This in vitro study was carried out in the Department of Conservative dentistry and Endodontics of Dayananda Sagar College of Dental Sciences, Bengaluru, Karnataka, India. Forty intact human maxillary first premolars indicated for extraction for orthodontic or periodontal reason were selected for the study and stored in saline [Table/Fig-1]. Carious, restored, hypoplastic and fractured teeth were avoided. The teeth were mounted with adjacent teeth in an artificial dental jaw model. Standardised conventional mesio-occlusal preparations were made on these teeth using No. 245 bur with an airotor hand piece. The dimensions of these preparations were 2 mm buccolingually and the gingival seat placed at the level of CEJ and 1.5 mm deep axially, verified using a periodontal probe. No. 245 bur was changed after every five preparations. Tofflemire matrix band retainer was adapted to the preparation to prevent gingival overhang of the restoration.

[Table/Fig-1]:

Forty premolars specimen.

Group1: Self-etch bonding agent (Scotch bond universal) was applied with an applicator tip to the entire cavity and light cured for 20 seconds. Cavities were then restored with composite resin and light cured for 30 seconds.

Group 2: About 0.8 mm thick liner of RMGIC was applied on the axial wall. Self etch bonding agent (Scotch bond universal) was applied to the entire cavity including the liner and light cured for 20 seconds followed by restoration with composite resin as in Group 1.

Group 3: About 0.8 mm thick liner of Biodentine (Septodont) was applied on the axial wall. Self etch bonding agent (Scotch bond universal) was applied to the entire cavity including the liner and light cured for 20 seconds followed by restoration with composite resin as in Group 2.

Group 4: About 0.8 mm thick liner of Theracal LC was applied on the axial wall. Self etch bonding agent (Scotch bond universal) was applied to the entire cavity including the liner and light cured for 20 seconds followed by restoration with composite resin as in Group 3.

After the restorations were completed, excess proximal flash was removed with a sharp hand scaler. The apical foramen was sealed by embedding the teeth in acrylic resin blocks and two coats of nail varnish applied to the entire tooth, except for the restoration and 1 mm around it [Table/Fig-2]. The specimens were then subjected to thermocycling at temperatures of 5±1°C and 55±1°C for 1000 cycles with 30 seconds dwell time to simulate oral conditions.

[Table/Fig-2]:

Mounted specimens after insertion in dye.

The specimens were then stored in humidor and immersed in 0.5% aqueous solution of rhodamine B dye for 24 hours. The dye was then rinsed off with water to remove excess dye. The portions of the root embedded in acrylic blocks were severed and the specimens were sectioned mesiodistally through the centre of restoration with a diamond disc under copious water spray [Table/Fig-3]. Buccal half of the tooth section was retained and the lingual section was discarded. The entire tooth section was polished with aluminium oxide paste to obtain high degree of flatness. The tooth restoration interface was observed for the extent of dye penetration using confocal laser scanning microscope (ZEISS LSM) of 10X magnification and scored according to the study done by Tredwin CJ et al., [11].

[Table/Fig-3]:

Sectioned specimens.

Scores (C J TREDWIN)

The scores obtained were statistically analysed by Kruskal-Wallis test. Later to find significant differences between different groups Mann-Whitney was carried out. SPSS version 22.0 software was used for the analysis. Level of significance is set at p-value <0.05.

Higher mean microleakage was recorded in Group 1 (no liner) followed by Group 2 (RMGIC), Group 4 (Theracal LC) and Group 3 (Biodentine) respectively. The difference in mean microleakage among the groups was found to be statistically significant [Table/Fig-4].

In order to find out between which pair of groups there exist a significant difference, Mann-Whitney U test was applied for pair-wise comparisons.

The difference in mean microleakage was found to be statistically significant between Group 1 and Group 2 (p<0.01), Group 1 and Group 3 (p<0.001), Group 1 and Group 4 (p<0.001), Group 2 and Group 3 (p<0.001) as well as between Group 2 and Group 4 (p<0.05). No significant difference was observed between Group 3 and Group 4 (p>0.05) [Table/Fig-5].

Deep carious lesions cause pulpal inflammation (i.e., pulpitis); if not managed, they may result in pulp necrosis and involvement of the periradicular tissues, with possible pain requiring endodontic treatment or extraction [12]. One of the major disadvantage of root canal therapy is the loss of proprioception which can lead to excessive transmission of occlusal load and can ultimately lead to catastrophic fractures. So, it is always preferred to maintain the vitality of teeth whenever possible [13]. The choice of restorative materials for deep dental caries depends on many factors like location, remaining dentin thickness etc.

The increase in demand for tooth coloured restorations along with the concerns regarding mercury toxicity has lead to a dramatic decrease in the use of amalgam restorations. Resin composites represent the materials most commonly used as an alternative to amalgam. It is largely due the aesthetic results, requirement of little to no preparation, acceptable longevity, and relatively low costs. However, it has relatively high surface roughness, low polishability, poor resistance to staining and poor bonding at the tooth restoration interface, which, in combination with polymerisation shrinkage resulted in margin degradation and microleakage [14].

As microleakage is a major drawback in composite restorations, over years researchers have been working on ways or methods to reduce polymerisation shrinkage, hence microleakage. Diverse materials and techniques have been developed to decrease the shrinkage like using non shrinking resins, modifying filler particles, using low elastic modulus liners, various layering techniques and altering the degree of conversion of monomer systems by varying curing source and technique [15].

Another technique used to counteract the contraction stresses is “sandwich technique” where a liner is used between the dentin and the composite restoration. They are of two types-closed sandwich and open sandwich technique. In a closed sandwich, liner is completely covered by restorative material, whereas in open sandwich the liner is exposed to the oral environment [16].

The liners commonly used in deep dental caries with remaining dentin thickness less than 0.5 mm are calcium hydroxide and calcium silicate cements like MTA, Biodentine and Theracal LC [17].

In this study, dye penetration in each sample was scored according to the scores given by Tredwin CJ et al., [11]. In the present study, the depth of gingival seat was 1.5 mm of which 0.8 mm was occupied by liner and 0.7 mm filled with composite (Group 2,3 and 4) and entire 1.5 occupied by composite in Group 1. Tredwin’s Score ‘0’ indicates no dye peneteration, Score ‘1’ indicates dye peneteration upto 1/3^rd gingival seat axially, which in this study is 0.5 mm marginal leakage in the gingival seat. So, leakage is completely beneath composite restoration in Score ‘1’. Similarly, Score ‘2’ indicates dye peneteration 1/3^rd to 2/3^rd gingival study axially which is about 0.3 mm marginal leakage in the gingival seat beneath the liner in this study. Tredwins Score ‘3’ denotes dye peneteration in excess of 2/3^rd gingival seat which means leakage more than 0.3 mm in the gingival seat beneath the liner in our study. Score ‘4’ indicates dye peneteration at the entire gingival seat upto the axial wall which in this study indicates leakage beneath entire 0.8 mm of liner in gingival seat and extending to the axial wall.

On evaluating the results of the study, the highest marginal leakage was seen in composite restoration placed directly on the dentine without any liner. All the specimens showed leakage score ‘4’ [Table/Fig-6]. Thus, the use of liner reduces microleakage in cervically placed deep Class II composite restorations.

[Table/Fig-6]:

Confocal microscopic images of all groups.

Lower scores in other Groups (Group 2,3,4) were attribuited to the presence of a liner. Liners are less rigid and could have a modulus of elasticity 20-30% lower than conventional hybrid composites. Liners overcome polymerisation shrinkage concept of elastic cavity wall [18].

Also, application of liners reduces the volume of composite in the cavity. By decreasing the bulk amount of resin used, volumetric shrinkage will be less which reduces the stress generated as well as microleakage [19].

Liner also prevents the composite resin from bonding with the dentin. The reduction in bonded composite surfaces decreases the configuration factor (C-factor) of the cavity. As a result, the tensile stress generated by the polymerisation contraction of the composite is also reduced [20].

Among the groups RMGIC showed statistically significant higher leakage values as compared to Theracal LC and Biodentine when used as a liner beneath closed sandwich restorations. About 40% of the samples with RMGIC liner showed Score ‘4’, next 40% showed Score ‘3’ and the remaining 20% showed Score ‘2’. So here majority of samples exhibited more leakage beneath RMGIC.

Some researchers state that, RMGIC bonds get distrupted with dentin, mainly in the initial stages of GIC maturation due to contraction forces which occur within polymerising composite resin. So, the polymerisation stress leads to pulling away of RMGIC from dentin and cementum during polymerisation of composite resin layer [21].

Monomers may diffuse through tubules and can have direct, toxic effects on pulpal cells [22]. Resin content in RMGIC also increases polymerisation shrinkage leading to increased risk of microleakage beneath these liners [23].

Stability, setting time is short enough to complete the whole procedure in single appointment unlike other calcium silicate based cements and has good compressive strength to withstand occlusal forces when overlaid by composite resin. Good marginal integrity of sandwich restorations filled with Biodentine and Theracal LC is also likely due to the ability of the calcium silicate materials to form hydroxyapatite crystals at the surface, when formed at the interface between the restorative material and the dentin walls, these crystals may contribute to the sealing efficiency of the material [24]. This explains the reason for lower leakage values seen beneath Biodentine and Theracal LC in Class II closed sandwich restoration in this study. Biodentine performs better even when margins are located in cementum. It forms tag like structures at the interface which is an advantageous property over RMGIC which is sensitive to moisture [25].

Among the calcium silicate cements used in the present study, Biodentine showed less leakage compared to Theracal but the difference was not statistically significant. Nearly, 90% of Biodentine samples showed Score ‘2’ leakage that is less than 0.3 mm microleakage beneath the liner. Remaining 10% showed Score ‘3’. Whereas, in Theracal LC 80% of samples showed Score ‘4’, 10% exhibited Score ‘3’ and the remaining 10% showed Score ‘4’ leakage respectively which means marginal leakage below the entire 0.8 mm liner gingivally and even extending to the axial wall was also seen with Theracal LC [Table/Fig-6]. This might be due to the presence of a resin matrix in Theracal LC which might have modified some of its properties [9]. Polymerisation shrinkage due to its resin content might be a contributary factor as well.

Within the limitations of the present in vitro study, calcium silicate cements proved to be a better option as a protective liner beneath deep Class II closed sandwich restoration.

The major limitation of this study is that it is an in vitro study and the effect of various factors in the oral environment on the liner cannot evaluated.

Theracal LC can be considered as a liquid apatite at your finger tips and because of its ease of handling, command set property and possibility of direct bonding to composite, it is gaining more popularity as a protective liner beneath deep restorations, over other calcium silicate cements [26]. However, long term in vivo studies are required to investigate the clinical implications of this material like the quality of calcific bridge formed, solubility, mechanical strength, its durability as well as the bonding to substrate and overlying restoration.

It can be concluded that application of a liner below a deep Class II composite restoration decreased microleakage. The microleakage scores were significantly lesser in those teeth where Biodentine, Theracal LC were used as a liner as compared to RMGIC. Theracal LC when used as a liner demonstrated similar microleakage scores compared to Biodentine. It can be concluded that Biodentine and Theracal may be considered as the material of choice as a liner in deep Class II cavities requiring pulp capping procedure. However, none of the materials and techniques were able to completely eliminate microleakage from gingival margins of Class II composite restorations.