6.3.1 Tooth Preparation

CAD/CAM and conventionally fabricated restorations share similar principles of tooth preparation. However, a number of adaptations are needed to accommodate the CAD/CAM system. Two recent publications assessed the quality of teeth prepared for zirconia FDP; they found that fewer than 7% of preparations fulfilled total occlusal convergence, undercuts, amount of occlusal reduction, axial height, and finish line design (Winkelmeyer et al. 2016, Güth et al. 2013b). However, a number of key principles required for CAD/CAM restorations – such as the presence or absence of rounded line angles, occlusal anatomy, and a smooth finish line – were not assessed in these studies. Recent data from 392 teeth prepared for monolithic zirconia crowns found that the majority of preparations had one or more errors that violated the principles of tooth preparation, and over 50% had an error specific to milled restorations, such as a rough, irregular, and stepped finish line and/or a non‐anatomical occlusal surface (Sadid‐Zadeh et al. 2021a) (Figure 6.3).

Finish line design has an impact on the marginal gap of CAD/CAM crowns; studies report an improved marginal fit when a modified shoulder is used. However, the marginal gap is not significantly impacted by the use of a chamfer finish line (Subasi et al. 2012, Euán et al. 2012, Re et al. 2014, Euán et al. 2014) or does a knife‐edge finish line (Cetik et al. 2017). Depending on the materials used, the finish line design may also impact the strength of CAD/CAM crowns. For example, CAD/CAM monolithic lithium disilicate crowns with a modified shoulder have a greater fracture strength than those with a chamfer finish line design. In addition to finish line design, the quality of the finish line is crucial to the marginal integrity of CAD/CAM crowns. A smooth and defined finish line produces a restoration with a smaller marginal gap than a rough finish line (Renne et al. 2012).

Research results regarding the impact of finish line curvature on the marginal fit of CAD/CAM crowns are conflicting; however, the results favor finish lines with less curvature, which leads to a crown with a smaller marginal gap when compared with more curvature of the finish line. Studies assessing the impact of finish line curvature on the marginal fit of crowns were performed using two CAD/CAM systems, and the results may not be assumed to apply to current CAD/CAM systems (Tao and Han 2009, Asavapanumas and Leevailoj 2013). It is crucial to assess the impact of finish line curvature on the marginal gap of CAD/CAM crowns fabricated with recently developed CAD/CAM systems.

The presence of occlusal anatomy has a significant effect on the resistance form of restorations. Studies have shown that CAD/CAM crowns fabricated on preparations with anatomical occlusal form have a significantly smaller marginal gap than those fabricated on preparations with non‐anatomical forms (Habib et al. 2014, Sadid‐Zadeh et al. 2020b) (Table 6.1). The amount of occlusal reduction necessary for CAD/CAM restorations though is similar to conventionally fabricated restorations and depends on the material being used. The material’s strength, opacity, and nature determine the amount of occlusal and axial reduction required, with bilayered restorations requiring more tooth reduction than monolithic ones. However, a well‐rounded occlusal line angle is required to compensate diameter of burs cutting the intaglio surface of the restorations. Otherwise, the restoration is over‐ or under‐milled impaction internal or marginal adaptation of the restoration (Figure 6.4).

6.3.2 Optical Scanners

For the digital workflow in fixed prosthodontics, prepared teeth are either directly recorded with an intraoral optical scanner (IOS) (see Chapter 2) or their replica, the working casts and dies, are recorded using a desktop optical scanner (DOS) (Figure 6.2). A number of studies have confirmed the accuracy of IOSs and DOSs, supporting their use in fixed prosthodontics. Research has evaluated the trueness and precision of different IOSs, revealing that their accuracy is up to par with conventional impressions for recording short‐span areas (Nedelcu et al. 2018, Ren et al. 2020); however, DOSs have better accuracy compared with IOSs for complete arch scans. The accuracy of virtual models captured by IOSs is impacted by factors such as operator experience, software version, scanning protocol, type of substrate, and the length of the arch being scanned (Ender and Mehl 2013a, 2013b, Müller et al. 2016, Medina‐Sotomayor et al. 2018, Dutton et al. 2020, Resende et al. 2021). However, the accuracy of virtual models produced by DOSs is influenced by camera resolution, the number of cameras, software version, and the accuracy of conventional impressions and the working casts and dies. For additional information on the accuracy of optical scanners, please refer to Chapter 2.

The quality of images produced by optical scanners may be tied to that of CAD/CAM restorations (Patzelt et al. 2014). However, it should not be generally assumed that CAD/CAM restorations fabricated using DOSs have better marginal and internal fit than those produced by using IOSs, or vice‐versa. As a result, even if studies have concluded that CAD/CAM single crowns and three‐unit partial FDP have a clinically acceptable marginal and internal fit (Abdel‐Azim et al. 2015, Pedroche et al. 2016, Su and Sun 2016, Shembesh et al. 2017, Arezoobakhsh et al. 2020), the results should be interpreted for the specific IOS, and DOS used, and the outcomes should not be extended to all the systems. Figure 6.5 shows examples of virtual models produced by the TRIOS (TRIOS3; 3ShapeA/S) and Planscan (Planmeca USA) intraoral scanners. CAD/CAM lithium disilicate crowns fabricated using the TRIOS scanner demonstrated better marginal and internal fit than those of the Planscan system (Sadid‐Zadeh et al. 2020a). However, this data cannot be applied to newer versions of IOSs from the same manufacturers. Providers must assess and compare clinical outcomes for specific IOSs to produce restorations with the best marginal and internal fit.

6.3.3 Computer‐Aided Design

Appropriate parameters in the computer‐aided design (CAD) software must be applied to the virtual restoration to ensure the optimum marginal and internal fit of the produced CAD/CAM crown. Despite the importance of these parameters, only a few studies assessing the marginal and internal fit of CAD/CAM restorations have reported details of the parameters used. A practitioner must learn the parameters defined in the CAD software used in their practice to ensure high‐quality work is produced for patient care. Defined parameters differ in varying CAD software. A simplified example of internal parameters can be found in PlanDesign software (Planmeca USA), wherein the “spacer thickness” and “marginal ramp” by default are set to 100 and 250 μm, respectively (Figure 6.6). In PlanDesign software, the spacer thickness is defined as the space between the intaglio surface of the designed restoration and the virtual die; the margin ramp is defined as the length of the margin on the designed restoration in close contact with the finish line of the virtual die. When these parameters are modified from the default to fabricate restorations with an acceptable resistance form, the marginal gap can be affected. For example, Renne et al. (2012) and Neves et al. (2014) applied the default settings of internal parameters to fabricate CAD/CAM lithium disilicate crowns, and manufactured crowns with a marginal gap of 62.3 and 66.9 μm, respectively, without assessing the resistance form of crowns. However, when Sadid‐Zadeh et al. (2020a) applied stringent internal parameters (die spacer = 50 μm; margin ramp = 0.8 mm) to fabricate restorations with a clinically acceptable resistance form, the marginal gap increased to 123.7 μm.

Figure 6.7 shows another example of the internal parameters for CAD software (Dental System; 3Shape A/S). The parameters for this software are more elaborate. Here, the die spacer is defined as “cement gap,” the gap at the margin is defined as “extra cement gap,” the distance from the finish line and the base of the axial wall is defined as the “distance to margin line,” and the area above the base of the axial wall, which ensures a smooth transition between the two parts of the interface that have different offsets is referred to as “smooth distance.” When Kale et al. (2016) assessed the impact of die spacers on the marginal gap of CAD/CAM crowns, they found that applying fewer than 50 μm of “cement gap” leads to a significantly larger marginal gap. The above examples present differences in internal parameters’ definitions and their importance in producing clinically acceptable marginal gaps and resistance forms.

6.3.4 Computer‐Aided Manufacturing

The compatibility of CAD software with CAM systems and its impact on the quality of produced crowns should also be considered. For example, Sadid‐Zadeh et al. (2020) reported that CAD/CAM‐fabricated single crowns using the Trios3 IOS and Dental System Software (3Shape/AS) are not compatible with PlanMill40 (Planmeca USA) when using standard milling strategy. Their pilot study found that this workflow results in restorations with a visually significant open margin. In addition to CAD and CAM compatibility, the type of milling machine, its number of axes, and milling strategy are also essential factors impacting the quality of CAD/CAM restorations.

Milling machines use various burs and strategies to detail the external and intaglio surface of restorations (Figure 6.8). Both the milling machine and its strategy may impact the marginal gap and the internal adaptation of the restoration. For example, a 3‐axis milling machine (PlanMill40; Planmeca USA) has two strategies for milling restorations: standard and detailed. The detailed strategy uses a diamond bur with a 1 mm diameter, while the standard strategy uses a 1.6 mm bur to cut the intaglio surface of restorations. As seen in Figure 6.9, the internal adaptation of the design restoration, specifically at the cusp tip, is affected by the milling strategy. Designed restorations simulated with the detailed milling strategy have better internal adaptation than ones simulated with the standard milling strategy. Also, restorations produced by the detailed strategy of PlanMill40 have a significantly smaller marginal gap and better resistance form than those fabricated with the standard milling strategy (Sadid‐Zadeh et al. 2020a).

The number of axes in a given milling machine may also affect the marginal gap of manufactured restorations. Hamza and Sherif (2017) assessed the marginal gap of zirconia crowns fabricated using five milling machines, comparing 4‐axis and 5‐axis models, and concluded that the 5‐axis milling machine produced significantly better marginal gaps. However, zirconia billets from varying manufacturers were used in this study, which makes it difficult to conclude if the effect is attributable to the milling machine or to the difference in materials. Conversely, Sadid‐Zadeh et al. (2019) found that a 3‐axis milling machine at the detailed milling strategy (PlanMill40; Planmeca USA) produced significantly better lithium disilicate restorations than a 5‐axis milling machine in the millin strategy of “cavity finishing 3‐2” with “additional internal preparation detail” (Roland DGA Corporation; Irvine, CA). However, these data cannot be applied across milling machines. These studies do, however, inform us that different milling machines produce varying marginal gaps, and the practitioner should be aware of this impact. Clinicians must assess a wide variety of factors to produce high‐quality CAD/CAM restorations.

6.4.1 Die Materials

The quality of FDP is determined by marginal and internal fit as well as appropriate occlusal and interproximal contacts (Kuhn et al. 2015, Mously et al. 2014, Kale et al. 2016, Chochlidakis et al. 2016, Anadioti et al. 2014, McLean and von Fraunhofer 1971). Accurate impressions are necessary for replicating the prepared and neighboring teeth, which are essential to ensure a precise fit of the fabricated restoration (Persson et al. 2008, Ender and Mehl 2013a, Cho et al. 2015, Güth et al. 2013a). An inaccurate impression or replica, on the other hand, leads to a restoration that can be compromised in the aspects of internal and marginal adaptation and/or interproximal and occlusal contacts.

Conventionally, the above aspects of fabricated restorations are assessed and modified on working casts and dies fabricated from the conventional definitive impression. When a partial digital workflow (Figure 6.1) is used, CAD/CAM restorations could also be assessed and modified on conventional working casts and dies; however, no physical model exists for the assessment and modification in the complete digital workflow (Figure 6.1). Apart from internal adaptation, vertical gap, and interproximal and occlusal contacts, which should be assessed and modified for all FDP, CAD/CAM FDP has integral horizontal discrepancies requiring modification (Holmes et al. 1989). The assessment and modification of CAD/CAM restoration fabricated through a complete digital workflow could be done at the chairside by the practitioner or at the laboratory side by the laboratory technician. Chairside modifications are time‐consuming and sometimes difficult due to the presence of surrounding soft and hard tissue. On the other hand, alterations in the laboratory side require CAD/CAM working casts and dies manufactured from digital impressions. Irrespective of the fabrication method, working casts and dies must:

• Replicate the prepared tooth’s geometry and that of the surrounding soft and hard tissue
• Be free of voids and distortions
• Allow adequate access and visualization of the margin
• Have high strength and good abrasion resistance
• Demonstrate dimensional stability over time

In partial digital workflow (Figure 6.1), conventional working casts and dies are recorded using a DOS. The variety of gypsum color appears to have no impact on the accuracy of dies or the fit of CAD/CAM crowns (Kim et al. 2015, Tan et al. 2018). It should be noted that in the conventional workflow, high‐expansion gypsum is needed to compensate for the shrinkage of the cast restoration; conversely CAD/CAM restorations may benefit from low‐expansion gypsum for creating a die with dimensional geometry closer to that of the prepared tooth.

In a complete digital workflow (Figure 6.1), virtual models are used to design and fabricate restorations. Virtual models have multiple advantages over conventional working casts and dies, including the ability to digitally manipulate the model by subtracting and recapturing digital impressions to resolve isolated errors (Figure 6.11), storing the model without the need for physical space, and ease of sharing data. A virtual model can be converted into an analog working cast and die by subtractive or additive techniques if necessary. Then, CAD/CAM working casts and dies may be used to assess and modify the restoration at the margin and interproximal/occlusal contacts.

Fabrication of complex shapes such as working casts and dies with subtractive technique, also referred to as milling, is difficult, as the axes of milling machines are limited (Liu et al. 2006). Furthermore, the use of subtractive technique, also referred to as three‐dimensional (3D) printing, results in significant waste, and errors may occur due to the diameter of burs used in milling machines (Örtorp et al. 2011). In contrast, additive technique (described in Chapter 3) transforms virtual model files into slice data and builds them layer by layer to create the desired shapes; minimal material is wasted, and features such as undercuts and complex geometry are easily replicated (Liu et al. 2006). Both the additive and subtractive techniques use resin to create CAD/CAM working casts and dies and produce dies with high strength and good abrasion resistance. Although casts fabricated via the additive technique have inferior accuracy compared with conventional gypsum models; FDP fabricated from these models have clinically acceptable marginal integrity and internal fit (Jang et al. 2020). Inaccuracies in CAD/CAM working casts and dies increase as the length of the arch being scanned and printed increases. This may be due to inaccuracy in digital scans, or the shrinkage of resin materials during the subtractive technique processing (Park et al. 2020). Despite these shortcomings, when a complete digital workflow is used to fabricate a FDP, 3D printed working cast and dies are used to assess and modify internal/marginal discrepancies and interproximal/occlusal contacts (Figure 6.12). Ongoing research continues to create 3D printed working casts and dies that are as – or more – accurate than their gypsum counterparts. For more information regarding 3D printing, refer to Chapter 3.

6.4.2 Pattern Materials

Patterns are conventionally formed freehand from wax for the fabrication of cast or pressed FDP. However, wax patterns can now be manufactured using the subtractive or additive technique (Figure 6.13). The pattern materials are mainly composed of polymers and wax, and CAD/CAM patterns have a number of advantages over conventional patterns; they are strong, unaffected by variations in temperature, they are time‐ and cost‐effective, and they do not rely on the laboratory technician experience. Wax patterns can be used for the fabrication of large or complex frameworks, where small distortions in a connector can compromise the fit of the entire framework. Subtractive fabrication of patterns is limited by factors such as the complexity of the restoration; however, the additive technique can be used to manufacture patterns with complex shapes. The CAD/CAM wax pattern can later be used to fabricate FDP by cast or pressed technique.

Three‐dimensional printed and milled patterns have demonstrated clinically acceptable marginal fit in fixed prosthodontics and offer similar or superior internal and marginal fit (Lövgren et al. 2017, Homsy et al. 2018) compared with conventionally fabricated patterns. In addition, they produce crowns with similar or better marginal fit compared with conventionally produced wax patterns or milled crowns (Örtorp et al. 2011, Ghodsi et al. 2019, Sadid‐Zadeh et al. 2019, Koutsoukis et al. 2015, Revilla‐León et al. 2018). However, it is crucial to mention that the outcome of a CAD/CAM wax pattern – and, consequently, a FDP – is dependent on the quality of tooth preparation, quality of the digital impression, internal parameter settings in the CAD software, and the quality of manufacturing. In addition, the quality of the definitive restoration fabricated through this workflow depends on the laboratory technician’s skill in casting or pressing the CAD/CAM pattern.

6.4.3 Restorative Materials

Several restorative materials are developed for the fabrication of CAD/CAM FDPs to be manufactured by additive or subtractive techniques. Additively fabricated fixed restorations are limited to polymers and metal alloys (refer to Chapter 3 for materials used in additive technology). However, dental ceramics fabricated through additive manufacturing are being explored and have demonstrated encouraging primary outcomes for fabricating ceramic FDP. (Dehurtevent et al. 2017, Wang et al. 2019, Methani et al. 2020, Revilla‐León et al. 2020)

In the subtractive technique, varying sizes of billets are produced under controlled conditions to increase the material’s homogeneity and reduce microcracks and voids, which weaken the mechanical property of the material. Then, restorations are milled from billets at the chairside or in the dental laboratory (Figure 6.14). Subtractive techniques may make use of polymethyl methacrylate (PMMA), composite resin, ceramics, polyetheretherketone (PEEK), and metal alloys. In subtractive technique restorative materials are milled in either a dry or wet‐milling environment. In dry milling, materials are machined without the need for a cooling agent. Dry milling is required for materials such as partially sintered zirconium oxide. In wet milling, coolant is sprayed during manufacturing to protect the burs and material against overheating. The latter technique is necessary for manufacturing metal alloys and glass ceramics.

6.4.3.1 Polymethyl Methacrylate

Polymethyl methacrylate is a synthetic material produced from the polymerization of methyl methacrylate. Interim FDPs can be fabricated from PMMA billets by the subtractive technique. Depending on the milling machine used, PMMA FDP may be manufactured in a dry‐ or wet‐milling environment; PMMA materials used with subtractive technique possess better physical and mechanical properties than conventionally processed PMMA, including color stability, water absorption, wear, microhardness, flexural strength, and fracture resistance (Rayyan et al. 2015, Stawarczyk et al. 2012, Stawarczyk et al. 2013b, Song et al. 2020). The cytotoxicity of milled PMMA is also lower than that of conventional PMMA (Souza et al. 2020). These properties support PMMA interim restorations fabricated by subtractive technique for use as long‐term interim restorations (Figure 6.15).

Polymethyl methacrylate billets are used to manufacture interim restorations by indirect or indirect‐direct technique. Indirect techniques in the fabrication of CAD/CAM interim restorations require a digital record of prepared teeth, design of restorations on the virtual models, and manufacturing of CAD/CAM restorations. Then, the restorations are luted on the prepared teeth. However, in the indirect–direct technique, a shell is designed on the virtual models with unprepared teeth; then, the shell is relined intraorally with the conventional PMMA after tooth preparation and luted on the prepared tooth (Figure 6.16). CAD/CAM interim FDP fabricated by the indirect technique (Figure 6.15) are reported to have better marginal integrity and fracture strength than those fabricated conventionally (Peñate et al. 2015, Edelhoff et al. 2016, Reymus et al. 2020, Wimmer et al. 2014, Alt et al. 2011, Stawarczyk et al. 2012, Abdullah et al. 2016, Abdullah et al. 2018, Peng et al. 2020). However, interim partial FDPs fabricated by an indirect–direct technique using CAD/CAM and conventionally fabricated shells have not shown a significant difference in the fracture strength of FDPs (Sadid‐Zadeh et al. 2021b).