Effect of bioactive and conventional liners on marginal sealing and dentin bond strength of Class V resin composite restorations: an in vitro study

Abstract

Objective

The objective of this in vitro study was to evaluate the effects of different cavity liners on microleakage and microtensile bond strength (µTBS) between composite restorations and dentin.

Methodology

Seventy-five extracted human third molars were randomly allocated into five groups for microleakage evaluation, eight teeth were assigned to each group, whereas seven teeth per group were used for microtensile bond strength testing. These Class V cavities were created. Meanwhile, seven samples were utilized for µTBS testing. Group 1 (control): Filtek Z250, Group 2: Filtek Supreme Flowable Restorative, Group 3: Glass Liner, Group 4: ACTIVA BioACTIVE Restorative, Group 5: TheraBase Ca. The specimens were exposed to 10,000 thermocycles (5–55 °C) and 50,000 mechanical loading cycles (100 N). For microleakage testing, the specimens were then immersed in a 2% methylene blue solution for precisely 24 h in order to facilitate penetration, and then buccolingual sections were taken and examined with a stereomicroscope. Sections were taken from the teeth and tested for µTBS using a universal testing machine at a crosshead speed of 0.5 mm/min. The data were then analyzed.

Results

Significant differences in microleakage scores were observed among the groups (p = 0.008). The control group demonstrated significantly lower microleakage than the flowable composite liner group, whereas no significant differences were observed among the remaining groups. µTBS values also differed significantly among the groups (p < 0.001). Group 3 (Glass Liner) demonstrated the highest bond strength values, whereas Group 5 (TheraBase Ca) demonstrated the lowest values. Overall, liner type did not appear to substantially influence composite microleakage, whereas dentin bond strength varied among the tested liner materials.

Conclusion

Within the limitations of this in vitro study, liner type did not appear to substantially influence composite microleakage, whereas dentin bond strength varied among the tested liner materials. Further studies are required to confirm these findings.

Graphical Abstract

Introduction

The increasing demand for esthetic dental treatments, together with advances in adhesive technology, has contributed to the widespread use of resin composite restorations in clinical practice [1]. Nevertheless, the occurrence of shrinkage resulting from polymerization is a significant drawback in clinical applications [2]. In instances where polymerization shrinkage stress surpasses the bond strength between the tooth and the restoration, this can result in fluid migration and bacterial infiltration into the spaces between the cavity walls and the restorative material [3]. The phenomenon of microleakage, defined as clinically imperceptible infiltration, has been associated with a variety of adverse outcomes, including postoperative hypersensitivity, marginal leakage, secondary caries, pulp inflammation, and necrosis [4].

A variety of techniques have been developed to mitigate polymerization shrinkage stress. These techniques include altering the chemical composition of the resin formulation, regulating light exposure, employing incremental layering techniques, and applying liners. The reliability of enamel bonding ensures that microleakage is not a significant concern in restorations when the margins are in enamel [5]. Reliable bonding to enamel generally minimizes microleakage when restoration margins are located entirely in enamel. However, when restoration margins extend below the cemento-enamel junction and involve dentin, microleakage and inadequate adhesion remain significant clinical challenges. When the restoration margins terminate on enamel, the bond is known to be effective, and the adhesive interactions with etched enamel contribute to consistent and satisfactory strength [6]. Nonetheless, when the cavity penetrates below the CEJ with the margins in dentin, microleakage, fissure development, and inadequate adhesion to dentin persist as significant problems, as previously mentioned. The inherent composition of dentin, marked by a high organic content, the presence of fluid, and odontoblastic processes, as well as the development of the smear layer, are elements that impede the formation of a robust and stiff bond [7]. The utilization of low-viscosity and low-modulus layering materials, such as glass ionomers, flowable resin liners, or certain dentin-bonding agents, has been proposed as a method to enhance the marginal integrity of composite restorations [8, 9]. The implementation of one of these intermediate layers prior to the placement of a packable resin composite has been shown to result in a substantial reduction in microleakage, thereby addressing a critical challenge to the continued success of direct posterior restorations. The utilization of flowable composites as liners within composites is strongly recommended, a choice that is largely influenced by the materials’ low viscosity, enhanced elasticity, and superior wettability [8]. Glass ionomer cements (RMGICs) are a class of versatile materials that have found application in the field of dentistry, particularly in the domains of restorative, base, or liner procedures. These materials are distinguished by their high degree of biocompatibility, the capacity to release fluoride, robust chemical adhesion to dental tissue, and minimal toxicity [10]. RMGICs utilized in the sandwich approach provide chemical adhesion to dentin, micromechanical adhesion to composite resin, pulp protection, and potential cariostatic properties [11]. ACTIVA BioACTIVE Restorative is composed of an advanced RMGIC, which contains a blend of diurethane monomers modified by the addition of hydrogenated polybutadiene (a synthetic rubber) and methacrylate-based monomers [12]. These materials contain bioactive fillers and have been reported to release and recharge calcium, phosphate, and fluoride ions. These materials have been observed to release and recharge with calcium, phosphate, and a greater concentration of fluoride in comparison to glass ionomers. Furthermore, they have the capacity to establish a chemical bond with teeth, thereby sealing cavities and preventing bacterial microleakage [13]. TheraBase (Bisco, Schaumburg, IL, USA) is a recently developed polymer that functions as a dual-cure and self-adhesive base-liner. TheraBase has been reported to release calcium and fluoride ions and to establish chemical interaction with dental tissues. It possesses radiopaque characteristics, facilitating straightforward assessment on radiographs [14].

No definitive method for eliminating polymerization shrinkage has been described in the literature. Although flowable composites and resin-modified glass ionomer cements have been extensively investigated as cavity liners, limited information is available regarding the performance of recently introduced bioactive liner materials such as ACTIVA BioACTIVE Restorative and TheraBase Ca when used beneath resin composite restorations. In particular, evidence comparing their effects on both marginal sealing and dentin bond strength under thermomechanically aged Class V restorations remains scarce. Therefore, the present study aimed to evaluate and compare the influence of conventional and bioactive liner materials on microleakage and dentin bond strength in Class V composite restorations after artificial aging. The null hypothesis was that different cavity liners would not influence the marginal sealing ability or dentin bond strength of composite restorations.

Methodology

Sample size calculation was performed using G*Power software (Version 3.1.9.6, Heinrich Heine University, Düsseldorf, Germany). Based on the microleakage data reported by Feiz et al. [11], an effect size of 0.84, a significance level of 0.05, and a statistical power of 95% were assumed, resulting in a required sample size of 40 teeth (8 per group). Similarly, based on the microtensile bond strength data reported by Tohidkhah et al. [1], an effect size of 0.839, a significance level of 0.05, and a statistical power of 95% were assumed, resulting in a required sample size of 35 teeth (7 per group).

The study protocol was approved by the Ethics Committee of the Faculty of Medicine at Erciyes University (Approval No. 2023/790, December 2023). 75 undamaged human third molars were collected, and following the meticulous removal of connective tissue, the teeth were thoroughly immersed in a 0.5% chloramine-T solution and subsequently stored at a temperature ranging from 4 °C to 7 °C. The restorative materials utilized in the investigation are presented in Table 1.

Table 1.

Materials of used in this study

Material	Manufacturer	Batch No	Composition
G-Premio Bond	GC, Japan	2307000123	MDP, 4-MET, MEPS, BHT, Acetone dimethacrylate resins, initiators, water
Filtek Z250	3M ESPE, USA	N954930	Matrix: Bis-GMA, Bis-EMA, UDMA, CQ Filler: Zr/Si. 75% wt Average particle size: 0.06 μm
Filtek Supreme Flowable Restorative	3M ESPE, USA	10349091	Matrix: Bis-GMA, TEGDMA, Procrylat resin Filler: Ytterbium trifluoride (YBF3), Silicon dioxide (SiO2), Zirconia 65% wt, 55% vol
Glass Liner	WP Dental, Germany	211330	Qualitative composition: Glass ceramic, glass ionomer powder, silica, camphorquinone, hexanediol dimethacrylate, Bis-GMA, BHT, DMTBA Quantitative composition: Fillers 65%; activators, accelerators and stabilisers 1%, dimethacrylates 34%
ACTIVA BioACTIVE Restorative	Pulpdent, USA	231227	Powder: diurethane dimethacrylate, bis (2-(methacryloyloxy) ethyl) Phosphate, barium glass, ionomer glass, sodium fluoride, colorants Liquid: polyacrylic acid/maleic acid copolymer. Fillers 56% wt.
TheraBase Ca	Bisco, USA	2200002988	Portland cement, ytterbium glass with barium, ytterbium fluoride and Bis-GMA, 10-MDP

BHT butylated hydroxytoluene, DMTBA dimethylaminobenzoic acid derivative, 10-MDP 10-methacryloyloxydecyl dihydrogen phosphate, MDP Methacryloyloxydecyl dihydrogen phosphate, 4-MET 4-methacryloxyethyl trimellitate, MEPS methacryloyloxyalkyl thiophosphate methacrylate, BHT butylated hydroxytoluene, Bis-EMA ethoxylated bisphenol A dimethacrylate, UDMA urethane dimethacrylate, CQ camphorquinone, Zr/Si zirconia/silica, YbF₃ ytterbium trifluoride, SiO₂ silicon dioxide, Bis-GMA bisphenol A-glycidyl methacrylate, TEGDMA triethylene glycol dimethacrylate

Cavity preparation for filling

Class V cavities were created on the buccal surface of the teeth utilizing a diamond bur (Frank Dental GmbH, Gmund, Germany) and a high-speed handpiece with water irrigation. The bur was substituted following the formation of every four cavities. The cavities were measured at uniform intervals, with widths measuring 2 millimeters, depths measuring 1.5 millimeters, and lengths measuring 3 millimeters. These measurements were taken over the cement-enamel junctions. The depth of the cavity was evaluated through the use of a periodontal probe. Subsequently, the teeth were methodically sorted into five distinct groups. The materials were then meticulously placed within the designated cavities, adhering strictly to the manufacturer’s guidelines: Group 1: The enamel was selectively demineralized with 35% phosphoric acid (Ultra-etch, Ultradent, USA). Subsequently, a coating of universal adhesive (G Premio Bond, GC, Japan) was applied to the cavity, followed by a 10-second curing process using an LED light curing unit (Valo Cordless, Ultradent Product, USA) at an intensity of 800 mW/cm² from a standardized distance of 1 mm. The irradiance was consistently assessed using a radiometer (Bluephase Meter II; Ivoclar Vivadent) to verify adequate intensity (800 mW/cm²). A2 Filtek Z250 (3M ESPE, St. Paul, MN, USA), a conventional nano-hybrid composite, was meticulously introduced into the cavity and exposed to light-curing for 20 s (control group). Group 2: The etching and bonding processes were performed in accordance with the protocol established for Group 1. A 1-mm layer of A2 Filtek Supreme Flowable Restorative (3M ESPE, St. Paul, MN, USA) was applied and exposed to light for 20 s for the purpose of curing. Following the application of Filtek Z250 into the remaining cavity, the material was subjected to light curing. Group 3: The etching and bonding processes were performed as in Group 1. A 1-mm layer of RMGIC (WP Dental, Barmstedt, Germany) was applied and light-cured for 20 s. The remaining cavity volume was restored with Filtek Z250 and light-cured according to the manufacturer’s instructions. Group 4: The etching and bonding processes were performed as in Group 1. The cavity was then filled with 1 mm A2 ACTIVA BioACTIVE Restorative (Pulpdent Corporation, Watertown, MA, USA) and light-cured for 20 s. After Filtek Z250 was applied into the rest of cavity and light cured. Group 5: Selective enamel etching was performed as in Group 1. A 1-mm layer of TheraBase Ca (Bisco, Schaumburg, IL, USA) was applied directly onto dentin according to the manufacturer’s instructions and light-cured for 20 s without prior dentin adhesive application. Subsequently, a universal adhesive (G-Premio Bond) was applied over the cured TheraBase Ca surface, and the remaining cavity was restored with Filtek Z250 XT composite resin and light-cured. The restorations were concurrently finished and polished with low-speed fine-grain diamond finishing burs, stones, silicone polishers (Frank Dental, Germany) and enough water.

Ageing process

The specimens were immersed in distilled water for 24 h. All restored teeth underwent thermocycling (ED, JULABO, Seelbach, Germany) for 10,000 cycles between 5 °C and 55 °C, with a dwell period of 30 s and a transfer time of 10 s [4]. The specimens were affixed to a chewing simulator (CS-4; SD Mechatronik, FeldkirchenWesterham, Germany) and underwent 50,000 cycles (100 N and 1.7 Hz) at ambient temperature (23 ± 1 °C) and 100% humidity [4]. A vertical load was exerted using a 3.2-mm stainless-steel spherical stylus at the center of the restorations. The specimens were submerged in distilled water during the aging process. The thermocycling and mechanical loading parameters were selected based on a previously published methodology to simulate clinically relevant thermal and mechanical aging conditions. Although no universally accepted conversion exists, 10,000 thermocycles have been suggested to approximate one year of clinical service, while 50,000 chewing cycles are commonly used to represent an initial period of functional loading in vitro.

Microleakage

Subsequently, two layers of nail polish were applied to the entire specimen, extending 1 mm beyond the edges of the restoration, and the apex was sealed with utility wax. Subsequently, the teeth were immersed in a 2% methylene blue solution at a temperature of 37 °C for a duration of 24 h. Following the removal of the dye, the teeth were meticulously rinsed and thoroughly cleaned with a bristle brush to eliminate surface coloration. Subsequently, the teeth were reintroduced to deionized water before being sectioned. The specimens were meticulously sectioned buccolingually through the restoration’s center using a diamond disk. The teeth were examined using a stereomicroscope (Leica Microsystems, Deerfield, IL, USA) at 40x magnification. The examiner evaluating dye penetration was blinded to the experimental groups. This instrument was connected to a computer for the purpose of evaluating the extent of dye infiltration [15]. 0: no leakage 1: leakage extending to 50% of the cervical wall (mild) 2: leakage involving the entire cervical wall, except the axial wall (moderate) 3: leakage involving the entire cervical wall, including the axial wall (severe).

Microtensile bond strength

A total of 35 extracted third molars were selected for the study, and the occlusal third of the teeth were sectioned using a slow-speed, water-cooled diamond saw (Struers, Detroit Road Westlake, OH, USA) to expose the dentin. The specimens were then divided into five groups, with each group containing seven specimens (n = 7), as previously described. Group 1, Group 2, Group 3, and Group 4: A single layer of G-Premio Bond Universal Adhesive was applied to the dentin specimen in the self-etch mode, in accordance with the manufacturer’s instructions. In the fifth group, TheraBase was applied to the tooth specimens without the use of a bonding agent. To address the absence of material in the dentin specimen, 1 mm of liner materials were applied, and the restorations were completed with 2 mm of composite. The teeth were meticulously embedded in acrylic resin. The dentin specimens were meticulously prepared using a high-speed, water-cooled disk (Struers, Detroit Road Westlake, OH, USA) to ensure precise division into multiple sections. A total of 70 sticks were obtained from each experimental group (approximately 10 sticks per tooth). Pre-testing failures (PTFs) were recorded separately for each group. The restored teeth were sectioned longitudinally and transversely to obtain beam-shaped specimens with dimensions of approximately 6 × 1 × 1 mm³. For µTBS testing, each specimen was attached to the testing jig using cyanoacrylate adhesive and subjected to tensile loading in a universal testing machine (MTD-33 500, SD Mechatronik GmbH, Feldkirchen, Germany) at a crosshead speed of 0.5 mm/min until failure. The microtensile bond strength was determined by dividing the failure force (N) by the sample base area (mm²), and it was represented in megapascals (MPa) (MPa = N/mm²). In instances where a specimen failed prior to undergoing adequate testing, a bond strength of 0 MPa was utilized for statistical analysis [16–18]. The precise count of pre-testing failures was accurately recorded. The failure modes were categorized under a stereomicroscope at 40 × into three groups [19]: Cohesive failure in dentin: The fracture occurs exclusively within the dentin matrix. Cohesive failure in composites: In the event of a fracture that occurs exclusively within the composite matrix. Adhesive failure is defined as the deterioration of the bond between the adhesive and the restorative material at the point of contact. Mixed: The occurrence of both adhesive and cohesive fractures simultaneously.

Statistical method

The data were analyzed with IBM SPSS Statistics for Windows, Version 31.0 (IBM Corp., Armonk, NY, USA). The study examined the compliance with normal distribution using the Shapiro-Wilk test. The Kruskal-Wallis H test was employed to assess variables that deviated from a normal distribution. Subsequently, multiple comparisons were analyzed with Dunn’s test. The Fisher exact test, with the Monte Carlo correction, was employed to examine the relationship between categorical variables. Frequency and percentage were used to represent categorical data. Quantitative data were represented using the mean ± standard deviation and median (minimum-maximum) methods. For µTBS analysis, the tooth was considered the experimental and statistical unit. Although multiple sticks were obtained from each tooth, the bond strength values of sticks originating from the same tooth were averaged to generate a single representative µTBS value. Statistical comparisons were subsequently performed using these tooth-level mean values (n = 7 teeth per group), thereby avoiding pseudo-replication. The total number of sticks and pre-testing failures were reported descriptively. The significance level was established at p < 0.05.

Results

Microleakage results

Table 2 presents the microleakage scores of the respective groups. A statistically significant difference in microleakage scores was observed among the groups (p = 0.008). Pairwise comparisons revealed that Group 1 demonstrated significantly lower microleakage scores than Group 2 (p < 0.05). No statistically significant differences were observed among the remaining groups. Group 1 exhibited the highest score 0 rate (87.5%), whereas Group 2 demonstrated the lowest score 0 rate (12.5%).

Table 2.

Distribution of microleakage scores among the experimental groups after thermomechanical aging

Microleakage scores	Groups					Total n = 40	Test Statistic	P
	Group 1 n = 8	Group 2 n = 8	Group 3 n = 8	Group 4 n = 8	Group 5 n = 8
Score 0	7 (87.5)b	1 (12.5)a	2 (25)ab	4 (50)ab	4 (50)ab	18 (45)	20.929	0.008 x
Score 1	0 (0)a	4 (50)a	5 (62.5)a	0 (0)a	1 (12.5)a	10 (25)
Score 2	1 (12.5)a	1 (12.5)a	1 (12.5)a	3 (37.5)a	3 (37.5)a	9 (22.5)
Score 3	0 (0)a	2 (25)a	0 (0)a	1 (12.5)a	0 (0)a	3 (7.5)

Bold values indicate statistically significant differences (p < 0.05)

^xFisher’s Exact Test with Monte Carlo correction; n(%) a-b There is no difference between groups with the same letter. Score 0 = no leakage; Score 1 = leakage extending to 50% of the cervical wall; Score 2 = leakage involving the entire cervical wall except the axial wall; Score 3 = leakage involving the entire cervical and axial walls.

Microtensile bond strength results

Table 3 summarizes the mean µTBS, standard deviation (SD), median, minimum (min), maximum (max), total number of specimens, and number of pre-testing failures (PTF) for each group. Statistically significant differences were observed among the groups (p < 0.001). Group 3 exhibited the highest µTBS values, whereas Group 5 exhibited the lowest values. No statistically significant difference was observed between Groups 1 and 2 (p > 0.05). Both groups demonstrated higher µTBS values than Group 4. The TheraBase Ca group exhibited 60 pre-testing failures, whereas no pre-testing failures were observed in the remaining groups.

Table 3.

Microtensile bond strength (µTBS) values, pre-testing failures (PTF), and intergroup comparisons

Groups	Mean ± SD	Median (min-max)	Teeth (n)	PTF/Total sticks	Test Statistic	p
Group 1	9.15 ± 2.77	8.66 (5.50–17.50)c	7	0/70	209.005	< 0.001 x
Group 2	9.29 ± 1.52	8.67 (7.16–12.17)c	7	0/70
Group 3	11.46 ± 2.30	11.33 (8–16.34)d	7	0/70
Group 4	7.27 ± 0.92	7.01 (5.83–9.50)b	7	0/70
Group 5	1.25 ± 3.20	0 (0–12.84 )a	7	60/70

Bold values indicate statistically significant differences (p < 0.05)

^xKruskall Wallis H Test; ^a-dGroups sharing the same superscript letter are not significantly different (p > 0.05). PTF: Pre-testing failure. Statistical analyses were based on tooth-level mean µTBS values (7 teeth per group). Multiple sticks were obtained from each tooth (70 sticks per group), and stick counts and PTFs are reported descriptively

Modes of failure

Table 4 presents the distribution of failure modes among the groups. A statistically significant association was observed between liner type and failure mode distribution (p < 0.001). Adhesive failure was the predominant failure mode in Groups 1, 3, 4, and 5, whereas cohesive failure within the restoration was most frequently observed in Group 2.

Table 4.

Distribution of failure modes among the experimental groups

Failure mode	Groups					Total	Test Statistic	p
	Group 1	Group 2	Group 3	Group 4	Group 5
Adhesive	47 (67.1) a	16 (22.9) b	48 (68.6) a	52 (74.3) a	7 (70) a	170 (58.6)	66.637	< 0.001 x
Mix	4 (5.7) c	20 (28.6) a	16 (22.9) a, b	5 (7.1) b, c	2 (20) a, b, c	47 (16.2)
R/Cohesive	17 (24.3) a, b, c	32 (45.7) b	5 (7.1) a, c	12 (17.1) c	1 (10) a, b, c	67 (23.1)
T/Cohesive	2 (2.9) a	2 (2.9) a	1 (1.4) a	1 (1.4) a	0 (0) a	6 (2.1)

Bold values indicate statistically significant differences (p < 0.05)

^xFisher’s Exact Test with Monte Carlo correction; n(%) ^a-c There is no difference between groups with the same letter

A Adhesive failure, C Cohesive failure, M Mixed failure

Discussion

In this in vitro study, we evaluated the effects of different cavity liners, including a flowable composite, resin-modified glass ionomer cement (RMGIC), ACTIVA BioACTIVE Restorative, and TheraBase Ca, on microleakage and the bond strength between composite restorations and dentin. The first null hypothesis was partially rejected because Group 2 exhibited significantly higher microleakage scores than Group 1, whereas no significant differences were observed among the remaining groups. Significant differences were also observed in dentin bond strength among the groups. Group 3 exhibited the highest µTBS values, whereas Group 5 exhibited the lowest values. However, no statistically significant difference was observed between Groups 1 and 2. Therefore, the second null hypothesis was also partially rejected.

The restoration of carious and noncarious cervical lesions remains a considerable challenge in clinical dentistry. One of the main difficulties associated with Class V cavities is achieving adequate isolation because the anatomical characteristics of the cervical region may limit the use of rubber dams and clamps [20]. Another important concern is microleakage, which occurs when the restorative material fails to establish an adequate seal with dentin or cementum, resulting in microscopic gaps at the tooth–restoration interface. These gaps may permit bacterial penetration between the cavity wall and the resin composite, potentially leading to postoperative hypersensitivity, secondary caries, pulpal complications, and ultimately restoration failure [21].

Previous studies have suggested that achieving an effective and durable marginal seal remains challenging when restoration margins are located on dentin or cementum. Packable resin composites, owing to their relatively high viscosity, may not adequately adapt to cavity walls and margins. Consequently, flowable resin-based materials, which exhibit lower viscosity, improved wettability, and greater elasticity, have been recommended as liners beneath packable composite restorations [22]. The use of low-modulus, low-viscosity intermediate materials, such as flowable composites, resin-modified glass ionomer cements, and filled adhesives, may act as stress-absorbing layers, thereby improving cavity adaptation, reducing void formation, and minimizing polymerization shrinkage stress at the tooth–restoration interface [4]. Previous studies have also reported that low-elastic-modulus liner materials may reduce polymerization shrinkage stress and gap formation by functioning as elastic stress-relieving layers beneath composite restorations [23]. Nevertheless, varying degrees of microleakage have been reported for all restorative materials [13]. Therefore, the present study aimed to evaluate the effectiveness of different liner materials in improving marginal sealing and dentin bond strength.

In the present study, the flowable composite liner group exhibited higher microleakage scores than the control group, whereas no significant differences were observed among the other liner materials. These findings suggest that the use of a flowable composite liner did not improve marginal sealing under the conditions evaluated. Flowable composite liners have been recommended as an initial increment at the cervical or gingival floor of restorations because of their low viscosity, improved wettability, and ability to adapt to cavity irregularities. A thin layer of flowable composite has also been reported to improve marginal adaptation and reduce void formation [24].

Previous studies have reported conflicting findings regarding the effect of flowable composite liners on microleakage. Some investigations have demonstrated reduced microleakage when flowable liners were used [7, 15], whereas others found no significant benefit [25, 26]. Similarly, several studies have reported results comparable to those observed in the present investigation and did not support the routine use of flowable composite liners for reducing microleakage [27, 28]. Moorthy et al. likewise reported that flowable composites used as base materials did not reduce cervical microleakage in Class II restorations compared with incrementally placed conventional composites [29].

The polymerization behavior of flowable composites may contribute to these findings. Owing to their lower filler content and higher proportion of resin matrix, flowable composites may exhibit greater volumetric shrinkage than more heavily filled restorative materials [30]. If there is more resin in flowable liners, they may have more shrinkage stress. Although their lower elastic modulus may partially compensate for polymerization stresses, shrinkage-related stresses at the tooth–restoration interface may still contribute to gap formation and microleakage [3, 31]. Previous studies have identified filler content, as well as the composition of the resin matrix, as important determinants of polymerization shrinkage [32, 33]. In particular, TEGDMA has a lower molecular weight and viscosity than Bis-GMA, facilitating filler incorporation but also increasing polymerization shrinkage because of its greater degree of cross-linking [34]. In the present study, the flowable composite liner was the only tested material containing TEGDMA. However, because no interfacial characterization was performed, any explanation relating material composition to the observed microleakage findings should be considered speculative and interpreted with caution.

Previous studies have reported favorable effects of RMGIC liners on marginal sealing. For example, one study demonstrated that the use of a light-cured glass ionomer liner significantly reduced microleakage in packable composite restorations [22]. Similarly, the ability of glass ionomer–based materials to reduce microleakage has been well documented [35]. Feiz et al. reported that resin-modified glass ionomer cement liners reduced microleakage compared with composite restorations placed without a liner [11]. In contrast, the present study did not identify statistically significant differences in microleakage among the RMGIC-based liner groups, including conventional RMGIC, ACTIVA BioACTIVE Restorative, and TheraBase Ca.

Differences among studies may be related to variations in restorative materials, cavity designs, aging protocols, and microleakage assessment methods. Previous investigations have suggested that material composition, elastic modulus, filler content, and polymerization behavior may influence microleakage outcomes [20, 36, 37]. However, in the present study, the RMGIC-based materials demonstrated numerically lower microleakage scores than the flowable composite liner group but did not differ significantly from the control group.

ACTIVA BioACTIVE Restorative has attracted interest because of its bioactive properties, ion release potential, and reported ability to interact with tooth structures [38–40]. Several studies have reported reduced microleakage with ACTIVA BioACTIVE Restorative compared with conventional restorative materials [13, 41]. However, other investigations have found no significant differences between ACTIVA BioACTIVE Restorative and conventional composite resins [42]. A study was conducted to assess the marginal microleakage of enamel and dentin with and without a bonding agent for ACTIVA BioACTIVE Restorative. The findings indicated that ACTIVA BioACTIVE without a bonding agent exhibited considerably greater microleakage at the enamel and dentin margins [1]. Direct comparison with the present findings is difficult because ACTIVA BioACTIVE Restorative was evaluated as a restorative material in those studies, whereas it was used as a liner beneath composite restorations in the present investigation.

To the best of our knowledge, no previous studies have specifically evaluated the effects of ACTIVA BioACTIVE Restorative and TheraBase Ca on microleakage when used as liner materials beneath composite restorations. Therefore, direct comparison with existing literature is limited. Although some liner materials demonstrated numerically lower microleakage scores than others, no statistically significant differences were observed among the liner groups. Consequently, any proposed mechanisms related to material composition, elastic modulus, fluoride release, or bioactivity should be considered speculative and interpreted with caution. The present findings do not provide sufficient evidence to conclude the superiority of any liner material with respect to microleakage performance.

In the present study, Group 3 (RMGIC) exhibited the highest µTBS values, followed by Groups 1 and 2, which did not differ significantly from one another, whereas Group 5 (TheraBase Ca) exhibited the lowest values. In addition, Group 5 demonstrated a markedly higher number of pre-testing failures than the other groups. Previous studies have suggested that bond strengths of approximately 17–20 MPa may be required to withstand polymerization stresses generated by resin composite restorations [43]. The bond strength values observed for TheraBase Ca in the present study were below this range and differed from those reported in some previous investigations [44, 45]. Differences in material composition, testing methodologies, and experimental conditions may partially explain these discrepancies.

A previous study evaluating silicate-based materials used for indirect pulp therapy reported that TheraBase exhibited predominantly cohesive failures and was interpreted as forming a strong bond with composite resin [46]. The authors attributed this finding to the presence of 10-methacryloyloxydecyl dihydrogen phosphate (10-MDP), which may facilitate chemical interaction with dental tissues. In contrast, the lowest µTBS values in the present study were observed in the TheraBase Ca group. One possible explanation is that TheraBase Ca was applied directly onto dentin without adhesive pretreatment, in accordance with the manufacturer’s instructions. Adhesive systems contain functional acidic monomers that interact with dentin, contribute to hybrid layer formation, and may improve interfacial bonding [47]. In addition, bonding agents have been reported to reduce polymerization shrinkage stress and help maintain marginal integrity [48]. Therefore, the lower µTBS values observed in the TheraBase Ca group may reflect both the intrinsic properties of the liner material and the differences in the adhesive application protocol. Consequently, the contribution of each factor cannot be distinguished within the design of the present study, and these findings should be interpreted with caution.

Several studies have evaluated the bond strength performance of ACTIVA BioACTIVE Restorative and RMGIC-based materials. Tohidkhah et al. reported that ACTIVA BioACTIVE Restorative used without a bonding agent demonstrated the lowest µTBS values among the tested restorative materials [1]. Other studies have found no significant differences in shear bond strength among ACTIVA BioACTIVE Restorative, bulk-fill composites, and conventional composites [42], whereas additional investigations reported similar bond strength values for ACTIVA BioACTIVE Restorative and RMGIC materials [49]. Furthermore, a study evaluating various liner materials found lower bond strength values for ACTIVA BioACTIVE Restorative than for RMGIC [50].

A recent systematic review comparing calcium silicate–based liners and RMGIC materials beneath resin composite restorations concluded that RMGIC generally demonstrates superior bond strength performance [51]. This finding has been attributed to its ability to establish chemical interaction with dental tissues, its relatively low elastic modulus, limited setting shrinkage, and stress-absorbing capacity [30, 35, 52, 53]. Nevertheless, factors such as resin composition, filler content, photoinitiator systems, and testing methodology may substantially influence bond strength outcomes [54].

The TheraBase Ca group exhibited a substantially higher number of pre-testing failures (60/70) than the other groups. Pre-testing failures are generally considered indicative of weak interfacial bonding and may therefore provide additional information regarding the integrity of the bonded interface. Although PTFs were assigned a value of 0 MPa in accordance with previous µTBS studies [16–18], the high failure rate in the TheraBase Ca group should be considered when interpreting the bond strength results. Therefore, the lower µTBS values observed in this group likely reflect not only reduced bond strength but also reduced interfacial reliability. Accordingly, the bond strength results of the TheraBase Ca group should be interpreted with particular caution.

A relationship between bond strength and failure mode has been reported in the literature [55]. Previous studies have suggested that higher bond strength values are generally associated with fewer adhesive failures and a greater proportion of mixed or cohesive failures [56]. Cohesive failures are often considered indicative of a stronger bonded interface because failure occurs within one of the substrates rather than at the interface itself [50].

In the present study, adhesive failure was the predominant failure mode in all groups except Group 2, although the differences among groups should be interpreted cautiously. The predominance of adhesive failures is consistent with the µTBS findings and may reflect relatively weaker interfacial bonding compared with the cohesive strength of the restorative materials or dentin. In contrast, Group 2 exhibited the highest proportion of cohesive failures within the restoration, suggesting that the bonded interface may have been stronger than the internal strength of the restorative material in some specimens.

The predominance of adhesive failures observed in most groups is consistent with the µTBS findings. Adhesive failures are generally associated with weaker interfacial bonding, whereas mixed and cohesive failures tend to occur when the bonded interface is stronger than one of the substrates. In the present study, the high frequency of adhesive failures, particularly in the TheraBase Ca group, corresponds with the lower bond strength values and the elevated number of pre-testing failures observed in this group. Together, these findings suggest reduced interfacial reliability in the TheraBase Ca group.

The significance of microleakage studies remains uncertain, as standardization for accuracy has not yet been fully established despite ongoing research in this field. None of the currently available methods has been universally accepted as the gold standard; therefore, further studies employing different assessment techniques are warranted. Furthermore, dye penetration analysis is inherently a semi-quantitative method and may not accurately reproduce the complex conditions encountered in the oral environment. Although it is widely used because of its simplicity, low cost, and reproducibility, the extent of dye penetration does not necessarily correlate with bacterial penetration or the long-term clinical performance of restorations. Consequently, the microleakage findings of the present study should be interpreted within the methodological limitations of dye penetration testing.

Another limitation of the present study is that failure mode assessment was performed solely for classification purposes under stereomicroscopic examination, and representative microscopic images were not available. In addition, although microleakage scoring was performed by a blinded examiner using predefined criteria, formal examiner calibration and intra-examiner reliability assessment were not conducted. Furthermore, no interfacial characterization or scanning electron microscopic (SEM) analysis was performed. Therefore, mechanistic explanations regarding bonding behavior, interfacial integrity, or failure mechanisms should be considered speculative and interpreted with caution.

This study was conducted under in vitro conditions, which may not fully reproduce the biological, thermal, and mechanical complexities of the oral environment. Moreover, only one adhesive strategy and one composite system were evaluated. Different adhesive protocols, restorative materials, and clinical conditions may influence the outcomes. Additionally, because TheraBase Ca was applied directly onto dentin without adhesive pretreatment in accordance with the manufacturer’s instructions, the lower µTBS values observed in this group may be attributable not only to the liner material itself but also to differences in the adhesive application sequence. Consequently, the lower µTBS values observed in this group may represent both material-related and protocol-related effects, and comparisons involving the TheraBase Ca group should therefore be interpreted cautiously.

Future studies employing advanced imaging techniques such as micro-computed tomography, confocal laser scanning microscopy, and SEM-based interfacial analyses may provide a more comprehensive understanding of microleakage, interfacial integrity, and bonding mechanisms. Consequently, the findings of the present study should be extrapolated to clinical situations with appropriate caution.

Conclusion

Within the limitations of this in vitro study, liner type did not appear to significantly influence composite microleakage overall, although the control group demonstrated lower microleakage than the flowable composite liner group. In contrast, dentin bond strength varied among the tested liner materials. The RMGIC liner group exhibited the highest µTBS values, whereas the TheraBase Ca group exhibited the lowest values. However, given the methodological limitations of the study and the limited statistically significant differences observed among the liner groups, these findings should be interpreted with caution and should not be considered definitive evidence of the superiority of any specific liner material. Further studies incorporating advanced interfacial characterization techniques and clinically relevant study designs are needed to validate and expand upon the present findings.

Abbreviations

BHT

Butylated hydroxytoluene

Bis-EMA

Ethoxylated bisphenol A dimethacrylate

Bis-GMA

Bisphenol A-glycidyl methacrylate

CEJ

Cemento-enamel junction

CQ

Camphorquinone

DMTBA

Dimethylaminobenzoic acid derivative

MDP

Methacryloyloxydecyl dihydrogen phosphate

MEPS

Methacryloyloxyalkyl thiophosphate methacrylate

MPa

Megapascals

µTBS

Microtensile bond strength

PTF

Pre-testing failures

RMGICs

Resin-modified glass ionomer cements

SiO₂

Silicon dioxide

TEGDMA

Triethylene glycol dimethacrylate

UDMA

Urethane dimethacrylate

YbF₃

Ytterbium trifluoride

Zr/Si

Zirconia/silica

4-MET

4-methacryloxyethyl trimellitate

10-MDP

10-methacryloyloxydecyl dihydrogen phosphate

Authors’ contributions

AND: investigation, methodology, data curation, formal analysis, writing – original draft; SA: funding acquisition, project administration, supervision, validation, resources, review and editing; NŞK: investigation, methodology, data curation, validation; YÜ: supervision, validation, resources, methodology, review and editing. All authors reviewed and approved the final version of the manuscript.

Funding

This study was supported by Erciyes University Scientific Research Projects (THD-2024-13532).

Data availability

The data that support the findings of this study are available from the corresponding author upon reasonable request.

Declarations

Ethics approval and consent to participate

This study protocol was approved by the Ethics Committee of Erciyes University Faculty of Medicine (Approval No. 2023/790, December 2023). All procedures involving human-derived materials were conducted in accordance with relevant guidelines and regulations and the principles of the Declaration of Helsinki. Written informed consent was obtained from all participants prior to tooth extraction and use of the extracted teeth for research purposes.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.