Color Stability of Polymer-Based Composite CAD/CAM Blocks: A Systematic Review

Background: This systematic review aims to evaluate the color stability of resin composite CAD/CAM blocks (CCB) when submitted to staining solutions. Methods: A systematic search was performed on different databases (Embase, Medline, Scopus, Web of Science). Search terms were: ‘polymer infiltrated’, ‘polymer-based’, ‘resin nanoceramic’, ‘resin ceramic’, ‘hybrid composite’, ‘hybrid ceramic’, ‘composite ceramic’, ‘resin infiltrated’, ‘CAD-CAM’, ‘CAD/CAM’, ‘color stability’, ‘staining’, ‘staining susceptibility’, ‘color change’, ‘color difference’. Inclusion criteria: in vitro articles published in the English language until 18 September 2022 without initial time restriction evaluating the color stability of CCB when submitted to staining solutions. Exclusion criteria: studies investigating color change induced by smoke or whitening treatments; studies not including a clinical evaluation of the results using the thresholds for color perceptibility and acceptability. Risk of bias assessment using the QUIN tool. Findings: Out of the 378 initially retrieved articles, 19 were included in this review. They investigated 17 different CCBs and different artificial staining by liquid protocols, including coffee, red wine, tea, and cola. CCBs exceeded clinical acceptability thresholds for color shift in 18 out of 19 studies, with a significantly higher color stability than conventional hybrid resin-based composites (RBCs), and a significantly lower color stability than ceramic materials. The identified differences in CCBs in color stability can be attributed to the material’s composition, but also to the heterogeneity of staining procedures. Interpretation and clinical implication: Clinicians should be aware that, although to a lower degree when compared to RBCs used in direct or indirect procedures, CCBs undergo color changes to a higher degree in comparison to ceramic materials.


Introduction
In less than 40 years, CAD/CAM technology has experienced constant hardware and software improvements that have resulted in easier use and more reliable clinical performance [1][2][3]. Along with progress in technology, new materials have been developed for CAD/CAM restorations, such as glass ceramics, zirconia, and composites, allowing clinicians the choice of different mechanical [4,5] and optical [6] properties. Although the use of direct resin-based composite (RBC) restorations is largely diffused and able to provide reliable and esthetic results both for anterior [7,8] and posterior [9] direct restorations, some drawbacks have been reported, such as weak mechanical properties [10] and lack of color stability.
When compared with RBCs used in direct restorations, indirect ones are characterized by higher mechanical properties and color stability, mainly due to the higher degree of 1. In vitro studies investigating color stability of composite CAD/CAM blocks; 2.
In vitro studies including artificial staining procedures by liquids; 3.
Studies using color difference clinical thresholds to analyze the color difference values; 4.
Publications in English language; Exclusion Criteria

5.
In vitro studies with a sample size of less than five test specimens in each subgroup; 6. In vitro studies investigating color stability of hybrid dental ceramic CAD/CAM blocks (polymer-infiltrated ceramic networks); 7.
In vitro studies investigating color stability of CAD/CAM materials for temporary restorations; 8.
Clinical trials, case reports, reviews, or animal studies; 9.
Papers analyzing color stability only with water aging/thermocycling procedures; 10. Papers analyzing color stability with whitening procedures; 11. Papers analyzing color stability with mouth rinses; 12. Papers analyzing color stability with smoking procedures;

Information Sources
One reviewer (GP) conducted a search for English language articles published in dental journals until 18 September 2022 in the following electronic databases: PubMed, Embase, Web of Science, Scopus. A manual search was also conducted.

Selection Process
For the selection of studies, two authors (G.P. and M.M.) independently reviewed titles and abstracts of the studies according to the inclusion criteria. Final inclusion of studies was based on screening and assessing full texts, and with consensus of the authors of the current review.

Data Items
An extraction form was used to collect retrieved data items: type of CCB, comparison with other materials, staining liquids, staining protocol, time of color assessment, type of spectrophotometer, color difference formula, specimens finished, specimen repolished, clinical thresholds, outcomes.

Study Risk of Bias Assessment
The risk of bias assessment used the QUIN tool (risk-of-bias tool for assessing in vitro studies conducted in dentistry) [34]. The study's quality assessment was conducted according to a fixed set of domains of bias (Clearly stated aims/objectives; Detailed explanation of sample size calculation; Detailed explanation of sampling technique; Details of comparison group; Detailed explanation of methodology; Operator details; Randomization; Method of measurement of outcome; Outcome assessor details; Blinding Statistical analysis; Presentation of results). QUIN final assessment was performed by categorizing each of the study features at 'low', 'medium', or 'high' risk of bias. Both reviewers (G.P. and F.D.P.) independently conducted the assessment, and any uncertainties or disagreements were then resolved by discussion.

Study Selection and Study Characteristics
The study selection process according to the PRISMA checklist is reported in Figure 1.
tion; Method of measurement of outcome; Outcome assessor details; Blinding Statistical analysis; Presentation of results). QUIN final assessment was performed by categorizing each of the study features at 'low', 'medium', or 'high' risk of bias. Both reviewers (G.P. and F.D.P.) independently conducted the assessment, and any uncertainties or disagreements were then resolved by discussion.

Study Selection and Study Characteristics
The study selection process according to the PRISMA checklist is reported in Figure  1. A total of 378 studies were identified through the initial database search. Following duplicates removal, 285 records were screened by title and abstract. During the screening process, 252 records were excluded as not relevant to the subject, and 33 were selected for full-text assessment. Finally, 19 studies were included in this systematic review as they met the inclusion criteria. A total of 378 studies were identified through the initial database search. Following duplicates removal, 285 records were screened by title and abstract. During the screening process, 252 records were excluded as not relevant to the subject, and 33 were selected for full-text assessment. Finally, 19 studies were included in this systematic review as they met the inclusion criteria.
Data from the retrieved papers were chronologically reported in two predefined data extraction forms (Tables 3 and 4).
The color change of all ceramic specimens was within the clinically acceptable range, except for the glazed Mazic Duro ceramic specimens immersed in carrot juice. However, the color difference of Vita Enamic and Mazic Duro was higher than that of feldspathic porcelain.

Artificial Staining Procedures
Artificial staining procedures were different among the retrieved studies. The staining solutions used were coffee (n = 15), red wine (n = 9), cola (n = 7), tea (n = 5), ginger (n = 2), and juice (n = 2). Specimens were immersed in the staining media for different time periods that ranged from 2 days to 12 weeks. Most of the studies renewed liquids at different intervals. The immersion was static and ran from some minutes per day to continuous immersion for the entire staining periods. Most of the studies kept the liquid temperature stable at 37 • C (n = 13). For other studies, the liquids were kept at room temperature (n = 2) or thermocycled (n = 2). In two studies, information on storage temperature was not reported.

Color Assessment
Seventeen studies out of the nineteen examined used a spectrophotometer to assess color change, while two used a spectroradiometer. Among the spectrophotometers, the most used (n = 8) was EasyShade (Vita Zahnfabrik, Bad Säckingen, Germany). In ten papers, the CIEDE2000 color difference formula was used, while in seven the CIELAB formula was used. In only two studies were both formulas used. Seventeen papers compared the color stability of CCB with other materials, while two compared different CCBs.

Surface Treatment
Sixteen of the retrieved studies performed surface finishing and polishing before staining procedures. These procedures were fairly uniform among the studies because fourteen out of sixteen used silicon carbide abrasive papers (and twelve with 1200 grit as a final step) and two used abrasive disks. Only one paper performed a repolishing step after the staining procedure.

Discussion
In recent years, the request for esthetic dental restorations has considerably increased. Today, RBCs are the most widely used materials for direct and for indirect restorative procedures because of their excellent esthetic and mechanical properties [39,40].
CCBs are claimed to provide better mechanical and optical properties than their traditional direct and indirect resin counterparts thanks to the benefits of the industrial production processes [12]. Amid other advantages, they are claimed to reduce one of the primary drawbacks of direct and indirect RBCs, which is color stability. This can compromise the esthetic outcomes of the restorations over time [41]. Despite their increasing use, very little is known on the color changes of CCBs. Therefore, the purpose of this review was to evaluate the color stability of CCBs.

Type of Material
Regarding the examined materials, Lava Ultimate was the most investigated, followed by Cerasmart. Acar et al. [35] reported that Lava Ultimate, after 5,000 thermocycles in coffee, showed color change values higher than the clinical acceptability threshold when compared to lithium disilicate and polymer-infiltrated ceramic network (Enamic). This result, despite the paper presenting a high risk of bias, could be related to the composition of the material: Lava Ultimate consists of a hydrophobic urethane-dimethacrylate (UDMA) and a hydrophilic triethylene-glycol-dimethacrylate (TEGDMA). TEGDMA is generally added to the composition of RBCs because it is more viscous than bisphenol-glycidylmethacrylate (Bis-GMA) and permits copolymerization, diluting Bis-GMA and increasing composite sculptability. TEGDMA, however, increases the hydrophilicity of the composite, resulting in an increased susceptibility to staining [42,43].
Lava Ultimate contains Bis-GMA and its ethoxylated form (Bis-EMA). Dental materials containing Bis-GMA show the highest degree of water sorption and, therefore, liquid dyes, when compared with those based on UDMA, TEGDMA, and BisEMA [44].
Al Amri et al. [36], Eldwakhly et al. [15], and Schürmann and Olms [17] confirmed Lava Ultimate's lower color stability when compared to ceramic materials. Furthermore, these papers show low or medium risk of bias; therefore, their findings could be considered reliable.
Three papers included in this review compared Lava Ultimate with Cerasmart. Two of them showed significant higher color stability for Cerasmart [14,19], while the third, with a lower risk of bias, reported no significant differences [36]. The higher Cerasmart color stability may be related to the absence of Bis-GMA in its composition, thus confirming that this monomer is responsible for water uptake and, therefore, possibly for discoloration.

Spectrophotometric Analysis and Clinical Thresholds
Spectrophotometric analysis allows for an objective color comparison. Color coordinates are measured, and differences are compared by CIELAB or CIEDE2000 formulas, which are the most frequently used to analyze color changes [45]. The Perceptibility Threshold (PT) refers to the magnitude of color difference that is visually detectable by the human eye, while the Acceptability Threshold (AT) corresponds to the magnitude of color difference that is considered clinically not acceptable [46]. The CIEDE2000 color difference formula [47] is considered to be a better indicator of human capability to detect perceptible and acceptable color differences, and was used by the majority of the papers (10 out of 19) included in this review. Differences detected by spectrophotometers can be evaluated from a statistical point of view or by utilizing the PT or AT. The statistical outcome of color measurements should always be integrated with PT and AT to validate clinical consistency. For this reason, papers that did not take into consideration clinical thresholds [46] for the interpretation of the results were not included in the current review. The most frequently used instrument used for color measurements was the spectrophotometer. Among the spectrophotometers, Easyshade (Vita Zahnfabrik) was the most widely used. This type of spectrophotometer is designed to be a clinical device (working in "tooth mode") and it is not recommended for in vitro testing. The results of studies performed with such a device should be cautiously evaluated, and a bench-top spectrophotometer should be preferred [48].

Staining Solution
The level of CCB color change, as with conventional RBCs, is closely related to the type of staining solution [49,50]. Depending on the staining liquid used to evaluate color stability, colorants can deposit either on the surface [51] or in the structure [52] of the tested material, or the liquid can induce staining, altering the surface because of low pH [49,50]. Probably due to its increased use among the population, coffee is the most investigated staining media in the papers evaluated in this review. Coffee induces staining through a yellow coloring pigment that is characterized by different polarities [30,53]. Red wine, the second most used solution in this review, has staining capability and contains alcohol that may lead to rough surfaces and, consequently, pigment adsorption [49,50,54]. It has also been reported that ethanol, contained in red wine, has a solvent effect on the monomers, increasing potential discoloration [55,56]. Other dyes investigated in the paper analyzed in the current review, such as tea, cola, energy drink, and juices, are responsible for CCB color changes, but to a lower degree compared to coffee and wine, generally below AT.

Effects of Surface Treatment on Discoloration
Some of the included studies investigated the effects of surface treatments on color stability of CCBs submitted to staining with colored dyes. Unlike conventional hybrid composites used in direct procedures, CCBs are not characterized by a surface-resin-rich layer, which could be responsible for higher color changes if not removed by finishing and polishing procedures [41]. However, it has been reported that after milling, finishing and polishing CCBs may reduce surface roughness and, therefore, staining [57]. Aydin et al. [23] reported that polishing a specimen's surface produced a lower color change when compared with a control group (unpolished). However, both groups exceeded the AT, confirming that clinical thresholds should always be referenced for a correct data interpretation. Sagsoz et al. [20] also confirmed that polishing is crucial for all materials, as differences were observed between polished and unpolished specimens. Moreover, the authors reported significant differences in color stability when different finishing and polishing systems were used, suggesting that each material requires a specific finishing/polishing system for optimal performance.

Conclusions
Based on the findings of this systematic review, and considering the risk of bias, the following conclusion can be drawn:

1.
Resin-based blocks for CAD/CAM procedures show higher color stability than direct or indirect (laboratory) RBCs; 2.
Resin-based blocks for CAD/CAM procedures show lower color stability than ceramic materials; 3.
The color stability of CCBs mainly depends on material composition and staining media, but finishing/polishing procedures also have an influence.