Effect of Fabrication Technique on the Marginal Discrepancy and Resistance of Lithium Disilicate Crowns: An In Vitro Study

Abstract

Purpose:

To evaluate the impact of fabrication technique on the marginal fit and resistance of lithium disilicate crowns.

Materials and Methods:

Twelve ivorine molars were prepared to receive lithium disilicate crowns. The preparations were digitally recorded using an intraoral scanner, and the crowns were designed following the anatomy of an unprepared tooth using a design software. The designed crowns were fabricated using 3 techniques: 1) milling from lithium disilicate blocks using a 3-axis milling machine (3XM), 2) milling from lithium disilicate blocks using a 5-axis milling machine (5XM), and 3) milling from resin-wax billet using a 5-axis milling machine, followed by heat-pressing the pattern into lithium disilicate (5XWP). For the control group, the wax patterns were fabricated by one lab technician, and the crowns were fabricated by heat-pressing the pattern into lithium disilicate (CWP). After sintering, the crowns were secured on their associated preparations using an elastomeric material. The marginal gap of each crown was then measured at 14 defined locations through analyses of 20× images captured with a stereomicroscope. The marginal integrity and resistance to rotation of each crown were assessed by 2 calibrated practitioners. Differences in outcomes by fabrication technique were assessed using Wilcoxon, Kruskal Wallis, and Fisher’s exact tests, as appropriate (α = 0.05).

Results:

Crowns fabricated using digital workflows (3XM, 5XM, 5XWP) had significantly smaller mean marginal gaps compared to the CWP group (p = 0.0001, p = 0.0002, p = 0.0001, respectively); however, 3XM group was the only group to exhibit significantly better marginal integrity than the CWP group (p = 0.0004). No significant difference (p = 0.6004) in the resistance to rotation of crowns was observed between groups.

Conclusions:

Choice of fabrication technique and instrument may impact the marginal discrepancy of lithium disilicate crowns; however, all fabrication techniques analyzed produced crowns with acceptable marginal discrepancies.

Computer-aided design and computer-aided manufacturing (CAD/CAM) technologies have advanced since they were first introduced for the fabrication of indirect restorations in the 1970s.¹ Despite these advances, the 3 functional components of CAD/CAM systems generally remain the same: an optical scanner, design software, and manufacturing hardware.

The impact of the choice of optical scanner on the accuracy of CAD/CAM crowns has been evaluated in numerous studies.^2–6 Use of different intraoral or lab-side scanners to digitize preparations for fabrication of zirconia and lithium disilicate crowns has been shown to result in a wide range of mean marginal gaps, varying between 14 and 207.8 μm.^2–6 However, the effects of the internal settings of the design software on the marginal adaptation are not well documented.

Few studies have reported the internal settings used in the design software and its impact on the marginal gap. Each design software has one or more internal setting parameters; however, the paramters may be defined with different terminologies. Neves et al reported that the marginal gap for lithium disilicate crowns fabricated using one commercial system was 66.9 μm with application of a die spacer of 100 μm; however, the margin ramp setting was not reported.⁷ A similar marginal gap (62.3 μm) was reported by Renne et al for the same workflow with a die spacer of 100 μm and a marginal ramp of 250 μm.⁸ In contrast, Sadid-Zadeh et al reported a larger marginal gap with the selection of a 50 μm die spacer and 0.8 mm marginal ramp.³ Only one study has compared the effects of internal settings on the marginal gap of CAD/CAM crowns.⁹ Kale et al used a digital design software to design zirconia crowns with a 25 μm cement space at the margin and a die spacer of 30, 40, or 50 μm. Results from this study showed that the crowns designed with the thickest die spacer had the smallest marginal gap.⁹

The utility of CAD/CAM-based processes has been supported by recent studies showing that lithium disilicate crowns fabricated using digital workflows have smaller marginal gaps than those fabricated using a heat-pressed technique.^10,11 However, there is still little known about impact of the specific aspects of the milling process, such as the number of axes in the milling machine, or the mode of milling used for fabrication, on the marginal gap and resistance of crowns. Hamza and Sherif compared the effects of 4- and 5-axis milling machines on the marginal discrepancy of monolithic zirconia crowns. Results from this study demonstrated that crowns fabricated using 5-axis milling machines had significantly smaller marginal gaps.¹² In addition, Sadid-Zadeh et al reported a significantly smaller marginal gap for lithium disilicate crowns milled in the detailed mode of milling, compared to the standard mode of milling.³ However, the effect of the milling machines with different numbers of axes on the marginal gap of lithium disilicate crowns remains to be explored.

Much like the different restoration materials used in dentistry, the accuracy of the marginal gap has been recognized as an important factor for defining the success of lithium disilicate crowns.^13,14 American Dental Association (ADA) specification No.8 recommends a film thickenss of 25 to 40 μm for luting agents, suggesting a minimum cement thickness at the margin of the cemented crowns.¹⁵ However, clinical evaluation of complete coverage crowns suggests that typical marginal gaps considered clinically acceptable for conventionally cemented crowns range between 120 and 150 μm.^16,17 These wider gaps may provide sufficient space for bacterial growth, leading to biological and mechanical problems.^18,19 As a result, the marginal gaps between 25 and 40 μm should be considered best practice and a clinical goal for dental practitioners.¹¹ In addition to the marginal gap, the internal fit of all ceramic crowns is a clinically important criteria that requires evaluation. Studies have shown that the greater the internal gap of an all-ceramic crown, the higher the fracture load.^20,21 Morover, the increased internal gap may impact the resistance of the of all-ceramic crowns.^22,23

This study evaluated the marginal gap and marginal integrity of lithium disilicate crowns manufactured using different milling techniques and compared the resulting values to those of conventional, heat-pressed crowns. The marginal gap was defined as the perpendicular measurement from the margin of the crown to the closest part of the finish line.¹⁶ Marginal integrity was defined as the tactile-visual evaluation of the marginal gap over 360° of the crown using an explorer. In addition, this study assessed the resistance of crowns by a simple “yes” or “no” answer to the question: “Does the crown resist the rotation?”²⁴ The null hypothesis was that the fabrication technique does not impact the marginal gap, marginal integrity, and resistance of lithium disilicate crowns.

MATERIALS AND METHODS

Specimen preparation

The number of specimens per group was defined based on a power analysis performed in a previous study.³ A single operator prepared 12 ivorine molar teeth (Kilgore International Inc.; Coldwater, MI), applying a 2 mm occlusal reduction following the occlusal anatomy of the tooth with a 1 mm modified shoulder finish line. The preparations were smoothed with a finishing bur, and the angles were rounded. Each preparation was then mounted on a typodont (200 Series; Kilgore International Inc.) and surface scanned 3 times using an intraoral scanner (Trios 3; 3Shape A/S; Copenhagen, Denmark). An unprepared tooth was scanned along with each preparation to represent the pre-preparation scan. The scanning procedures were standardized and performed following the manufacturer’s instructions. Crowns were designed for each preparation (3Shape Dental System; 3ShapeA/S) following the anatomy of the pre-preparation scan. The following internal settings were applied for the design of the crown: (1) a die spacer (defined as “Extra Cement Gap” in the software) of 50 μm, (2) a distance to margin (defined as “Smooth Distance”) of 0.2 mm, (3) a marginal gap (defined as “Cement Gap” in the software) of 25 μm, and (4) a finish line width (defined as “Distance to Margin Line”) of 0.8 mm. The outcome of the abovementioned procedures was 3 sets of the designed crowns, standard tessellation language (STL) files, for each preparation.

Two sets of STL files were used to manufacture lithium disilicate crowns in 2 different techniques. The first 2 groups of crowns were fully fabricated using digital workflows. Crowns in the 3XM group were milled from lithium disilicate (IPS e.max CAD; Ivoclar Vivadent, Amherst, NY) using a 3-axis milling machine (PlanMill 40; Planmeca, Helsinki, Finland) in the detailed mode. In this milling mode, the PlanMill 40 uses a diamond rotary instrument (Two Striper, Premier Dental; Plymouth, PA) 1 mm in diameter to mill the intaglio surface of the crown. The second set of designs (5XM) were milled from lithium disilicate (IPS e.max CAD) using a 5-axis milling machine (Zenotech Select Hybrid; Ivoclar Vivadent) in the “e.max 3–2– crown 5× cavity” milling mode. The Zenotech Select Hybrid uses 3 diamond rotary instruments (Wieland Dental + Technik GmbH & Co. KG, Pforzheim, Germany) with 3, 2, and 1 mm diameters in sequential order to mill the intaglio surface of the crown in this milling mode. After fabrication of the crowns, the crowns were inserted into their assigned preparations, and over-contoured margins were modified by trying-in the crowns. No modifications were made to the intaglio surfaces of the crowns. The crowns were then crystallized using a sintering oven (Programat CS2; Ivoclar Vivadent) following the manufacturer’s instructions.

The third set of STL files (5XWP) were used to mill wax patterns using a 5-axis milling machine (DG Shape DXW-52DC, Roland DGA Corporation; Irvine, CA) from a resin-wax billet (ArgenWax; Argen Corp., San Diego, CA). The manufacturing was performed in the “cavity finishing 3–2” with “additional internal preparation detail” milling mode. The DG Shape DXW-52DC uses three carbide rotary instruments (Roland DGA Corp., Irvine, CA) with diameters of 2, 1, and 0.6 in sequential order to mill the intaglio surface of the crowns in this milling mode. After milling, the wax pattern was used to fabricate lithium disilicate (IPS e.max press; Ivoclar Vivadent) crowns with heat-pressed technique following the manufacturer’s instructions.

For the control group, lithium disilicate crowns were fabricated using conventional heat-pressed technique (CWP). Vinyl polysiloxane (VPS) adhesive (Ivoclar Vivadent) was applied to stock trays (Dentsply Sirona, York, PA) and allowed to dry for a period of 3 minutes. Then, VPS impression material (Heavy Body and Extra Light Body Virtual XD; Ivoclar Vivadent) was used to obtain 12 impressions from 12 preparations following the manufacturer’s instructions while the preparations were placed in a mandibular typodont (200 Series). An impression was recorded from the mandibular typodont with an unprepared molar using a stock impression tray (Dentsply Sirona) and irreversible hydrocolloid material (Jeltrate; Dentsply Sirona) according to the manufacturer’s recommendations. Each impression was then poured in Type 5 dental stone (Jade Stone; Whip Mix Corp., Louisville, KY) according to the manufacturer’s recommendations. A single technician fabricated wax patterns following the anatomy of an unprepared tooth, then the crowns were fabricated from lithium disilicate ingots (IPS e.max Press) using the heat-pressed technique according to the manufacturer’s instructions. After fabrication, the crowns were inserted into their assigned preparations, and over-contoured margins were modified as needed; however, no modifications were made to the intaglio surfaces of the crowns in any of the heat-pressed groups.

Evaluation of restorations

The marginal gap was defined as the perpendicular measurement from the margin of the crown to the closest part of the finish line.¹⁶ Crowns were secured on their respective preparations using an elastomeric material (Fit Checker Blue; GC America, Alsip, IL) with application of finger pressure for 1 minute.²⁵ They were then placed under a constant weight of 200 g for 1 minute at room temperature. A stereo optical microscope (SMZ-U; Nikon Instruments Inc., Tokyo, Japan) was used to capture images (20×) of the buccal, lingual, mesial, and distal sides of each preparation and its associated crown. To standardize the imaging, 4 jigs were fabricated for the buccal, lingual, mesial, and distal sides of the unprepared tooth using VPS (Lab Putty; Coltene/Whaledent Inc., Cuyahoga Falls, OH) as described previously.⁸ ImageJ software (NIH, Bethesda, MD) was used to measure the marginal gap of each crown at 4 points on each of the buccal and lingual surfaces and 3 points on each of the mesial and distal surfaces, yielding a total of 14 data points per crown (Fig 1). The average marginal gap across 14 points measured on each crown was defined as the marginal gap (μm).

Fig. 1.

Image captured by optical microscope (left); measurement made at margin (right).

Marginal integrity and resistance of the crowns were also evaluated while they were placed on their associated preparation. Marginal integrity (Table 1) was defined as the tactile-visual evaluation of the marginal gap over 360° of the crown using an explorer. Crown resistnce (Table 1) was defined as the resistance to rotation around the x- or y-axis.³ Each crown was randomly assigned a letter of A, B, C, or D. Two calibrated and blinded clinicians evaluated the resistance to rotation and marginal integrity of crowns using 3× magnification glasses (Orascoptic; Middleton, WI) and a new explorer (Hu-Friedy Mfg. Co. LLC; Chicago, IL). The clinicians also scored the crowns from 1 to 4, with 1 representing the best clinical outcome, and 4 representing the worst. When there was disagreement, the evaluators discussed the scores until agreement was reached.

Table 1.

Definitions of clinically evaluated parameters

Resistance of the restoration	1. Clinically acceptable: No rotation around the x- or y-axis
	2. Not clinically acceptable: Rotation around the x- and/or y-axes
Marginal integrity	1. Excellent: Explorer moves smoothly during occluso-gingival movement at the margin
	2. Clinically acceptable: Explorer catches in a limited location during occluso-gingival movement, but marginal integrity is clinically acceptable
	3. Not clinically acceptable: Marginal integrity is not clinically acceptable
Clinical ranking	Rank the crowns for each preparation from 1 to 4, with 1 representing the best choice.
	1. ______2. ______ 3. ______4. ______

Statistical analysis

All statistical analyses were performed using SAS (v9.4; SAS Institute Inc., Cary, NC) with at a significance level of 0.05 (α = 0.05). The Kruskal-Wallis test was used to assess the significance of differences in the marginal gap between fabrication techniques. Pairwise tests of the least square means between each of the techniques were performed using the Tukey-Kramer adjustment for multiple comparisons. The Kruskal-Wallis test was also used to analyze the marginal integrity data, followed by Wilcoxon pairwise tests with Bonferroni adjustment for multiple comparisons to compare specific differences between fabrication techniques.

A Fisher’s exact test was used on the categorial data gathered from the assessment of crown resistance, followed by application of a logistic regression model that incorporated pairwise tests of the least square means to assess differences in the resistance of crowns fabricated using different techniques. P-values were adjusted via the Bonferroni correction for multiple comparisons of pairwise tests. The highest-ranked workflow for each preparation was reported descriptively.

RESULTS

The mean and standard deviation of the marginal gap for each fabrication technique are listed in Table 2. Application of the Kruskal-Wallis test revealed a significant overall difference across the 4 fabrication techniques, and pairwise comparisons demonstrated significant differences (3XM vs 5XM, p = 0.0003; 3XM vs 5XMP, p = 0.0175; 3XM vs CWP, p = 0.0001; 5XM vs CWP, p = 0.0002; 5XMP vs CWP, p = 0.0001) between all pairs except the 5XM and 5XWP pair (p = 0.6). The marginal gap of the CWP group was significantly larger than that of any digital design group.

Table 2.

Marginal gap (mean ± standard deviation [SD]) for complete coverage crowns fabricated using each digital workflow

Marginal Gap	3XM	5XM	5XWP	CWP
Mean ± SD (μm)	77.2 ± 17a	94.1 ± 32b	88.9 ± 20b	113.3 ± 26c

Means labeled with different letters significantly differ from one another (P < 0.05).

Table 3 shows the distribution of scores for the marginal integrity and resistance to rotation of crowns by fabrication technique. Analysis of the marginal integrity data using the Kruskal-Wallis test revealed a significant overall difference (P = 0.0001) across the fabrication groups, and pairwise comparisons showed that the only significant difference was between the 3XM and CWP groups (P = 0.0004). Neither the Fisher’s exact test for overall differences (P = 0.6) nor the pairwise comparisons showed any significant differences in the resistance of crowns for the different fabrication techniques. Clinical rankings from 2 blinded clinicians revealed that lithium disilicate crowns from the 3XM group were always selected as the first choice, and crowns from the CWP group were ranked as the last choice 67% of the time (Fig 2).

Table 3.

Contingency table summarizing the marginal integrity and resistance distributions of crowns from each group

		3XM	5XM	5XWP	CWP
Marginal integrity	Excellent	12	3	4	0
	Clinically acceptable	0	9	7	12
	Not clinically acceptable	0	0	1	0
Resistance to rotation	Clinically acceptable	12	12	11	10
	Not clinically acceptable	0	0	1	2

Fig. 2.

Clinical evaluation and ranking of crowns by fabrication technique. Restorations were ranked from 1 to 4. A score of 1 represents the best rank.

DISCUSSION

The aim of this study was to evaluate the effect of different fabrication techniques on the marginal gap and the resistance of crown. An elastomeric material, selected for its low film thickness, was used to secure the crowns on their associated prepared teeth for evaluation.¹⁶ This method allows for use of a single preparation to evaluate the marginal fit of crowns fabricated using different techniques. An advantage of this technique over the replica technique is that the measurement of the marginal gap, through a combination of microscopy and the putty-ruler assembly, is more reproducible due to consistency of location and improved accuracy in the vertical direction; however, this technique is limited by the fact that, as in the replica technique, the marginal gap is only measured at preset points, potentially missing marginal discrepancies in other areas. To compensate for this limitation, the marginal integrity of the crowns was also evaluated across 360° of the crown margin by 2 evaluators. As a result, any undetected unacceptable marginal gaps not recorded by digital photography were likely to be noted in this evaluation by practitioners.³

The null hypothesis of this study, that fabrication technique does not impact the marginal gap of lithium disilicate crowns, was rejected. Crowns fabricated using digital workflows, 3XM, 5XM, and 5XWP, were shown to have significantly smaller mean marginal gaps (77.2–94.1 μm) than those fabricated using conventional technique (CWP; 113.3 μm). The results of this study align with a recent study by Ng et al, who found that lithium disilicate crowns manufactured by digital workflow resulted in a smaller marginal gap (60 μm) than crowns fabricated with conventional technique (74 μm).¹⁰ However, the die material, intraoral scanner, design software, and milling unit used in Ng et al’s study differed from that used in this study.¹⁰ Furthermore, crowns fabricated with conventional technique in this study had a larger marginal gap (113.3 μm) compared to the crowns fabricated with the same technique in Ng et al’s study (74 μm).¹⁰ This discrepancy could be due to differences in the level of experience of the lab technicians and/or the use of different working cast materials. In the present study, Type 5 gypsum, which has an expansion percentage of 0.18%, was used. In contrast, Ng et al employed Type 4 gypsum, which has an expansion percentage of 0.09%.¹⁰

Results from the present analysis better align with those of Abdel-Azim et al, who found marginal gaps of 89 and 112 μm for lithium disilicate crowns fabricated by digital and conventional workflows, respectively, although a different digital workflow was used.⁵ Alqahtani et al reported that the marginal gaps of lithium disilicate crowns fabricated using conventional techniques were significantly larger than those of crowns fabricated using a digital workflow similar to the 5XM workflow.²⁶ The present study also revealed that lithium disilicate crowns fabricated using conventional workflows had significantly higher marginal gaps than those fabricated using digital workflows. The digital workflow (similar to the 5XM group from this study) marginal gap reported by Alqahtani et al was 60 μm, which was smaller than that of the data from this study (94.1 μm). This difference may be because Alqahtani et al scanned the working cast, as opposed to a direct scan of the preparation in the present study. In addition, Alqahtani et al²⁶ did not report the internal settings used for crown design or the milling strategy, so changes in the internal settings⁹ or milling strategy³ may have contributed to the difference in marginal gaps. Moreover, Alqahtani et al evaluated the marginal gaps at 4 points, rather than the 14 points evaluated in the present study, which could also contribute to the discrepancy.²⁶

In addition to marginal gap, crown resistance, defined as rotation around the x- or y-axis, was also evaluated. The data revealed that 100% of the crowns in 3XM and 5XM groups had a clinically acceptable resistance of crowns, and that 83% and 94% of the crowns in 5XWP and CWP groups, respectively, had a clinically acceptable crown resistance. There was no statistically significant difference between these values. The small sample number may have contributed to this result. The few specimens that exhibited unacceptable resistance of crowns in the 5XWP and CWP groups may have arisen from lab processing errors associated with the heat-pressing procedures. The insignificant differences in the resistance of crowns, as well as the marginal integrity of the crowns, may have contributed strongly to the selection of crowns fabricated using the 3XM and 5XM workflows as the first and second choices, respectively, for clinical rankings.

Previously, Hamza and Sherif compared the effects of 4- vs. 5-axis milling machines on the marginal discrepancy of monolithic zirconia crowns. Their results suggested that crowns fabricated using 5-axis milling machines had significantly smaller marginal gaps compared to those generated by 4-axis milling machines.¹² However, different types of zirconia crowns were fabricated using the different types of milling machines in Hamza et al’s study, so the difference in marginal discrepancy could be related to either the number of axes used for milling or the type of zirconia material.^12,27 In the present study, the same lithium disilicate material was used for crowns milled with either 3- or 5-axis milling machines. The 3-axis milling machine produced crowns with smaller marginal gaps and standard deviations (77.2 ± 17 μm) compared to the 5-axis milling machine (94.1 ± 32 μm). This difference may be because the 3-axis milling machine used in this study usess a 1 mm diameter diamond bur to cut the internal surface of the crown. In contrast, the 5-axis milling machine used in this study uses 3 mm, 2 mm, and 1 mm diameter burs in sequence to cut the intaglio surface of the crown, which may contribute to the discrepancy at the margin. This difference in burs may also underlie the observation that the 5XM and 5XWP groups did not significantly differ from one another, as multiple diameters of burs were used to cut the intaglio surfaces of the crowns in both groups.

A limitation of this study is that a 4-axis milling machine was not used to mill lithium disilicate crowns for comparison. Similarly, milling machines from different manufacturers and with different bur sizes were not explored. The data presented in this study solely provide information about the 3- and 5-axis milling machines used in this study, and it cannot be exteneded to the milling machines from different manufacuturers. Therefore, future studies should evaluate the marginal gaps of lithium disilicate crowns produced using a wider variety of milling machines and procedures in order to devise an optimal strategy that ensures that marginal gaps fall within an acceptable range.

CONCLUSIONS

1. The 3XM fabrication technique produced the smallest marginal gap (77.2 μm) of all of the workflows examined.

2. All of the fabrication techniques produced a crown with an acceptable marginal gap, marginal integrity, and resistance to rotation.