The study protocol was approved by the Faculty of Science, Mahidol University Animal Care and Use Committee (MUSC-ACUC), Bangkok, Thailand. Thirty-two 8-week-old male Wistar Albino rats (Nomura Siam International, Bangkok, Thailand) weighing 180–220 g were used in this study. The rats received general anesthesia through an intraperitoneal injection of 20 mg/kg of tiletamine hydrochloride and zolazepam (Zoletil 50, Virbac, Bangkok, Thailand) and 5 mg/kg of xylazine hydrochloride (X-Lazine, MacroPhar, Bangkok, Thailand).

Forty right and left maxillary first molars in 20 rats were randomly allocated into 8 experimental groups according to the 1) site of exposure (occlusal or mesial surface), 2) pulp capping material (WMTA [lot no. 0000144254] or Bio-MA [lot no. 01/2017)]), and 3) the observation period (1 or 7 days [n = 5 per group]). Six maxillary first molars served as negative controls (intact teeth with no preparation), while 12 maxillary first molars served as positive controls (pulp exposure without capping material) at the 2 follow-up periods (n = 3 per group). Each rat was fixed on the operating board in a supine position, and the mouth was kept open with a mouth prop. Teeth were cleaned with a micro applicator brush and disinfected with a cotton pellet soaked in 2% chlorhexidine gluconate. Under the aid of magnifying loupes attached with fiber-optic light (×2.5 magnification, Zeiss, Oberkochen, Germany), a sterile 0.8 mm slow-speed round diamond bur was used to prepare a cavity at the mesial half of the occlusal surface or the middle third of the mesial surface, according to the randomization. The depth of the cavity was approximately half the diameter of the bur. Each bur was changed after 2 cavity preparations. During the preparation, the cutting area was irrigated with copious saline solution (0.9% NaCl) to prevent heat generation.

Once the bur had penetrated the dentin, the dental pulp was then exposed with a sterile sharp endodontic explorer (DG16, Dental USA Inc., Mc Henry, IL, USA). Pulpal bleeding was controlled by pressing with a sterile saline-soaked cotton pellet and paper points. The exposure areas were directly capped with one of the pre-sterilized capping materials that was prepared according to the manufacturer's instructions and carried to the exposure sites using the tip of an endodontic explorer. In the contralateral tooth, the pulp was capped with the other pulp capping material. In the positive control groups, a sterile piece of polytetrafluoroethylene (PTFE; Teflon tape) was placed at the exposure site instead of a capping material, to prevent direct contact between pulpal tissue and the coronal restoration. All cavities were restored with light-cured glass-ionomer cement (GC Fuji II LC, GC, Tokyo, Japan), and the cusp tips of occluding teeth were ground to minimize occlusal forces.

At 1 and 7 days after direct pulp capping, the animals were sacrificed by an overdose of carbon dioxide gas. The whole maxilla was collected and fixed in 10% neutral buffered formalin for 24 hours at room temperature, and demineralized in 20% formic acid for 3 days. Then, the maxilla was vertically sectioned into 2 halves, washed in running tap water overnight, dehydrated through ascending concentrations of ethanol, and embedded in paraffin blocks. The 3-µm thick serial sections were cut through the molar area in the mesio-distal direction using a sliding microtome (RM2255, Leica Biosystems, Wetzlar, Germany) to include the entire pulp chamber and the exposed area. The slides were numbered, and the sections were stained with hematoxylin and eosin to evaluate the pulp tissue response and reparative dentin formation. Gram bacterial staining with the modified Brown and Brenn technique was used to confirm the absence or presence of bacteria in the capped pulp.

The stained sections were blindly evaluated by a trained investigator who was previously calibrated with an experienced pathologist. Then, the histopathological evaluation was performed twice at a 2-week interval to control intra-observer reliability (Cohen kappa ≥ 0.8). Under a light microscope (Olympus BX53, Olympus, Tokyo, Japan) at ×100, ×200, and ×400 magnifications, all samples were evaluated and scored in terms of: 1) inflammatory cell infiltration, 2) reparative dentin formation, 3) bacterial penetration, and 4) impaction of dentin fragments into the pulp, according to the criteria modified from Liu et al. [9], as presented in Table 1. In addition, the widest dimension of exposure was measured at ×40 magnification and recorded in micrometers as the pulp exposure size in each sample.

The data were statistically analyzed with the level of significance set at 0.05 (IBM SPSS statistics software version 19, IBM, Armonk, NY, USA). The degree of inflammatory cell infiltration and reparative dentin formation between the two capping materials were compared using the Wilcoxon matched-pair signed rank test. Data from different exposure sites or observation periods were compared using the Mann-Whitney U test. The presence of dentin-fragment impaction was descriptively analyzed. The size of pulpal exposures was analyzed using the independent sample t-test.

Table 2 presents categorizations of the pulpal response, in terms of inflammatory cell infiltration and reparative dentin formation, to direct pulp capping with 2 calcium-silicate cements and positive controls. At 1 and 7 days, the intact teeth (negative control group) showed an intact odontoblastic layer and no inflammation of the pulp tissue. At 1 day, a mild to moderate acute inflammatory pulpal response was observed in the uncapped teeth (positive control group, Figure 1A, 1D, 1G, and 1J). Direct pulp capping with WMTA and Bio-MA showed a similar acute inflammatory response in the pulp tissue. Most pulps with either occlusal (Figure 1B-1C and 1E-1F) or mesial exposure (Figure 1H-1I and 1K-1L) exhibited a mild inflammatory response with a focal accumulation of polymorphonuclear leukocytes (PMNs) and congested blood vessels. Local disruption of the odontoblastic layer was observed at the exposure site in all samples.

At 7 days, a moderate to severe chronic inflammatory response was observed in the positive control groups (Figure 2A, 2D, 2G, and 2J). WMTA and Bio-MA capping on occlusal exposures showed variable pulpal responses, ranging from the absence of inflammation to moderate chronic inflammation (Figure 2B-2C and 2E-2F). WMTA and Bio-MA capping on mesial exposures mostly exhibited no inflammatory response or just a few inflammatory cells (Figure 2H-2I and 2K-2L), although a moderate inflammatory response was observed in a sample of each group.

No significant difference in the inflammatory response was observed between the 2 exposure sites or between the 2 calcium-silicate cements at either 1 day or 7 days (p > 0.05, Figure 3A). The inflammatory responses were also not significantly different between the 2 observation periods (p > 0.05, Figure 3A).

At 1 day, no hard tissue barrier was formed in any samples of the capped or uncapped (positive control) groups (Figure 1).

At 7 days, diffused calcification was observed at a distance below the exposure area in the positive control groups (Figure 2A, 2D, 2G, and 2J). The sites that were capped with WMTA and Bio-MA presented a layer of newly-generated mineralized matrix subjacent to the capping material. Four of the 5 samples with mesial exposure presented a continuous dentin bridge that completely covered the exposure site (Figure 2H-2I and 2K-2L). One sample showed moderate hard tissue deposition that incompletely covered the exposure. In contrast, only one sample with occlusal exposure in each of the WMTA or Bio-MA groups exhibited a continuous dentin bridge. The majority of samples in the WMTA and Bio-MA groups showed incomplete, moderate hard tissue deposition (Figure 2B-2C and 2E-2F). All dentin bridges in the experimental groups were formed with minor defects of cell inclusions and aligned with a layer of odontoblast-like cells underneath. The formation of hard tissue barriers at this observation period was significantly different from that observed at 1 day (p < 0.01). No significant difference was noted in reparative dentin formation between the 2 exposure sites, or between the 2 calcium-silicate cements (p > 0.05, Figure 3B).

At 1 day, all samples in the experimental and positive control groups exhibited no bacteria in the pulp or contained only few bacteria next to the exposure areas. At 7 days, most of the samples were free of bacteria, except 5 samples that contained bacteria in a limited area of coronal pulp below the dentin bridge.

Small dentin fragments were generally observed in most of the experimental and positive control groups in the 1-day samples, regardless of the site of exposure (Figure 1). At 7 days, the impacted dentin fragments were surrounded and merged with the newly mineralized bridge (Figure 2).

The pulp exposure sizes ranged from 207.5 ± 67.5 µm to 261.3 ± 104.4 µm in the occlusal exposure groups, and from 150.9 ± 41.2 µm to 181.2 ± 101.4 µm in the mesial exposure groups. The overall mean values for the occlusal and mesial groups were 234.5 and 166.8 µm, respectively. The difference in the exposure size between the 2 groups was 33.7%. However, there was no significant difference in the size of the pulp exposures among the experimental and positive control groups (p > 0.05, Table 3).