Dental and composite resin discoloration induced by different hydraulic calcium silicate-based cements: two-year in vitro assessment

Abstract Few long-term studies assess the discoloration induced by hydraulic calcium silicate-based cement on dental structures. In addition, as far as we know, no long-term study has assessed the discoloration induced by these cement on composite resin. Objective This in vitro study aimed to assess, during a period of two years, the discoloration potential of different hydraulic calcium silicate-based cements (hCSCs) on the enamel/dentin structure and composite resin restoration. Methodology A total of 40 enamel/dentin discs were obtained from bovine incisors, and 40 composite resin discs (10 mm in diameter × 2 mm thick) were fabricated. A 0.8 mm-deep cavity was made in the center of each disc and filled with the following hCSCs (n=10): Original MTA (Angelus); MTA Repair HP (Angelus); NeoMTA Plus (Avalon); and Biodentine (Septodont). An initial color measurement was performed (T0 - baseline). After 7, 15, 30, 45, 90, 300 days, and two years, new color measurements were performed to determine the color (ΔE00), lightness (ΔL’), chroma (ΔC’), hue differences (ΔH’), and whiteness index (WID). Results For enamel/dentin, the ΔE00 was significant among groups and periods (p<0.05). NeoMTA Plus had the greatest ΔE00. The NeoMTA Plus group had the greatest ΔE00 after two years for composite resin. Significant reduction in lightness was observed for all groups after two years (p<0.05). The most significant WID values were observed after 30 days for Biodentine (enamel/dentin) and MTA Repair HP groups (composite resin) (p<0.05). Conclusions The hCSCs changed the colorimetric behavior of both substrates, leading to greater darkening over time. The Bi2O3 in the Original MTA seems relevant in the short periods of color change assessment.


Introduction
Hydraulic calcium silicate-based cements (hCSCs) have been widely used in Dentistry as repair biomaterials due to their excellent mechanical, chemical, and biological properties. [1][2][3] The precursor of this class of cement was the Mineral Trioxide Aggregate (MTA), initially developed as a retro filling material and as a root perforation sealing. 1,2 However, due to its good features, its clinical application quickly expanded beyond these treatments. [1][2][3] Despite the success achieved by MTA in several clinical situations, studies have reported that this hCSC causes discoloration of dental tissues in a short period of time. 4,5 Therefore, since the introduction of MTA in the early 1990s, the cement composition has been modified to improve its physical-chemical properties and minimize the adverse effects on dental esthetics. [1][2][3][4][5] The initial version of MTA (grey MTA) had a high amount of iron oxide (Fe 2 O 3 ) in its composition, 6 which led to intense crown discoloration, especially in the cervical portion of the tooth. 4,5,7 Despite the introduction of a new MTA (white MTA) with a lower concentration of Fe 2 O 3 , it also promoted tooth discoloration. 4,5,7 Many studies have suggested that the MTA radiopacifier, the bismuth oxide (Bi 2 O 3 ), was responsible for tooth discoloration [8][9][10] since this compound destabilizes when in contact with the sodium hypochlorite solution, which is used for root canal preparation. 9,11 In addition, the amino acids in the collagen fibrils of the dentin organic matrix destabilize the Bi 2 O 3 , leading to a dark precipitate deposition. 10 Despite the changes in the MTA composition to improve its esthetical performance, it still promotes some degree of tooth discoloration throughout time. 4,5,[7][8][9][10][11] Therefore, new hCSCs have been developed to overcome this limitation of MTA. 12 The MTA Repair HP (high plasticity) (Angelus, Londrina, PR, Brazil) is an evolution of the original MTA cement, with improved rheological properties due to the addition of an emulsifying agent to the distilled water used for its manipulation. 13,14 Additionally, the Bi 2 O 3 was replaced by the calcium tungstate (CaWO 4 ) as a radiopacifier. 13,14 The NeoMTA Plus (Avalon Biomed Inc. Bradenton, FL, USA) is mixed with a water-based gel that results in good handling characteristics to the cement. Its radiopacifier is tantalum oxide (Ta 2 O 5 ). 15,16 Finally, the Biodentine (Septodont, Saint-Maur-des-Fossés, France) is claimed as a permanent bulk dentin substitute with improved mechanical properties compared to the original MTA. 17,18 Its radiopacifier is the zirconium oxide (ZrO). 17,18 Despite some success in avoiding tooth discoloration by hCSCs in short periods, it is a consensus among researchers that hCSCs promote changes in the color behavior of the dental structures in the long term, but the real mechanism of this discoloration is still uncertain. 4,5,[7][8][9][10][11]19 Możyńska, et al. 5 (2017) have reported that some hCSCs have a high potential for hard tissue discoloration, whereas some others promoted a small color change, which was nearly invisible to the human eye.
In addition, little scientific evidence exists on the potential discoloration induced by hCSCs in resinous materials, such as bonding agents and composite resin restorations. 20,21 Therefore, this in vitro study aimed to perform a two-year assessment of the discoloration potential of four different hCSCs (original MTA, MTA Repair HP, NeoMTA Plus, and Biodentine) on the enamel/ dentin structure and composite resin restoration. The tested null hypothesis was that the types of cement would not induce color (ΔE 00 ), lightness (ΔL'), chroma (ΔC›), and hue (ΔH›) differences on the enamel/dentin structure and composite resin, in addition to changes in the whiteness index (WI D ).

Experimental design
To determine the discoloration potential of hCSCs, a 4×7×2 experimental design was adopted, with the hCSCs (four levels), storage periods (7, 15, 30, 45, 90, 300 days, and two   The coordinates L*a*b* were recorded at the end of color measurements and the mean values were estimated. Following the study by de Jesus,et al. 7 (2021), the teeth were distributed into four groups (n=10) according to their colorimetric data to avoid significant variation among specimens from the same experimental group. For this, the data provided by the color measurements were correlated to the color scale guides (VITA Classical scale guide and VITA 3D Master guide) provided by the VITA Easyshade Advance 4.0 device. The experimental groups were formed with a balanced distribution of teeth with the same colorimetric references (e.g., three teeth color A2, three teeth color B1, two teeth color B2, and two teeth color A3 in the same experimental group). 7 After this distribution, the crown of each tooth was sectioned at the cementoenamel junction using a double-sided diamond saw (Buehler, Lake Bluff, IL, USA) coupled to a metallographic cutter (Isomet 1000; Buehler) operating at 300 RPM under copious watercooling. Next, a trephine drill (Neodent, Curitiba, PR, Next, the discs' surfaces, enamel, and dentin, were   with an A3 color composite resin, the preliminary color measurement for a uniform distribution of the specimens based on their colorimetric data was unnecessary. As described for the enamel/dentin discs, a 0.8 mm-deep cavity was made in the center of the composite resin discs, which was filled with the hCSCs to be tested (n=10). At the end of the hCSCs setting period, the cavities were restored with composite resin, following the same adhesive protocols previously described.

Color measurement
After the enamel/dentin and composite resin discs were restored, an initial color measurement (T0baseline) was performed at the center of the enamel surface and at the center of the surface opposite to the cavity containing the hCSCs for composite resin specimens, following previous descriptions. Once again, a neutral color condensation silicone matrix (Zetaplus; Zhermack) was fabricated and coupled to the VITA Easyshade Advance 4.0 device tip for proper measurement at the central area of each specimen.
The color measurement followed the CIE L*a*b* system. According to this system, the L* coordinate represents an object lightness, ranging from absolute black (0) to absolute white (100). The a* coordinate represents the green to red value rates (-90 to 70), and the b* coordinate represents the rate from blue to yellow (-80 to 100). 22 At the end of the initial color measurement (T0 -baseline), the discs were stored in artificial saliva Dental and composite resin discoloration induced by different hydraulic calcium silicate-based cements: two-year in vitro assessment

Statistical analysis
The normal distribution of data was confirmed by the Shapiro-Wilk test. The level of significance was set at α = 0.05. Color (ΔE 00 ) differences for enamel/ dentin and composite resin groups were evaluated by Two-way ANOVA (hCSCs × time). The lightness (ΔL'), chroma (ΔC'), and hue (ΔH') differences were evaluated by repeated-measures ANOVA. The WI D values were evaluated by One-way ANOVA, comparing the initial references of each group (7 days - Table 3) with the tested periods. Multiple comparisons were performed by post hoc Tukey's HSD test (α=0.05).
The statistical analysis was performed using SPSS 21.0 for Windows software (SPSS Inc., Chicago, IL, USA).

Results
Color (ΔE 00 ), lightness (ΔL'), chroma (ΔC'), and hue differences (ΔH') are presented in Tables 1 and 2 for the enamel/dentin and composite resin groups, respectively. The L*a*b* standard averages (L= 92.38; a= 1.07; and b= 26.22) were used as references for the comparisons in enamel/dentin groups and periods, whereas the L*a*b* standard averages (L= 85.95; a= 0.12; and b= 26.42) were used as references for the comparisons in composite resin (groups and periods). Therefore, the comparison of these differences among groups and periods can also be seen in Figure 2 (for enamel/dentin groups) and Figure 3 (for composite resin groups). In addition, the WI D values of the groups and periods may be seen in Table 3 and Figure 4.

Despite the multiple comparisons test showing similar
statistical values among the other evaluated periods (7,15,30,45,90, and 300 days), an increasing trend in color differences over time was noted.
When lightness (ΔL'), chroma (ΔC›), and hue differences (ΔH›) were evaluated by the repeatedmeasures ANOVA, significant lightness differences were found for the groups and periods tested (p<0.05). Interactions between groups × periods were not significant (p=0.67) ( Table 4)  to the other groups tested (p=0.036). However, MTA Repair HP and Biodentine groups were different from each other (α=0.05). Although the periods of analysis are not significant (p=0.28), all groups presented an increase in chroma values, with a trend toward red (a+). Concerning the hue, all groups had a trend toward blue (b-). The differences found were significant among groups (p<0.05) and periods (p<0.05). Interactions between groups × periods were not significant (p=0.57). The Original MTA and MTA Repair HP groups had the smallest hue differences, with similar statistical behavior (α=0.05). The greatest variations were observed after two years but with statistical similarity at the 300-day period (α=0.05).
For the composite resin specimens, groups, periods, and interactions between groups × periods were significant (p<0.05). The smallest color differences were observed after 15 days in the Biodentine group (1.5∆E 00 ), and the greatest differences were found in the NeoMTA Plus group, after two years (5.9∆E 00 ).
The greatest color differences were observed (α=0.05) after 300 days and two years, when statistical similarity was found between Original MTA and NeoMTA Plus groups . Differences in lightness were affected by groups, periods, and interactions between groups × periods (p<0.05) ( Table 4). The statistical similarity was observed between Original MTA and NeoMTA Plus groups (α=0.05) and MTA Repair HP and Biodentine groups (α=0.05), with a more evident lightness reduction for the NeoMTA Plus group. The lightness reduction was more evident at the most advanced periods (90 days, 300 days, and two years), which were statistically similar (α=0.05).
Groups and periods were statistically significant when chroma differences were analyzed (p<0.05).

Statistical similarity was also observed between MTA
Dental and composite resin discoloration induced by different hydraulic calcium silicate-based cements: two-year in vitro assessment   Repair HP and NeoMTA Plus groups (α=0.05), with an increase in chroma for all groups, especially for the Original MTA group, with a tendency towards red (a+). Chroma increased over time, with statistical similarity at the 90-day, 300-day, and two-year periods (α=0.05). Regarding the hue differences, repeatedmeasures ANOVA identified a significant difference between the groups and periods evaluated (p<0.05).
The greatest hue differences were found for the Biodentine group (-3.1∆H') after two years. Tukey's multiple comparisons test found these differences in the Biodentine group and identified statistical similarities among the other experimental groups.
The greatest variations were found after two years; however, this period differed from all other periods tested (α=0.05).
Regarding the WI D index, the greatest values were observed after 30 days for the Biodentine group (20.0 WI D -enamel/dentin) and the MTA Repair HP group (19.6 WI D -composite resin). It was possible to observe that WI D values decreased over time in these groups, especially after two years (Table 3 and Figure 3). Regarding the enamel/dentin specimens, ANOVA identified differences among the tested periods (p<0.05). Groups and interactions between groups × periods did not show statistical differences (p=0.57 and p=0.85, respectively). The periods of 300 days and two years were statistically similar (α=0.05).
The periods of 7,15,30,45, and 90 days were also similar (α=0.05). When the composite resin specimens were analyzed, groups and periods were statistically significant (p<0.05). The Tukey's test identified statistical similarity between Original MTA and MTA Repair HP groups, with the greatest WI D values, as well as between the periods of 90, 300 days, and two years (α=0.05), and 7, 15, and 30 days (α=0.05).

Discussion
This 2-year assessment in vitro study aimed to investigate the discoloration induced by four hCSCs (Original MTA, MTA Repair HP, NeoMTA Plus, and Biodentine) on the enamel/dentin structure and a composite resin restoration. According to the results, the null hypothesis was rejected since the different hCSCs changed the color, lightness, chroma, hue, and whiteness index of teeth and composite resin over time.
Although this study aimed to simulate the clinical use of different hCSCs, as a laboratory study, it has some limitations, such as the use of bovine teeth to assess discoloration. Human teeth are the most used substrate for dental research. 27,28 However, large quantities of human teeth in proper conditions for use may be challenging to obtain, primarily due to ethical concerns. 27,28 Many studies have already adopted human and bovine teeth to evaluate the potential discoloration induced by hCSCs. 4,[7][8][9][10] The similar results obtained by both models ensure the reliability of using bovine teeth in studies involving esthetics demand. 4,7,29,[32][33][34] In addition to the bovine teeth used to fabricate the enamel/dentin discs, composite resin specimens were fabricated and used to test the potential discoloration of hCSCs. Some studies assessed the pre-application of a bonding agent to dentin to prevent tooth discoloration caused by hCSCs. 20,21 However, to our knowledge, no studies have assessed the discoloration promoted by these types of cement on composite resin. Amongst their wide range of clinical applications, 2 hCSCs may be used as pulpcapping materials, coming into direct contact with resinous-based restorative materials. It should also be highlighted that there is a clinical trend towards performing vital pulp therapies and regenerative procedures in a single appointment. These treatment modalities may expose composite resins to hCSCs before they finish their setting reactions. 35 For this reason, their potential discoloration effect on these resinous-based materials must be tested. Therefore, standardized cavities were made at the center of dental and composite resin discs to mimic a pulp-capping clinical scenario.
Regarding the colorimetric analysis methodology, in the present study, the color differences (∆E 00 ) were estimated from laboratory measurements, and the CIEDE2000 color difference system was used to validate the results. 23 The CIEDE2000 system allows a better adjustment compared to the CIE L*a*b* An essential revision to the ISO/TR 28642 Dentistry-Guidance on color measurement was performed by Paravina, et al. 36 (2015). The authors provided a 50:50% PT and 50:50% AT under simulated clinical conditions, using both the CIE L*a*b* and CIEDE 2000 systems. 36 The threshold values reported for the CIEDE2000 system were 0.8 and 1.8 ΔE 00 for PT and AT, respectively. In the present study, the color differences found for all tested hCSCs over time were visually perceptible (color differences >0.8), as proposed by Paravina, et al. 36 (2015). Based on their previous study, 36 Paravina,et al. 24 (2019)  after two years in the composite resin specimens.
The color variation over time, regardless of the hCSCs tested and the substrate experimental condition, had greater significance after two years of assessments (Tables 1 and 2). Furthermore, this color variation did not follow an ascending pattern throughout the experiment. 37 This fact may be related to the polychromatic nature of both substrates tested 38 and the non-linear release of chemical compounds by the hCSCs during the different experimental periods. 39,40 These two variables probably played a vital role in the esthetics behavior of the tested cements. [38][39][40] As the precursor of the hCSCs, the Original MTA is a gold-standard material. [8][9][10] Thus, in this laboratory experiment, it was used as a positive control. The Original MTA has Bi 2 O 3 as radiopacifier. 1 The high atomic number of Bi (Z = 83) ensures greater radiopacity to this hCSC, even at lower concentrations (20% by weight). 1 However, this high atomic number has a deleterious effect on the hydration mechanism of MTA. 1 After the cement powder is mixed with water, the Bi 2 O 3 is incorporated into the hydrated phase of the cement.
A structure composed of ettringite, monosulphate, and hydrated bismuth calcium silicate is formed, which dampers the microstructure of MTA and negatively affects the physical-chemical and biological properties of this cement. [1][2][3] In addition, the remaining Bi 2 O 3 may undergo oxy-reduction or oxidation, forming metallic bismuth and bismuth carbonate, respectively. [41][42][43] According to several studies, the dark precipitate created after these two phenomena leads to tooth discoloration over time. [41][42][43] We also highlight that pre-etching of the dentin surface with 37% phosphoric acid was performed during the enamel/dentin disc restoration with composite resin. The amino acids present in the exposed collagen fibrils of the dentin organic matrix contributed to the Bi 2 O 3 destabilization, also leading to the dark precipitate deposition. [8][9][10]   It is essential to mention that reduction in lightness contributes to color difference (ΔE 00 ). 4 On the other hand, the MTA Repair HP group had a conflicting colorimetric behavior concerning this parameter, demonstrating how the mechanism of discoloration induced by hCSCs remains unknown. 7,45,46 Despite the similarity in the main components of the tested hCSCs, 45,46 the tested cements presented distinct colorimetric patterns over time for both substrates. Subtle differences in some chemical components of the tested cements suggest changes in their colorimetric behavior over time. 45,46 This phenomenon demonstrated the multi-factorial nature of the discoloration induced by hCSCs. 45,46 The findings in our study cover a gap in the current literature, especially for resinous materials. However, further studies assessing the discoloration potential of hCSCs are still needed to achieve proper esthetic needs for this class of repair biomaterials.

Conclusions
Within the limits of a laboratory study, the following conclusions may be drawn: 1. The tested hydraulic calcium silicate-based cement induced significant changes in the colorimetric behavior of the dental structures and composite resin over time, especially lightness.
2. The presence of Bi 2 O 3 as a radiopacifier must be considered when the discoloration potential of hydraulic calcium silicate-based cement is assessed in shorter-term studies.
3. The replacement of Bi 2 O 3 by other radiopacifiers to reduce dental and composite resin discoloration is not supported by experiments with more extended periods of analysis.