The purpose of this study was to comparatively assess the seven-year clinical performance of a one-bottle etch-and-rinse adhesive with resin composite (RC) and resin-modified glass ionomer (RMGI) restorations in noncarious cervical lesions.
One operator placed 70 restorations (35 restorations in each group) in 30 patients under rubber dam isolation without mechanical preparation. The restorations were directly assessed by two independent examiners, using modified US Public Health Service criteria at baseline and 6, 12, 24, 60, and 84 months. The obtained data were tabulated and statistically analyzed using the Fisher and McNemar tests. A difference was significant if p<0.05.
Twenty patients were available for recall after seven years (66.6%), and 25 RC and 26 RMGI restorations out of 70 restorations were evaluated. Excellent agreement was registered for all criteria between examiners (κ≥0.85). Alfa and bravo scores were classified as clinically acceptable. The McNemar test detected significant differences within RC restorations between baseline and seven-year evaluations for anatomic form, marginal integrity, and retention (p<0.05). For RMGI restorations, a significant difference was identified for marginal integrity (p<0.05). As to material comparison, the Fisher exact showed a better retention performance for RMGI restorations than for RC restorations (p<0.05). Twelve composite restorations were dislodged (52.0% retention) and three ionomer restorations were lost (88.5% retention). The cumulative success rate for RC and RMGI was 30% and 58.1%, respectively.
After seven years of service, the clinical performance of RMGI restorations was superior to that of the adhesive system/resin composite restorations in this study.
About a quarter of the population has noncarious cervical lesions, and these lesions are significantly more prevalent at older ages, with premolars being the most affected teeth.1
Although the dentin characteristics of surfaces from abrasion/abfraction lesions may inhibit dentin etching and the formation of hybrid layers,2 clinical bonding effectiveness has been shown in noncarious cervical lesions located mainly in dentin.3 Therefore, it is expected that the clinical performance of restorations placed in noncarious cervical lesions would be more critical, especially in long-term evaluations.
Numerous studies have investigated the clinical behavior of noncarious cervical lesions restored with resin composite and glass ionomer cement. A large number of clinical trials are short-term evaluations. In general, in up to three years of follow-up, both resin composite and glass ionomer restorations have good performance.4-11 The main causes of failure of this type of restoration are low retention rates for resin composite restorations5,6,12-14 and poor color stability for glass ionomer restorations.15,16
Studies with longer evaluation periods show that resin composite materials tend to fail with time. Retention rates of such restorations are markedly lower when compared with glass ionomer restorations.17,18 A 13-year follow up of class V noncarious cervical lesions verified a continuous degradation of the bond with a wide variation, independent of the adhesion strategy used, with the best retention for restorations performed with resin-modified glass ionomer cement (RMGI) and with adhesive restorations using four-step etch-and-rinse systems.3
A review of prospective clinical trials showed that restorations with glass ionomer resulted in the highest success rate in regard to retention if compared with adhesive systems and resin composite restorations.18
Long-term evaluations are especially interesting when considering the behavior of one-bottle or primer-adhesives. Currently, some problems have been identified with this category of adhesive system, such as permeability19 and hydrolysis. Although a tendency exists toward adhesives with simplified application procedures, simplification appears to induce a loss of effectiveness in restorations of noncarious cervical lesions.18
Long-term clinical trials are certainly needed, because they remain the ultimate way to collect scientific evidence on the clinical effectiveness of restorative treatments. Thus, the aim of this study was to assess the seven-year clinical performance of a two-step etch-and-rinse adhesive system and resin composite system in comparison with a RMGI restorative material in noncarious cervical lesions in a prospective study.
METHODS AND MATERIALS
Thirty volunteers (18 to 50 years of age) were properly instructed about the condition and objectives of the study and signed an informed consent and authorization form in order to participate in this investigation, following the guidelines of the local institutional review board (approved on February 24, 2000). Inclusion criteria were good oral hygiene, no periodontal disease or deleterious habits, and presence of at least two noncarious cervical lesions, following the split-mouth group design.
No lesions were less than 1 mm deep. Each patient randomly received at least one adhesive system/resin composite and one resin-modified glass ionomer restoration, for a total of 70 restorations. All types of class V noncarious lesions were included in the present study. As the association of abrasion, erosion, and abfraction is usually observed, all situations were considered for inclusion. Although the degree of dentin sclerosis was not considered an exclusion criterion, highly mineralized pigmented lesions were not observed.
Each patient received at least two or a multiple of two restorations. Each restorative material was randomly allocated to one of the randomized cervical lesions until the two treatments were present in the same subject and in equal numbers. The randomization process was performed by placing all the selected teeth on a list and assigning treatment according to a predefined sequence: 1) resin composite and 2) RMGI.
Restorative procedures were carried out by an experienced clinician, who was submitted to a calibration process that consisted of performing 10 repeated restorations of each material under direct supervision of the project's coordinator. The operator's questions were addressed and consensus was obtained during the calibration session. All restorations were placed under rubber dam isolation, and no cavity preparation was carried out. Enamel margins were not beveled and no mechanical retention was performed. Twenty-six patients received two restorations, three patients received four restorations, and one patient received six restorations.
Half of the cavities were restored with a two-step etch-and-rinse adhesive system and a resin composite (Excite/Tetric Ceram, Ivoclar Vivadent, Schaan, Liechtenstein) and the other half with the RMGI (Vitremer, 3M ESPE, St Paul, MN, USA) (Table 1), according to the manufacturer's instructions, as follows:
Enamel was etched for 30 seconds and dentin for 15 seconds with 37% phosphoric acid gel (Total Etch, Ivoclar Vivadent), washed for 30 seconds, and dried gently. Absorbent paper was used to remove the excess water. One coat of the two-step etch-and-rinse adhesive system (Excite, Ivoclar Vivadent) was applied to the visibly moist dentin surface and brushed gently for 10 seconds. An air blast was applied to facilitate the evaporation of the alcohol solvent and the primer adhesive was light-cured for 20 seconds. Resin composite (Tetric Ceram, Ivoclar Vivadent) increments were inserted and light-cured for 40 seconds using a calibrated light-curing unit (XL 3000, 3M ESPE) at 600 mW/cm2 with a curing radiometer (Demetron/Kerr, Orange, CA, USA). Composite excess was immediately removed using a No. 12 blade. Finishing and polishing was performed one week later using 12-fluted tungsten carbide burs (FG Bur, KG Sorensen, Brazil), abrasive cups (Enhance, Dentsply, USA), and disks (Sof-Lex polishing disks, 3M ESPE);
The RMGI (Vitremer, 3M ESPE) was used according to the manufacturer's instructions. Primer was applied for 30 seconds to the surface of the lesion using a micro-brush. Light-curing was conducted for 20 seconds using the same calibrated device that was used for resin composite restorations. The RMGI was manipulated in a 1:1 powder-to-liquid ratio and was inserted using disposable tips and a syringe (Centrix, Shelton, CT, USA). The restoration was light-cured for 40 seconds and excess material was immediately removed using a No. 12 surgical blade. Finishing and polishing were carried out similarly to resin composite restorations, one week after the restorations were placed.
Two independent and calibrated examiners other than the operator were responsible for the clinical evaluations. Calibration procedures were carried out using picture slides representing each condition to be assessed in the study. A double-blind design was originally assigned. In cases where the two examiners disagreed on a rating, both reexamined the restoration and arrived at a final joint decision. Modified US Public Health Service criteria20 were used to evaluate retention, marginal integrity, marginal discoloration, anatomic form, and secondary caries (Table 2) at baseline and 6, 12, 24, 60, and 84 months. The baseline rating was carried out one week after restoration, immediately after the finishing and polishing procedures.
Interexaminer agreement was assessed using κ. Excellent agreement was registered between both examiners for all criteria (κ≥0.85). Intragroup comparisons between baseline and other evaluation periods with the same material were performed by the McNemar test (p<0.05). Intergroup comparisons to identify differences between restorative materials at each period were conducted by Fisher exact tests (p<0.05). The cumulative survival rate was determined following the life table method.
Recall rates registered were 100% for baseline, 6, and 12 months; 93.3% for two years; 73.3% after five years; and 66.6% after seven years. The registered recall rates overcame the minimum requirement of the American Dental Association (ADA) guidelines21 for sample size in clinical trials for restorative materials, which state that a minimum of 20 patients must be available for the two-year recall and 15 patients must be examined at the four-year evaluation.21
Data for retention, marginal integrity, marginal discoloration, anatomic form, and secondary caries of all evaluation periods are presented in Table 3. Numbers in parentheses indicate the total number of restorations classified as clinically acceptable (alfa and bravo ratings) in each evaluation period versus the total number of restorations assessed at that time.
Given that all patients were available for recall at 6 and 12 months, all 70 restorations could be assessed. Although four resin composite restorations were lost at six months and one more was lost at the one-year assessment, no significant differences were detected in all criteria for both materials at these periods.
At the two-year recall, two patients could not be found, so 66 restorations were assessed. Seven composite restorations were lost, whereas no ionomer restoration was lost. When comparing resin composite and RMGI restorations at the two-year recall, the only significant difference was for the retention criterion (p=0.011) in favor of RMGI restorations.
At the five-year recall, 22 volunteers returned for evaluation. Of 70 restorations, 55 were evaluated (27 resin composite restorations and 28 resin-modified glass ionomer restorations). Sixteen composite restorations were regarded as failures at this evaluation period (51.5% retention), whereas only one ionomer restoration was lost (96.4% retention). Significant differences were found for resin composite restorations between baseline and the five-year recall for marginal integrity and retention. For RMGI restorations, no significant differences were identified for all criteria. When comparing both materials, significant differences for retention after five years were found.
At the seven-year recall, 20 volunteers returned for evaluation. Of 70 restorations, 51 were evaluated (25 resin composite restorations and 26 RMGI restorations). Twelve composite restorations were considered as failures at this evaluation period due to retention (52.0% retention), whereas three ionomer restorations were lost (88.5% retention). Greater retention was found after seven years if compared with five years for resin composite restorations because one patient who was not evaluated at five years returned after seven years.
The cumulative survival rates of retention for both tested groups were 95.8% and 63.7% for RMGI restorations (Vitremer) and 2-step etch-and-rinse adhesive and resin composite restorations (Excite/Tetric Ceram), respectively (Table 4).
The distribution of restorations regarding location and type of tooth for both tested groups is shown in Table 5. The number of restorations that failed with respect to retention is also cited in Table 5. The ages of the patients whose restorations failed for retention were 30, 35, and 50 for resin-modified glass ionomer restorations; and they were 18, 27, 32, 35, 36, 37, 40, 45, and 50 years for resin composite restorations.
The McNemar test detected significant differences for resin composite restorations between baseline and the seven-year recall for retention (p=0.002). For RMGI restorations, a significant difference was identified for marginal integrity (p<0.05, Figure 1). When comparing both materials, the Fisher exact test pointed out a significantly better performance for RMGI restorations than for resin composite restorations with respect to retention (p<0.01).
Noncarious class V clinical trials remain the ultimate testing method for the assessment of bonding effectiveness.22 Due to a lack of inherent macro-mechanical retention, adhesion is the most important factor in the retention of restorations in cervical abrasion/erosion lesions.23 Microtensile adhesion tests have shown a correlation with marginal discoloration of clinical studies on class V nonretentive restorations; however, no significant correlation was found with the clinical index, retention rate, or marginal integrity.24
Adhesive materials such as resin composites, RMGI cements, or polyacid-modified resin composites have been mostly used to restore noncarious cervical lesions.18 This study investigated the clinical performance of a two-step etch-and-rinse/resin composite system and RMGI cement. Retention was considerably lower for resin composite restorations after seven years when compared with RMGI restorations. These results are in accordance with earlier research studies.3,23,25
Because no cavity preparation was performed, including no enamel beveling, the true dentin bonding capacity of the restorative systems could be evaluated. Several research studies have used this experimental model to assess the clinical performance of materials in noncarious cervical lesions.9,12,17
The loss rate of composite restorations has always been a matter of interest. The low retention rate of resin composites is possibly due to the degradation of the adhesive bond. With former adhesive systems, lost composite restorations composed up to 80% of the total restorations after four years.17 Other studies confirmed that two-step etch-and-rinse adhesives have shown satisfactory retention rates (80% to 93%) during the first two years of evaluation,4,26,27 but in the present study the number of lost restorations after seven years of clinical service was considerable. Evidence exists now that two-step etch-and-rinse adhesives are regarded as permeable membranes after polymerization because they lack a comparatively more hydrophobic bonding resin layer.19 Therefore, they allow the continuous transudation of dentinal fluid and do not provide a hermetic seal of deep dentin,19 which may accelerate the degradation of the adhesive interface.
Differences of clinical success have been observed according to the type of adhesive system used. High retention rates of etch-and-rinse adhesives after a 24-month evaluation of noncarious cervical lesions were detected, as long as the clinician rubbed the adhesives vigorously onto the dentin surfaces.7 In noncarious class V lesions, high retention rates were observed in a 13-year clinical evaluation of two three-step etch-and-rinse adhesives, with values ranging between 85% and 94% depending on the resin composite used.28 Van Dijken and Pallesen29 observed that restorations placed with a self-etching primer (Xeno III) and a resin composite (Tetric Ceram) or a modified poly-acid resin composite (Dyract AP) in noncarious cervical lesions showed acceptable clinical rates of retention to dentin surfaces after seven years regardless of the restorative material used. After eight years of clinical service, the clinical effectiveness of a mild two-step self-etch adhesive (Clearfil SE) was excellent.30 Selective phosphoric acid-etching of the enamel margins had only some minor positive effects in marginal defects/discolorations.30
A statistically significant difference was observed in a systematic review18 of glass ionomer versus two-step etch-and-rinse in noncarious restorations. Glass ionomer restorations showed the highest success rate in regard to retention with a 1.9% mean annual loss of retention; whereas, the mean annual retention loss of two-step etch-and-rinse adhesive systems was 6.2%.18 Similar observations have been recorded31 with glass ionomers' retention loss of 2% ± 2%. However, another systematic review32 about different adhesive systems used for the restoration of noncarious cervical lesions concluded that there is not enough evidence to support one adhesive or bonding strategy over another for the treatment of this type of lesion.
Another challenge is the characteristic of noncarious cervical lesions. Sclerotic dentin and tubule occlusion by mineral crystals are very often present. Additionally, many parts of the wedge-shaped cervical lesion contain a hypermineralized surface that resists acid etching.2
The effectiveness of cervical restorations is most often expressed as retention loss relative to the observation time. Hence, retention is one of the most important criteria and is often used to assess the longevity of a restorative material. Retention loss is a robust objective criterion that is affected by a low degree of variability between different examiners in contrast with other clinical criteria. The latest guidelines of the ADA for submission of dentin and enamel adhesive materials require as provisional acceptance a retention rate of at least 95% of the restorations placed at the six-month recall.21 To obtain full acceptance, retention of 90% after 18 months is required. Unfortunately, the guidelines have no requirements for the long-term durability of adhesive systems. In a systematic review, 96% of glass ionomer and 51% of two-step etch-and-rinse class V restorations got the full ADA acceptance in the observation period of the studies varying between one-half and six years.18 These results are in accordance with the data obtained in the present study.
High retention rates of Vitremer restorations may be attributed to their adequate mechanical properties and better adhesion to dental tissues.33,34 Vitremer possesses two adhesion mechanisms: first, an auto-adhesive capacity by forming ionic bonds between the carboxyl groups of polyalkenoic acid and hydroxyapatite35 ; and second, micromechanical interlocking of the polymer.36 In addition, it is suggested that the clinical retention of an adhesive restoration depends not only on the retention capacity of the adhesive system used but also on the viscoelastic properties of the restorative material tested.37 It has been reported that the elastic modulus of glass ionomer cements is more similar to enamel and dentin than adhesives are.38 It has been claimed that flexural deformation of a tooth in the cervical region is at least partly absorbed by restorative material. The materials with a modulus of elasticity similar to that of a tooth when used in cervical restorations tend to bend more like a tooth structure when subjected to a masticatory load and may flex and be retained.39 The combination of these factors may have been responsible for the higher retention rate of glass ionomer restorations in the present study.
Although RMGI cements have shown good retention rates in the literature, the major problem with this material is poor color stability.16,23,25,40 Sidhu41 reviewed some of the existing literature on the clinical performance of RMGI cements and observed an adequate performance in terms of retention. Secondary caries as well as postoperative sensitivity are not a concern for this type of restorative procedure.41 However, this is not necessarily true for marginal characteristics, surface properties, and color stability.41 In cases where esthetics is essential, an association of RMGI cement and resin composite has been an optional restorative therapy. The combination of these restorative materials is a good alternative and seems to minimize restoration loss.12,37,42
The overall better behavior of ionomer restorations is in accordance with previous papers on short-term4,15,43-44 or long-term evaluations.40,45 Few clinical trials suggested a limited longevity for RMGI restorations compared with resin composite restorative materials.15,46 This difference observed between the present study and other previous studies may be due to cavity or type of lesion. Additionally, the shape and size of restorations, operator variability, occlusal factors, bonding capacity of the restorative system, application and curing technique used, and other factors during aging of the restoration such as temperature and pH cycles in the mouth47 could account for the differences between the studies. In a five-year evaluation of class V restorations placed by UK general practitioners, different factors were associated with increased longevity. Greater durability of these restorations could be achieved by improving operator skills, followed by cavity preparation and appropriate material handling.48 Additionally, Brackett and others49 observed reduced percentages of clinical success of noncarious class V lesions restored by inexperienced operators due to limitations in the manufacturer's instructions of self-etching adhesives. In the present study, the restorations were performed by only one operator, in order to decrease the effect of the operator factor. This allows ranking of the restorative materials tested but does not give information about variations in durability of the materials due to operator variability.
The cumulative survival rates at the 5- to 7-year follow-up interval in the present study were 95.8% and 63.7% for RMGI restorations and resin composite restorations, respectively. These data are in agreement with long-term studies performed with similar materials, in which better performance of glass ionomer materials was also observed.3 The comparison of the present data with studies50 of equivalent evaluation periods revealed similarity to retention rates of resin composite restorations (range, 60% to 74%).
Despite the limitations of this study, the overall clinical performance of resin-modified glass ionomer restorations (Vitremer) was superior to the two-step etch-and-rinse adhesive and resin composite (Excite/Tetric Ceram) restorations after seven years of evaluation.
Conflict of Interest
The authors of this manuscript certify that they have no proprietary, financial, or other personal interest of any nature or kind in any product, service, and/or company that is presented in this article.
Eduardo B Franco,DDS, MS, PhD, professor, Department of Operative Dentistry, Endodontics and Dental Materials, Bauru School of Dentistry, University of São Paulo, Bauru, Brazil