been associated with occlusal contact, voids, or weaknesses in specific locations in the material (Leinfelder, 1985). There is considerable literature on techniques to measure the deterioration of restorations. The most widely used was developed by Ryge and is often referred to as the Ryge or USPHS criteria (Ryge and Snyder, 1973). This evaluation is performed directly in the mouth with a dental explorer and mirror. Generally, restorations are judged as (A) excellent, (B) acceptable, (C) unsatisfactory, or (D) unsatisfactory and requiring immediate replacement. Other authors have modified the USPHS system or compared restorations with standards at chairside. The most sophisticated direct-measurement technique employs laser holography. The most widely used indirect evaluation techniques rely on impressions and casts or photographs of the restorations. With amalgams, the photographic techniques are the most widely used. Originally developed by Mahler (Mahler and Marantz, 1979a), blackand-white photographs of the subject restorations are compared with a set of five standards and scored accordingly, or they may be ranked. Alternatively, impressions and casts of the restorations can be evaluated with a scanning electron microscope, sectioned and viewed optically, measured directly, ranked, or measured by stereophotogrammetry or laser holography. The most popular system for evaluation of posterior composites uses dental stone models compared with a set of standards. The subject of clinical evaluation techniques has been reviewed in more detail elsewhere (Goldberg, 1986). In the context of this symposium, it is important to point out that both dental amalgam and resin composite restorations continue to deteriorate with time, although the rate of deterioration decreases with time. This is shown in Fig. 1, which illustrates the increasing marginal gap in various amalgam alloys with time. Note the rapid initial marginal deterioration followed by continued loss of material, but at a slower rate. The differences among materials shown in Fig. 1, even at one year, are statistically significant and could readily be detected in a practical clinical setting. Interestingly, this pattern of wear is similar for composites, although the amount and location of material loss differ from those for amalgam. The mechanism responsible for marginal deterioration has received considerable attention because of
8
15 ,-OM
r
7 6 •ON
-OD
Time (years)
Fig. 1 —Marginal fracture of amalgam restorations vs. time. Restorations were scored by comparison of black-and-white photographs with a set of standard photographs which establish 11 categories of marginal fracture. Alloys: M, Micro (L.D. Caulk); P, Spheraloy (Kerr); O, Optaloy (L.D. Caulk); A, Aristaloy (Engelhard); N, New True Dentalloy (S.S. White); D, Dispersalloy (Western Metallurgical) (from Mahler and Marantz, 1979a).
the potentially significant public health impact of improved amalgam alloys (Leinfelder and Lemons, 1988). The mercury-tin (gamma-two) component of the alloy is the weakest (Fairhurst and Ryge, 1962) and the most corrosion-prone (Marek et al., 1976), and is the phase responsible for marginal deterioration of conventional amalgam alloys. Removal of this phase by the incorporation of 13 to 30% copper leads to the socalled "gamma-two-free" alloys which have greater resistance to marginal deterioration (Mahler et al., 1970). The metallurgy, corrosion, and mechanical properties of dental amalgam have been reviewed recently by Mahler (1988). Most importantly for the current discussion, it has been demonstrated that all dentists, regardless of their skill level, benefit from the improved alloys (Mahler and Marantz, 1979b). Of course, factors other than the alloy affect the rate of marginal deterioration. These include operator performance, type of tooth, and patient effects. Furthermore, these effects may be additive. In summary, amalgam restorations continue to deteriorate with time, and there are numerous clinical factors known to affect this deterioration. The question for this symposium is "at what point does this deterioration provide justification for replacement of the restoration?" MARGINAL DETERIORATION AND SECONDARY CARIES The most widely cited work in this area is that of j0rgensen and Wakumoto (1968), who studied lower
Adv Dent Res June 1990
GOLDBERG
16
first permanent molars extracted for orthodontic purposes. The subjects ranged in age from 9 to 15 years. One of the important aspects of this population, not often noted, is that the children were considered to have relatively high caries activity. Amalgam restorations in these teeth were examined for the degree of deterioration and the presence of secondary (recurrent) caries. Lesions were defined by examination of the restorations at a magnification of 72X and identified as those which included loss of tissue, irregular demarcation of the defect, and a discolored or decalcification zone with reduced enamel translucency. The results are plotted in Fig. 2, which shows the percentage of restorations with recurrent caries as a function of the size of the marginal defect. The increase in likelihood of recurrent caries with the increasing size of the defect is apparent. There were no secondary carious lesions associated with defects under 50 |xm in size. This work is frequently cited as justification for replacing defective restorations as a preventive measure. j0rgensen and Wakumoto stratified their data for various locations on the occlusal surface. The effect shown in Fig. 2 was most prominent at the anatomic grooves. At other locations, the association between recurrent caries and marginal fracture was not as strong. This suggests that the presence of the groove plays a role in the association. Alternatively, it is possible that a lesion was initially present in the groove and was not excavated at the time of cavity preparation. It also suggests that plaque lodged in these grooves could influence the probability of a carious lesion at those sites, a suggestion supported by the data of others (Goldberg et al., 1981), to be discussed below. Derand et al. (1990) studied the relationship between the amalgam-enamel gap and the presence of recurrent caries with an in vitro model. Using cariesfree molars, they cut sections of enamel and dentin and positioned them in an apparatus which allowed
fV
/>
60-
yS
IOUS
50
o
an amalgam restoration to be placed with a controlled gap size as shown in Fig. 3. Gap widths were established at 0, 30, 40, 60, and 80 \xm. Model carious lesions were induced by inoculation with S. mutans in sucrose- and glucose-supplemented culture media. The presence of lesions was detected by microradiography. No enamel caries was noted, in contrast to the J0rgensen and Wakumoto (1968) in vivo study. Dentin caries was present in all of the specimens cultured in the sucrose media with gaps of 30 |xm or more. In the glucose group,, dentin caries was present only for the largest gaps, 60 and 80 |xm. This evidence supports the case for the increasing likelihood of recurrent caries with marginal breakdown as well as the concept of a threshold gap size for initiation of recurrent caries. However, in both this case and the J0rgensen and Wakumoto study, there was a high degree of caries activity. Goldberg et al. (1981) used a cross-sectional experimental design to study the relationship among recurrent caries, size of the marginal defect, and oral hygiene status. Eighty-seven volunteers, from 17 to 61 years old, were examined. Marginal quality of the restorations was scored at chairside by direct comparison of the restorations to the Mahler set of photographic standards. Oral hygiene status was measured by plaque and gingival indices. The presence of caries was judged with the USPHS criteria modified to include scoring of chalky non-cavitated areas ("white spots") and radiographs. Tooth tissues, unless chalky, were not firmly probed, so as to avoid potential damage shown to increase risk for caries (Bergman and Linden, 1969; Ekstrand et al., 1987). A total of 914 restored teeth with 1556 restored surfaces was examined. Of the restored surfaces, 17.7% had marginal carious sites, compared with 2.1% of the surfaces with primary lesions. This demonstrates that secondary caries is an important disease phenomenon among adults. Its prevalence is probably underestimated because it is not included in standard indices such as DMFS and because white-spot lesions are not recorded under the USPHS criteria. Representative data from this study for smooth surfaces are shown in Fig. 4. The plot confirms that there is an increasing likelihood of recurrent caries with increasing size of the marginal gap. However, it also indicates that oral
40 30 20 10 100
200
300
400
500
Marginal Fracture (\i)
Fig. 2—Percentage of restorations with recurrent caries vs. size of the marginal defect (from Jergensen and Wakumoto, 1968).
Fig. 3 —Schematic diagram of the apparatus used by Derand et al. (1990) to vary gap size (G) with in vitro caries model. E, enamel; D, dentin; A, amalgam.
Vol. 4
DETERIORATED MARGINS AND SECONDARY CARIES
hygiene plays a significant role in determining the* likelihood of recurrent caries. For example, for those restorations with a modified marginal quality score of 2, the decrease in oral hygiene from G.I. = 1 to G.I. = 2 had a more dramatic effect on the likelihood of recurrent caries than changing the marginal quality score by one unit. On the occlusal surfaces, the likelihood of recurrent caries initially increased with marginal quality score, but then remained constant. On these surfaces, the oral hygiene effect was more pronounced at the lower marginal quality scores. The authors drew the following conclusions: (1) Both oral hygiene and marginal quality have an effect on the risk of recurrent caries; (2) either oral hygiene or marginal integrity is the predominant correlate of recurrent decay, depending on the severities and the anatomic location; (3) there is a difference in the strength of association among recurrent caries, oral hygiene, and marginal integrity for occlusal and smooth surfaces; (4) clinical decisions about re-restoration of non-carious teeth should be made with these parameters considered. Although the above studies support an association between amalgam marginal gaps and recurrent caries, clearly the etiology is complex and the correlation may be an oversimplification. Factors such as the geometry of the gap, whether it is open to the dentin, the sizes and nature of relevant bacteria and other molecules, capillary action, the presence of pellicle, and whether the original lesion was completely removed—all likely have an impact on disease progress. As demonstrated above, general intra-oral factors (such as oral hygiene status, diet, caries history, biochemistry of the saliva, etc.) further complicate the mechanism. These factors may, in part, explain contradictory evidence found by Soderholm et al. (1989). They studied the relationship between gap width and caries for 68 surfaces of 35 extracted posterior teeth restored with amalgam. The presence or absence of caries wras determined clinically, and marginal gaps were measured on scanning electron micrographs. Subsequently, the restorations were removed, and the presence and location of carious lesions were determined with visual examination and hand instruments. Interestingly, there was no statistically significant difference between the gap sizes at the carious and caries-free sites, although the mean gap size for the carious group was actually smaller. The poor validity and specificity for detection of caries with just an explorer and visual examination were also demonstrated. Apparently, such examinations "overemphasize (the importance of) defective margins rather than detect true carious lesions". Hamilton et al. (1983) also concluded that marginal fracture was not a predictor of the longevity of amalgam restorations. In this study, one dentist placed restorations using two different alloys. The patients were from 18 to 57 years of age, with a mean of 31. All received routine dental care. The USPHS evalu-
17
50
01 a 40-
Gi«5
.22
S .5
30 H
1
20-
"8 10-
Marginal quality (modified scale)
Fig. 4 —Percentage of filled smooth surfaces with marginal caries vs. size of the marginal defect. Data are from combined buccal, lingual, mesial, and distal surfaces, stratified by gingival index score. Definitions for marginal caries and marginal quality scores are in the original article (from Goldberg et ah, 1981).
ation system was used to evaluate the restorations at baseline and at one, two, three, four, five, and 10 years. There was a statistically significant difference in the amounts of marginal deterioration of the two alloys. However, there was no difference in the replacement rate due to materials-related reasons, which included factors such as related caries, bulk fracture, and open margins. There was no difference in the replacement rate despite differences in marginal deterioration. This conclusion is consistent with studies by Bailit et al. (1979) which showed that numerous factors, in addition to marginal deterioration, could be related to differences in replacement rate. Clearly, it is important to differentiate between "replacement rate" and "recurrent caries", which was the outcome variable measured in all of the works cited earlier. The two may not necessarily be consistent. In reviewing all of the above data, as well as the data presented in the other papers, it seems as though the question being addressed at this symposium should not be one of leaving in a defective amalgam vs. possible unnecessary replacement. The issues should be understanding the disease process and development of improved treatments. Materials should be able to restore esthetically as well as to prevent recurrence of caries. Newer materials, especially resin
18
GOLDBERG
composites, although not uniformly indicated for posterior uses, do hold the promise of providing improvement in treatment outcome. Dentistry should not be satisfied until it can provide a satisfactory longterm, esthetic restorative treatment while at the same time effectively halting the disease process. Not only do the materials available today have greater resistance to deterioration in the oral environment, but also the possibility of bonding to both enamel and dentin has dramatic implications for the issue of recurrent caries. SUMMARY It would seem reasonable to summarize the literature on deterioration of restorative materials and draw the following conclusions: The deterioration of restorative materials continues with time, but at a decreasing rate. There are many factors which affect the rate of deterioration, including the material, type of tooth, operator, and manipulative factors. These effects may be additive. Improved materials and integrity of the material-tooth interface probably reduce recurrent caries regardless of the other risk factors. The likelihood of secondary caries probably increases with increasing marginal deterioration, at least in caries-active populations, although there is evidence to the contrary. A threshold minimum gap size is required for initiation of secondary caries. Oral hygiene is a determinant of recurrent caries. Marginal deterioration alone is not necessarily a predictor of need for a restoration to be replaced. Replacement rate is not necessarily a good indicator of the frequency of disease. Improved methods for diagnosing recurrent caries are needed. Not enough is known about the etiology and risk factors for recurrent caries now to provide definitive guidance for practitioners. More detailed studies are warranted. REFERENCES ANUSAVICE, K.J., Ed. (1989): Quality Evaluation of Dental Restorations: Criteria for Placement and Replacement, Chicago: Quintessence Publishing Co., Inc. BAILIT, H.L.; CHIRIBOGA, D.; GRASSO, J.; DAMUTH, L.; and WILLEMAIN, T.R. (1979): A New Intermediate Dental Outcome Measure: Amalgam Replacement Rate, Med Case 17:783-786. BERGMAN, G. and LINDEN, L. (1969): The Action of the Explorer on Incipient Caries, Svensk Tandlek T 62:629-634. DERAND, T.; BIRKHED, D.; and EDWARDSSON, S. (1990): Secondary Caries Related to Various Marginal Gaps Around Amalgam Restorations in vitro, Swed Dent J (submitted) EKSTRAND, K.; QVIST, V.; and THYLSTRUP, A. (1987): Light
Adv Dent Res June 1990 Microscope Study of the Effect of Probing in Occlusal Fissures, Caries Res 21:368-374. FAIRHURST, C.W. and RYGE, G. (1962): X-ray Diffraction Investigation of the Sn-Hg Phase in Dental Amalgam. In: Advances in X-ray Analysis, Vol. 5, W. M. Muelles, Ed., New York: Plenum Press. GOLDBERG, J. (1986): Dental Materials: Clinical Evaluation. In: Encyclopedia of Materials Science and Engineering, M.B. Bener, Ed., Oxford: Pergamon Press, pp. 1088-1093. GOLDBERG, J.; TANZER, J.; MUNSTER, E.; AMARA, J.; THAL, F.; and BIRKHED, D. (1981): Cross-sectional Clinical Evaluation of Recurrent Enamel Caries, Restoration Marginal Integrity, and Oral Hygiene Status, / Am Dent Assoc 102:635-641. HAMILTON, J.C.; MOFFA, J.P.; ELLISON, J.A.; and JENKINS, W.A. (1983): Marginal Fracture Not a Predictor of Longevity for Two Dental Amalgam Alloys: A Ten-Year Study, / Prosthet Dent 50:200-202. J0RGENSEN, K.D. and WAKUMOTO, S. (1968): Occlusal Amalgam Fillings: Marginal Defects and Secondary Caries, Odont Tidskrift 76:43-53. KUSY, R.P. and LEINFELDER, K.F. (1977): Pattern of Wear in Posterior Composite Restorations, / Dent Res 56:544. LEINFELDER, K.F. (1985): Evaluation of Clinical Wear of Posterior Composite Resins. In: Posterior Composite Resin Dental Restorative Materials, G. Vanherle and D.C. Smith, Eds., Netherlands: P. Szulc Publishing Co., pp. 501-509. LEINFELDEIVK.F. and LEMONS, J.E. (1988): Clinical Restorative Materials and Techniques, Philadelphia: Lea and Febiger. MAHLER, D.B. (1988): Research on Dental Amalgam, Adv Dent Res 2:71-82. MAHLER, D.B. and MARANTZ, R.L. (1979a): The Effect of Time on the Marginal Fracture Behavior of Amalgam, / Oral Rehabil 6:391-398. MAHLER, D.B. and MARANTZ, R.L. (1979b): The Effect of the Operator on the Clinical Performance of Amalgam, / Am Dent Assoc 99:38-41. MAHLER, D.B.; TERKLA, L.G.; VAN EYSDEN, J.; and REISBICK, M.H. (1970): Marginal Fracture versus Mechanical Properties of Amalgam, / Dent Res 49:1452-1457. MAREK, M.; HOCKMAN, R.F.; and OKABE, T. (1976): In vitro Corrosion of Dental Amalgam Phases, / Biomed Mater Res 10:789804. MERRETT, M.C.W. and ELDERTON, R.J. (1984): An in vitro Study of Restorative Dental Treatment Decisions and Dental Caries, Br Dent J 157:128-133. PHILLIPS, R.W.; AVERY, D.R.; MEHRA, R.; SWARTZ, M.L.; and McCUNE, R.J. (1972): Observations on a Composite Resin for Class II Restorations: Two-year Report, / Prosthet Dent 28:164169. RYGE, G. and SNYDER, M. (1973): Evaluating the Clinical Quality of Restorations, / Am Dent Assoc 87:369-377. SODERHOLM, K.J.; ANTONSON, D.E.; and FISCHLSCHWEIGER, W. (1989): Correlation Between Marginal Discrepancies at the Amalgam/Tooth Interface and Recurrent Caries. In: Proceedings, Quality Evaluation of Dental Restorations: Criteria for Placement and Replacement, K.J. Anusavice, Ed., Chicago: Quintessence Publishing Co., Inc., pp. 95-108.
GOLDBERG
Department of Prosthodontics, School of Dental Medicine, University of Connecticut Health Center, Farmington, Connecticut 06032 Adv Dent Res 4:14-18, June, 1990
ABSTRACT his article reviews the literature related to marginal deterioration of amalgam restorations and the risk for secondary caries. Background information on this characteristic material failure is provided, inT cluding how it is measured clinically, its rate of progress with time, and the responsible microstructural mechanisms. The association between degree of marginal fracture and recurrent or secondary caries is reviewed in clinical and in vitro studies. Earlier clinical investigations and a more recent laboratory study demonstrated a strong correlation between gap size and recurrent caries for a caries-active environment, supporting replacement of defective restorations for preventive reasons. However, more recent clinical data show the complex and equivocal nature of this association. Poor specificity and validity for detection of recurrent caries with an explorer and mirror further complicate interpretation of some trials and decision-making in clinical practice. Improved methods of diagnosing recurrent caries and further studies of the factors influencing its initiation and progression are needed.
INTRODUCTION This paper examines available data on the relationship between the deterioration of amalgam restoration margins and the risk for secondary caries. Restorative resin composites are also discussed, but related more to mode of failure than risk of secondary caries. Relatively few data are available related to secondary, marginal, or recurrent caries of prosthodontic procedures, so this area is not reviewed here. The paper covers several aspects of the problem. First, there is a brief discussion of how restorative material deterioration is measured clinically. Second, data are presented on the characteristics of this deterioration and its rate. Third, microstructural mechanisms responsible for clinically observed amalgam failures are discussed. Finally, specific evidence concerning the relationship between marginal deterioration and recurrent caries is reviewed. Because all restorative materials deteriorate with time, another way of looking at this question is to ask "at what Presented at a symposium "Re-restoration of Teeth: Materials Failure, Secondary Caries, and Socio-economic Factors'', March 9,1988, during the 66th General Session of the International Association for Dental Research, Montreal, PQ, Canada 14
point does the deterioration justify replacement?" Other papers in the symposium address other factors which affect the decision to replace a restoration. MARGINAL DETERIORATION OF AMALGAM RESTORATIONS For amalgam, the common mode of chronic clinical deterioration is marginal ditching or dyscrasia, most commonly observed at the cavo-surface margins (Leinfelder and Lemons, 1988). Clinicians are familiar with the wide range of marginal gaps routinely observed. In practice, this amalgam-enamel interface is typically evaluated with an explorer and mirror. Margins into which the explorer can penetrate are sometimes referred to as "open" or possibly "defective". Traditional teaching and practice associate such imperfections, especially when also accompanied by "sticking" of the explorer, with marginal or recurrent caries. However, this relationship is equivocal (Merrett and Elderton, 1984; Soderholm et al., 1989), as discussed below and at a previous symposium (Anusavice, 1989). Although the marginal areas of amalgam restorations do deteriorate with time, there is little overall
Vol. 4
DETERIORATED MARGINS AND SECONDARY CARIES
loss of material from the occlusal surface or from surfaces distant from margins. In contrast, resin composite restorations usually exhibit general loss of material across the entire occlusal surface (Kusy and Leinfelder, 1977). This results in exposed enamel walls at the cavo-surface margin, but the geometry is such that it may not be scored as "defective marginal adaptation7', according to some clinical evaluation criteria (Phillips et al, 1972; Ryge and Snyder, 1973). Some composites, particularly in posterior application, develop excessive localized wear which has been associated with occlusal contact, voids, or weaknesses in specific locations in the material (Leinfelder, 1985). There is considerable literature on techniques to measure the deterioration of restorations. The most widely used was developed by Ryge and is often referred to as the Ryge or USPHS criteria (Ryge and Snyder, 1973). This evaluation is performed directly in the mouth with a dental explorer and mirror. Generally, restorations are judged as (A) excellent, (B) acceptable, (C) unsatisfactory, or (D) unsatisfactory and requiring immediate replacement. Other authors have modified the USPHS system or compared restorations with standards at chairside. The most sophisticated direct-measurement technique employs laser holography. The most widely used indirect evaluation techniques rely on impressions and casts or photographs of the restorations. With amalgams, the photographic techniques are the most widely used. Originally developed by Mahler (Mahler and Marantz, 1979a), blackand-white photographs of the subject restorations are compared with a set of five standards and scored accordingly, or they may be ranked. Alternatively, impressions and casts of the restorations can be evaluated with a scanning electron microscope, sectioned and viewed optically, measured directly, ranked, or measured by stereophotogrammetry or laser holography. The most popular system for evaluation of posterior composites uses dental stone models compared with a set of standards. The subject of clinical evaluation techniques has been reviewed in more detail elsewhere (Goldberg, 1986). In the context of this symposium, it is important to point out that both dental amalgam and resin composite restorations continue to deteriorate with time, although the rate of deterioration decreases with time. This is shown in Fig. 1, which illustrates the increasing marginal gap in various amalgam alloys with time. Note the rapid initial marginal deterioration followed by continued loss of material, but at a slower rate. The differences among materials shown in Fig. 1, even at one year, are statistically significant and could readily be detected in a practical clinical setting. Interestingly, this pattern of wear is similar for composites, although the amount and location of material loss differ from those for amalgam. The mechanism responsible for marginal deterioration has received considerable attention because of
8
15 ,-OM
r
7 6 •ON
-OD
Time (years)
Fig. 1 —Marginal fracture of amalgam restorations vs. time. Restorations were scored by comparison of black-and-white photographs with a set of standard photographs which establish 11 categories of marginal fracture. Alloys: M, Micro (L.D. Caulk); P, Spheraloy (Kerr); O, Optaloy (L.D. Caulk); A, Aristaloy (Engelhard); N, New True Dentalloy (S.S. White); D, Dispersalloy (Western Metallurgical) (from Mahler and Marantz, 1979a).
the potentially significant public health impact of improved amalgam alloys (Leinfelder and Lemons, 1988). The mercury-tin (gamma-two) component of the alloy is the weakest (Fairhurst and Ryge, 1962) and the most corrosion-prone (Marek et al., 1976), and is the phase responsible for marginal deterioration of conventional amalgam alloys. Removal of this phase by the incorporation of 13 to 30% copper leads to the socalled "gamma-two-free" alloys which have greater resistance to marginal deterioration (Mahler et al., 1970). The metallurgy, corrosion, and mechanical properties of dental amalgam have been reviewed recently by Mahler (1988). Most importantly for the current discussion, it has been demonstrated that all dentists, regardless of their skill level, benefit from the improved alloys (Mahler and Marantz, 1979b). Of course, factors other than the alloy affect the rate of marginal deterioration. These include operator performance, type of tooth, and patient effects. Furthermore, these effects may be additive. In summary, amalgam restorations continue to deteriorate with time, and there are numerous clinical factors known to affect this deterioration. The question for this symposium is "at what point does this deterioration provide justification for replacement of the restoration?" MARGINAL DETERIORATION AND SECONDARY CARIES The most widely cited work in this area is that of j0rgensen and Wakumoto (1968), who studied lower
Adv Dent Res June 1990
GOLDBERG
16
first permanent molars extracted for orthodontic purposes. The subjects ranged in age from 9 to 15 years. One of the important aspects of this population, not often noted, is that the children were considered to have relatively high caries activity. Amalgam restorations in these teeth were examined for the degree of deterioration and the presence of secondary (recurrent) caries. Lesions were defined by examination of the restorations at a magnification of 72X and identified as those which included loss of tissue, irregular demarcation of the defect, and a discolored or decalcification zone with reduced enamel translucency. The results are plotted in Fig. 2, which shows the percentage of restorations with recurrent caries as a function of the size of the marginal defect. The increase in likelihood of recurrent caries with the increasing size of the defect is apparent. There were no secondary carious lesions associated with defects under 50 |xm in size. This work is frequently cited as justification for replacing defective restorations as a preventive measure. j0rgensen and Wakumoto stratified their data for various locations on the occlusal surface. The effect shown in Fig. 2 was most prominent at the anatomic grooves. At other locations, the association between recurrent caries and marginal fracture was not as strong. This suggests that the presence of the groove plays a role in the association. Alternatively, it is possible that a lesion was initially present in the groove and was not excavated at the time of cavity preparation. It also suggests that plaque lodged in these grooves could influence the probability of a carious lesion at those sites, a suggestion supported by the data of others (Goldberg et al., 1981), to be discussed below. Derand et al. (1990) studied the relationship between the amalgam-enamel gap and the presence of recurrent caries with an in vitro model. Using cariesfree molars, they cut sections of enamel and dentin and positioned them in an apparatus which allowed
fV
/>
60-
yS
IOUS
50
o
an amalgam restoration to be placed with a controlled gap size as shown in Fig. 3. Gap widths were established at 0, 30, 40, 60, and 80 \xm. Model carious lesions were induced by inoculation with S. mutans in sucrose- and glucose-supplemented culture media. The presence of lesions was detected by microradiography. No enamel caries was noted, in contrast to the J0rgensen and Wakumoto (1968) in vivo study. Dentin caries was present in all of the specimens cultured in the sucrose media with gaps of 30 |xm or more. In the glucose group,, dentin caries was present only for the largest gaps, 60 and 80 |xm. This evidence supports the case for the increasing likelihood of recurrent caries with marginal breakdown as well as the concept of a threshold gap size for initiation of recurrent caries. However, in both this case and the J0rgensen and Wakumoto study, there was a high degree of caries activity. Goldberg et al. (1981) used a cross-sectional experimental design to study the relationship among recurrent caries, size of the marginal defect, and oral hygiene status. Eighty-seven volunteers, from 17 to 61 years old, were examined. Marginal quality of the restorations was scored at chairside by direct comparison of the restorations to the Mahler set of photographic standards. Oral hygiene status was measured by plaque and gingival indices. The presence of caries was judged with the USPHS criteria modified to include scoring of chalky non-cavitated areas ("white spots") and radiographs. Tooth tissues, unless chalky, were not firmly probed, so as to avoid potential damage shown to increase risk for caries (Bergman and Linden, 1969; Ekstrand et al., 1987). A total of 914 restored teeth with 1556 restored surfaces was examined. Of the restored surfaces, 17.7% had marginal carious sites, compared with 2.1% of the surfaces with primary lesions. This demonstrates that secondary caries is an important disease phenomenon among adults. Its prevalence is probably underestimated because it is not included in standard indices such as DMFS and because white-spot lesions are not recorded under the USPHS criteria. Representative data from this study for smooth surfaces are shown in Fig. 4. The plot confirms that there is an increasing likelihood of recurrent caries with increasing size of the marginal gap. However, it also indicates that oral
40 30 20 10 100
200
300
400
500
Marginal Fracture (\i)
Fig. 2—Percentage of restorations with recurrent caries vs. size of the marginal defect (from Jergensen and Wakumoto, 1968).
Fig. 3 —Schematic diagram of the apparatus used by Derand et al. (1990) to vary gap size (G) with in vitro caries model. E, enamel; D, dentin; A, amalgam.
Vol. 4
DETERIORATED MARGINS AND SECONDARY CARIES
hygiene plays a significant role in determining the* likelihood of recurrent caries. For example, for those restorations with a modified marginal quality score of 2, the decrease in oral hygiene from G.I. = 1 to G.I. = 2 had a more dramatic effect on the likelihood of recurrent caries than changing the marginal quality score by one unit. On the occlusal surfaces, the likelihood of recurrent caries initially increased with marginal quality score, but then remained constant. On these surfaces, the oral hygiene effect was more pronounced at the lower marginal quality scores. The authors drew the following conclusions: (1) Both oral hygiene and marginal quality have an effect on the risk of recurrent caries; (2) either oral hygiene or marginal integrity is the predominant correlate of recurrent decay, depending on the severities and the anatomic location; (3) there is a difference in the strength of association among recurrent caries, oral hygiene, and marginal integrity for occlusal and smooth surfaces; (4) clinical decisions about re-restoration of non-carious teeth should be made with these parameters considered. Although the above studies support an association between amalgam marginal gaps and recurrent caries, clearly the etiology is complex and the correlation may be an oversimplification. Factors such as the geometry of the gap, whether it is open to the dentin, the sizes and nature of relevant bacteria and other molecules, capillary action, the presence of pellicle, and whether the original lesion was completely removed—all likely have an impact on disease progress. As demonstrated above, general intra-oral factors (such as oral hygiene status, diet, caries history, biochemistry of the saliva, etc.) further complicate the mechanism. These factors may, in part, explain contradictory evidence found by Soderholm et al. (1989). They studied the relationship between gap width and caries for 68 surfaces of 35 extracted posterior teeth restored with amalgam. The presence or absence of caries wras determined clinically, and marginal gaps were measured on scanning electron micrographs. Subsequently, the restorations were removed, and the presence and location of carious lesions were determined with visual examination and hand instruments. Interestingly, there was no statistically significant difference between the gap sizes at the carious and caries-free sites, although the mean gap size for the carious group was actually smaller. The poor validity and specificity for detection of caries with just an explorer and visual examination were also demonstrated. Apparently, such examinations "overemphasize (the importance of) defective margins rather than detect true carious lesions". Hamilton et al. (1983) also concluded that marginal fracture was not a predictor of the longevity of amalgam restorations. In this study, one dentist placed restorations using two different alloys. The patients were from 18 to 57 years of age, with a mean of 31. All received routine dental care. The USPHS evalu-
17
50
01 a 40-
Gi«5
.22
S .5
30 H
1
20-
"8 10-
Marginal quality (modified scale)
Fig. 4 —Percentage of filled smooth surfaces with marginal caries vs. size of the marginal defect. Data are from combined buccal, lingual, mesial, and distal surfaces, stratified by gingival index score. Definitions for marginal caries and marginal quality scores are in the original article (from Goldberg et ah, 1981).
ation system was used to evaluate the restorations at baseline and at one, two, three, four, five, and 10 years. There was a statistically significant difference in the amounts of marginal deterioration of the two alloys. However, there was no difference in the replacement rate due to materials-related reasons, which included factors such as related caries, bulk fracture, and open margins. There was no difference in the replacement rate despite differences in marginal deterioration. This conclusion is consistent with studies by Bailit et al. (1979) which showed that numerous factors, in addition to marginal deterioration, could be related to differences in replacement rate. Clearly, it is important to differentiate between "replacement rate" and "recurrent caries", which was the outcome variable measured in all of the works cited earlier. The two may not necessarily be consistent. In reviewing all of the above data, as well as the data presented in the other papers, it seems as though the question being addressed at this symposium should not be one of leaving in a defective amalgam vs. possible unnecessary replacement. The issues should be understanding the disease process and development of improved treatments. Materials should be able to restore esthetically as well as to prevent recurrence of caries. Newer materials, especially resin
18
GOLDBERG
composites, although not uniformly indicated for posterior uses, do hold the promise of providing improvement in treatment outcome. Dentistry should not be satisfied until it can provide a satisfactory longterm, esthetic restorative treatment while at the same time effectively halting the disease process. Not only do the materials available today have greater resistance to deterioration in the oral environment, but also the possibility of bonding to both enamel and dentin has dramatic implications for the issue of recurrent caries. SUMMARY It would seem reasonable to summarize the literature on deterioration of restorative materials and draw the following conclusions: The deterioration of restorative materials continues with time, but at a decreasing rate. There are many factors which affect the rate of deterioration, including the material, type of tooth, operator, and manipulative factors. These effects may be additive. Improved materials and integrity of the material-tooth interface probably reduce recurrent caries regardless of the other risk factors. The likelihood of secondary caries probably increases with increasing marginal deterioration, at least in caries-active populations, although there is evidence to the contrary. A threshold minimum gap size is required for initiation of secondary caries. Oral hygiene is a determinant of recurrent caries. Marginal deterioration alone is not necessarily a predictor of need for a restoration to be replaced. Replacement rate is not necessarily a good indicator of the frequency of disease. Improved methods for diagnosing recurrent caries are needed. Not enough is known about the etiology and risk factors for recurrent caries now to provide definitive guidance for practitioners. More detailed studies are warranted. REFERENCES ANUSAVICE, K.J., Ed. (1989): Quality Evaluation of Dental Restorations: Criteria for Placement and Replacement, Chicago: Quintessence Publishing Co., Inc. BAILIT, H.L.; CHIRIBOGA, D.; GRASSO, J.; DAMUTH, L.; and WILLEMAIN, T.R. (1979): A New Intermediate Dental Outcome Measure: Amalgam Replacement Rate, Med Case 17:783-786. BERGMAN, G. and LINDEN, L. (1969): The Action of the Explorer on Incipient Caries, Svensk Tandlek T 62:629-634. DERAND, T.; BIRKHED, D.; and EDWARDSSON, S. (1990): Secondary Caries Related to Various Marginal Gaps Around Amalgam Restorations in vitro, Swed Dent J (submitted) EKSTRAND, K.; QVIST, V.; and THYLSTRUP, A. (1987): Light
Adv Dent Res June 1990 Microscope Study of the Effect of Probing in Occlusal Fissures, Caries Res 21:368-374. FAIRHURST, C.W. and RYGE, G. (1962): X-ray Diffraction Investigation of the Sn-Hg Phase in Dental Amalgam. In: Advances in X-ray Analysis, Vol. 5, W. M. Muelles, Ed., New York: Plenum Press. GOLDBERG, J. (1986): Dental Materials: Clinical Evaluation. In: Encyclopedia of Materials Science and Engineering, M.B. Bener, Ed., Oxford: Pergamon Press, pp. 1088-1093. GOLDBERG, J.; TANZER, J.; MUNSTER, E.; AMARA, J.; THAL, F.; and BIRKHED, D. (1981): Cross-sectional Clinical Evaluation of Recurrent Enamel Caries, Restoration Marginal Integrity, and Oral Hygiene Status, / Am Dent Assoc 102:635-641. HAMILTON, J.C.; MOFFA, J.P.; ELLISON, J.A.; and JENKINS, W.A. (1983): Marginal Fracture Not a Predictor of Longevity for Two Dental Amalgam Alloys: A Ten-Year Study, / Prosthet Dent 50:200-202. J0RGENSEN, K.D. and WAKUMOTO, S. (1968): Occlusal Amalgam Fillings: Marginal Defects and Secondary Caries, Odont Tidskrift 76:43-53. KUSY, R.P. and LEINFELDER, K.F. (1977): Pattern of Wear in Posterior Composite Restorations, / Dent Res 56:544. LEINFELDER, K.F. (1985): Evaluation of Clinical Wear of Posterior Composite Resins. In: Posterior Composite Resin Dental Restorative Materials, G. Vanherle and D.C. Smith, Eds., Netherlands: P. Szulc Publishing Co., pp. 501-509. LEINFELDEIVK.F. and LEMONS, J.E. (1988): Clinical Restorative Materials and Techniques, Philadelphia: Lea and Febiger. MAHLER, D.B. (1988): Research on Dental Amalgam, Adv Dent Res 2:71-82. MAHLER, D.B. and MARANTZ, R.L. (1979a): The Effect of Time on the Marginal Fracture Behavior of Amalgam, / Oral Rehabil 6:391-398. MAHLER, D.B. and MARANTZ, R.L. (1979b): The Effect of the Operator on the Clinical Performance of Amalgam, / Am Dent Assoc 99:38-41. MAHLER, D.B.; TERKLA, L.G.; VAN EYSDEN, J.; and REISBICK, M.H. (1970): Marginal Fracture versus Mechanical Properties of Amalgam, / Dent Res 49:1452-1457. MAREK, M.; HOCKMAN, R.F.; and OKABE, T. (1976): In vitro Corrosion of Dental Amalgam Phases, / Biomed Mater Res 10:789804. MERRETT, M.C.W. and ELDERTON, R.J. (1984): An in vitro Study of Restorative Dental Treatment Decisions and Dental Caries, Br Dent J 157:128-133. PHILLIPS, R.W.; AVERY, D.R.; MEHRA, R.; SWARTZ, M.L.; and McCUNE, R.J. (1972): Observations on a Composite Resin for Class II Restorations: Two-year Report, / Prosthet Dent 28:164169. RYGE, G. and SNYDER, M. (1973): Evaluating the Clinical Quality of Restorations, / Am Dent Assoc 87:369-377. SODERHOLM, K.J.; ANTONSON, D.E.; and FISCHLSCHWEIGER, W. (1989): Correlation Between Marginal Discrepancies at the Amalgam/Tooth Interface and Recurrent Caries. In: Proceedings, Quality Evaluation of Dental Restorations: Criteria for Placement and Replacement, K.J. Anusavice, Ed., Chicago: Quintessence Publishing Co., Inc., pp. 95-108.
