Evaluation of Fracture Resistance in Aqueous Environment under Dynamic Loading of Lithium Disilicate Restorative Systems for Posterior Applications. Part 2 Matilda Dhima, DMD, MS,1 Alan B. Carr, DMD, MS,2 Thomas J. Salinas, DDS,3 Christine Lohse, MS,4 Lawrence Berglund, BS,5 & Kai-An Nan, PhD6 1
Former Assistant Professor of Dentistry, Mayo Clinic College of Medicine, and Chief Resident, Prosthodontics and Maxillofacial Prosthetics, Division of Prosthetic and Esthetic Dentistry, Department of Dental Specialties, Mayo Clinic, Rochester, MN Currently: Maxillofacial Prosthodontist, University of Pittsburgh School of Dental Medicine, Department of Prosthodontics, Pittsburgh, PA 2 Professor of Dentistry, Mayo Clinic College of Medicine, Chair, Consultant, Department of Dental Specialties, Mayo Clinic, Rochester, MN 3 Professor of Dentistry, Mayo Clinic College of Medicine, Consultant, Department of Dental Specialties, Mayo Clinic, Rochester, MN 4 Statistician, Department of Biostatistics, Mayo Clinic, Rochester, MN 5 Engineer, Department of Biomechanical Testing, Mayo Clinic, Rochester, MN 6 Professor of Biomechanical Engineering, Mayo College of Medicine, Consultant, Department of Orthopedic Surgery, Mayo Clinic, Rochester, MN The article is associated with the American College of Prosthodontists’ journal-based continuing education program. It is accompanied by an online continuing education activity worth 1 credit. Please visit www.wileyhealthlearning.com/jopr to complete the activity and earn credit.
Keywords Lithium disilicate; ceramic; dynamic testing; bonding; load; crown; tooth; indenter; aqueous; digital technology. Correspondence Dr. Matilda Dhima, University of Pittsburgh School of Dental Medicine, Department of Prosthodontics, 3501 Terrace Street Suite 3189, Pittsburgh, PA 15261. E-mail: [email protected] This study was funded in part by the Greater New York Academy of Prosthodontics (GNYAP) Research Grant. The authors deny any conflicts of interest. Accepted August 21, 2013 doi: 10.1111/jopr.12124
Abstract Purpose: The goals of part 2 of the study presented here were 1) to assess whether there is a difference in failure mode of different thicknesses (2.0, 1.5, 1.0, and 0.5 mm) of anatomically standardized full contour monolithic lithium disilicate restorations for posterior teeth, and 2) to assess if there is a difference among various crown thicknesses when these restorations are subjected to dynamic load forces common for posterior teeth. Materials and Methods: Four groups (n = 10), each with a different thickness of anatomically appropriate all-ceramic crowns, were to be tested as established from the statistical analysis of the preliminary phase. Group 1: 2.0 mm; group 2: 1.5 mm; group 3: 1.0 mm; group 4: 0.5 mm. The specimens were adhesively luted to the corresponding die, and underwent dynamic cyclic loading (380 to 390 N) completely submerged in an aqueous environment until a failure was noted by graphic recording and continuous monitoring. Results: There was a statistically significant difference of the fatigue cycles to failure among four groups (p < 0.001; Kruskal-Wallis test). The mean number of cycles to fail for 2.0 mm specimens was 17 times more than the mean number of cycles to fail for 1.0 mm specimens and 1.5 times more than the mean number of cycles to fail for 1.5 mm specimens. The 0.5 mm specimens failed with one cycle of loading. A qualitative characteristic noted among the 2.0 mm specimens was wear of the area of indenter contact followed by shearing of the material and/or crack propagation. Conclusion: Based on the findings of this study, it may be reasonable to consider a crown thickness of 1.5 mm or greater for clinical applications of milled monolithic lithium disilicate crowns for posterior single teeth.
Several studies1-3 have investigated in vitro fracture resistance and origin of failure of simulated posterior teeth for all-ceramic crowns. The fracture of veneering porcelain and/or ceramic coping is the most commonly reported major complication, often requiring restoration remakes.4 The strength of all-ceramic restorations is dependent on the ceramic material(s) used, core/veneer bond strength, crown thickness, and restoration design.5,6
Currently, no clear evidence suggests the appropriate amount of tooth reduction for all-ceramic restorations. The guidelines for metal-ceramic restorations recommend a minimum of 2.0 mm reduction of tooth structure on functional cusps to accommodate such crown design. With significant tooth reduction of this sort, there is an increased risk of tooth sensitivity, dentin exposure, and risk for postoperative sensitivity or pulpal inflammation.
C 2014 by the American College of Prosthodontists Journal of Prosthodontics 23 (2014) 353–357
353
Dynamic Loading Lithium Disilicate for Posterior Applications
The introduction of all-ceramic crowns for anterior teeth has allowed for less tooth preparation, which offers the possibility of maintaining some of the enamel surface. This approach is advantageous for bonding such restorations chemically and mechanically, increasing their strength, decreasing postoperative sensitivity and risk for pulpal inflammation. A recent clinical comparison of the performance between thermally pressed monolithic lithium disilicate single crowns with nanofluorapatite-layered lithium disilicate and metal-ceramic crowns showed that beyond 2 and 3 years, gradual roughness and wear of the layered lithium disilicate crowns was found.6 The findings of part 17 of this multiphase study presented here identified lithium disilicate as the material of interest to be investigated under cyclic dynamic fatigue based on the following reasons. It was sufficiently resistant for restoration of posterior teeth with mean static failure loads of 743.1 ± 114.3 N and the narrowest range of standard deviation. Additionally, digital die and restoration fabrication by scanning technology allows for specimen standardization and ease of fabrication. Upon completion of a literature review on clinically observed maximum peak forces,8-12 it was noted that maximum peak forces are present only for a fraction of a second for every chewing cycle, and continually change in magnitude and direction. As a result, it was decided that specimens be subjected dynamically (mouth motion fatigue: indenter contact-load-slide liftoff) to a constant load slightly less than mean maximum bite forces of 500 N for posterior teeth.9,10,13 This represents almost 75% of the lowest forces (536 N) found in part 1 for lithium disilicate.7 Since lithium disilicate is readily produced, it is especially applicable for restoration of posterior teeth where sufficient reduction can be achieved. As with other bonded ceramic restorations, it is of interest to investigate the minimum thickness required for resistance to fracture. It was determined that four thicknesses of anatomically standardized full contour restorations for a mandibular molar be evaluated with sample sizes as established from the statistical analysis of the preliminary phase. The hypothesis of part 2 of the study was that there is no difference in fracture mode for various thicknesses of anatomically appropriate full-contour lithium disilicate single crowns when subjected to dynamic load forces common for posterior teeth. The aims of part 2 of the study presented here were 1) to assess whether there is a difference in failure mode of different thicknesses (2.0, 1.5, 1.0, and 0.5 mm) of anatomically standardized full-contour restorations for posterior teeth, and 2) to assess if there is a difference among various crown thicknesses when these restorations are subjected to dynamic load forces common for posterior teeth.
Materials and methods The material of interest identified from preliminary testing of four restorative systems in part 17 was monolithic lithium disilicate, based on mean failure loads within the range of loads for posterior teeth, narrowest standard deviation, ability to standardize fabrication within and among groups by scanning technology, and ease of specimen fabrication. In addition, the preliminary testing suggested that a load in the range of 350 to 400
354
Dhima et al
N would be appropriate for dynamic testing. Four groups (n = 10), each with a different thickness of anatomically appropriate all-ceramic crowns, were to be tested as established from the statistical analysis of the preliminary phase. Group 1: 2.0 mm; group 2: 1.5 mm; group 3: 1.0 mm; group 4: 0.5 mm. For each group, specimens were fabricated as follows. A dentoform (Columbia Dentoform, Long Island City, NY) was equilibrated with rotary instruments. A clear stent (X-Wide Coping & Temp C & B Material; Buffalo Dental, Syosset, NY) was fabricated for the mandibular arch to allow measuring of the reduction on all tooth surfaces using a periodontal probe (Patterson Dental, St. Paul, MN). For each group, one mandibular right first molar was reduced uniformly as determined for each group. Chairside CEREC 4.0 Software (Sirona Dental Systems, New York, NY) was used to scan the dentoform model virtually for the four preparations representing each group (Fig 1). From the virtual model, a die was fabricated to represent the reduced tooth in each group. A total of 40 dies (n = 10/ group) were milled with 3M Paradigm MZ100 Block (3M ESPE, St. Paul, MN) for CEREC using the corresponding milling unit, MCXL (Sirona Dental Systems). Next, the CEREC 4.0 chairside software was used to digitally design a crown for each of the dies representing every group. Based on this digital design, a total of 40 lithium disilicate crowns (n = 10 per group) were milled with IPS e.max CAD (Ivoclar Vivadent, Amherst, NY). The milled crowns underwent crystallization, glazing, and finishing following manufacturer’s guidelines. Each crown’s intaglio surface was treated with 4.5% hydrofluoric acid according to manufacturer’s guidelines. For each group, crowns were adhesively luted to corresponding dies with resin cement (RelyX cement; 3M ESPE) based on manufacturer’s guidelines. The specimens were stored in an aqueous environment for 126 days. The crowns in all groups were subjected to a mouth motion fatigue test: indenter contact-load-slide liftoff. The sliding phase of molars for posterior teeth starts with the inner inclines of the maxillary buccal cusps’ eccentric contact with mandibular buccal cusps, moving through centric then lifting off. According to Sakaguchi et al,14 the average length of the sliding path of a first molar is approximately 0.5 mm. The duration of forces is known to be 0.25 to 0.33 seconds.15 The indenter contacted the specimen on the inner incline of the buccal cusp, sliding 0.5 mm lingually down the cusp beginning at 0.5 mm lingual to the cusp tip. The specimens underwent dynamic cyclic loading (380 to 390 N) in an aqueous environment until a failure was noted by graphically recording and continuous monitoring. The loading cycles were recorded with MTS software (MTS Corp., Eden Prairie, MN). The selection of material composition and size of the testing indenter was as stated in part 1.7 To ensure consistency, the indenter surface was inspected regularly under magnification and light to ensure no visual changes in indenter surface area. The restorations were also monitored visually for any surface changes. The graphical recordings were further evaluated to assess number of cycles and where a change in the pattern of cycles was noted. Failure was assessed visually and R
R
C 2014 by the American College of Prosthodontists Journal of Prosthodontics 23 (2014) 353–357
Dhima et al
Dynamic Loading Lithium Disilicate for Posterior Applications
Milled die
Scanning of preparation for die fabrication
Tooth reduction on model
Scanning of preparation for crown fabrication
Milled crown
Figure 1 Specimen preparation steps from dentoform to digital fabrication of dies and anatomically correct crowns.
Indenter contact-loadslide lift-off model
Crack propagation radiating from indenter contact area
Wear and chipping on indenter contact area
Figure 2 Location of indenter contact and failure characteristics.
graphically and defined as fracture or a crack on the restoration. Data were summarized using means, standard deviation (SD), range (minimum to maximum), and interquartile range (25th to 75th percentile). Comparisons among groups were evaluated using Kruskal-Wallis and Wilcoxon rank sum tests.
Results Overall, there was a statistically significant difference of the fatigue cycles to failure among the four groups (p < 0.001; Kruskal-Wallis test) (Table 1). When evaluating pairwise comparisons, the p-values for comparisons of number of cycles to
C 2014 by the American College of Prosthodontists Journal of Prosthodontics 23 (2014) 353–357
355
Dynamic Loading Lithium Disilicate for Posterior Applications
Dhima et al
Table 1 Failure characteristics under dynamic testing Group (n = 10) 1 (2.0 mm) 2 (1.5 mm) 3 (1.0 mm) 4 (0.5 mm)
Number of cycles to failure Mean ± SD (Range) 23,520 14,756 1282 1
± ± ± ±
2730 (20,268 to 28,995) 2196 (11,477 to 18,138) 231 (1023 to 1702) 0 (1 to 1)
failure for group 1 versus 2, group 1 versus 3, and group 2 versus 3 were all
Former Assistant Professor of Dentistry, Mayo Clinic College of Medicine, and Chief Resident, Prosthodontics and Maxillofacial Prosthetics, Division of Prosthetic and Esthetic Dentistry, Department of Dental Specialties, Mayo Clinic, Rochester, MN Currently: Maxillofacial Prosthodontist, University of Pittsburgh School of Dental Medicine, Department of Prosthodontics, Pittsburgh, PA 2 Professor of Dentistry, Mayo Clinic College of Medicine, Chair, Consultant, Department of Dental Specialties, Mayo Clinic, Rochester, MN 3 Professor of Dentistry, Mayo Clinic College of Medicine, Consultant, Department of Dental Specialties, Mayo Clinic, Rochester, MN 4 Statistician, Department of Biostatistics, Mayo Clinic, Rochester, MN 5 Engineer, Department of Biomechanical Testing, Mayo Clinic, Rochester, MN 6 Professor of Biomechanical Engineering, Mayo College of Medicine, Consultant, Department of Orthopedic Surgery, Mayo Clinic, Rochester, MN The article is associated with the American College of Prosthodontists’ journal-based continuing education program. It is accompanied by an online continuing education activity worth 1 credit. Please visit www.wileyhealthlearning.com/jopr to complete the activity and earn credit.
Keywords Lithium disilicate; ceramic; dynamic testing; bonding; load; crown; tooth; indenter; aqueous; digital technology. Correspondence Dr. Matilda Dhima, University of Pittsburgh School of Dental Medicine, Department of Prosthodontics, 3501 Terrace Street Suite 3189, Pittsburgh, PA 15261. E-mail: [email protected] This study was funded in part by the Greater New York Academy of Prosthodontics (GNYAP) Research Grant. The authors deny any conflicts of interest. Accepted August 21, 2013 doi: 10.1111/jopr.12124
Abstract Purpose: The goals of part 2 of the study presented here were 1) to assess whether there is a difference in failure mode of different thicknesses (2.0, 1.5, 1.0, and 0.5 mm) of anatomically standardized full contour monolithic lithium disilicate restorations for posterior teeth, and 2) to assess if there is a difference among various crown thicknesses when these restorations are subjected to dynamic load forces common for posterior teeth. Materials and Methods: Four groups (n = 10), each with a different thickness of anatomically appropriate all-ceramic crowns, were to be tested as established from the statistical analysis of the preliminary phase. Group 1: 2.0 mm; group 2: 1.5 mm; group 3: 1.0 mm; group 4: 0.5 mm. The specimens were adhesively luted to the corresponding die, and underwent dynamic cyclic loading (380 to 390 N) completely submerged in an aqueous environment until a failure was noted by graphic recording and continuous monitoring. Results: There was a statistically significant difference of the fatigue cycles to failure among four groups (p < 0.001; Kruskal-Wallis test). The mean number of cycles to fail for 2.0 mm specimens was 17 times more than the mean number of cycles to fail for 1.0 mm specimens and 1.5 times more than the mean number of cycles to fail for 1.5 mm specimens. The 0.5 mm specimens failed with one cycle of loading. A qualitative characteristic noted among the 2.0 mm specimens was wear of the area of indenter contact followed by shearing of the material and/or crack propagation. Conclusion: Based on the findings of this study, it may be reasonable to consider a crown thickness of 1.5 mm or greater for clinical applications of milled monolithic lithium disilicate crowns for posterior single teeth.
Several studies1-3 have investigated in vitro fracture resistance and origin of failure of simulated posterior teeth for all-ceramic crowns. The fracture of veneering porcelain and/or ceramic coping is the most commonly reported major complication, often requiring restoration remakes.4 The strength of all-ceramic restorations is dependent on the ceramic material(s) used, core/veneer bond strength, crown thickness, and restoration design.5,6
Currently, no clear evidence suggests the appropriate amount of tooth reduction for all-ceramic restorations. The guidelines for metal-ceramic restorations recommend a minimum of 2.0 mm reduction of tooth structure on functional cusps to accommodate such crown design. With significant tooth reduction of this sort, there is an increased risk of tooth sensitivity, dentin exposure, and risk for postoperative sensitivity or pulpal inflammation.
C 2014 by the American College of Prosthodontists Journal of Prosthodontics 23 (2014) 353–357
353
Dynamic Loading Lithium Disilicate for Posterior Applications
The introduction of all-ceramic crowns for anterior teeth has allowed for less tooth preparation, which offers the possibility of maintaining some of the enamel surface. This approach is advantageous for bonding such restorations chemically and mechanically, increasing their strength, decreasing postoperative sensitivity and risk for pulpal inflammation. A recent clinical comparison of the performance between thermally pressed monolithic lithium disilicate single crowns with nanofluorapatite-layered lithium disilicate and metal-ceramic crowns showed that beyond 2 and 3 years, gradual roughness and wear of the layered lithium disilicate crowns was found.6 The findings of part 17 of this multiphase study presented here identified lithium disilicate as the material of interest to be investigated under cyclic dynamic fatigue based on the following reasons. It was sufficiently resistant for restoration of posterior teeth with mean static failure loads of 743.1 ± 114.3 N and the narrowest range of standard deviation. Additionally, digital die and restoration fabrication by scanning technology allows for specimen standardization and ease of fabrication. Upon completion of a literature review on clinically observed maximum peak forces,8-12 it was noted that maximum peak forces are present only for a fraction of a second for every chewing cycle, and continually change in magnitude and direction. As a result, it was decided that specimens be subjected dynamically (mouth motion fatigue: indenter contact-load-slide liftoff) to a constant load slightly less than mean maximum bite forces of 500 N for posterior teeth.9,10,13 This represents almost 75% of the lowest forces (536 N) found in part 1 for lithium disilicate.7 Since lithium disilicate is readily produced, it is especially applicable for restoration of posterior teeth where sufficient reduction can be achieved. As with other bonded ceramic restorations, it is of interest to investigate the minimum thickness required for resistance to fracture. It was determined that four thicknesses of anatomically standardized full contour restorations for a mandibular molar be evaluated with sample sizes as established from the statistical analysis of the preliminary phase. The hypothesis of part 2 of the study was that there is no difference in fracture mode for various thicknesses of anatomically appropriate full-contour lithium disilicate single crowns when subjected to dynamic load forces common for posterior teeth. The aims of part 2 of the study presented here were 1) to assess whether there is a difference in failure mode of different thicknesses (2.0, 1.5, 1.0, and 0.5 mm) of anatomically standardized full-contour restorations for posterior teeth, and 2) to assess if there is a difference among various crown thicknesses when these restorations are subjected to dynamic load forces common for posterior teeth.
Materials and methods The material of interest identified from preliminary testing of four restorative systems in part 17 was monolithic lithium disilicate, based on mean failure loads within the range of loads for posterior teeth, narrowest standard deviation, ability to standardize fabrication within and among groups by scanning technology, and ease of specimen fabrication. In addition, the preliminary testing suggested that a load in the range of 350 to 400
354
Dhima et al
N would be appropriate for dynamic testing. Four groups (n = 10), each with a different thickness of anatomically appropriate all-ceramic crowns, were to be tested as established from the statistical analysis of the preliminary phase. Group 1: 2.0 mm; group 2: 1.5 mm; group 3: 1.0 mm; group 4: 0.5 mm. For each group, specimens were fabricated as follows. A dentoform (Columbia Dentoform, Long Island City, NY) was equilibrated with rotary instruments. A clear stent (X-Wide Coping & Temp C & B Material; Buffalo Dental, Syosset, NY) was fabricated for the mandibular arch to allow measuring of the reduction on all tooth surfaces using a periodontal probe (Patterson Dental, St. Paul, MN). For each group, one mandibular right first molar was reduced uniformly as determined for each group. Chairside CEREC 4.0 Software (Sirona Dental Systems, New York, NY) was used to scan the dentoform model virtually for the four preparations representing each group (Fig 1). From the virtual model, a die was fabricated to represent the reduced tooth in each group. A total of 40 dies (n = 10/ group) were milled with 3M Paradigm MZ100 Block (3M ESPE, St. Paul, MN) for CEREC using the corresponding milling unit, MCXL (Sirona Dental Systems). Next, the CEREC 4.0 chairside software was used to digitally design a crown for each of the dies representing every group. Based on this digital design, a total of 40 lithium disilicate crowns (n = 10 per group) were milled with IPS e.max CAD (Ivoclar Vivadent, Amherst, NY). The milled crowns underwent crystallization, glazing, and finishing following manufacturer’s guidelines. Each crown’s intaglio surface was treated with 4.5% hydrofluoric acid according to manufacturer’s guidelines. For each group, crowns were adhesively luted to corresponding dies with resin cement (RelyX cement; 3M ESPE) based on manufacturer’s guidelines. The specimens were stored in an aqueous environment for 126 days. The crowns in all groups were subjected to a mouth motion fatigue test: indenter contact-load-slide liftoff. The sliding phase of molars for posterior teeth starts with the inner inclines of the maxillary buccal cusps’ eccentric contact with mandibular buccal cusps, moving through centric then lifting off. According to Sakaguchi et al,14 the average length of the sliding path of a first molar is approximately 0.5 mm. The duration of forces is known to be 0.25 to 0.33 seconds.15 The indenter contacted the specimen on the inner incline of the buccal cusp, sliding 0.5 mm lingually down the cusp beginning at 0.5 mm lingual to the cusp tip. The specimens underwent dynamic cyclic loading (380 to 390 N) in an aqueous environment until a failure was noted by graphically recording and continuous monitoring. The loading cycles were recorded with MTS software (MTS Corp., Eden Prairie, MN). The selection of material composition and size of the testing indenter was as stated in part 1.7 To ensure consistency, the indenter surface was inspected regularly under magnification and light to ensure no visual changes in indenter surface area. The restorations were also monitored visually for any surface changes. The graphical recordings were further evaluated to assess number of cycles and where a change in the pattern of cycles was noted. Failure was assessed visually and R
R
C 2014 by the American College of Prosthodontists Journal of Prosthodontics 23 (2014) 353–357
Dhima et al
Dynamic Loading Lithium Disilicate for Posterior Applications
Milled die
Scanning of preparation for die fabrication
Tooth reduction on model
Scanning of preparation for crown fabrication
Milled crown
Figure 1 Specimen preparation steps from dentoform to digital fabrication of dies and anatomically correct crowns.
Indenter contact-loadslide lift-off model
Crack propagation radiating from indenter contact area
Wear and chipping on indenter contact area
Figure 2 Location of indenter contact and failure characteristics.
graphically and defined as fracture or a crack on the restoration. Data were summarized using means, standard deviation (SD), range (minimum to maximum), and interquartile range (25th to 75th percentile). Comparisons among groups were evaluated using Kruskal-Wallis and Wilcoxon rank sum tests.
Results Overall, there was a statistically significant difference of the fatigue cycles to failure among the four groups (p < 0.001; Kruskal-Wallis test) (Table 1). When evaluating pairwise comparisons, the p-values for comparisons of number of cycles to
C 2014 by the American College of Prosthodontists Journal of Prosthodontics 23 (2014) 353–357
355
Dynamic Loading Lithium Disilicate for Posterior Applications
Dhima et al
Table 1 Failure characteristics under dynamic testing Group (n = 10) 1 (2.0 mm) 2 (1.5 mm) 3 (1.0 mm) 4 (0.5 mm)
Number of cycles to failure Mean ± SD (Range) 23,520 14,756 1282 1
± ± ± ±
2730 (20,268 to 28,995) 2196 (11,477 to 18,138) 231 (1023 to 1702) 0 (1 to 1)
failure for group 1 versus 2, group 1 versus 3, and group 2 versus 3 were all
