Effect of surface treatment on glass ionomers in sandwich restorations: a systematic review and meta-analysis of laboratory studies
Article information
Abstract
Objectives
This study aimed to evaluate the effect of different surface treatments on the bond strength between new or aged glass ionomers (GI) and resin composites in sandwich restorations.
Methods
A comprehensive search was conducted in three databases to identify studies focusing on the bond strength of new or aged GIs and resin composites in laboratory settings. The selected studies were assessed for potential biases based on predetermined criteria. Additionally, a meta-analysis was performed using three studies.
Results
A total of 29 studies were included, with 24 investigating the bond strength of new GIs and five focusing on GI repair. Three studies were included in the meta-analysis (with a 95% confidence interval) which revealed no significant difference in the mean MPa values of resin-modified glass ionomer (RMGI) treated with phosphoric acid or Er,Cr:YSGG laser before the application of an etch-and-rinse adhesive. Surface treatment was found to be crucial for achieving optimal bonding between GI and resin composite, regardless of the GI’s condition.
Conclusions
The combination of mechanical and chemical surface treatments does not significantly affect the bond strength between new RMGI and composite. However, for GI repair, it is recommended to use both treatments to enhance the bond strength.
INTRODUCTION
The use of glass ionomer cement (GIC) in dental restoration is highly regarded for its strong bond with enamel and dentin, sustained fluoride release, and thermal expansion similar to dentin [1,2]. However, GICs have limited aesthetics and lower abrasion resistance compared to resin composites, limiting their use in high-stress areas. To address this, the “sandwich technique” or “bilayered technique” was developed [3].
The lamination technique uses two restorative materials to create a single restoration, aiming to combine their physical and aesthetic benefits [4]. It pairs the dentin-adhesion and fluoride release of GIC with the aesthetics and polishability of resin composite for optimal results [5]. This method is ideal for deep cavities or those with undermined areas, where resin composite is preferred [5]. When restoration margins contact with dentin, applying lamination over GIC enhances adhesion and reduces microleakage [2]. Conventional and RMGI can be used, differing in adhesion, setting reaction, and moisture sensitivity [6].
The weak cohesive strength of GIC, combined with the limited chemical interaction between GIC and resin composite resulting from their distinct chemical reaction mechanisms, reduces the bond strength between the two materials [7]. To improve this, RMGI, which includes polymerizable functional groups, replaced conventional GIC, enhancing bond strength through both chemical and mechanical bonding [8].
In sandwich restorations, the bond between GIC and resin composite is crucial for retention, durability, and sealing [7]. Inadequate bonding is a primary cause of failure, often leading to caries and restoration breakdown [9]. Factors affecting the bond include the base material’s tensile strength, bonding agent viscosity, and the polymerization shrinkage and bonding capacity of the overlay composite [10].
Despite the use of sandwich restorations, the water-based nature and brittleness of GIC can lead to fractures or wear, necessitating repairs by adding fresh material [11]. Repair is needed in cases of technique errors or unaesthetic marginal staining. While resin composite repair is well-studied and routine in minimally invasive dentistry [12], the protocol for repairing GI-based restorations with resin composite remains unclear.
The literature examines various chemical and mechanical surface treatments for GICs in sandwich restorations or resin composite repairs [13]. Early studies recommended etching GICs to improve bonding with composite resin [14], but recent research warns that etching may weaken GICs and cause cracks [15]. It is suggested that GICs may already have sufficient surface roughness for bonding, and excessive etching can damage the material [15]. Studies on bonding agents show mixed results, with some recommending mild self-etch adhesives [16], while others favor GI-based adhesives [7].
There is ongoing debate about the necessary mechanical surface treatment for aged GICs to ensure a durable bond with resin composite. Studies have used various methods for mechanical roughening, such as bur use [17] and air abrasion [13]. Additionally, alternative treatments like photodynamic therapy (PDT) and laser treatments (Er,Cr:YSGG and Nd:YAG) have shown promising results [18].
Given the variety of adhesive systems, advancements in GIs, and differences in adhesion and setting mechanisms, no consensus exists on the optimal surface treatment for GICs in sandwich restorations. This includes cases where GICs are bonded or repaired with resin composite. This review aims to assess the literature to identify the most effective surface treatment (chemical or mechanical) for new or aged GIs, focusing on bond strength in resin composite sandwich restorations.
METHODS
Using the PICO (Participant, Intervention, Comparator, and Outcome) framework, this systematic review aimed to address the following research questions [19]:
1. What is the most effective surface treatment (mechanical and/or chemical) (C) for achieving optimal bond strength (O) when bonding (I) new GI and resin composite in sandwich restorations (P)?
2. What is the most effective surface treatment (mechanical and/or chemical) (C) for achieving optimal bond strength (O) when bonding (I) aged GI and either resin composite or GI in sandwich restorations (P)?
The systematic review procedure was registered with Prospero and assigned the identification number CRD42024614235. This review strictly followed the protocols outlined in the recently revised PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) guidelines [20]. To acquire pertinent data, a comprehensive search was conducted across three electronic databases: the National Library of Medicine (MEDLINE/PubMed), Scopus, and ScienceDirect. The specific search methodologies used for each database are detailed hereafter.
#1 AND #2 AND #3 AND #4 AND #5
#1 “glass ionomer” OR “ionomer” OR “ionomers” OR “glass ionomer restoration” OR “glass ionomer cement”
#2 “glass ionomer repair” OR “dental restoration repair” OR “aged glass ionomer”
#3 “sandwich restoration” OR “bilayered restoration” OR “bilayered resin composite” OR “sandwich technique” OR “laminate technique”
#4 “surface treatment” OR “chemical treatment” OR “mechanical treatment” OR “laser treatment” OR “surface roughening” OR “bonding agents” OR “adhesives”
#5 “bond strength” OR “shear bond strength” OR “microshear bond strength” OR “tensile bond strength” OR “microtensile bond strength”
The following filters were applied to each database:
a. Pubmed
Year: 2010-2024
Article type: Comparative study, evaluation study, and observational study
Language: English
b. ScienceDirect
Year: 2010-2024
Article type: Research articles
Subject area: Medicine and Dentistry
Language: English
c. Scopus
Year: 2010-2024
Subject area: Dentistry and materials science
Document type: Limited to article
Language: limited to English
Non-internet sources were manually checked to ensure comprehensive coverage. The retrieved articles from the searches were transferred into EndNote X7.7 software (Clarivate Analytics, Philadelphia, PA, USA), which was utilized to detect and eliminate duplicate entries.
Search strategy
The comprehensive search strategy, collaboratively developed by the review team, underwent peer review by a second information specialist. Eligibility criteria for the studies were evaluated by each author based on three factors: the relevance of the title, the abstract, and the full text. The laboratory studies included in this review focused on investigating the required surface treatment (mechanical and/or chemical) for bonding new or aged glass ionomers (GIs) and resin composites in sandwich restorations, with a specific emphasis on bond strength. The exclusion criteria included: review papers, clinical trials, and case reports, implant-supported restorations, primary and endodontically treated teeth, studies that evaluated the bonding of pulp capping materials to GIs, and studies that evaluated the GI/resin composite interface when bonded to dentin or in a cavity. To be included in the review, all three authors had to reach a consensus. Any disagreements were resolved either through consensus or discussions with a senior researcher.
Risk of bias assessment
In order to evaluate potential bias, two researchers independently evaluated the research utilizing criteria established in earlier systematic reviews of laboratory experiments [2,21]. If a criterion was specifically addressed in the research, it was labeled as “yes” to indicate its presence. Conversely, if the information was absent or a criterion was not detailed, it was marked as “no.” Studies that documented 1 to 3 criteria were deemed to possess a significant risk of bias, those documenting 4 or 5 criteria were considered to have a moderate risk of bias, and studies detailing 6 or 7 criteria were categorized as having a low risk of bias.
Statistical analysis
The meta-analysis focused on studies that compared the shear bond strength values in RMGI (Fuji II LC) treatments when either phosphoric acid or Er,Cr:YSGG laser was used before the application of an etch-and-rinse adhesive. Information such as sample sizes, averages, and standard deviations (SD) in MPa were extracted from these studies and analyzed using Stata for Windows, ver. 17 (StataCorp. LLC, College Station, TX, USA), with a confidence level of 95%. The average MPa data following both surface treatments were calculated using either fixed-effect or random-effects models, which were selected based on statistical assessments for heterogeneity outcomes. The heterogeneity of the data was assessed through the Q homogeneity test, with significance determined at p < 0.05. The thresholds for interpreting I2 are as follows: 0% to 40%, heterogeneity might not be considered significant; 30% to 60%, heterogeneity may indicate moderate levels; 50% to 90%, heterogeneity may suggest substantial levels; and 75% to 100%, heterogeneity is considered considerable. A funnel plot, along with Egger’s test, was later employed to evaluate the possibility of publication bias.
RESULTS
Search results
Electronic searches, utilizing the aforementioned filters for each database, were conducted and identified 11,007 published articles. The authors independently evaluated the titles and abstracts of the search results. However, 10,960 studies were excluded for one or more of the following reasons: they were not related to dental research, were case reports or clinical trials, focused on implant-supported restorations, primary teeth, endodontically treated teeth, or indirect restorations.
Forty-seven studies underwent full-text assessment to determine eligibility. Subsequently, based on the inclusion and exclusion criteria, 18 studies were excluded. These excluded studies focused on assessing the bonding of pulp capping materials to GIs or the bond strength of the GI/resin composite interface when bonded to dentin or within a cavity. Ultimately, 29 studies met the originally specified inclusion criteria for this review. The stages of the search process are depicted in the flowchart (Figure 1).
A flowchart depicting the search process, modified from the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) Statement.
Data extraction
The current review evaluated 29 studies that investigated the effect of surface treatment approach (either mechanical or chemical) on new or aged GIs, particularly in terms of bond strength in resin composite sandwich restorations. Twenty-four studies were included for the first review question, and five studies were included for the second question. Extracted data from the studies are summarized in Tables 1 to 4.
Assessment of sample size, surface treatment types, and test types in the included studies
Scientific categories, brand names, and placement techniques of restorative materials used in the included studies
Detailed specifications for surface treatments used in the included studies
Assessment of bond strength testing methodologies in the included studies
Analysis of study design and methodology
After conducting a thorough evaluation of the included studies, the primary parameters examined in the analysis of study design and methodology were as follows: (1) control groups, (2) finishing and polishing (F/P) prior to bonding, (3) type of GI and the composite material used, (4) surface treatment method, (5) aging before testing, and (6) testing methods.
These parameters were consistent across the included studies for both the first and second research questions. However, the second research question also considered an additional parameter: the technique used to age the tested GIs.
Bonding new glass ionomer and resin composite in sandwich restorations
1. Control groups
Fourteen studies included a control group (without the use of the tested mechanical or chemical surface treatments) [10,13,16,18,22–31], while 10 studies did not include a separate control group [14,15,32–39] (Table 5). Nine of the controlled studies did not involve any surface treatment prior to bonding with resin composite [10,13,22,23,25–29,31]. In contrast, the remaining studies included a control group that underwent some form of surface treatment, either using an adhesive [16,24,30] or a phosphoric acid etchant [18].
Risk of bias assessment
2. Finishing and polishing before bonding
Four studies performed F/P on the surface of new GI before bonding procedures [14,26,33,37], while one study performed F/P only on the surface of conventional GI and not on the tested RMGI [32]. Nineteen studies did not perform any F/P on the surface of the GI before bonding [10,13,15,16,18,22–25,27–31,34–36,38,39] (Table 3).
3. Type of glass ionomer and overlying composite
All included studies tested RMGIs except four studies [14,15,28,39]. Nine studies did not test conventional GIC [10,18,22,23,25–27,30,35], while the remaining studies did. Eleven studies included and compared both RMGIs and conventional GICs [13,16,24,29,31–34,36–38]. Only two of the included studies tested glass hybrid [16,36]. Five of the included studies tested nanofilled composite [10,16,22,25,39], while nine studies tested nanohybrid composite [10,18,24,26,27,29,32,33,37]. Seven studies used microhybrid composite [13–15,28,30,36,38], while three studies used silorane-based composite [32,34,37]. One study used packable resin composite [23], and one study tested nanoceramic resin composite [28]. Only one study included bulk-fill composites [10], and one used submicron-filled resin composite [31] (Table 2).
4. Type of surface treatment
Nineteen studies included chemical surface treatment [10,14–16,22,23,25,26–29,33–39], while five studies tested both mechanical and chemical surface treatments [13,18,24,30,31].
Seventeen studies used etch-and-rinse adhesives for chemical bonding between GIs and composites [10,14,16,18,22–27,30,31,33,35,36,38,39]. Similarly, 16 of the included studies used various self-etch adhesives, either two-step, one-step, or two-component self-etch adhesives [10,13,14–16,22,23,26,27,32,33,35–39]. Three studies used silorane-based adhesive [32,34,37], additionally, four studies used universal adhesives with different adhesive strategies [10,16,29,31]. Two studies included a group in which the surface treatment consisted solely of phosphoric acid, without any adhesive application [28,29]. One of the included studies used a universal primer and a coating material as a surface treatment for the tested GIs before bonding to resin composite [16]. Five of the included studies used mechanical surface treatment in combination with chemical treatment [13,18,24,30,31]. The mechanical treatment included laser (Er,Cr:YSGG and Nd-YAG) [18,24,30], air abrasion [13,18,30,31], and Methylene blue photosensitizers activated with photodynamic therapy [18]. These mechanical treatments were used as a replacement for the phosphoric acid etching. It is worth mentioning that none of the studies that used any type of mechanical treatment did any sort of F/P for the specimens before receiving the treatment (Table 3).
5. Aging before testing
Four of the included studies tested the specimens after 48 hours of water storage [16,18,33,35], one study made 1 and 6 months of water storage [14], and four studies performed thermal cycling for 5,000 [10,30], and 500 [28,37] cycles. The rest of the included studies tested the specimens after 24 hours of water storage without any aging (Table 4).
6. Testing methods
Four of the included studies performed micro shear bond strength (μSBS) tests [14,26,36,38], while one study performed a tensile test [25], and one performed a microtensile bond strength test (μTBS) [29]. The rest of the included studies performed SBS tests. The cross-head speed was 0.5 mm/min in 13 of the included studies [10,16,18,22,24,26,27,29,32–36], while 10 studies used a cross-head speed of 1 mm/min [13–15,25,28,30,31,37–39]. Additionally, one study utilized a cross-head speed of 3 mm/min [23] (Table 4). Fifteen studies examined the failure mode after testing using either an optical microscope [13,14,18,30,33], a stereomicroscope [10,16,24–26,28,31,35,37], or a binocular microscope [16]. Two studies did not specify the examination technique [28,32]. The magnifications used differed among the studies and ranged from ×2.5 to ×40 (Table 4).
Bonding aged glass ionomer and either resin composite or glass ionomer in sandwich restorations
1. Control groups
One study included a control group (without the use of the tested chemical or mechanical surface treatments) [40], while four studies did not include a separate control group [12,17,41,42] (Table 5).
2. Finishing and polishing before bonding
Three studies performed F/P on the surface of aged GI before any bonding procedures, one before the aging of the GI [17], and the other two after the aging of the specimens and immediately before bonding [12,42]. Two studies did not perform F/P on the surface of the GI [40,41] (Table 3).
3. Type of glass ionomer and overlying composite
Three of the included studies tested RMGIs [12,41,42], while the other two studies tested glass hybrid [17,40]. Two of the included studies tested nanofilled resin composite [12,42], microfilled resin composite [12], while one used microhybrid [40], and one tested nanoceramic resin composite [17]. Two studies used both GI or RMGI and resin composite as a repair material [12,40], while two studies used only RMGI as a repair material [41,42], and one included only resin composite for repair [17] (Table 2).
4. Technique of aging the tested GI
The aging techniques for the tested GIs varied before the bonding procedures, including water storage [12], thermal cycling [40], both water storage and thermal cycling [41,42], and brushing simulation and thermal cycling [17] (Table 4).
5. Type of surface treatment
Four of the included studies tested chemical surface treatments [12,17,41,42], while only one study included both mechanical and chemical surface treatments [41]. Two studies used etch-and-rinse adhesives for chemical bonding between GIs and composites [12,41], while one of the included studies used a one-step self-etch adhesive [12]. Three of the included studies tested universal adhesives [17,40,42]. Two of the included studies used dentin conditioner [12,40], one used a nanofilled RMGI primer [12], and one study used silane [17] as a chemical surface treatment for the surface of the aged GI. One of the included studies used wet 800-grit silicon carbide paper for roughening the surface of the aged GI, referred to as “sanding,” and considered it a separate mechanical surface treatment [41] (Table 3).
6. Aging before testing
None of the included studies performed any aging before testing (Table 4).
g. Testing methods
One of the included studies performed a μSBS test [17], while two studies performed μTBS tests [40,42]. The other two included studies performed SBS tests. The cross-head speed in the included studies ranged from 0.5 mm/min in one study [17], 0.75 ± 0.25 mm/min [12], and 1 mm/min [40,42]. One study did not specify the cross-head speed used [41]. All of the included studies examined the failure mode using either a light microscope [12], a stereomicroscope [17,40,42], or scanning electron microscopy (SEM) [41]. The magnifications used varied across the studies and ranged from ×2 to ×40 (Table 4).
Analysis of results
Following a thorough evaluation of the studies included, the key parameters examined during the result analysis were as follows: (1) comparison of various types of GIs; (2) comparison of different surface treatment methods, including control, mechanical, and chemical treatments; (3) universal adhesives; (4) assessment of the type of resin composite or repair material used in general, specifically for studies related to the second research question; and (5) failure mode analysis.
Bonding new glass ionomer and resin composite in sandwich restorations
1. Comparing different types of glass ionomers
Regardless of the surface treatment employed, seven studies consistently demonstrated that RMGIs exhibited statistically significantly higher bond strength values when bonded to resin composites compared to conventional GIC [13,16,24,29,31,33,36]. Conversely, three studies reported comparable bond strength values between RMGIs and conventional GICs [32,37], including nanofilled RMGI [34]. Two studies reported statistically significantly higher bond strength for RMGI compared to glass hybrid [16,29]. However, their findings regarding glass hybrid compared to conventional GI were contradictory. One study reported higher bond strength for glass hybrid compared to conventional GICs [16], while the other found no significant difference between them [29]. Two studies found that the type of RMGI used significantly affected the bond strength results [13,38], and another two studies found that the type of conventional GIC also affected the results [14,38].
2. Control vs surface treatment
Seven studies indicated a statistically significant difference in the chemical treatment of the surface of RMGI [10,22,23,26,27,29,31], conventional GIC [13,29], and glass hybrid [29] compared to the control groups, which received no chemical or mechanical treatment. However, one study reported the opposite outcome for RMGI [13]. Furthermore, one study found comparable bond strength values between the control and chemically treated RMGI [25].
3. Effect of combining mechanical and chemical surface treatments
Three studies investigating the bond strength of conventional GIC yielded contradictory results. One study [24] found no statistically significant difference in bond strength when combining mechanical treatment (laser) with etch-and-rinse adhesive, regardless of whether acid etching was used or not. Another study found that combining both treatments (mechanical sandblasting) resulted in lower bond strength compared to using the chemical treatment alone [31]. On the other hand, another study [13] reported a statistically significant difference in bond strength when air abrasion was used as a mechanical treatment before chemical treatment with self-etch adhesive. Furthermore, the use of acid etching enhanced the bond strength to the resin composite.
Three of the included studies compared different mechanical surface treatments (laser [24,30], sandblasting [31], and roughening by bur [30] with chemical treatment alone (etch-and-rinse adhesive and universal adhesive). One study found that laser treatment resulted in significantly higher bond strength than acid etching or the control group [24]. Another study found that sandblasting or bur treatment of RMGI resulted in a statistically significant difference compared to other treatment methods [30]. However, another study reported higher bond strength values for the chemical treatment alone (universal adhesive used in both adhesive strategies) on the surface of RMGI compared to combining chemical and mechanical treatments (sandblasting) [31]. Additionally, one of the included studies found that using Er,Cr:YSGG laser and acid etching before applying the etch-and-rinse adhesive resulted in significantly higher bond strength values compared to photodynamic therapy, Nd-YAG laser, or sandblasting before adhesive application [18].
4. Etch-and-rinse vs self-etch adhesives
Six studies compared the effect of etch-and-rinse vs self-etch adhesives on conventional GIC [14,33,36,38,39], and glass hybrid [16]. Among these, three studies found no significant difference [33,36,38], while the other three reported higher bond strength values for conventional GIC [14,39] and glass hybrid [16] with the use of self-etch adhesive surface treatment.
On the other hand, nine studies evaluated the same comparison on RMGI. Among these, three studies found no significant difference between the two adhesive types [33,36,38], while four studies concluded that the use of self-etch adhesive yielded higher bond strength values [22,23,27,35]. One study found no significant difference when the RMGI was bonded to bulk-fill composites, while the self-etch adhesive performed better when the RMGI was bonded to incremental composite [10]. Additionally, one study reported no significant difference between etch-and-rinse and one-step self-etch, while two-step self-etch showed higher bond strength values [26].
5. The use of phosphoric acid solely or etch-and-rinse adhesive without phosphoric acid
One of the included studies compared the use of phosphoric acid without adhesive application on the surfaces of conventional GIC, RMGI, and glass hybrid materials. The findings revealed that applying only phosphoric acid, as opposed to no surface treatment, led to higher bond strength values exclusively for RMGI. However, when phosphoric acid was combined with a universal adhesive, significantly higher bond strength values were observed for all tested materials compared to acid etching alone [29].
In another study, etch-and-rinse adhesive was used with or without phosphoric acid etchant on the surfaces of conventional GIC and RMGI. The researchers found no significant difference when phosphoric acid was applied before adhesive application for 15 seconds. However, the difference became statistically significant when both the acid and adhesive were applied for 30 seconds or longer [33].
6. Comparing different self-etch adhesives
For the conventional GIC, two studies compared a two-step silorane-based adhesive with a two-step methacrylate-based adhesive [32,37]. The first study found no significant difference in bond strength when compared to the strong self-etch adhesive [32], while the other study found that the mild self-etch adhesive had significantly higher bond strength values [37]. Additionally, one study did not find a significant difference between two-step and one-step adhesives [14]. Three of the included studies found statistically significantly higher bond strength values when using a mild self-etch adhesive compared to intermediate and strong self-etch [15,36,39], and two of them found a significant difference between intermediate and strong self-etch adhesives [15,39].
For RMGI, the same findings observed with conventional GIC were reported in both included studies [32,37]. One study found that a mild two-step self-etch had a statistically significant difference compared to the one-step self-etch [26]. Regardless of the number of steps, one study found that mild self-etch had significantly higher bond strength than intermediate and strong self-etch adhesives [36]. Furthermore, one of the included studies found comparable bond strength values between one-step and two-step intermediate self-etch adhesives [35].
7. Universal adhesive
A study evaluated the application of a universal adhesive on the surface of glass hybrid material using both the etch-and-rinse and self-etch techniques, comparing its performance to that of traditional etch-and-rinse and self-etch adhesives. The findings revealed that the universal adhesive performed similarly in both modes and produced results comparable to the self-etch adhesive. Furthermore, the universal adhesive exhibited significantly higher bond strength values than the etch-and-rinse adhesive [16].
Three studies compared the use of universal adhesive on the surfaces of RMGI [10,31], conventional GIC [31], and glass hybrid [29] in both bonding modes. Two of these studies reported comparable bond strength values when using the universal adhesive in both modes [10,31], which were also comparable to those of the self-etch adhesive [10]. However, the etch-and-rinse mode of the universal adhesive differed from the etch-and-rinse adhesive among the tested resin composite groups. The other study reported higher bond strength values for the etch-and-rinse mode of the universal adhesive tested compared to the self-etch mode of the same adhesive, regardless of the material tested [29].
8. Type of composite
Two separate studies examined the performance of methacrylate-based composites in comparison to silorane-based composites [32,37], but their findings were inconsistent. One study concluded that there was no statistically significant difference between the bond strength of the two materials [32]. In contrast, the other study determined that the methacrylate-based composite exhibited significantly higher bond strength values than its silorane-based counterpart [37].
In another investigation, the performance of nanohybrid and nanofilled bulk fill composites was contrasted with that of nanofilled incremental composite [10]. The outcomes were found to vary based on the type of adhesive used. When both etch-and-rinse and self-etch adhesives were employed, the bond strength values for both bulk fill composites surpassed those of the incremental composite. In the case of the universal adhesive category, the nanohybrid composite demonstrated the highest bond strength values. Furthermore, a separate study compared nanofilled, nanoceramic, and microhybrid composites, uncovering notably superior bond strength values for the nanoceramic composite compared to the other two varieties [28].
9. Effect of aging
Among the studies analyzed, only one study investigated the influence of aging on the bond strength between the examined GI and resin composite [14]. The results revealed that following 6 months of water immersion, the bond strength values decreased compared to the immediate testing conducted after 24 hours. Nevertheless, this discrepancy did not reach statistical significance, except for the specific etch-and-rinse adhesive that was tested.
10. Failure mode analysis
Six studies [13,14,16,18,24,26] reported that the most commonly observed failure mode in the tested GI was cohesive. Two studies [14,16] found no statistically significant differences in failure modes among the tested groups. One study [33] observed that the predominant failure mode in conventional GIC was mixed, whereas in RMGI, it was cohesive. Another study [25] reported that adhesive failure was predominant when RMGI was bonded using an etch-and-rinse adhesive. In contrast, a separate study [35] found that the predominant failure modes were adhesive and mixed for self-etch adhesives, while cohesive failure was more common for RMGI with etch-and-rinse adhesive.
One study [37] reported that the cohesive failure mode was predominant when using a methacrylate-based restorative system, whereas the adhesive failure mode was predominant with a silorane-based restorative system. In another study [10], adhesive failure was observed when no surface treatment was applied between the tested GI and resin composite. However, when a universal adhesive was used in etch-and-rinse mode, cohesive failure was observed across all groups with different surface treatments. Similarly, another study [30] reported varying failure modes depending on the surface treatment: adhesive failure without any surface treatment, cohesive failure with laser treatment, and mixed failure when air abrasion was performed prior to adhesive application. Finally, one study [31] found a higher incidence of cohesive failure compared to adhesive failure, which correlated with increased bond strength.
Meta-analysis results
Owing to the variations in experimental setups, only three studies fulfilled the criteria for inclusion in the meta-analysis [18,24,30]. In one of these selected studies, the data were originally presented in Newton and were later converted to MPa by dividing the values by the area of the bonded surfaces. Essential details such as sample sizes, means, SD, and mean differences with 95% confidence intervals (CIs) for each study are summarized in Table 6. The combined sample size for both the phosphoric acid and Er,Cr:YSGG laser-treated groups amounted to 40 each.
Results of meta-analysis
As depicted in Table 7, a random-effects model was utilized due to the presence of heterogeneity (p < 0.001). The calculated overall mean difference was 1.555 (95% CI, –2.857 to 5.968). This overall mean difference did not yield statistical significance, as indicated by a p-value of 0.490 (significance was set at < 0.05). The outcome of this meta-analysis suggested no substantial variance in the aggregated mean MPa values of RMGI surfaces treated with either phosphoric acid or Er,Cr:YSGG laser before the application of an etch-and-rinse adhesive (Table 6, Figure 2). Notably, the funnel plot and Egger’s test at 95% CIs hinted at potential publication bias (p = 0.043) (Figure 3).
Test for heterogeneity among included studies in the meta-analysis
Forest plots of the meta-analysis. CI, confidence interval; MD, mean difference; SD, standard deviation.
A funnel plot used to evaluate the possibility of publication bias. CI, confidence interval.
Bonding aged glass ionomer and either resin composite or glass ionomer in sandwich restorations
1. Comparing different types of glass ionomers
Only one study examined nanofilled RMGI and RMGI. However, the results for each material were analyzed independently, and no direct comparison between the two materials was made [12].
2. Control vs surface treatment
Among the included studies, only one study compared a control group of glass hybrid being repaired either with glass hybrid or resin composite, along with chemical surface treatment. This study found statistically significantly higher bond strength values in the surface treatment groups [40].
3. Effect of combining mechanical and chemical treatment
One of the studies included in the analysis examined the effectiveness of combining mechanical treatment (sanding) with chemical treatment, specifically acid etching alone or acid etching followed by the application of a bonding agent [41]. The results of this study varied depending on the aging conditions applied to the samples. During the initial 5 minutes of RMGI setting, surface treatment had no impact on the bond strength results. However, when the specimens were immersed in water and subjected to thermal cycling, the group that underwent both mechanical and chemical treatment, including the application of the bonding agent, exhibited the highest bond strength values. Another study reported that combining roughening with a bur before chemical treatment with a universal adhesive demonstrated a statistically significant difference compared to using only the chemical treatment (universal adhesive) alone. However, it showed comparable values to using universal adhesive with prior phosphoric acid etchant [40].
4. The use of phosphoric acid solely or etch-and-rinse adhesive without phosphoric acid
One of the studies included in the analysis observed that applying an acid etchant prior to using either an etch-and-rinse or self-etch adhesive did not produce a statistically significant improvement in bond strength when compared to the use of the adhesives alone for nanofilled RMGI and conventional RMGI, respectively [12]. In contrast, a different study reported that treating the RMGI surface with sanding and acid etching alone significantly reduced bond strength values compared to when an adhesive was applied after these surface treatments [41].
5. Universal adhesives
Three studies utilized a universal adhesive as the chemical surface treatment in their respective research [17,40,42]. In one study, the adhesive was not compared with any other chemical treatment [17]. However, in the second study, the use of the adhesive was compared in two bonding modes: when bonding glass hybrid to resin composite and when applying a repair material of glass hybrid with prior application of polyacrylic acid primer vs using the self-etch mode [40]. The results indicated that utilizing the universal adhesive in the etch-and-rinse mode, along with the prior application of polyacrylic acid on the surface of the aged glass hybrid, resulted in significantly higher bond strength values compared to the self-etch mode of the universal adhesive. The third study did not find a statistically significant difference when using the universal adhesive in either bonding strategy when bonding RMGI to the repair composite [42].
6. Type of repair material
One of the studies included in this analysis compared the use of nanofilled RMGI and RMGI as repair materials for the same material, alongside nanofilled and microfilled resin composites [12]. The findings indicated higher bond strength values for the nanofilled composite compared to the nanofilled RMGI. Furthermore, no statistically significant difference was observed between the RMGI and microfilled resin composite. In another study, statistically significantly higher bond strength values were found for microhybrid resin composite when compared to glass hybrid as a repair material for an aged glass hybrid [40]. Contrary to previous findings, another included study reported higher bond strength values for repairing RMGI using the same material compared to nanofilled resin composite, irrespective of the adhesive mode of the universal adhesive used for bonding to the composite [42].
7. Failure mode analysis
Three of the studies included in the analysis reported adhesive failure as the predominant mode [17,40,42], whereas one study identified cohesive failure as the predominant mode in the aged GI [12]. However, when repairing nanofilled RMGI with the same material, the predominant failure mode was adhesive. Another study reported that premature failure was more frequent when the universal adhesive was used for bonding aged RMGI to the overlying composite in self-etch mode [42].
DISCUSSION
The authors of this review intentionally limited the inclusion of studies to those published within the past decade. This decision was driven by the rapid advancements in dental adhesives and restorative materials. Consequently, earlier research focusing on older generations of restorative materials and adhesives, which may no longer be commercially available, was excluded.
The inclusion of laboratory studies in this review was necessitated by the limited number of clinical studies conducted in the past decade that specifically examined various surface treatments for GIs prior to resin composite restorations. Notably, the primary objective of the available clinical studies was to compare different types of GIs [5,43], rather than to evaluate surface treatments.
Most of the studies included in this analysis utilized SBS testing due to its easier specimen preparation and alignment during measurements [18]. Previous research revealed that SBS testing is considered reliable and less questionable compared to μTBS when examining GIs [44]. However, it has been suggested that SBS evaluation may be less reliable than μTBS evaluation because the adhesive interface in μTBS analysis is relatively small, leading to more uniform stress distribution and better access to the true interfacial bond strength [40]. It should be noted that during sectioning for μTBS testing, there is a risk of premature micro-cracking of samples [44]. In contrast to μTBS testing, the procedure and preparation of samples for μSBS testing are comparatively simpler. Scholars have observed that reducing the sample diameter in μSBS tests can mitigate stress formation and minimize potential errors [44].
Initial observations from the review suggest that the bond strength between GIC and glass hybrid materials with resin composites is relatively lower when compared to the bond strength between RMGI and resin composites. The bonding mechanisms between GIC and resin composite predominantly rely on micromechanical retention, which is influenced by surface irregularities, roughness, and porosities [16]. In contrast, RMGI is thought to involve chemical bonding, resulting in a stronger bond. The incorporation of resin constituents in RMGI has been shown to effectively enhance both bond strength and mechanical properties [13].
Regardless of the type of GI, using surface treatment (either mechanical or chemical) resulted in a stronger bond compared to no treatment. When applying resin composite over various GIC bases, the weaker bond strength between the two materials can be explained by the relatively lower cohesive strength of GIC compared to the resin composite [10]. This difference may also be attributed to the high viscosity of the resin composite, which impedes proper flow on the surface of conventional GIC unless a wetting agent is used [29]. Furthermore, the distinct chemical properties and reactions of these materials are likely to significantly contribute to adhesion failure without surface treatment [10].
The recent examination uncovered contradictory outcomes in the comparison of etch-and-rinse adhesives with self-etch adhesives on GIC surfaces. Certain studies included in the review rejected the use of acid etching and provided the rationale for their findings. They argued that the application of 37% phosphoric acid during acid etching leads to the degradation of the underlying layers of the GIC matrix, thereby diminishing the cohesive strength of the material [14,39]. Consequently, this reduction in cohesive strength can have a detrimental effect on the bond strength between the composite and GIC. It is important to highlight that the porosity created on the GIC surface by phosphoric acid differs from that induced by self-etch adhesives [14,39]. Conversely, other studies that did not observe a significant distinction between the two adhesive types or favored etch-and-rinse adhesives proposed an alternative rationale for their results. They suggested that acid etching of the GIC solely weakens the surface of the cement, leading to cohesive failure within this weakened region rather than affecting the interfacial resin bond strength [33,38].
Regarding the combination of mechanical and chemical surface treatments, the results of the current review did not find evidence supporting a synergistic effect in terms of bonding new GI to resin composite. However, it is possible that this combination could increase the bond strength of aged GIs. One possible explanation for this is that the use of a bur or SiC paper may not only increase the roughness and irregularities on the surface of RMGI, thereby enhancing the micromechanical bond strength [30], but it may also expose glass particles in the ‘old’ material that could react with the polyacrylic acid in the ‘new’ material (in the case of using GI for repair) or unreacted methacrylate monomers that could interact with the overlying adhesive and resin composite (in the case of using resin composite for repair) [12]. However, this may not be critical for newly placed RMGI, as they already contain unreacted glass particles and methacrylate groups on their surface.
The predominant RMGI used in the studies was Fuji II LC, prompting a meta-analysis to examine the impact of surface treatments (mechanical or chemical) on RMGI surfaces before applying an etch-and-rinse adhesive. The findings indicated no significant difference in bond strength to resin composite when comparing phosphoric acid treatment with the Er,Cr:YSGG laser.
The Er,Cr:YSGG laser, with a wavelength of 2790 nm, has a high affinity for water and strong absorption capacity for hydroxyl groups (OH), enabling effective surface modifications. Since RMGI primarily consists of fluoroaluminosilicate glass and water-based polymeric acid, it is hypothesized that the laser-induced micro-irregularities and increased porosity, enhancing composite adhesion [18]. These laser-induced porosities and irregularities may resemble those created by phosphoric acid, explaining the lack of a significant difference between the two techniques. Notably, one of the studies in the meta-analysis used a microhybrid resin composite instead of the nanohybrid composite used in the other two studies. An earlier systematic review demonstrated comparable clinical performance among nanofilled, nanohybrid, and microhybrid composites [45].
While asymmetry in the funnel plot suggests potential publication bias, caution is warranted when interpreting these results due to the limited inclusion of only three studies. The reliability of such tests generally improves with a minimum of ten studies in a meta-analysis [46]. With fewer studies, the statistical power is significantly reduced, making it difficult to distinguish between random variation and true asymmetry in the data [46].
The acidity of the self-etch adhesives tested in the included studies is likely a key factor influencing the outcomes. An increase in the acidity of self-etch adhesives has been associated with reduced bond strength values of resin composites, regardless of the type of GI. One possible explanation for this is that stronger acids lead to greater neutralization of cations and the formation of weaker structures, such as salt crusts, which ultimately result in weaker bonding [15].
Additionally, other studies have suggested that highly acidic self-etching adhesives contain elevated solvent levels to facilitate the full ionization of acidic monomers. As a result, after solvent evaporation, the adhesive layer becomes thinner, potentially leading to insufficient polymerization due to the formation of an air-inhibited layer. This can cause the accumulation of unpolymerized acidic monomers within the layer [36,39].
The results of the current review did not provide conclusive evidence to support a specific bonding protocol when using universal adhesive for bonding GIs to resin composites. However, it is highly probable that a chemical bond is formed between the universal adhesive and GIs through interactions between the functional monomers present in the adhesive and the calcium ions (or strontium ions) found in the GIs matrix [16]. Furthermore, the hydrophilic nature of the functional groups facilitates their penetration into the hydrophilic matrix. Other interactions may occur between the acrylic and itaconic acid, as well as between the silane present in the universal adhesive and GIs [10,16].
In one of the studies reviewed, the decline in bond strength noted following 6 months of water immersion could be associated with the deterioration of GIC caused by the plasticizing effect of water on the resin [14]. Furthermore, persistent by-products from the bonding procedure, like uncured monomers and degradation products arising from resin hydrolysis, could also contribute to the bond weakening [47].
Several studies focusing on failure modes have highlighted the common occurrence of cohesive failure in GIC. This outcome is frequently observed in adhesive assessments of GICs. While some sources suggest that cohesive fracture within the substrate indicates stronger bond strength [48], other research has not found a definitive correlation between fracture mode and SBS value [49].
The tendency for GIC to exhibit cohesive failure is likely due to limitations in its physical characteristics. However, bond failure is a complex event influenced by various factors. These factors may include uneven stress distribution during testing, micro-porosities within the cement that can serve as potential stress points, differences in the setting reactions between materials, and the curing contraction of the resin composite, which could lead to detachment of the GIC from the margins [14].
It is important to note that the majority of the included studies employed macro-bond strength testing, which reported a higher occurrence of cohesive and mixed failures [17,50]. Conversely, in μSBS tests, where the surface area is reduced, force can be applied from a more specific area targeting the entire interface, making adhesive failures more common [44].
During the research phase of the current review, there was a lack of studies aimed at investigating GI repair, particularly concerning the recent types of GIs. Additionally, the included studies exhibited a high degree of heterogeneity in terms of study design, testing parameters, and the use of restorative materials and chemical treatments that were not based on manufacturer instructions. Furthermore, most of the included studies did not subject the samples to any form of aging before testing, raising concerns about the durability of the bonding interface between different GIs and composites. Heintze [50] argues that immediate bond tests do not yield clinically relevant results, nor do they show meaningful outcomes after 24 hours. Instead, such tests should be viewed merely as baseline values, serving as a reference point to assess the reduction in bond strength due to stress and extended storage.
Another point that warrants discussion is that most of the studies did not perform any form of F/P prior to surface treatment for the new GI restoration, while half of the studies included GI repair did perform such procedures. Research indicates that F/P significantly influence the surface properties of GIs [51]. This raises the question of whether the results of the included studies would have been influenced by the pre-surface treatment of F/P, especially considering that two of the included studies examined the effect of using a bur in combination with chemical treatment [30,40].
The review encountered several limitations, including the exclusion of non-English publications and inconsistencies in test materials and methodologies. Additionally, a significant number of the studies lacked an aging process before testing, and some deviated from the manufacturer-recommended material usage. It is suggested that laboratory bond strength assessments should replicate clinically relevant conditions to enhance the applicability of findings. Another notable limitation was the absence of examiner blinding, with only one study explicitly mentioning blinding procedures. While discrepancies in surface treatments among the included studies likely contributed to result variations and limited the number of studies suitable for quantitative analysis, the review attempted to address this heterogeneity by categorizing treatments into specific groups (eg, chemical, mechanical, or combined).
The authors recommend conducting more detailed investigations to determine the most suitable surface treatment for different GIs, particularly the more recent types. This research should adhere to internationally accepted standards to ensure standardization and validation, enabling comparisons across studies. Further studies are needed to explore the use of universal adhesives in both bonding modes, in order to identify the preferred method for bonding with GIs.
CONCLUSIONS AND CLINICAL RECOMMENDATIONS
Regarding the first research question:
1. Evidence suggests that the use of RMGIs is preferable to conventional GICs for sandwich restorations.
2. Surface treatment of the GI is essential, regardless of the type of GI or the chosen surface treatment method.
3. There is no conclusive evidence indicating that combining mechanical and chemical surface treatments produces a synergistic positive effect on the bond strength of RMGIs to resin composites.
4. No conclusive evidence to support the preference for self-etch adhesives, particularly one-step systems, over etch-and-rinse adhesives in terms of bond strength between GI and resin composite. This also applies to the two bonding modes of universal adhesives.
5. The acidity of self-etch adhesives may have a greater effect on bond strength results than the number of steps involved in the self-etch adhesive technique.
6. There is no conclusive evidence indicating that the type of resin composite used affects the results.
Regarding the second research question:
1. Surface treatment is necessary for aged GIs before repair.
2. Mechanical treatment (eg, roughening) prior to applying an adhesive enhances the bond strength. Additionally, using the etch-and-rinse mode of universal adhesive on the surface of aged GIs may improve the bond strength to the repair material in comparison to self-etch modes.
3. Resin composite is recommended for repairing conventional GIC and glass hybrid restorations. However, RMGI can also be used for repairing aged RMGI.
Notes
CONFLICT OF INTEREST
No potential conflict of interest relevant to this article was reported.
FUNDING/SUPPORT
None.
DATA SHARING STATEMENT
The datasets are not publicly available but are available from the corresponding author upon reasonable request.
References
