The Clinical Significance of Marginal Fit

The Clinical Significance of Marginal Fit

Thomas D. Larson, D.D.S., M.S.D.*:

Clinical evaluation of restoration acceptability includes marginal adaptation, restoration and preservation of anatomic form, color match, cavosurface discoloration, and presence  of marginal caries. This paper will review what is known about marginal fit of all materials relative to their acceptable fit. Some explanation will be given of how material properties affect marginal fit, what the expected longevity of different materials is relative to marginal fit, and how marginal fit affects development of secondary or marginal caries. Marginal  fit is assumed to affect restoration longevity by either encouraging or discouraging microleakage and development of secondary or marginal caries.
Marginal fit and its effect on  the health of the gingival tissues have been reviewed in an earlier paper on polishing and will not be reviewed here.1 Is there some correlation between marginal fit of the various  materials used in restorative dentistry and this assumption? 
This paper will review the literature found on Medline and present the science behind the clinical significance of  marginal fit utilizing in vitro studies where necessary and in vivo studies whenever possible. Numerous authors replicating these studies are used  to improve reliability and replicability. 

Diagnosis of Marginal Fit
Gunnar Ryge was the first to devise a quality assessment instrument for clinical evaluation of dental restorations (USPHS criteria).2,3 This quality assessment instrument has been used worldwide to evaluate the surface, color, anatomic form, and marginal integrity of clinical restorations in clinical studies since that time. The categories used to evaluate the  restorations are described as (1) alpha: meets all standards (excellent in all respects), and (2) bravo: observe at next visit (shows minor deviations from ideal, nevertheless  acceptable). Both alpha and bravo are clinically acceptable. Two additional categories are described as clinically unacceptable: charlie: replace for prevention (to avoid the  likelihood of future damage), and delta: replace statim (damage now being done, requires immediate replacement). Criteria are used to describe what is seen clinically that places  the evaluation into these various categories. Most dental schools have used such an evaluation scheme since the early 1970s to evaluate the quality of work the dental students perform. 
The criteria used to describe marginal fit are as follow. “Alpha = No visible evidence of a crevice along the margin into which the explorer will penetrate”, and for anterior restorations, “no discoloration on the margin between the restoration and the tooth. Bravo = Visible evidence of a crevice along the margin into which the explorer will penetrate”,  and for anterior restorations, “discoloration on the margin between the restoration and the tooth structure. Charlie = Dentin or base is exposed”, and for anterior restorations,  “discoloration has penetrated along the margin of the restorative material in a pulpal direction. Delta = Mobile or fractured restoration OR caries contiguous with the margin or the  restoration OR tooth structure fractured.“2
This review will look at these criteria to decide whether what was described by Ryge in 1973 is still accurate with regard to marginal fit for all materials. 
Evaluation of Marginal Fit
Use of a dental explorer has been used for the diagnosis of caries and for marginal evaluation for more than a century. But the explorer presents some significant problems of  inaccuracy of judgment for both categories. Explorers have a low sensitivity in identifying carious from non-carious fissures and will frequently stick in non-carious fissures.4,5  Diagnosis of caries is best seen visually because there is a clear relationship between the external appearance of caries and its internal development.6,7 
Christensen tested the use of an explorer in detecting acceptable results of marginal fit for visually accessible and visually inaccessible margins in vitro using cast gold onlay  restorations. He found that visible proximal surfaces were judged to be clinically acceptable if the gap was from 9-34 microns; least acceptable margins on the proximal started  with gap sizes at 26 micron and larger. Occlusal marginal gaps were judged clinically acceptable if the gap size was from 2-51 microns; the least acceptable occlusal margin had a  gap size starting at 39 micron. The visually inaccessible gingival surfaces were judged to be clinically acceptable if the margin fell between 34-119 microns. The least acceptable  margin started with gaps of 74 micron. His finding was that visual examination with use of an explorer was more accurate in determining clinical acceptability than the use of the  explorer alone.8 
Comparing the clinical evaluation of crown margins both supra- and sub-gingival, prosthodontists and dental students agreed on the findings of what is clinically unacceptable margin fit. However, the students had more disagreement with their earlier evaluation when the evaluation was repeated six months later compared to the prosthodontists.9  Explorer sharpness has a significant impact on judging the clinical acceptability of sub-gingival margins. A sharp explorer can identify an opening of 36 microns with 95% of the  people using it. Dull explorers have difficulty in detecting marginal gaps of this size.10,11 The sharp explorer can detect horizontal defects of marginal fit accurately, but are  unable to detect vertical discrepancies in marginal fit.12 Marginal fit can be most accurately assessed using a combination of explorer and visual examination.13 The difference  between an alpha and a bravo rating using the USPHS criteria is 170 ± 3 microns in a clinical study of posterior composite restorations.14 

Restoration Longevity and Reasons for Failure
The failure of a restoration is a complex judgment, with numerous reasons provided for restoration replacement. One of the most interesting findings is that the age of replaced  restorations is shortest for the group of clinicians with the least experience and longest for the group of clinicians with >_ 30 years experience.15 With greater clinical experience  comes more certain knowledge of when, e.g., a marginal discrepancy is significant and requires replacement of the restoration and the ability to rate the risk of doing nothing  versus replacement. 
Using the longevity of restorations and the reasons for restoration replacement can tell us some significant patterns of judgments relative to marginal fit and secondary caries or  marginal caries. Annual failure rates for posterior, stress-bearing restorations are reported in the Table I. These values are from two separate studies, and the spread of failures  leads to the conclusion that many factors can cause failure. In these two meta reviews of clinical studies, the most common reasons for replacement were secondary caries, fracture, marginal discrepancies, wear, and postoperative sensitivity.16,17 
Indirect restorations exhibit significantly greater longevity than direct restorations.72  
Longevity has been reported in a different fashion, looking at survival rate, using data from one dentist in Belgium with samples of 722 amalgam restorations, 115 composite  restorations, and 89 crowns placed in 428 adults. Twentyeight percent of the restorations had failed in total, due to fracture of the restoration (8%), secondary caries (6%), or  fracture of the cusp (5%). Failures occurred in premolar teeth (34%) more often than molars (27%), and occurred in 28% of amalgams, 30% of resins, and 24% of the crowns.  The survival times are more than 14.6 years for crowns, 12.8 years for amalgam restorations, and 7.8 years for composite restorations.18  In a study of anterior hybrid composite  restorations, 44 restorations were assessed after eight years, with a 73% survival with all categories of evaluation using USPHS criteria in alpha or bravo categories.19 Posterior 
composites have been shown to have a 74.2% survival at ten years. The greatest risk of failure was due to bulk fractures and partial loss of material and marginal caries.20,21 
Larger direct restorations tend to last fewer years than smaller restorations.22,23,24 Premolar restorations tend to last longer than molar restorations.25 
Finnish studies of longevity report on the age of restoration when they are being redone. Secondary caries is the most common reason for replacement, found in 36% of composite restorations, 52% of glass ionomer restorations, and 41% for amalgam restorations. Fracture of  the restorations occurred in 23% of composite, 11% of glass ionomer,  and 22% of amalgam restorations. Lost restorations occurred in 16%. Median age for replacement was 15 years for amalgam, six years for composite, and seven years for glass  ionomer.26,27
A survey of a single dentist’s practice (Dr. R.V. Tucker) evaluating high quality cast gold restorations up to 52 years of age showed survival rates of 97% at nine years, 90.3% at  20 years, 94.9% at 25 years, 96.9% at 39 years, and 94.1% for restorations more than 40 years old. On the other end of the spectrum, in a dental school population, gold restorations placed from 1963 to 1993 in 890 patients numbered 3,518 gold restorations. One hundred eleven of the restorations were lost. Ten-year survival rates were 76.1%  for occlusal inlays, 88.3% for MO inlays, 83.4% for DO inlays, 87.5% for MOD inlays, and 86.1% for partial crowns. This averages to 85.7% for all restorations.29 
Marginal quality  and development of secondary caries are clearly very significant factors in deterioration of restorations. In a meta-review of clinical studies using 58 studies looking at amalgam Class I and II, composite resin, glass ionomer, and cast gold restorations, the median survival rate for these materials is reported among these studies as  shown in Table II.30 

Secondary or Marginal Caries Secondary or marginal caries are carious lesions found adjacent to existing restoration margins. It is the most common reason for replacement of restorations, as seen in Table II.31 However, the diagnosis of marginal caries is a difficult decision. The lesion could be residual caries left by the previous operator that has  continued to develop. The lesion could also be an active or arrested lesion. The lesion could be newly formed, and if it is, it will develop an outer lesion and a wall lesion.32 Wall  lesions are difficult to determine, whether using an explorer, or transillumination, or X-rays. 
An in vivo study looked at the marginal crevices of amalgam restorations and measured their width. Samples of the plaque from these sites were evaluated for type and quantity of bacteria. The study included frankly carious margins as a control (Figure 1). Plaque samples from wide marginal crevices (>0.4 mm) harbored significantly more mutans  streptococci and lactobacillus than clinically intact margins or narrow crevices (<0.4 mm). These wide marginal crevices also resembled the bacterial makeup of the frankly carious  margins. There was no correlation to discoloration at the margin and development of a carious lesion (Figures 2, 3). The authors recommend preventive replacement of amalgams  with wide crevices.33,34,35,36 
Using the same approach with composite restorations, marginal stains could not reliably predict caries, though composite restorations with stained margins generally have more  mutans streptococcus. Margins with wide crevices (>0.4 mm) have significantly greater amounts of bacteria than narrow crevices (<0.4 mm) but significantly fewer mutans streptococcus than frankly carious lesions. The authors recommend replacing composite restorations only when a frankly carious lesion is present at the margin, as marginal staining is unreliable in predicting development of wall lesions of caries (Figure 3, 4, 5).37,38
An in vitro study measured the required marginal gap size to develop caries in a highly cariogenic environment in sucrose baths. They determined that where the marginal gap was  >30 microns, carious lesions could be detected.39 Oral hygiene plays a key role in the ability of a lesion to form regardless of marginal crevice size. If the environment is highly  cariogenic, marginal caries is likely to form with increasing frequency as the crevice size increases.40,41,42 Beneath amalgam and composite restorations are mostly anaerobic  bacteria (≈ 85-86%) and some aerobic bacteria (≈14-15%). Composite material may have up to eight times more bacteria present than amalgam restorations.43 
Marginal caries is reported to be most prevalent — i.e., 80-90% at the gingival margins — regardless of the type of restorative material.44 

Material Effects on Marginal Fit
Amalgam restorations undergo marginal deterioration as a result of their corrosion in the mouth. A study using 165 extracted or exfoliated teeth with amalgam restorations  examined them to determine the correlation between marginal deterioration and amalgam corrosion. Samples with poor marginal integrity have large quantities of tin-rich and tin- chlorine-rich corrosion with little to no gamma-2 phase remaining. None of these samples were high copper amalgams. These margins were also characterized by obvious fracture  and cracking at the cavosurface margin. Amalgam restorations with good marginal integrity had none of these microscopic cracks or fractures and little sign of corrosion. Six  samples with good marginal integrity did show corrosion products at the margins, indicating corrosion by itself might not be the only important factor in marginal deterioration.45  High copper amalgams have been shown to decrease marginal deterioration over time due to lower creep and less corrosion, causing less marginal fracture and marginal  deterioration. The highest survival rates among the different amalgam alloys are found with zinccontaining high copper amalgam, at 85% after 13 years.46,47 Deterioration of amalgam margins is not related to their ultimate survival. In clinical studies looking at surface roughness, surface tarnish, marginal staining, and marginal fracture, amalgam  restorations lasted as long as 9-11 years after being evaluated with significant marginal stains and fractures, with a median survival of 3.4-3.6 years before replacement, once so identified.48 
Composite resin restorations all undergo polymerization shrinkage, which affects resistance to microleakage and marginal discoloration. Dentinal bonding is not as effective as  enamel bonding at preventing cavosurface discoloration and deterioration of margins.49 While microleakage can be minimized by the use of incremental placement techniques,50,51 occlusal loads and stress from mastication can increase microleakage.52 While it is possible to find bacterial penetration associated with microleakage, and to simulate caries  formation artificially around restoration margins,53,54,55 there is less certainty that microleakage actually causes secondary caries. Secondary caries is thought to be associated with marginal gaps, voids, and openings along the margins of existing restorations.56,57This infection is thought to be no different from that of a primary caries lesion in dentin, except it is located at the faulty margin of a restoration. 
Different composite materials have different marginal defects. Microfilled composites usually exhibit crevices due to marginal fracture; small particle hybrids show evidence of wear and crevice formation; coarse filler composites demonstrate noticeable wear; fine textured composites show fracture of excess composites at margins.58 Marginal integrity of posterior composite restorations is improved with smaller cavities; greater bulk of composite resin at the margin, especially in functional cusp areas; and well-finished margins  without over filling.59 There is question as to whether marginal deterioration and marginal discoloration are important predictors of the failure of composite restorations.60 
Crown margins, whether cast castable porcelain, do not have a uniform fit around the circumference of the margin.61,62 Marginal fit of indirect restorations is affected by use of  die relief, degree of preparation taper, design of the margin, and method of cementation. The Table III demonstrates the marginal fit of materials, as taken from a number of studies.63,64,65,66,67,68,69 
Choice of cement can affect the seating of an indirect restoration. For example, the fit of a cast coping before cementation is 30 microns (median 23 microns), whereas when  cemented with zinc phosphate cements, the fit is 86 microns (median 63 microns). The fit of the coping when cemented with glass ionomer is 47 microns (median 44 microns).70  The use of adhesive luting agents can also affect the final marginal fit of an indirect restoration. In an in vitro study, Class I porcelain inlays were luted using dual-cure resin  cements (Calibra, Choice, and RelyX ARC). Epoxy replicas were made after cementation and analyzed by SEM. When used as a selfcure (without light cure), these materials showed no difference in stress. When used in dual cure (with light activation), higher stress occurred at the margins, with an average gap 22% greater in dentin than enamel,  with wide variations along the margins, but no significant difference between the cements. Choice and RelyX ARC showed greater microleakage in dual cure than self-cure.71  Fatigue loading of ceramic inlays luted with two resin cements and one glass- ionomer cement shows that the glass ionomer cement is unable to resist the deterioration at the margin caused by the fatigue loading. The resin cements, being stronger adhesives, were able to resist this force, and demonstrated no significant marginal deterioration.72  
Choice of material for indirect restoration can affect marginal fit. Porcelain inlays in vitro, when they could be seated, compared to cast gold inlays, show a marginal gap of 26  microns before luting, compared to 15 microns for cast gold.73 Castable ceramic inlays and onlays show a seven percent failure rate after four years. In 79% of these 96  restorations, there are significant marginal deficiencies regardless of the adhesive luting cement used (Figure 6).74 The marginal fit in vitro of a porcelain-fusedto- metal crown  was compared to pressedto- metal crown and to a pressed ceramic crown. There is no significant difference in the fit as measured at four points around each crown with the PFM  at 94± 41 micron; the ceramic pressed to metal crown at 88± 29 micron; and the pressed ceramic at 81± 25 micron.75 CAD-CAM ceramic crowns made of alumina oxide were  analyzed for marginal fit both in vitro and in vivo. In vitro the fit is 30 microns before and after luting. However, in the deepest part of the chamfer, the fit was 135 microns. In  vivo, one of 300 restorations failed within five years due to fracture, with an additional six percent with small porcelain fractures that could be polished. Within the five-year study  period, 1.8% of the margins were unacceptable.76 
Direct composite inlay and onlay restorations (inlay and onlay restorations formed in the mouth and then luted using a resin cement) were compared to direct composite resin  restorations (restorations formed directly in the tooth and bonded during placement) after 11 years of clinical service. The direct inlay/onlay composite material has a 17.7% failure compared to 27.3% of the direct composites. The main reasons for failure of the inlay/onlay composite restorations are fracture = 8.3%; occlusal wear = 4.2%; and  marginal caries = 4.2%, with eight replacements and the others repaired with direct composite resin. The reasons for the failure of the direct composite resin restorations are  fracture = 12.1%; occlusal wear = 6.1%; and marginal caries = 9.1%, with five restorations replaced, while the rest were repaired. Higher rates of failure occurred in molars  compared to premolars. Marginal adaptation was considered acceptable after 11 years, with marginal ditching observed in only a few inlay restorations.77 CAD-CAMproduced indirect composite inlays and onlays after three years show only slight change in marginal adaptation due to the exposure and wear of the resin luting material.78 

Assuming that the contour of a restoration is acceptable, and that no overhang exists, especially at the gingival margin, the marginal gap size may only be significant in highly  cariogenic environments. That said, note that the most common area for marginal caries is on the gingival margin, an area most difficult to judge the marginal fit. Problems in this  area are related to difficulty in isolation from oral fluids while restoring, as well as control of contour for both direct and indirect restorations, preventing overhangs and open  margins. Material properties can also create difficulties in the gingival area. Direct Class II composite restorations, for example, are made impossible when the gingival margin is  located too far from the occlusal to permit adequate light curing. Deep gingival margins are difficult to isolate while making impressions for indirect restorations of all kinds. 
When faced with these circumstances, it may be best to step back and decide if crown lengthening may not present a better opportunity for success, or that the choice of a  different material is required. Mastery of many isolation techniques is required in these circumstances. Maintaining supragingival margins when possible makes judging the fit more  accurate. The actual gap size in many healthy patients may be irrelevant except for esthetic considerations if the restoration is bonded and sealed. However, if the fit of an  indirect restoration is sloppy, the resin cement used to bond the restoration has been shown to develop significant wear and can create an esthetic failure by staining. Marginal fit  is most significant in development of marginal caries when the fit is inadequate — i.e., the gap is larger rather than smaller, with some material differences. In these circumstances, it is recommended to repair carious margins that are accessible rather than replace the entire restoration (Figure 7).79,80 
In conclusion, while marginal fit as envisioned by Ryge is still an important consideration, the criteria used to judge when to replace restorations have changed according to the  material we judge as we have begun to bond all of our restorations and we have become cognizant of the risks for caries development. 
1. Larson TD. Why do we polish? Northwest Dent 2011 May-June;90(3):17-22.
2. Ryge G, Snyder M. Evaluating the clinical quality of restorations. J Am Dent Assoc 1973 Aug;87(2):369-77.
3. Ryge G. Clinical criteria. Int Dent J 1980 Dec;30(4):347-58.
4. Penning C, van Amerongen JP, Seef RE, ten Cate JM. Validity of probing for fi ssure caries diagnosis. Caries Res 1992;26(6):445-9.
5. Tveit AB, Espelid I, Fjelltveit A. Clinical diagnosis of occlusal dentin caries. Caries Res 1994;28(5):368-72. 
6. Ekstrand KR, Kuzmina I, Bjørndal L, Thylstrup A. Relationship between external and histologic features of progressive stages of caries in the occlusal fossa. Caries Res  1995;29(4):243-50. 
7. Chan DC. Current methods and criteria for caries diagnosis in North America. J Dent Educ 1993 Jun;57(6):422-7.
8. Christensen GJ. Marginal fit of gold inlay castings. J Prosthet Dent 1966 Mar-Apr;16(2):297-305.
9. Bronson MR, Lindquist TJ, Dawson DV. Clinical acceptability of crown margins versus marginal gaps as determined by pre-doctoral students and prosthodontists. J Prosthodont  2005 Dec;14(4):226-32. 
10. Baldissara P, Baldissara S, Scotti R. Reliability of tactile perception using sharp and dull explorers in marginal opening identifi cation. Int J Prosthodont 1998 Nov- Dec;11(6):591-4.
11. Rappold AP, Ripps AH, Ireland EJ. Explorer sharpness as related to margin evaluations. Oper Dent 1992 Jan-Feb;17(1):2-6.
12. Hayashi M, Wilson NH, Ebisu S, Watts DC. Influence of explorer tip diameter in identifying restoration margin discrepancies. J Dent 2005 Sep;33(8):669-74. Epub 2005 Mar 5.
13. Jahangiri L, Wahlers C, Hittelman E, Matheson P. Assessment of sensitivity and specifi city of clinical evaluation of cast restoration marginal accuracy compared to  stereomicroscopy. J Prosthet Dent 2005 Feb;93(2):138-42. 
14. Hayashi M, Wilson NH, Watts DC. Quality of marginal adaptation evaluation of posterior composites in clinical trials. J Dent Res 2003 Jan;82(1):59-63.
15. Mjor IA, Dahl JE, Moorhead JE. Age of restorations at replacement in permanent teeth in general dental practice. Acta Odontol Scand 2000 Jun;58(3):97-101. 
16. Hickel R, Manhart J. Longevity of restorations in posterior teeth and reasons for failure. J Adhes Dent 2001 Spring;3(1):45-64. 
17. Manhart J, Chen H, Hamm G, Hickel R. Buonocore Memorial Lecture. Review of the clinical survival of direct and indirect restorations in posterior teeth of the permanent  dentition. Oper Dent 2004 Sep-Oct;29(5):481-508.
18. Van Nieuwenhuysen JP, D’Hoore W, Carvalho J, Qvist V. Long-term evaluation of extensive restorations in permanent teeth. J Dent 2003 Aug;31(6):395-405.
19. Millar BJ, Robinson PB, Inglis AT. Clinical evaluation of an anterior hybrid composite resin over 8 years. Br Dent J 1997 Jan 11;182(1):26-30.
20. Gaengler P, Hoyer I, Montag R, Gaebler P. Micromorphological evaluation of posterior composite restorations - a 10-year report. J Oral Rehabil 2004 Oct;31(10):991-1000. 
21. Collins CJ, Bryant RW, Hodge KL. A clinical evaluation of posterior composite resin restorations: 8-year fi ndings. J Dent 1998 May;26(4):311-7. 
22. Miyamoto T, Morgano SM, Kumagai T, Jones JA, Nunn ME. Treatment history of teeth in relation to the longevity of the teeth and their restorations: Outcomes of teeth  treated and maintained for 15 years. J Prosthet Dent 2007 Mar;97(3):150-6.
23. Gilthorpe MS, Mayhew MT, Bulman JS. Multilevel survival analysis of amalgam restorations amongst RAF personnel. Community Dent Health 2002 Mar;19(1):3-11. 
24. da Rosa Rodolpho PA, Cenci MS, Donassollo TA, Loguercio AD, Demarco FF. A clinical evaluation of posterior composite restorations: 17-year fi ndings. J Dent 2006  Aug;34(7):427-35. Epub 2005 Nov 28.
25. Janus CE, Unger JW, Best AM. Survival analysis of complete veneer crowns vs. multisurface restorations: a dental school patient population. J Dent Educ 2006 Oct;70(10):1,098-104. 
26. Forss H, Widstrom E. From amalgam to composite: selection of restorative materials and restoration longevity in Finland. Acta Odontol Scand 2001 Apr;59(2):57-62. 
27. Forss H, Widstrom E. Reasons for restorative therapy and the longevity of restorations in adults. Acta Odontol Scand 2004 Apr;62(2):82-6. 
28. Donovan T, Simonsen RJ, Guertin G, Tucker RV. Retrospective clinical evaluation of 1,314 cast gold restorations in service from 1 to 52 years. J Esthet Restor Dent  2004;16(3):194-204.
29. Stoll R, Sieweke M, Pieper K, Stachniss V, Schulte A. Longevity of cast gold inlays and partial crowns - a retrospective study at a dental school clinic. Clin Oral Investig 1999 
30. Downer MC, Azli NA, Bedi R, Moles DR, Setchell DJ. How long do routine dental restorations last? A systematic review. Br Dent J 1999 Oct 23;187(8):432-9.
31. Mjör IA, The reasons for replacement and the age of failed restorations in general dental practice. Acta Odontol Scand. 1997 Jan;55(1):58-63.
32. Kidd EA. Caries diagnosis within restored teeth. Adv Dent Res 1990 Jun;4:10-3. 
33. Kidd EA, Joyston-Bechal S, Beighton D. Marginal ditching and staining as a predictor of secondary caries around amalgam restorations: a clinical and microbiological study. J  Dent Res 1995 May;74(5):1,206-11. 
34. Kidd EA, O’Hara JW. The caries status of occlusal amalgam restorations with marginal defects. J Dent Res 1990 Jun;69(6):1,275-7. 
35. Mjör IA, Toffenetti F. Secondary caries: a literature review with case reports. Quintessence Int 2000 Mar;31(3):151. 
36. Hodges DJ, Mangum FI, Ward MT. Relationship between gap width and recurrent dental caries beneath occlusal margins of amalgam restorations. Community Dent Oral  Epidemiol 1995 Aug;23(4):200-4. 
37. Kidd EA, Beighton D. Prediction of secondary caries around tooth-colored restorations: a clinical and microbiological study. J Dent Res 1996 Dec;75(12):1,942-6.
38. Rezwani-Kaminski T, Kamann W, Gaengler P. Secondary caries susceptibility of teeth with long-term performing composite restorations. J Oral Rehabil 2002 Dec;29(12):1,131- 8.
39. Dérand T, Birkhed D, Edwardsson S. Secondary caries related to various marginal gaps around amalgam restorations in vitro. Swed Dent J 1991;15(3):133-8.
40. Goldberg J, Tanzer J, Munster E, Amara J, Thal F, Birkhed D. Cross-sectional clinical evaluation of recurrent enamel caries, restoration of marginal integrity, and oral hygiene  status. J Am Dent Assoc 1981 May;102(5):635-41.
41. Goldberg AJ. Deterioration of restorative materials and the risk for secondary caries. Adv Dent Res 1990 Jun;4:14-8.
42. Mjör IA, Qvist V. Marginal failures of amalgam and composite restorations. J Dent 1997 Jan;25(1):25-30.
43. Splieth C, Bernhardt O, Heinrich A, Bernhardt H, Meyer G. Anaerobic microfl ora under Class I and Class II composite and amalgam restorations. Quintessence Int 2003 Jul- Aug;34(7):497-503.
44. Mjör IA. The location of clinically diagnosed secondary caries. Quintessence Int 1998 May;29(5):313-7. 
45. Port RM, Marshall GW. Characteristics of amalgam restorations with variable clinical appearance. J Am Dent Assoc 1985 Apr;110(4):491-5.
46. Letzel H, van ‘t Hof MA, Marshall GW, Marshall SJ. The infl uence of the amalgam alloy on the survival of amalgam restorations: a secondary analysis of multiple controlled  clinical trials. J Dent Res 1997 Nov;76(11):1,787-98.
47. Smales RJ, Gerke DC, Hume WR. Clinical behaviour of high-copper amalgams with time, site, size and class of cavity preparation. J Dent 1990 Feb;18(1):49-53.
48. Smales RJ, Webster DA. Restoration deterioration related to later failure. Oper Dent 1993 Jul-Aug;18(4):130-7. 
49. Berry TG, Osborne JW. Dentin bonding vs. enamel bonding of composite restorations: a clinical evaluation. Dent Mater 1989 Mar;5(2):90-2.
50. Opdam NJ, Feilzer AJ, Roeters JJ, Smale I. Class I occlusal composite resin restorations: in vivo post-operative sensitivity, wall adaptation, and microleakage. Am J Dent 1998  Oct;11(5):229-34.
51. Müllejans R, Lang H, Schüler N, Baldawi MO, Raab WH. Increment technique for extended Class V restorations: an experimental study. Oper Dent 2003 Jul-Aug;28(4):352-6. 
52. Retief DH. Do adhesives prevent microleakage? Int Dent J 1994 Feb;44(1):19-26. 
53. Purton DG, Rodda JC. Artifi cial caries around restorations in roots. J Dent Res 1988 May;67(5):817-21.
54. Kersten HW, Moorer WR. Particles and molecules in endodontic leakage. Int Endod J 1989 May;22(3):118-24.
55. Coleman AJ. Macromolecular leakage beneath full cast crowns. Part I: The diffusion of lipopolysaccharide and dextran. J Prosthet Dent 1995 Aug;74(2):187-97.
56. Mjor IA, Toffenetti F. Secondary caries: a literature review with case reports. Quintessence Int 2000 Mar;31(3):165-79.
57. Goldberg AJ. Deterioration of restorative materials and the risk for secondary caries. Adv Dent Res 1990 Jun;4:14-18. 
58. Bryant RW, Marzbani N, Hodge KV. Occlusal margin defects around different types of composite resin restorations in posterior teeth. Oper Dent 1992 Nov-Dec;17(6):215-21.
59. Fukushima M, Setcos JC, Phillips RW. Marginal fracture of posterior composite resins. J Am Dent Assoc 1988 Oct;117(5):577-83.
60. Hayashi M, Wilson NH. Marginal deterioration as a predictor of failure of a posterior composite. Eur J Oral Sci 2003 Apr;111(2):155-62.
61. Mitchell CA, Pintado MR, Douglas WH. Nondestructive, in vitro quantifi cation of crown margins. J Prosthet Dent 2001 Jun;85(6):575-84.
62. Holmes JR, Sulik WD, Holland GA, Bayne SC. Marginal fi t of castable ceramic crowns. J Prosthet Dent 1992 May;67(5):594-9.
63. Kern M,Schaller HG, Strub JR. Marginal fit of restorations before and after cementation in vivo. Int J Prosthodont 1993 Nov-Dec;6(6):585-91 
64. Sulaiman F et al. A comparison of the marginal fit if In-Ceram, IPS Empress, and Procera crowns. Int J Prosthodont 1997 Sep- Oct;10(5):478-84
65. Beschnidt SM, Strub JR. Evaluation of marginal accuracy of different all-ceramic crown systems after simulation in the artifi cal mouth. J Oral Rehabil 1999 Jul;26(7):582-93
66. Bindl A, Windisch S, Mormann WH. Fullceramic CAD/CIM anterior crowns and Copings. Int J Comput Dent 1999 Apr;2(2):97-111.
67. Tinschert J et al. Marginal fit of alumina and zirconia-based fi xed partial dentures produced by CAD/CAM system. Oper Dent 2001 Jul-Aug;26(4):367-74.
68. Nakamura T et al. Marginal and internal fit of Cerec 3 CAD/CAM all-ceramic crowns. Int J Prosthodont 2003 May-Jun;16(3):244-8 
69. Yeo IS, Yang JH, Lee JB. In vitro marginal fit of three all ceramic crown systems. J Prosthet Dent 2003 Nov;90(5):459-64.
70. Kern M, Schaller HG, Strub JR. Marginal fit of restorations before and after cementation in vivo. Int J Prosthodont 1993 Nov-Dec;6(6):585-91.
71. Braga RR, Ferracane JL, Condon JR. Polymerization contraction stress in dualcure cements and its effect on interfacial integrity of bonded inlays. J Dent 2002 Sep-Nov;30(7-8):333-40.
72. Abdalla AI, Davidson CL. Marginal integrity after fatigue loading of ceramic inlay restorations luted with three different cements. Am J Dent 2000 Apr;13(2):77-80. 
73. Christensen GJ, Brown TA, Babb R. Accuracy of fit of direct-firing porcelain inlays. J Prosthet Dent 1969 Jul;22(1):46-53.
74. Krämer N, Frankenberger R, Pelka M, Petschelt A. IPS Empress inlays and onlays after four years — a clinical study. J Dent 1999 Jul;27(5):325-31.
75. Goldin EB, Boyd NW, Goldstein GR, Hittelman EL, Thompson VP. Marginal fit of leucite-glass pressable ceramic restorations and ceramicpressed- to-metal restorations. J Prosthet Dent 2005 Feb;93(2):143-7.
76. Naert I, Van der Donck A, Beckers L. Precision of fi t and clinical evaluation of all-ceramic full restorations followed between 0.5 and 5 years. J Oral Rehabil 2005 Jan;32(1):51- 7.
77. van Dijken JW. Direct resin composite inlays/onlays: an 11 year follow-up. J Dent 2000 Jul;28(5):299-306.
78. Fasbinder DJ, Dennison JB, Heys DR, Lampe K. The clinical performance of CAD/CAMgenerated composite inlays. J Am Dent Assoc 2005 Dec;136(12):1714-23.
79. Mjör IA. The reasons for replacement and the age of failed restorations in general dental practice. Acta Odontol Scand 1997 Jan;55(1):58-63. 
80. Mjör IA, Gordan VV. Failure, repair, refurbishing and longevity of restorations. Oper Dent 2002 Sep-Oct;27(5):528-34. 
*Dr. Larson is Associate Professor, Department of Restorative Sciences, Division of Operative Dentistry, University of Minnesota School of Dentistry, 8-450 Moos Tower, 515  Delaware Street S.E., Minneapolis, MN 55455. Phone is (612) 624-5998; email is