Comparative Analysis of CAD-CAM Workflow Variations on the Marginal and Internal Gaps and Fatigue Behavior of Ceramic and Resin Composite Dental Crowns

Abstract

Objectives To analyze the marginal/internal gap and the fatigue behavior of crowns made of two different materials, using four combinations of a digital workflow—two intraoral scanners (IOSs) and two milling machines.

Materials and Methods Crowns were made considering three factors: IOS (a confocal microscopy-based scanner: TRIOS 3—TR; or a combination of active triangulation and dynamic confocal microscopy: Primescan—PS), milling machines (four-axis: CEREC MC XL—CR or five-axis: PrograMill PM7—PM), and restorative material (lithium disilicate—LD or resin composite—RC) (n = 10). The bonding surface of each crown was treated and bonded to each respective glass fiber-reinforced epoxy resin die using a dual-cure resin cement. A computed microtomography analysis was performed to access marginal/internal gap. The specimens were subjected to a cyclic fatigue test (20 Hz, initial load = 100 N/5,000 cycles; step size= 50 N/10,000 cycles until 1,500 N, then if specimens survived, the step size was increased to 100 N/10,000 cycles).

Statistical Analysis For data analysis, three-way analysis of variance and Kaplan–Meier with log-rank (Mantel–Cox) test were performed (α = 0.05).

Results TR resulted in a smaller axial-occlusal angle and occlusal gap, and five-axis milling resulted in a smaller marginal gap, axial-occlusal angle, and occlusal gap. Angled points and occlusal surface showed a tendency for overmilling. RC crowns displayed higher survival rates and a more pronounced topography compared with LD independently of the scanning and milling method. LD crowns produced with a five-axis milling machine resulted in higher fatigue performance and rougher topography compared with a four-axis machine.

Conclusion RC crowns displayed better fatigue behavior compared with LD, while LD benefited from a five-axis machine for improved survival probability.

Keywords

CAD-CAM - fatigue - intraoral scanner - lithium disilicate - resin composite

Introduction

Current chairside dentistry relies on scanning devices to reconstruct the tooth preparation digitally, software to process the data and design the restoration, and milling units.[1] [2] [3] This technology can perform an indirect restoration in a single visit using computer-aided design and computer-aided manufacturing (CAD-CAM). This approach generates a standardized, reliable, efficient, and less technical sensitivity method compared with conventional ones, which depend heavily on the manual skills of the professional.[2] However, there is a wide range of variables in the available CAD-CAM systems that can impact their performance.[3] [4]

For scanning, a light is projected to capture the image and transfer it to software, which will process the data into standard triangle language format.[5] Intraoral scanners (IOSs) may differ in scan depths, recording the image by static or dynamic images, scanning principles, and being open or closed systems.[3] [6] [7] [8] [9] The impact of different scanning methods and depths on their accuracy is significant, and thus the marginal and internal gaps of tooth-supported restorations may be affected.[4] This remains a debated topic in the literature, requiring well-designed studies for clarification,[4] especially considering already validated IOS as a control condition. After designing the restoration, a subtractive milling unit is used to fabricate it. This involves a device attached to diamond or carbide burs that rotate and cut the block into a planned shape at high speed. The units may use three-, four-, or five-axis machines, which may impact on the marginal and internal gaps along with the restoration trueness.[3] [10] [11] The accuracy of the milling will also depend on the burs' size and grit size[11] [12] [13] and the material of choice.[5]

A recent scoping review showed that the milling axis, along with the number and characteristics of the instruments used to mill, can impact the final aspect of a CAD-CAM milled restoration.[14] However, one topic still poorly explored in the literature is the different responses to the milling of various microstructures.[14] There is a large scale of CAD-CAM materials that may be used in a chairside context for single-crown restorations.[15] The literature is already established regarding the success of lithium disilicate (LD) ceramic (a glass-ceramic material) for single crown rehabilitations.[16] However, as a brittle material, it is susceptible to tensile stresses and is especially sensitive to CAD-CAM hard milling, which results in microcracks and residual stresses.[17] [18] As an alternative, resin-based indirect materials stand out considering the lower elastic modulus (closer to the dentin), facilitated adjustment, milling procedure, and intraoral repairs.[19] [20] These differences in microstructure may respond differently to the same CAD-CAM workflow for crowns.

As a consequence of the aforementioned variations, diverse marginal and internal gaps can be observed.[21] [22] Poorly adapted restorations have increased cement thickness at the tooth-restoration interface, potentially resulting in detrimental effects on the restoration's mechanical behavior.[23] Despite some studies that have shown that a cement space of up to 300 µm does not negatively affect the fatigue behavior of leucite crowns,[24] studies have shown that regions with greater discrepancy can induce different load concentrations, making the restoration more susceptible to fracture.[25] Additionally, the potential misfit may negatively impact the periodontium, favoring biological failure in the long term.[26] Therefore, it is important to consider that the scientific literature lacks high-level evidence regarding how variations in the CAD-CAM workflow can affect the marginal and internal gaps and the fatigue behavior of glass-ceramic and resin composite (RC) single crowns.

Based on such assumptions, this study aims to evaluate four combinations of two different scanning devices (one that uses confocal microscopy—TRIOS 3 and the other that uses a combination of active triangulation and dynamic confocal microscopy—Primescan) and two milling machines (a four- and a five-axis), used in the manufacture of single crowns with two different CAD-CAM materials (an LD-based ceramic and a RC). The tested hypotheses were that: (1) there would be no difference regarding the marginal gap of the crowns, independently of scanning device, milling machine, or restorative material; (2) there would be no difference in the internal gap of the crowns, independently of the tested context; and (3) there would be no difference of fatigue behavior, independently of the tested context.

Materials and Methods

Study Design

This in vitro study is composed of eight groups, considering three factors: (1) the intraoral scanning device; (2) the milling machine; and (3) the restorative material used to fabricate the single crown, considering the marginal/internal gap, and fatigue behavior, as described in [Fig. 1].

Fig. 1 Adapted typodont model with the glass fiber-reinforced resin epoxy preparation in position for the scanning procedure and study design. (CR, CEREC MC XL; LD, lithium disilicate; PM, PrograMill PM7; PS, Primescan; RC, resin composite; TR, TRIOS 3.

Specimen Fabrication

Eighty glass fiber-reinforced epoxy resin (Protec Produtos Técnicos Ltda., Brazil) dies were fabricated as the substrate of the eight experimental groups (n = 10). Crown-shaped preparations were made to simulate a superior first molar tooth prepared for a full crown with a ferrule shoulder preparation (depth = 1.2 mm and radii = 0.5 mm) ([Fig. 1]). Each die was individually scanned by a single trained operator (R.O.P.) using the different IOS systems: confocal microscopy-based IOS (TRIOS 3, 3Shape, Denmark), or IOS that uses a combined technology of active triangulation associated with dynamic confocal microscopy (CEREC Primescan AC, Dentsply Sirona, United States). First, the scanning was performed on the die positioned within the model ([Fig. 1]) and then a standard crown was placed on top of it for a second scan to design the restoration using a biogeneric copy.

Each restoration was designed in the systems' specific software—for the 3Shape, the TRIOS Design Studio 21.2.3 (3Shape), and the Primescan the CEREC 5.1.3 (Dentsply Sirona). A 1.2-mm thickness along the restoration and a planned cement space of 120 µm were used in both, considering the manufacturer's standard instructions. The crowns' milling was performed in the units according to the experimental design: a four-axis machine that uses two burs at 42,000 rpm (CEREC MC XL, Dentsply Sirona) or a five-axis machine that uses three burs at 60,000 rpm (PrograMill PM7, Ivoclar AG, Liechtenstein). Two different restorative materials were used—a LD glass-ceramic (IPS e.max CAD, Ivoclar AG) or a RC (Tetric CAD, Ivoclar AG). Milling was performed in fine mode, using the mandatory burs for each machine (Step Bur 12S Ø = 1.35 mm and Cylinder Pointed Bur 12S Ø = 1.75 mm, both with 65 µm grit size, for CEREC MC XL and PrograMill tool red g2.8, g2.0, and g1.0, where the numbers correspond to bur diameters, for PrograMill PM7). Specificities of the milling, such as cooling liquid and instruments' condition were respected according to the particularities of each system. Afterward, LD crowns were crystallized according to the manufacturer's recommendations (speed crystallization in the Programat CS4, Ivoclar AG).

Bonding Procedure

The restorations were cleaned in an ultrasonic bath with 70% alcohol for 5 minutes. The LD crowns were etched with 4.9% hydrofluoric acid (IPS Ceramic Etching-gel, Ivoclar AG) for 20 seconds, washed with air/water spray for 30 seconds, air-dried, and the silane-containing coupling agent (Porcelain Silane, B.J.M. Laboratories Ltd) was actively applied for 15 seconds and allowed to react for another 45 seconds. The RC restorations were sandblasted with 50 µm aluminum oxide (10 mm distance, 1 bar pressure; Ossido di Alluminio, Henry Schein, United States) and cleaned in an ultrasonic bath with 70% alcohol for 5 minutes. Afterward, an adhesive system (Adhese Universal, Ivoclar AG) was actively applied for 20 seconds, followed by a light air jet, without light curing as recommended by the restorative material manufacturer (Ivoclar AG). The epoxy resin preparations were etched with 5% hydrofluoric acid for 60 seconds,[25] washed with air/water spray for 30 seconds, air-dried, and received the application of an adhesive layer as previously described. After the surface treatments, the simplified crowns were adhesively bonded to their respective dies using a dual-cure resin cement (Variolink Esthetic DC, Ivoclar AG) with a standardized load of 5 N. The cement excesses were removed, and the set was light-cured (Starlight Uno, Mectron, Italy) for 40 seconds on each side (0, 90, 180, 270 degrees, and on the top).

Marginal and Internal Gaps

Afterward, the marginal and internal gap measurements were collected using computed microtomography (SkyScan 1172 Micro-CT, Bruker, United States) with the parameters: voltage = 100 kV, current = 100 µA, source-object distance = 89.510 mm, source-detector distance = 217.578 mm, pixel binning = 9.01 µm, exposure time/projection = 790 ms, aluminum and copper (Al + Cu) filter, pixel size = 14.82 µm, averaging = 5, and rotation step = 0.6 degrees.[13] [27] The images were reconstructed (NRecon, Bruker) with the following parameters for LD/RC: smoothening = 0/2, misalignment compensation = 5/4.5, ring artifacts reduction = 10/2, beam-hardening correction = 30/40%. The Data Viewer (Bruker) software was used to generate the cut of the sagittal and coronal images. Then, three selected images per cut were analyzed in ImageJ (1.53t, National Institutes of Health) considering the same regions of interest (ROI) for all images ([Fig. 2]). For marginal gap, a vertical measurement was evaluated, while for internal gap, seven different locations were measured, resulting in a total of 54 measurements per specimen (three images × two cuts × nine ROI; [Fig. 2]).[28]

Fig. 2 Representative image of measurements points in the computed microtomography slices.

Fatigue Test

The bonded assemblies were tested in a cyclic accelerated fatigue test to assess the fatigue behavior.[18] In an electromechanical testing machine (Instron ElectroPuls E3000, Instron, United States), a stainless steel hemispherical piston (Ø = 40 mm) was used to apply the load in the center of the restoration's occlusal region in distilled water. An adhesive tape (110 µm) was used between the restoration and the piston to enhance contact. Cyclic loads were applied at a frequency of 20 Hz,[29] with an initial load of 100 N for 5,000 cycles followed by increments of 50 N every 10,000 cycles until failure was detected or a threshold of 1,500 N was reached. In case of survival up to 1,500 N, the increment was increased to 100 N for every 10,000 cycles, until failure or completion of the test at 2,800 N. At the end of each step, the specimens were transilluminated and visually inspected for failure detection (cracks/fractures). Fatigue failure load (FFL) and number of cycles for failure (CFF) were collected for statistical analysis. Afterward, the specimens underwent fractographic analysis using a stereomicroscope, from which representative specimens were selected and examined using scanning electron microscopy (SEM; VEGA-3G; Tescan, United States) to characterize the failure characteristics. Additionally, a topographic analysis was conducted on an additional specimen per restorative material and milling machine. These specimens were sputter coated with gold and analyzed using SEM.

Statistical Analysis

A power estimation was performed (G*Power software 3.1.9.6, Germany) using a one-way analysis of variance (ANOVA) post hoc power analysis based on α = 0.05, sample size = 80, and effect size = 3.36 using the FFL means and the mean standard deviation. Specific statistical software (IBM SPSS Software v.21, IBM and Statistix 10, Analytical Software, United States) were used to perform analyses of the different tested outcomes (α = 0.05). For marginal/internal gap, after guaranteeing the parametric and homoscedastic data through Shapiro–Wilk (p > 0.05) and Levene's tests (p > 0.05), respectively, a three-way ANOVA with Tukey's post hoc test was conducted. For survival analysis, the Kaplan–Meier test and log-rank (Mantel–Cox) test were performed, along with three-way ANOVA to evaluate each factors' impact on the fatigue behavior. Pearson's linear correlation analysis was performed by comparing the occlusal gap and the FFL.

Results

The power was calculated with a noncentrality parameter λ = 904.7166, critical F = 2.14, numerator df = 7, denominator df = 72, estimated in 1 − β err prob = 1.00 (100%).

Considering the three-way ANOVA for marginal gap ([Tables 1] [2] [3]; [Fig. 3]), no difference was found between the IOSs. The PM milling machine and RC restorations resulted in a lower marginal gap compared with CR and LD (p < 0.05). Regarding the cervical-axial gap, no difference was found between the IOS and milling machines; however, RC crowns showed a smaller gap compared with LD (p < 0.05). In the axial wall gap, no difference was found between the IOS and the restorative materials; however, the CR milling machine resulted in a smaller gap compared with PM. Although the restorative material did not impact the axial occlusal gap results, differences were found regarding the IOS and the milling machine used, whereas TR and PM devices induced smaller gaps compared with each counterpart. The occlusal space was significantly affected by all three factors. The TR scanner, PM machine, and LD crowns showed smaller occlusal gaps compared with each counterpart.

Fig. 3 Survival plot according to the fatigue failure load (A) and number of cycles for failure (B) for each group, obtained by Kaplan–Meier and log-rank (Mantel–Cox) tests. CR, CEREC MC XL; LD, lithium disilicate; PM, PrograMill PM7; PS, Primescan; RC, resin composite; TR, TRIOS 3.

Table 1 Three-way ANOVA for the different measuring points for marginal and internal gaps and fatigue failure load considering the factors intraoral scanner, milling machine, restorative material used, and their respective association

Marginal gap						Axial-occlusal angle
Factors	SS	df	MS	F	p-Value	Factors	SS	df	MS	F	p-Value
Scanner device	3,383	1	3,383	1.63	0.202	Scanner device	70,813	1	70,813	47.05	0.000
Milling machine	155,194	1	155,194	74.82	0.000	Milling machine	25,823	1	25,823	17.16	0.000
Material	270,010	1	270,010	130.17	0.000	Material	1,709	1	1,709	1.14	0.287
Scanner milling machine	14,214	1	14,214	6.85	0.009	Scanner milling machine	69,309	1	69,309	46.05	0.000
Scanner material	29,128	1	29,128	14.04	0.001	Scanner material	5,905	1	5,905	3.92	0.048
Milling machine material	49,536	1	49,536	23.88	0.000	Milling machine material	6,484	1	6,484	4.31	0.038
Scanner milling machine material	814	1	814	0.39	0.531	Scanner milling machine material	46,607	1	46,607	30.97	0.000
Cervical-axial						Occlusal space
Factors	SS	df	MS	F	p -Value	Factors	SS	df	MS	F	p -Value
Scanner device	3,230	1	3230	2.95	0.086	Scanner device	274,229	1	274,229	95.20	0.000
Milling machine	658	1	658	0.60	0.438	Milling machine	429,783	1	429,783	149.20	0.000
Material	37,463	1	37,463	34.20	0.000	Material	15,974	1	15,974	5.55	0.019
Scanner milling machine	105,483	1	105,483	96.30	0.000	Scanner milling machine	17,100	1	17,100	5.94	0.015
Scanner material	381	1	381	0.35	0.555	Scanner material	46,001	1	46,001	15.97	0.001
Milling machine material	6,694	1	6,694	6.11	0.014	Milling machine material	9,408	1	9,408	3.27	0.071
Scanner milling machine material	11,077	1	11,077	10.11	0.001	Scanner milling machine material	174,003	1	174,003	60.41	0.000
Axial wall						Fatigue failure load
Factors	SS	df	MS	F	p -Value	Factors	SS	df	MS	F	p -Value
Scanner device	3,103	1	3,103	3.80	0.051	Scanner device	40,500	1	40,500	0.70	0.404
Milling machine	9,959	1	9,959	12.20	0.001	Milling machine	612,500	1	612,500	10.65	0.002
Material	814	1	814	1.00	0.318	Material	3.864	1	3.864	672.20	0.000
Scanner milling machine	162,917	1	162,917	199.55	0.000	Scanner milling machine	72,000	1	72,000	1.25	0.267
Scanner material	13,756	1	13,756	16.85	0.000	Scanner material	4,500	1	4,500	0.08	0.780
Milling machine material	2,112	1	2,112	2.59	0.108	Milling machine material	364,500	1	364,500	6.34	0.014
Scanner milling machine material	59,189	1	59,189	72.50	0.000	Scanner milling machine material	162,000	1	162,000	2.82	0.097

Abbreviations: ANOVA, analysis of variance; df, degree of freedom; F, F-value; MS, mean squares; SS, sum of squares.

Note: Significant differences (< 0.05) are marked in bold.

Table 2 Mean (in µm) and standard deviation for pairwise comparisons for marginal and internal gap data considering the factors IOS, milling machine, and restorative material isolated

Factor		Marginal gap	Cervical-axial angle	Axial wall	Axial-occlusal angle	Occlusal space
Scanner	TR	69.97 (48.75)A	130.71 (33.07)A	78.25 (34.88)A	135.29 (35.06)A	139.70 (50.05)A
	PS	66.22 (53.19)A	127.04 (37.70)A	74.66 (30.42)A	152.46 (45.69)B	187.50 (77.98)B
Milling machine	CR	80.81 (62.34)B	128.05 (38.75)A	73.23 (29.31)A	149.06 (49.16)B	193.52 (74.72)B
	PM	55.38 (31.67)A	129.70 (31.91)A	79.67 (35.62)B	138.69 (31.53)A	133.68 (48.49)A
Restorative material	LD	84.87 (61.40)B	135.12 (37.86)B	75.53 (26.63)A	145.21 (39.10)A	157.83 (66.17)A
	RC	51.33 (29.66)A	122.63 (31.78)A	77.37 (37.91)A	142.54 (43.95)A	169.37 (72.72)B

Abbreviations: CR, CEREC MC XL; LD, lithium disilicate; IOS, intraoral scanner; PM, PrograMill PM7; PS, Primescan; RC, resin composite; TR, TRIOS 3.

Note: Different capital letters according to three-way analysis of variance with Tukey's post hoc test for each measuring point separately (α = 0.05).

Table 3 Mean comparison, SD, and 95% CI of groups in terms of marginal and internal gap measurements (in µm)

Groups	Marginal gap (n = 120)		Cervical-axial angle (n = 120)		Axial wall (n = 120)		Axial-occlusal angle (n = 120)		Occlusal space (n = 60)
	Mean (SD)	95% CI	Mean (SD)	95% CI	Mean (SD)	95% CI	Mean (SD)	95% CI	Mean (SD)	95% CI
TR-CR-LD	96.37 (70.59)C	83.61–109.13	127.03 (41.23)B	119.58–134.49	66.63 (21.35)AB	62.77–70.49	140.16 (41.84)BC	132.59–147.72	182.28 (54.64)D	168.17–196.40
TR-CR-RC	61.32 (38.88)B	54.29–68.34	111.77 (29.78)A	106.38–117.15	57.37 (19.21)A	53.90–60.85	123.79 (34.98)A	117.47–130.10	145.02 (32.00)BC	136.75–153.28
TR-PM-LD	66.11 (32.50)B	60.23–71.98	148.14 (18.49)C	144.80–151.48	80.46 (28.35)CD	75.33–85.58	138.04 (22.45)ABC	133.98–142.10	105.15 (36.11)A	95.82–114.48
TR-PM-RC	56.11 (32.07)B	50.31–61.90	135.90 (27.64)BC	130.90–140.90	108.54 (41.97)E	100.96–116.13	139.15 (35.91)BC	132.66–145.64	126.33 (39.55)AB	116.12–136.55
PS-CR-LD	113.17 (78.72)C	98.94–127.40	136.28 (43.58)BC	128.40–144.15	80.96 (29.45)CD	75.64–86.28	155.43 (51.46)D	146.13–164.74	184.37 (75.73)D	164.80–203.93
PS-CR-RC	52.40 (21.28)AB	48.55–56.25	137.12 (33.76)BC	131.01–143.22	87.97 (34.48)D	81.73–94.20	176.86 (50.59)E	167.71–186.00	262.42 (73.75)E	243.37–281.47
PS-PM-LD	63.83 (34.57)B	57.58–70.08	129.04 (39.40)B	121.92–136.16	74.08 (24.33)BC	69.68–78.48	147.20 (32.63)CD	141.30–153.10	159.52 (59.67)CD	144.10–174.93
PS-PM-RC	35.48 (13.02)A	33.13–37.84	105.73 (21.88)A	101.78–109.69	55.62 (22.21)A	51.60–59.63	130.36 (31.63)AB	124.64–136.08	143.70 (37.83)BC	133.93–153.47

Abbreviations: CI, confidence interval; CR, CEREC MC XL; LD, lithium disilicate; PM, PrograMill PM7; PS, Primescan; RC, resin composite; SD, standard deviation; TR, TRIOS 3.

Note: Similar superscript letters in each column indicate statistically significant similarity according to three-way analysis of variance and Tukey's post hoc test (α = 0.05).

It is important to highlight that several points exhibited discrepancies compared with the planned cement space of 120 µm, ranging from a 70% reduction in the marginal gap for PS-PM-RC to a 118.68% increase in the occlusal gap for PS-CR-RC. The angled areas (cervical-axial and occlusal-axial angle regions) and occlusal space tended to be overmilled (higher values than what was planned), while the marginal and axial wall gaps were below 120 µm in both materials.

The fatigue behavior of the different tested groups is shown in [Table 4] and [Fig. 4]. According to the three-way ANOVA ([Table 1]), the IOS used did not influence the FFL of the set; however, both milling machines and restorative materials affected the fatigue behavior. Considering the RC restorations, scanners and milling machines resulted in a similar FFL and CFF (p > 0.05). As for the LD, the PM milling machine resulted in a higher FFL compared with the CR one (p < 0.05), independently of the IOS used. The fractographic analysis showed that the failure origin was similar among the groups, located in the center of the crown's bonding surface and related to the milling procedure ([Fig. 4]). It is noticeable that the rougher surface pattern generated by the PM milling machine compared with the CR one seems to introduce a smoother topography ([Fig. 5]). Microcracks were visualized in the LD crowns margin, in contrast to the smooth margin in the RC crowns, independently of the milling machine ([Fig. 5]). There was no linear correlation between the occlusal space and the FFL (p = 0.971; [Fig. 6A]) independently of the tested groups ([Fig. 6B]).

Fig. 4 Representative fractographic analysis of the different materials—lithium disilicate (A, B) and resin composite (C, D). The white arrow indicates the failure origin in the crowns' bonding surface, while dashed arrows indicate the direction of crack propagation (DCP). The yellow arrows indicate the presence of an arrest line at the upper third of the fractured surface on the ×200 image, corroborating to the stepwise propagation of the crack.

Fig. 5 Topographic features of the intaglio surface of the tested groups in different regions. It is possible to notice that the RC crowns (D, E, F, J, K, L) presented a wavy pattern as a consequence of grinding with the milling diamond burs, which is not visible in the LD crowns (A, B, C, G, H, I) that are noticeably smoother, however, with marginal chipping areas (C, I).

Fig. 6 Pearson's linear correlation analysis of the measured occlusal space with the fatigue failure load of the different groups (A). It is noticeable that there is no correlation between these factors, independently of the groups (B). CR, CEREC MC XL; LD, lithium disilicate; PM, PrograMill PM7; PS, Primescan; RC, resin composite; TR, TRIOS 3.

Table 4 Mean, SD, and 95% CI of FFL and number of CFF according to FFL of the different tested groups

Groups	FFL		CFF
	Mean (SD)	95% CI	Mean (SD)	95% CI
TR-CR-LD	790 (51)C	753–827	143,000 (10,328)C	135,612–150,388
TR-CR-RC	2,420 (168)A	2,299–2,541	380,000 (15,811)A	368,689–391,311
TR-PM-LD	1,130 (370)B	865–1,395	201,000 (61,500)B	157,006–244,994
TR-PM-RC	2,310 (129)A	2,218–2,402	366,000 (12,866)A	356,795–375,204
PS-CR-LD	880 (82)C	821–939	161,000 (16,465)C	149,221–172,779
PS-CR-RC	2,300 (313)A	2,076–2,524	362,000 (28,304)A	341,752–382,247
PS-PM-LD	1,160 (212)B	1,008–1,311	217,000 (42,374)B	186,687–247,312
PS-PM-RC	2,490 (354)A	2,237-2,743	386,000 (32,812)A	362,527–409,473

Abbreviations: CCF, cycles for failure; CI, confidence interval; CR, CEREC MC XL; FFL, fatigue failure load; LD, lithium disilicate; PM, PrograMill PM7; PS, Primescan; RC, resin composite; SD, standard deviation; TR, TRIOS 3.

Note: Similar superscript letters in each column indicate statistically significant similarity according to Kaplan–Meier and log-rank (Mantel–Cox) tests (α = 0.05).

Discussion

The results of this study showed that no differences were found between the IOSs on the single crowns' marginal gap. However, the milling machine and the restorative material presented a significant impact on the crowns' marginal gap. Therefore, the first hypothesis was rejected. The internal gap was partially affected by the IOS, the milling machines, and restorative materials, thus, the second hypothesis was rejected. Despite that the IOS did not affected the crowns' fatigue behavior, the milling machine and the restorative material significantly affected; therefore, the third hypothesis was rejected.

Biological complications are a common finding in clinical evaluations of single crowns.[15] In this sense, the marginal gap is highly important, as a poor fit may facilitate biofilm aggregation, potentially accelerating the occurrence of secondary caries or periodontal disease. Despite that there is no consensus, an appropriate marginal threshold that is commonly used in the literature is a 120-µm gap,[5] [21] [30] [31] [32] which was respected by all tested groups. The marginal gap for ceramic crowns reported in the literature ranges from 28 and 123 µm.[31] [33] [34] In this study, a 120-µm marginal gap was used for comparison considering the wide use in the literature and the lack of standardized and qualified clinical studies to support otherwise. Another factor regarding this topic that it is important to state is that the majority of the articles on this topic do not consider the measure after cementation, which is known to affect the restoration marginal gap.[30] In this study, the IOS did not influence the marginal gap, consistent with other studies.[7] [8] Primescan is a relatively new IOS compared with the already established TRIOS 3, and, in theory, it offers the advantages of two combined principles and a larger field of acquisition.[4] Only two studies evaluated the fit after using these devices: one assessed a three-unit zirconia restoration,[8] while the other examined an intraradicular preparation,[7] both reporting no significant difference. In contrast, in our study, TR resulted in a smaller axial-occlusal angle and occlusal gap compared with the PS. Nevertheless, for the other measuring points (cervical-axial angle and axial wall), the IOS devices showed similar gaps.

The milling machine had an important role in the marginal gap.[5] This can be attributed to important differences between the devices: a four-axis, utilizing two burs operating at 42,000 rpm[11] versus a five-axis using three instruments with varying geometries[10] at 60,000 rpm. Five-axis machines tend to yield more precise restorations due to the additional axis, coupled with the use of different burs, which move in multiple directions.[10] [11] Despite it did not affect the cervical-axial angle gap, the five-axis resulted in a smaller marginal gap, axial-occlusal angle, and occlusal gap compared with the four-axis. The angles are regions inherently more challenging to reproduce through milling,[13] as evidenced by overmilling (excessive wear compared with what was planned) in these areas. Previous literature has assessed these devices in surface accuracy, with contradictory results.[10] [11] Thus, further studies are required to gain a better understanding of how these devices may influence single crowns marginal and internal gaps.

The RC restorations showed a smaller marginal gap and cervical-axial angle; however, there was no difference for axial wall and axial-occlusal angle gap when comparing the materials. In contrast, LD presented a smaller occlusal gap. As reported by a previous study,[35] LD crowns may exhibit a larger marginal gap after crystallization, which could explain why the RC presented smaller marginal gaps, as they do not require any postmilling processes. Notably, the LD displayed marginal chipping areas, in contrast to the smoother margins seen in RC crowns ([Fig. 6]), which could also influence in a higher misfit in this area.[19] Furthermore, these materials have inherent differences in microstructures and mechanical properties, which impact the machinability.[27] [36] Since RC is a softer material, it facilitates milling in more complex areas, while the higher LD brittleness may lead to chipping.[17] [19] However, when considering a flat surface, such as the occlusal design, the lower elastic modulus of RC may result in a higher volume of material removed during milling.[37] Considering the inherent limitations of using simplified crowns, it may be stated that in the context of more complex preparations, a resin-based material with lower brittleness and chipping may result in smaller marginal gap.[9]

Subsequently, considering the above-mentioned differences between the tested materials, one could anticipate variations in the resultant fatigue behavior. Independently of the IOS and milling device used, the RC restorations exhibited superior fatigue behavior compared with LD ones, consistent with previous studies and with the noticeable differences in microstructure and mechanical properties between materials.[20] RC materials are known by their elastic behavior and viscoelastic deformation, requiring a higher load to failure compared with the brittle nature of LD ceramic.[38]

Notably, among the RC groups, there was no significant difference, although it is essential to emphasize that the PS-PM-RC group demonstrated a 50% survival rate of the fatigue test ([Fig. 4]). Otherwise, when considering the LD crowns, the use of a five-axis milling machine resulted in better fatigue behavior compared with the four-axis ([Table 4]; [Fig. 4]). Since no significant linear correlation was found between occlusal space and FFL, this may be due to the milling topographic patterns or other characteristics involved on differences between milling systems. Properly balancing the characteristics of the defect population and the contact of the surface with resin cement is crucial.[18] [39] Thus, despite the potential for milling defects to be a failure origin, as shown in the fractographic analysis ([Fig. 5]), a rough surface is beneficial for increasing bonding/micromechanical interlocking and, consequently, fatigue behavior, achieved through optimal filling by the resin cement.[23] Since the five-axis introduced a rougher topography compared with the four-axis ([Fig. 6]), this may explain the improved mechanical performance. Another possible explanation that could be explored in future studies is a potential resultant residual stress or subsurface damage of the different milling machines, which could affect the restoration fatigue behavior.[17] [18]

As an in vitro study, this experiment has inherent limitations. The primary limitations include the use of single crowns with simplified anatomy and the absence of an aging protocol. Additionally, there is a lack of established ideal gap standards in the literature, particularly considering novel fabrication methods and the use of resin cement. Consequently, the current literature on the gaps in CAD-CAM indirect restorations is still based on conventional fabrication methods and zinc phosphate cement. There is a need for clinical studies to evaluate current fabrication methods and their marginal and internal gaps in clinical survival.

Conclusion

The following conclusions were drawn:

1. The different IOSs did not influence the marginal, cervical-angle, and axial wall gaps. However, TRIOS 3 resulted in a smaller axial-occlusal angle and occlusal gap.

2. While the four-axis milling machine resulted in a smaller axial wall gap, the five-axis resulted in a smaller marginal gap, axial-occlusal angle, and occlusal gap.

3. The RC resulted in smaller marginal and cervical-axial gaps, while the LD resulted in a smaller occlusal gap.

4. The IOSs and milling machines did not impact the fatigue behavior of RC crowns, which presented higher survival rates compared with the glass-ceramic.

5. For the LD crowns, a five-axis milling machine resulted in better fatigue behavior compared with a four-axis one.

Conflict of Interest

None declared.

Acknowledgment

We especially thank Ivoclar AG for donating a part of the used materials. We emphasize that the supporting institutions did not have any role in the study design, data collection or analysis, the decision to publish, or in preparing the manuscript.

Authors' Contribution

R.O.P. contributed to conceptualization, data curation, methodology, formal analysis, investigation, visualization, and writing—original draft. L.S.d.R. and R.V.M. to methodology, formal analysis, investigation, and writing—review and editing. A.B. to methodology, software, formal analysis, investigation, and writing—review and editing. J.P.M.T. to methodology, investigation, and writing—review and editing. L.F.V. to conceptualization, methodology, investigation, writing—review and Editing, and funding acquisition. C.J.K. to methodology, investigation, writing—review and editing. N.S. to conceptualization, methodology, software, investigation, writing—review and editing, and supervision. G.K.R.P. to conceptualization, methodology, investigation, writing—review and editing, supervision, project administration, and funding acquisition.

• References

• 1 Sulaiman TA. Materials in digital dentistry-a review. J Esthet Restor Dent 2020; 32 (02) 171-181
• 2 Blatz MB, Conejo J. The current state of chairside digital dentistry and materials. Dent Clin North Am 2019; 63 (02) 175-197
• 3 Alghazzawi TF. Advancements in CAD/CAM technology: options for practical implementation. J Prosthodont Res 2016; 60 (02) 72-84
• 4 Pilecco RO, Dapieve KS, Baldi A, Valandro LF, Scotti N, Pereira GKR. Comparing the accuracy of distinct scanning systems and their impact on marginal/internal adaptation of tooth-supported indirect restorations. A scoping review. J Mech Behav Biomed Mater 2023; 144: 105975
• 5 Goujat A, Abouelleil H, Colon P. et al. Marginal and internal fit of CAD-CAM inlay/onlay restorations: a systematic review of in vitro studies. J Prosthet Dent 2019; 121 (04) 590-597.e3
• 6 Ting-Shu S, Jian S. Intraoral digital impression technique: a review. J Prosthodont 2015; 24 (04) 313-321
• 7 Dupagne L, Mawussi B, Tapie L, Lebon N. Comparison of the measurement error of optical impressions obtained with four intraoral and one extra-oral dental scanners of post and core preparations. Heliyon 2023; 9 (02) e13235
• 8 Özal Ç, Ulusoy M. In-vitro evaluation of marginal and internal fit of 3-unit monolithic zirconia restorations fabricated using digital scanning technologies. J Adv Prosthodont 2021; 13 (06) 373-384
• 9 de Andrade GS, Luz JN, Tribst JPM. et al. Impact of different complete coverage onlay preparation designs and the intraoral scanner on the accuracy of digital scans. J Prosthet Dent 2024; 131 (06) 1168-1177 PubMed
• 10 Kirsch C, Ender A, Attin T, Mehl A. Trueness of four different milling procedures used in dental CAD/CAM systems. Clin Oral Investig 2017; 21 (02) 551-558
• 11 Bosch G, Ender A, Mehl A. A 3-dimensional accuracy analysis of chairside CAD/CAM milling processes. J Prosthet Dent 2014; 112 (06) 1425-1431
• 12 Alajaji NK, Bardwell D, Finkelman M, Ali A. Micro-CT evaluation of ceramic inlays: comparison of the marginal and internal fit of five and three axis CAM systems with a heat press technique. J Esthet Restor Dent 2017; 29 (01) 49-58
• 13 Seo D, Yi Y, Roh B. The effect of preparation designs on the marginal and internal gaps in Cerec3 partial ceramic crowns. J Dent 2009; 37 (05) 374-382
• 14 Pilecco RO, Machry RV, Baldi A. et al. Influence of CAD-CAM milling strategies on the outcome of indirect restorations: a scoping review. J Prosthet Dent 2024; 131 (05) 811.e1-811.e10
• 15 Mazza LC, Lemos CAA, Pesqueira AA, Pellizzer EP. Survival and complications of monolithic ceramic for tooth-supported fixed dental prostheses: a systematic review and meta-analysis. J Prosthet Dent 2022; 128 (04) 566-574
• 16 Aziz A, El-Mowafy O, Tenenbaum HC, Lawrence HP, Shokati B. Clinical performance of chairside monolithic lithium disilicate glass-ceramic CAD-CAM crowns. J Esthet Restor Dent 2019; 31 (06) 613-619
• 17 Curran P, Cattani-Lorente M, Anselm Wiskott HW, Durual S, Scherrer SS. Grinding damage assessment for CAD-CAM restorative materials. Dent Mater 2017; 33 (03) 294-308
• 18 Pilecco RO, Dalla-Nora F, Guilardi LF. et al. In-lab simulation of CAD/CAM milling of lithium disilicate glass-ceramic specimens: effect on the fatigue behavior of the bonded ceramic. J Mech Behav Biomed Mater 2021; 121 (March): 104604
• 19 Tsitrou EA, Northeast SE, van Noort R. Brittleness index of machinable dental materials and its relation to the marginal chipping factor. J Dent 2007; 35 (12) 897-902
• 20 Mueller B, Pilecco RO, Valandro LF, Ruschel VC, Pereira GKR, Bernardon JK. Effect of immediate dentin sealing on load-bearing capacity under accelerated fatigue of thin occlusal veneers made of CAD-CAM glass-ceramic and resin composite material. Dent Mater 2023; 39 (04) 372-382
• 21 McLean JW, von Fraunhofer JA. The estimation of cement film thickness by an in vivo technique. Br Dent J 1971; 131 (03) 107-111
• 22 Mangano F, Gandolfi A, Luongo G, Logozzo S. Intraoral scanners in dentistry: a review of the current literature. BMC Oral Health 2017; 17 (01) 149
• 23 May LG, Kelly JR, Bottino MA, Hill T. Effects of cement thickness and bonding on the failure loads of CAD/CAM ceramic crowns: multi-physics FEA modeling and monotonic testing. Dent Mater 2012; 28 (08) e99-e109
• 24 Scherrer SS, de Rijk WG, Belser UC, Meyer JM. Effect of cement film thickness on the fracture resistance of a machinable glass-ceramic. Dent Mater 1994; 10 (03) 172-177
• 25 Kelly JR, Rungruanganunt P, Hunter B, Vailati F. Development of a clinically validated bulk failure test for ceramic crowns. J Prosthet Dent 2010; 104 (04) 228-238
• 26 Ercoli C, Caton JG. Dental prostheses and tooth-related factors. J Periodontol 2018; 89 (Suppl. 01) S223-S236
• 27 de Paula Silveira AC, Chaves SB, Hilgert LA, Ribeiro APD. Marginal and internal fit of CAD-CAM-fabricated composite resin and ceramic crowns scanned by 2 intraoral cameras. J Prosthet Dent 2017; 117 (03) 386-392
• 28 Nawafleh NA, Mack F, Evans J, Mackay J, Hatamleh MM. Accuracy and reliability of methods to measure marginal adaptation of crowns and FDPs: a literature review. J Prosthodont 2013; 22 (05) 419-428
• 29 Velho HC, Dapieve KS, Rocha Pereira GK, Fraga S, Valandro LF, Venturini AB. Accelerated loading frequency does not influence the fatigue behavior of polymer infiltrated ceramic network or lithium disilicate glass-ceramic restorations. J Mech Behav Biomed Mater 2020; 110: 103905
• 30 Peroz I, Mitsas T, Erdelt K, Kopsahilis N. Marginal adaptation of lithium disilicate ceramic crowns cemented with three different resin cements. Clin Oral Investig 2019; 23 (01) 315-320
• 31 Sadid-Zadeh R, Katsavochristou A, Squires T, Simon M. Accuracy of marginal fit and axial wall contour for lithium disilicate crowns fabricated using three digital workflows. J Prosthet Dent 2020; 123 (01) 121-127
• 32 Abdel-Azim T, Rogers K, Elathamna E, Zandinejad A, Metz M, Morton D. Comparison of the marginal fit of lithium disilicate crowns fabricated with CAD/CAM technology by using conventional impressions and two intraoral digital scanners. J Prosthet Dent 2015; 114 (04) 554-559
• 33 Rizonaki M, Jacquet W, Bottenberg P, Depla L, Boone M, De Coster PJ. Evaluation of marginal and internal fit of lithium disilicate CAD-CAM crowns with different finish lines by using a micro-CT technique. J Prosthet Dent 2022; 127 (06) 890-898
• 34 Mostafa NZ, Ruse ND, Ford NL, Carvalho RM, Wyatt CCL. Marginal fit of lithium disilicate crowns fabricated using conventional and digital methodology: a three-dimensional analysis. J Prosthodont 2018; 27 (02) 145-152
• 35 Kim JH, Oh S, Uhm SH. Effect of the crystallization process on the marginal and internal gaps of lithium disilicate CAD/CAM crowns. BioMed Res Int 2016; 2016: 8635483
• 36 Duan Y, Griggs JA. Effect of elasticity on stress distribution in CAD/CAM dental crowns: glass ceramic vs. polymer-matrix composite. J Dent 2015; 43 (06) 742-749
• 37 Coldea A, Fischer J, Swain MV, Thiel N. Damage tolerance of indirect restorative materials (including PICN) after simulated bur adjustments. Dent Mater 2015; 31 (06) 684-694
• 38 Lohbauer U, Belli R, Ferracane JL. Factors involved in mechanical fatigue degradation of dental resin composites. J Dent Res 2013; 92 (07) 584-591
• 39 May MM, Fraga S, May LG. Effect of milling, fitting adjustments, and hydrofluoric acid etching on the strength and roughness of CAD-CAM glass-ceramics: a systematic review and meta-analysis. J Prosthet Dent 2022; 128 (06) 1190-1200

Publication History

Article published online:
12 March 2025