INTRODUCTION

Clinical conditions such as caries lesions, discoloration, diastemas, crown fractures, and misaligned teeth may cause an undesirable esthetic appearance and smile.¹ Two main esthetic treatment options are available to solve these problems: indirect ceramic and direct resin-based composite (RBC) restorations.² With recent developments in adhesive dentistry materials and techniques, direct RBC restoration is now considered a good minimally, or even noninvasive, option.^3,4 Compared to indirect ceramic restorations, direct RBCs have the advantages of single-visit treatment, less preparation time, durability, and repairability.⁴ However, periodic checks are mandatory to ensure durability of resin composites, and more of these checks are needed as compared to ceramics. In cases of fracture or chipping of the composite, a simple repair protocol is used to extend the life of the original restoration.⁵ Direct restorations reportedly have successful short-term clinical results.⁶ Appropriate indications, effective isolation, good optical and mechanical properties of the resin composite, operator experience and skill, accurate shade selection, successful finishing and polishing of the direct restorations, and frequent checkups are needed for long-lasting functional and esthetic outcomes.^4,7 In the literature, there is a lack of long-term evidence of the clinical efficacy of direct RBC restorations placed in the anterior teeth. The most common reason for failure of direct composite build-ups is fracture of the RBC.^6,8 According to previous studies, the 3- to 5-year anterior restoration survival rate varies between 79% and 89%.^5,6,9 Limited longevity has been reported for composite laminate veneers due to their susceptibility to staining, wear, and fracture.¹⁰ However, the potential influence of chemical and physical properties of the resin composite, the size of restoration, and patient- and dentist factors still remain to be determined, especially in long-term clinical trials. A need for information regarding the potential factors influencing long-term failure clearly remains. Kubo and others¹¹ investigated the main factors associated with the longevity of Class III–V composite restorations, including cavity type, gender, age, dentist factors, and the requirement for retreatment. Dentist factors, cavity type, and retreatment significantly influenced the survival rate.

Two types of veneers comprised of different microhybrid resin composite materials were compared by Gresnigt and others¹², and no significant difference in 3- or 5-year survival was found. A meta-analysis of prospective studies on anterior composite restorations reported median survival rates of 95% and 90% for Class II and IV restorations, respectively, after 10 years.¹³ Since esthetics is one of the main concerns regarding anterior teeth, some researchers have recommended using resin composites with a smaller filler size (nanofilled composites) to produce a smoother surface.¹⁴ However, a systematic review comparing nanofilled and submicron composites to microhybrid composites reported no improvement in surface smoothness with use of nanofilled composites.¹⁵

The aim of this clinical study was to determine the mid-term survival rate of Class IV composite restorations of maxillary anterior teeth and to investigate the possible reasons for failure.

METHODS AND MATERIALS

Study Design and Participants

Patients who had received Class IV restoration(s) of maxillary anterior teeth were selected for this 4-year clinical follow-up study. Fifty patients (22 males and 28 females; total of 168 Class IV restorations) aged between 18 and 56 years (mean age, 31.1 years) were included. A flow diagram of the restorations is shown in Figure 1. All patients provided written informed consent before the restorative procedures. Class IV restorations of maxillary teeth, which were conducted at least 4 years ago, were included. Baseline (1 week) and 1-, 2-, 3-, and 4-year follow-up data were evaluated by two experienced restorative dentistry specialists. The distribution of the restorations according to the layering technique, composite filler type, and tooth number is provided in Table 1.

View Figure

• Download as Powerpoint
• Download Figure

Table 1: Distribution of Class IV Restorations According to Layering Type, Composite Filler Type, and Tooth Number

View Fullscreen

Restorative Procedures

All restorations were performed using the same procedures by a restorative dentistry specialist in an academic university clinic. A silicone key constructed via either diagnostic wax-up or direct mock-up was used for all restorations. The principle of minimally invasive dentistry guided the preparations. Before tooth preparation, shade matching was performed with the button technique using a digital camera (D750 ; Nikon, Tokyo, Japan) with a 105-mm macro lens (Nikon), the R1C1 Wireless Close-up Speedlight System (Nikon), a flash mounting bracket (Owlbrckt C, Torun & Torun, Ankara, Turkey), and the polar_eyes cross-polarization filter (PhotoMed International, Van Nuys, CA, USA). The most appropriate enamel and dentin shades of the selected composite resin were placed on the labial surfaces of the adjacent teeth and polymerized, following which photographs were taken and evaluated by the operator. The selected shades and initial dental photographs were recorded for each patient. Following shade selection, teeth #5–12 were isolated with a rubber dam (Nic Tone, Bucharest, Romania). The teeth to be restored were retracted with either dental floss or rubber dam clamps (Hygenic Brinker Clamps; Coltene, Altstatten, Switzerland). Minimally invasive removal of the former restorations was performed in all cases. Then, 45° beveling of the buccal surface of the fracture lines (including all enamel and up to half of the exposed dentin) was performed using a red diamond needle-shaped bur (Dentsply Maillefer, Ballaigues, Switzerland) under constant water cooling. Following this, 35% phosphoric acid gel (Ultra-Etch, Ultradent Products, Inc, South Jordan, UT, USA) was applied to all of the prepared enamel surfaces for 30 seconds, water-rinsed for 10 seconds, and gently spray-dried. Then a one-step adhesive system (G-Premio Bond, GC Corporation, Tokyo, Japan) was applied to all of the etched enamel and nonetched dentin surfaces, as per the manufacturer’s instructions. The adhesive was left undisturbed for 10 seconds after rubbing and air-dried under maximum air pressure for 5 seconds. Polymerization was performed using a light-curing unit (wave length: 430–480 nm; Elipar DeepCure-S LED, 3M Oral Care, Maplewood, MN, USA), applied for 20 seconds at a light intensity of 1370 mW/cm² with an irradiated diameter of 10 mm.

Regarding the restorative materials and technique, a microhybrid RBC (Universal Shade Essentia, GC Corporation) with a chameleon effect was used for monochromatic Class IV restorations (n=76). A combination of microhybrid and nanohybrid RBCs (Medium Dentin [MD] and Light Enamel [LE] Essentia shades, GC Corporation) were used for polychromatic Class IV restorations (n=92). The brand, type, manufacturer, and chemical compositions of the materials are listed in Table 2. Mono- or polychromatic layering was performed according to the necessity for incisal translucency of the incisal edge. If the adjacent or symmetrical tooth had these features, the polychromatic layering technique (PLT) was considered. All restorations were gradually built-up under silicone index guidance. The incremental layering technique (≤2-mm thickness) was used for monochromatic layering of the microhybrid resin and for polychromatic layering of the micro/nanohybrid resin. A nanohybrid translucent shade (LE) was used to mimic the natural enamel tissue, whereas microhybrid, opaque, and chromatic shades (MD) were used to mimic the natural dentin tissue. Marginal walls of the restorations were completed using self-contoured, kidney-shaped posterior metal matrix bands (No. 1298, Tor VM, Moscow, Russia). All shades were polymerized for 20 seconds at an irradiation of 1370 mW/cm². The light intensity of the curing unit was evaluated before each restoration using a radiometer (Hilux Curing Light Meter, Benlioglu Dental, Ankara, Turkey). The final labial surface layers of the restorations were polymerized under a glycerin gel cover (Air Barrier, GC Corporation) to eliminate the oxygen inhibition layer.

Table 2: Brands, Types, Manufacturers, and Chemical Compositions of the Materials

View Fullscreen

The final occlusion was adjusted by protrusive and lateral movements of the mandible. Interproximal surfaces were polished with interdental polishing strips (Epitex strips; GC Corporation) with three different grits (medium #500, fine #800, and extrafine #1200). Labial and incisal embrasures were adjusted using aluminum oxide-embedded abrasive polishing discs (Sof-Lex, 3M Oral Care) with three different grains (medium [40 μm], fine [24 μm], and superfine [8 μm]) under dry conditions at 15,000 rpm, as per the manufacturer’s instructions. Finishing of the restorations was performed using a 12-blade bur (Diatech, Dental AC, Heerbrugg, Switzerland) at 30,000 rpm under water cooling. Diamond particle-embedded medium- and fine-grit rubber wheels (Twist Dia; Kuraray Noritake Dental, Tokyo, Japan) were operated at 10,000 rpm without water cooling to polish the labial surfaces. Additional polishing was performed using a medium-grit diamond bur (Diatech), operated horizontally at 5000 rpm. All patients were scheduled for repolishing 24 hours later. Only the high-shine (fine-grit) polisher was used for final surface polishing. Patient’s medical/dental histories, as well as dental photographs, and radiography records, if necessary, were collected at the 1-week and 1-, 2-, 3-, and 4-year follow-up appointments. No repair or repolishing procedure was performed at any of follow-up visits.

RESULTS

In total, the outcomes of 156 teeth were acceptable after 4 years, and the overall survival (OS) rate was 92.86%. The failure rates for the monochromatic layering technique (MLT) and PLT were 9.2% and 5.4%, respectively. The survival rate for the first year was 99% for MLT and 100% for PLT, and 99% overall; the respective rates for the second year were 96%, 98%, and 97%, while those for the third year were 93%, 97%, and 95%, and those for the fourth year were 91%, 95%, and 93%. Restorations requiring any repair or replacement were considered as failures. Repolishing was not performed during the follow-up period. Of the 168 restorations, 12 (7.14%) were failures. No restoration had more than one clinically unacceptable score, and no patients were lost to follow-up, so the number of unacceptable scores was equal to the number of failed restorations. The reasons for failure included fractured restoration (n=7), marginal discoloration (n=2), color mismatch (n=1), surface roughness (n=1), and caries (n=1). Fracture occurred in seven restorations (4.2% of the 7.1% of restorations that failed) and was the most common reason for failure in both the MLT (3.9%) and PLT (4.3%) groups. Only 2 teeth (2.6%) in the MLT group and 21 (22.3%) in the PLT group showed no detectable changes (score of 0). In 145 (86.3%) restorations, at least one change was detected (score of 1–4). Postoperative sensitivity (USPHS score of 1) was noted in 20 restorations (11.9%) in only eight patients, all at baseline (1 week after the restoration); all of these were considered recovered at the first-year follow-up.

Figure 2 shows the Kaplan-Meier survival curves for the restorations performed with the two layering techniques. The MLT (microhybrid RBC) and PLT (micro/nanohybrid RBC) groups showed no significant difference in mean survival time (p=0.343). The mean survival time was 46.026 and 46.957 months for the MLT and PLT groups, respectively. According to the chi-squared (χ²) test, the proportion of acceptable USPHS scores did not differ between the two layering techniques in any year (p≥0.05); this was also the case in the Cochran Q test analysis (p≥0.05). There was no significant difference in failure rate among years in the MLT group (p≥0.05), whereas, in the PLT group there was a significant difference, attributable to the rate in the second year (p=0.042). For all restorations, second-year scores (p=0.018) differed significantly from the first-, third-, and fourth-year scores (p=0.433, p=0.151, p=0.302, and p<0.05, respectively).

View Figure

• Download as Powerpoint
• Download Figure

Cox regression analysis of the restorations was also performed to evaluate the effect of four independent variables (tooth number, patient age and gender, and layering technique). None of the variables were associated with survival (p≥0.05) (Table 3).

Table 3: Cox Regression Analysis of the Restorations Regarding the Independent Variables

View Fullscreen

DISCUSSION

In this clinical follow-up study, the long-term performance of maxillary Class IV composite restorations was investigated. The restoration outcomes using two layering techniques (MLT and PLT) were compared over a 4-year period. The methodology employed has been used in many other clinical studies.^7,11,17,18 Also, the clinical evaluations in our study were performed by two independent observers, similar to previous studies.^1,19,20 Challenges to clinical studies include standardization of indications and treatment protocols, achieving operator agreement, and dealing with missing data.¹⁶ As the cases in this study were to be used in laboratory demonstrations for undergraduate and postgraduate students, a high-quality operator was required; the operator was a university instructor with 15 years of clinical experience specializing in restorative dentistry. The modified USPHS criteria were used for evaluating the selected restorations in this study, allowing for a standardized and detailed longitudinal assessment of the restorations. Although there is a need for a more definitive method,²¹ this evaluation method has nevertheless been used in many clinical studies, to which our results could thus be compared.^{11,12,16,22,23}

Substantial intraobserver (κ=0.74 and κ=0.77) and interobserver (κ=0.67) agreements were obtained for the fully blinded observers. The observers agreed with all the unacceptable scores, and the disagreements were not related to the acceptable/unacceptable decisions. Disagreements occurred regarding the acceptable scores of 0 and 1 for marginal discoloration, color match, and surface roughness criteria. It might be difficult to make a decision between these scores using only dental photographs, particularly for segmental deterioations from ideal in some of these criteria. Previously, Peumans and others⁹ reported that the photographic evaluation may mask imperfections on the restoration surface and thereby lead to misjudgements, especially for the assessment of color match. Therefore, the obtained minor disagreements were considered to not effect the targeted outcomes of the present study. The evaluations of the first observer were taken into consideration for the statistical analyses.

A high OS rate of 92.84% after 4 years was obtained for the Class IV restorations in this study. For survival analysis, the Kaplan-Meier method is the gold standard and was therefore used in this study instead of the log-rank test, which has limited utility for analyzing multivariate datasets. Nevertheless, an additional survival analysis (Cox regression) of the restorations was performed. Four independent variables (tooth number, patient age and gender, and layering technique) were not associated with survival (p≥0.05) (Table 3). The OS rate observed in our study was higher than that in the studies of Coelho-de-Souza and others¹ (3.5-year survival rate of 80.1%), Frese and others³ (5-year survival rate of 84.6%), Lempel and others¹⁶ (7-year survival rate of 88.3%), Khayatt and others²⁴ (7-year survival rate of 85%), and Gresnigt and others¹² (3.5-year survival rate of 87.5%). This may be because all of the restorations in our study were done by a single restorative dentistry specialist under the same clinical conditions, and not by undergraduate students or inexperienced operators (which can affect restoration longevity).^20,21 Although all of the patients in this study were dental school applicants with low socioeconomic status, this was not associated with negative outcomes in our study, unlike some other clinical studies.^1,21

In this study, a microhybrid RBC (81 wt%) was used in conjunction with the MLT, and a combination of microhybrid and nanohybrid RBC (76 wt%) in conjunction with the PLT. Coelho-de-Souza and others¹ reported that microfilled composite veneers had better surface gloss, color matching, anatomical, marginal adaptation, and surface staining properties compared to universal composites. However, nanofilled resin was not evaluated. Accordingly, gloss retention and polishability were previously reported to be better for resins including nanofiller (0.005-0.01 μm) compared to those including microfiller (0.01-0.1 μm).^14,25 More incisal chipping and a 3.7-fold higher risk of failure were reported by Lempel and others¹⁶ for build-up restorations of anterior teeth when using microhybrid resin compared to nanofilled resin. Massano and others²³ reported good clinical performance of Class III and Class IV restorations using nanofilled resin over a 2.7-year period (failure rate of 2.4%). However, microfilled resins were also reported to have the advantages of high surface hardness and high resistance to wear, fracture, and shrinkage.^19,26,27 As well as the size, both the shape and amounts of particles were reported to affect the performance of resin composites.⁷

In the present study, the respective failure rates when using the MLT and PLT were 9.2% and 5.4%. The Kaplan-Meier survival analysis revealed that the mean survival durations of the restorations were not different at 46.026 and 46.957 months, respectively, (p=0.343) nor were the survival rates for each year or the OS rate (based on the proportion of acceptable USPHS scores; p≥0.05). Both RBCs (microhybrid and nanohybrid) applied to the top surface layer of the restorations exhibited satisfactory performance. While nanofilled composites only use nanosized particles, nanohybrids combine nano- and microsized particles, similar to microhybrid composites.¹⁴ Thus, becoming microhybrid or nanohybrid is directly related to the distribution of the nano- and microparticles.²⁶ Moraes and others¹⁴ reported that microhybrid and nanofilled composites with similar matrix components yielded similar polymer network structures and thus similar hardness despite noticeable differences in filler size. According to their results, the behavior of nanohybrid composites was more similar to that of microhybrid, rather than nanofilled composites. Also, in previous studies, nano- and microhybrid resins were reported to have similar physical characteristics, depending on the filler content.^14,27 The RBCs used in this study were micro- and nanohybrids of the same brand, with similar contents including almost the same filler type and the same amount of filler particles (81 wt% for both) (Table 2). Therefore, this similarity might be the reason for no significant difference between the restorative materials, for marginal discoloration, color matching, surface roughness, and restoration wear in our study. It may also explain the lack of difference in mean survival duration at any time point or in OS between the two resin composites used. In addition, performing the restorations by a single specialist, under the same restoration protocols and clinical conditions, might be related to the similar clinical performances of the restorations with different RBC materials.²⁰

The restoration failure rates of the two layering techniques used in this study did not differ by time point (p≥0.05), except for the rate at the 2-year follow-up in the PLT group (p=0.042). Considering both groups together, the failure rate at the 2-year follow-up was different to those of the other time points (p=0.018). Two unacceptable “fractured restoration” USPHS scores during the second year may explain this result. Fracture occurred in seven restorations and was the most common reason for failure when using either the MLT (3.9%) or PLT (4.3%). These results were similar to those of previous clinical trials.^5,21

Fracture and chipping were the most frequent reasons for failure in microhybrid anterior RBC restorations in the studies of Frese and others,³ van Dijken and others,⁸ Coelho-de-Souza and others,¹ Gresnigt and others,¹² and Milosevic and Burnside.²⁸ In their systematic review Heintze and others¹³ reported that Class IV restorations, including of the incisal edge, had a higher risk of failure compared to Class III restorations. All of the fractured restorations had a USPHS score of 2 (“partial fracture in restoration>1/4”). The fracture rate was not statistically different between the two layering techniques (p≥0.05). No fracture was observed using either technique during the first 2 years of follow-up.

Marginal discoloration, the second most common reason for failure, was observed in two cases in each layering technique group, all of which had a USPHS score of 2 (“obvious staining could not be polished away”). The rate of marginal discoloration was not significantly different between the layering techniques (p≥0.05). Heintze and others¹³ and Lempel and others¹⁶ reported that adhesion to enamel and 37% phosphoric acid etching were important for good sealing and prevention of discoloration. Selective enamel etching was performed for all restorations in our study. Additionally, 45° beveling of the labial surface of the teeth prior to conditioning the enamel was performed, to ensure that the transition between the restoration and enamel was not visible. Beveling prevents marginal staining^16,23,29 and improves fracture resistance at the tooth-restoration interface.^16,29

Color mismatch was the least common reason for failure in this study and the rate thereof did not differ between the two layering techniques (p≥0.05). Only one restoration in the MLT group had an unacceptable (“slight mismatch in color or shade”) USPHS color mismatch score; 51 restorations in the MLT group and 32 in the PLT group had a score of 1 (“good color match”) during the 4-year follow-up period. Nasim and others²⁶ reported that the rate of discoloration was the highest for nanofilled RBCs among the microhybrid and microfilled RBCs tested. Tekçe and others³⁰ reported similar findings in vitro. Superficial degradation of restorative materials and absorption of staining agents are responsible for discoloration.¹⁶ Vichi and others²⁵ reported that low triethylene glycol dimethacrylate (TEGDMA) content in the resin matrix may limit water uptake and, by extension, the color variation induced by absorption of the staining solution. In this study, both RBCs contained TEGDMA, which may explain the staining results. Additionally, filler particle type, size, and distribution are important physical properties of composite fillers²⁷ and may affect color stability. A previous study reported that smaller filler particle size led to low visual opacity,²⁷ while, in another study, it decreased staining and enhanced esthetics.²⁵ Lempel and others¹⁶ reported no long-term positive effects of nanoparticles on color stability or surface gloss in vivo. This was supported by a recent systematic review, which concluded that nanofilled and submicron RBCs did not yield superior color stability or gloss retention outcomes compared to microhybrids.¹⁵ In addition to material factors, patient factors (such as diet) and operator factors (operating environment, isolation, adhesion, finishing and polishing protocols, and recall frequency) may also influence RBC staining outcomes. In our study, the experienced operator, standardized restorative technique, and high patient motivation may have been responsible for the very low rate of color mismatch failures.

The surface roughness USPHS score was unacceptable only in one case, at the 4-year follow-up in the MLT group. In total, 36 restorations performed using the MLT, and 32 using the PLT, had a score of 1 (“slightly rough or pitted”) on the surface roughness USPHS criterion during the 4-year follow-up period. Repolishing was not performed for any restoration. There was no statistically significant difference in surface roughness between the two layering techniques (p≥0.05). Caries related to the restoration was considered unacceptable in only one case and no significant difference in caries was found between the two layering techniques (p≥0.05).

While clinical examinations were performed only during the fourth year of follow-up, postoperative sensitivity data were obtained from the medical histories of the patients at baseline; 20 restorations (11.9%) in only eight patients had a score of 1 for this USPHS criterion, all of which had recovered at the first-year follow-up. As the etching of dentin with phosphoric acid is considered a risk factor for postoperative sensitivity,³¹ the use of the selective etch technique in this study may explain the low postoperative sensitivity scores, which also showed no difference between the layering techniques (p≥0.05). In accordance with the results of Gresnigt and others¹² and Lempel and others¹⁶ regarding restoration wear, no wear was detected in our cases.

Some researchers have suggested that the failure criteria should be revised, where some repaired restorations remain functional and therefore should not be considered as complete failures.^16,32,33 Those studies concluded that if repaired restorations are not classified as failures, annual failure rates would drop, such that repairability could be considered as a predictor of better survival of RBC restorations.^16,34 Frese and others³ classified repaired cases as restoration survival rather than failure. Reparability of the RBC materials was considered the most important factor in extending the life of their restorations, which had a functional survival rate of 100%. Van de Sande and others³⁴ reported 69% survival and 2.4% annual failure rates for Class III and IV restorations, respectively, when repair was not considered as failure, compared to 64% and 2.9%, respectively, when it was considered as failure. Composite repair is a suitable alternative to Class III–IV and veneer restorations, since it may increase the survival rate of anterior restorations.^16,34 However, in our study, restorations needing repair, retreatment, or even repolishing were considered as failures. Considering this, the 92.84% OS rate can be considered very high.

There were some limitations to our survival analysis, including the relatively low number of cases, mid-term follow-up period, and lack of generalizability, as only one operator was involved. The results of survival analyses for different dental materials should be interpreted with care, as the numbers of cases (including failures) and follow-up periods tend to be limited.¹ Demarco and others⁵ noted a lack of long-term clinical results regarding the performance of anterior RBC restorations in a systematic review. The reasons for this include poor patient compliance and follow-up visit attendance.²² Regarding our results, in case of a long-term evaluation period, perceptible major differences might have occurred in nano- and microhybrid restorations, therefore similar longevity outcome might have also changed. Recently, Dietschi and others²⁰ identified several factors influencing outcomes in a systematic review including; patient hygiene, caries risk, age, socioeconomic status, operator characteristics, treatment environment, observation period, and evaluation method. Use of composite filler materials and the type of curing light had little to no impact on clinical success at any time point, whereas treatment environment and number of operators affected the restoration failure rate. According to their results, a single operator yielded the optimum results.²⁰ In the present study, all restorations were performed by a single operator at the same clinic under consistent conditions and using the same materials. However, the effect of operator’s skill, experience, and the operation environment still remained unclear. The outcomes of this study represent patients without bruxism. Therefore, the high success rate might also be associated with patients with low-risk factors. In spite of that, some patients might have developed slight or severe bruxism during its course, and this was not assessed. Differences among patients in oral parafunctions, malocclusion, dietary habits, and oral hygiene might have also affected the outcome. Because bruxism is a self-reported behavior that is difficult for patients to identify, diagnoses based on patient histories can be inaccurate. There is evidence that bruxism is a major risk factor for fracture.⁸ However, in many other clinical trials, bruxism was not associated with survival. Coehlo-de-Souza and others¹ reported no correlation between tooth fracture and the longevity of build-up restorations. In the study of Milosevic and Burnside,²⁸ bruxism was not associated with tooth fracture or restoration failure. Further clinical long-term studies are needed to assess the effect of bruxism on survival rate. Also, studies including more than one operator, larger sample size, and a variety of RBC materials are necessary to verify the findings of this study.

CONCLUSIONS

From this study, the following conclusions were drawn:

1. Class IV direct composite resin restorations showed good clinical outcomes, with a survival rate of 92.84% after 4 years.

2. Use of both the MLT and PLT for Class IV anterior restorations provided acceptable durability, with mean survival periods of 46 and 47 months, respectively.

3. Fracture was the most common reason for restoration failure in both the MLT (3.9%) and PLT (4.3%) groups.

4. Micro-/nanohybrid composite restorations showed a slightly higher survival rate (94.6%) than the microhybrid composite restorations (90.8% survival), but the difference was not statistically significant.

5. Monochromatic layered microhybrid and polychromatic layered micro-/nanohybrid Class IV restorations showed no significant difference in optical properties over the 4-year study period.