1. Introduction
The American Association of Endodontists defines a vertical root fracture as a fracture in the root, whereby the fractured segments are incompletely separated; it may occur buccal–lingually or mesial–distally; it may cause an isolated periodontal defect(s) or sinus tract; and it may be radiographically evident [
1]. There have been descriptions of both incomplete and complete root fractures. However, complete root fractures are more common. VRF can also occur in teeth with vital pulp, but it is more common in teeth that have undergone root canal treatment [
2]. VRF typically develops on the buccal–lingual aspect of the tooth and is more common in males and individuals with strong masticatory force and parafunctional habits, as well as those with attrited teeth [
3,
4].
Root fracture is also the third most common reason for tooth loss after periodontal disease and dental caries [
5]. Endodontically treated teeth show a higher incidence of VRF, as reported by Morfis AS et al. [
6] with 3.69%, Touré B et al. [
7] with 13.4%, and Bornstein MM et al. [
8] with 31.8%. This substantial variation in VRF frequency may be attributed to heterogeneity in the evaluation methods and variations in diagnosis and classification. Traumatic dental injuries occur as a result of accidental falls (41.2%), physical assault (23.4%), sports-related incidents, and road traffic accidents, accounting for 7.5% and 4.8%, respectively [
9], and could also lead to root fractures. If the VRF reaches the periodontal ligament and extends into the oral cavity, it allows food particles and bacteria to enter the fractured area. An inflammatory process is initiated, causing the destruction of the periodontal ligament, loss of alveolar bone, and formation of granular tissue [
10]. Radiographic attributes include periodontal ligament widening, vertical bone loss, and peri-radicular bone loss. The ‘classic’ J-shaped radiographic lesion of a long-standing VRF and a ‘halo’ radiolucency involving the furcation region are often observed in VRF presentation [
11]. Fracture origins initiate at the CEJ and extend towards the axial plane, leading to bone loss in the mesial or distal crests. A V-shape bone loss resulted from lateral bone resorption over time, with the root apex serving as the vertex. The shape of the bone loss is V-shaped, with a wider area of loss at the crestal bone and a narrower area extending downwards toward the apex. This occurs when there is a discrepancy between the amount of bone loss in the buccal and lingual bone plates [
12]. It also appears as angular bone loss that extends from the marginal bone to the fracture line [
13] or as a step-like bone defect associated with an oblique extension of the fracture line, without involving the apical portion [
14]. In their respective reports, Floratos et al. [
15] and Agarwal et al. [
16] detailed the utilization of a fractured root hemi-section and a bone graft to treat periodontal bone loss associated with VRF. Dedericet et al. [
17] employed a CO
2 laser to achieve the reunification of separated tooth structures, with acceptable bone levels observed during a one-year follow-up. Fidel et al. [
18] opted for endodontic intervention, post, and core build-up, followed by orthodontic extrusion to manage the VRF case.
Researchers [
19] strongly recommend treatment protocols that aim to preserve the remaining tooth structure by following conservative approaches. Researchers have proposed a conservation technique involving the use of adhesive resin cement to reattach vertically fractured pieces as an alternative approach. According to Unver S et al. [
20], intentional replantation with improved periodontal pocket depth and bone resorption resulted in a positive outcome for reattached teeth. The VRF reattachment treatment protocol showed a 59% success rate after five years. Calcium-silicate cement (Biodentine) has also been proven effective by Hadrossek PH et al. [
21] with 24-month follow-up results. However, its clinical applicability is limited due to its long setting time, discoloration, low biomechanical properties, and cost. Furthermore, it has been proposed that mineral trioxide aggregate (MTA) could effectively retain root fragments [
22]. Cyanoacrylate was used by Oliet et al. [
23] to re-cement the mesiodistal fractured segments extra-orally and replant them. Even though these teeth were mostly functional, their long-term prognosis is uncertain. Barkhordar et al. [
24] successfully regenerated alveolar bone by bonding fractured fragments using glass ionomer cement as a root canal sealer along with calcium hydroxide dressing. Several authors [
25,
26] utilized self-etching dual-cure resin cement to bond teeth with VRF, resulting in asymptomatic teeth with bone regeneration observed during a 2- to 4-year follow-up period.
While several authors have explored different agents for rebonding fractured segments and promoting periodontal and alveolar bone regeneration, the impact of these agents on the fractured resistance of teeth with vertical root fractures requires further evaluation. Furthermore, associated bone defects can alter biomechanics and, in turn, affect fracture resistance. Previous studies [
27] have demonstrated the influence of alveolar bone height on stress distribution in endodontically treated teeth.
Studies on the efficacy of rebonding fractured segments in teeth with vertical root fractures, particularly a comprehensive comparison of fracture resistance using different bonding agents and associated bone loss, are limited. Further research is required to enhance the evidence supporting the effectiveness of reattachment treatment protocols for VRF teeth. The objective of this in vitro study was to assess the impact of various bonding agents and bone defects on the fracture resistance of rebonded endodontically treated teeth with vertical root fractures. The null hypotheses were that the various bonding agents and bone defects would not affect the fracture resistance and type of failures in rebonded VRF teeth.
2. Materials and Methods
One hundred premolar teeth with a single root were collected from the oral surgery department for the purpose of research. These teeth were extracted due to therapeutic reasons such as periodontal or orthodontic treatments. The teeth specimens were chosen based on the specified criteria outlined in
Table 1. The calculus and periodontal tissues were removed using an ultrasonic scaler. They were then placed in a 10% formalin solution for two weeks for disinfection, followed by storage in distilled water at room temperature. The patients consented to the use of their extracted teeth for research purposes, and the research protocol was approved by the ethical review board at the College of Dentistry, King Khalid University in KSA (IRB/KKUCOD/ETH/2022-23/010).
2.1. Preparation of the Samples
The tooth was carefully sectioned 3 mm above the junction of the cementum and enamel using a low-speed diamond disk saw and water coolant (Isomet, Buehler Ltd., Lake Bluff, IL, USA). Root canal treatment was then performed following the established protocol, with final shaping completed using the ProTaper Next X2 file (Dentsply Sirona, Charlotte, NC, USA). The root canals were disinfected with 3% NaOCI in 5 ml quantities, irrigated with saline, and dried with paper points. Gutta-percha master cone and sealer were used to fill the root canals, while a temporary filling (Cavit, 3 M ESPE, Seefeld, Germany) was used to fill the coronal access. The teeth were then stored at 37 °C with 100% humidity for one week. A 2 mm groove was created on both proximal sides of the root using a diamond disc. Plaster forceps were used to divide the root along the tooth’s axis, imitating a vertical root fracture. Ten tooth samples were kept as a control group and not subjected to the fracture procedure. The fractured root halves were conditioned with 10% citric acid for 10 s, rinsed with distilled water, and dried. The teeth samples were randomly divided into three groups of 30 samples each and cemented using cyanoacrylate, dual-cure resin cement (RelyX™ Universal, 3 M ESPE, St. Paul, MN, USA) and resin-modified glass ionomer luting cement (RelyX™ Luting Plus, 3 M ESPE, St. Paul, MN, USA) according to the manufacturer’s instructions.
The tooth samples were placed in an acrylic block made of auto-polymerized polymethyl methacrylate with the help of a dental surveyor. When implanted inside, a 3 mm tooth structure above the acrylic block was kept intact. Before the acrylic resin hardens, a silicone rubber mold in the shape of the vertical bone defect is pressed into the auto-cured acrylic resin alongside the crackline, resulting in the replication of a vertical bone defect. The bone defects of the three groups varied in shape and size. The first group had angular-shaped defects, and the size of the defects was similar on both proximal surfaces (5 mm deep, 3 mm wide). The second group had step-like formations with measurements of 5 mm depth and 3 mm width. The third group had V-shaped defects measuring 5 mm deep and 3 mm wide at the coronal area on one proximal surface (
Figure 1). Based on prior research findings [
28,
29], with an effect size (d) of 1.4, α set at 0.05, and a power of 1 − β at 0.80, the sample size was determined to be 10 per sub-group, resulting in a total of 30 specimens per group (
Figure 2). The calculation for the sample size utilized G* Power software (version 3.1; the University of Dusseldorf) [
30].
The post space was prepared using progressively larger drills including Gates Glidden and a Peeso reamer. The final shaping of the post space was accomplished with a custom post drill, ensuring that a 5 mm apical gutta-percha seal was maintained. The post space was irrigated with 3% sodium hypochlorite followed by a normal saline solution. All samples were prepared with a post space length of 15 mm. The teeth were restored using a conventional FRC post (RelyX Fiber post, 3 M ESPE, Maplewood, MN, USA) and cemented with self-adhesive luting cement (RelyX Unicem, 3 M ESPE, Maplewood, MN, USA) without prior radicular dentin treatment. FRC posts were carefully sectioned to achieve a consistent coronal length of 6 mm above the remaining tooth structure. The hybrid composite material (Clearfil Photocore, Kuraray America Inc., New York, NY, USA) was utilized for the core reconstruction. The core had a consistent coronal height of 6 mm and a six-degree taper. The composite core was standardized using a parallel milling machine (Bravo, Mariotti, Forli, Italy). The height was determined with a digital caliper (Mitutoyo, Tokyo, Japan). Coronal preparation included establishing a consistent 1 mm chamfer finish line and a 2 mm circumferential ferrule on the remaining tooth structure. Full-coverage metal alloy coping was fabricated for the tooth samples using nickel–chromium alloys (Wiron 99, BEGO Bremer, Bremen, Germany) following the conventional dental casting method. Nickel–chromium metal copings were cemented with glass ionomer-type I luting cement (
Figure 3).
2.2. Fracture Resistance Test
All of the samples were subjected to 10,000 cycles of thermocycling in a thermal cycling machine (1100, SD Mechatronik, Pleidelsheim, Germany). The cycles consisted of dwell times of 30 s with temperatures ranging from 5 °C to 55 °C. After that, the samples were exposed to a humid environment for 24 h before testing. Each sample was mounted on a universal testing machine (Instron-5965, Norwood, MA, USA), and a compressive force was applied at 30-degree angulation over the palatal cusps using a 4 mm round tip loading jig at a crosshead speed of 1 mm per minute (
Figure 4). The force at fracture was recorded for each sample. Fractographic analysis using optical microscopy (Hirox, Hackensack, NJ, USA) was conducted on the samples to identify fracture properties [
31]. Failures of the radicular dentin were classified as unfavorable, while cracks or fractures above the cementum–enamel junction (CEJ) were categorized as favorable [
32].
2.3. Statistical Analysis
The statistical analysis was performed using SPSS 22.0 software. Fracture loads obtained from the experiments were examined using a two-way ANOVA with a significance level of 0.05 and a 95% confidence interval, taking into account variables such as the adhesive agent and patterns of bone loss, as well as their combined effect. A Tukey HSD post hoc analysis was then conducted at a significance level of 0.5%.
3. Results
Table 2 shows the results of a study that tested the strength of reattached teeth using different bonding agents and bone defect patterns. The control group, which had no vertical fractures in the teeth samples, had the highest vertical fracture resistance with a value of 1203.66 ± 54.1 N. Among the bonding agents tested, the experimental groups that used resin-modified glass ionomer luting cement (RMGI) showed the highest fracture resistance across all bone defect patterns. The RMGI luting cement groups had a fracture resistance of 1018.43 ± 112.32 N for angular bone defects, followed by 916.82 ± 111.61 N and 663.40 ± 47.50 N for step-shaped and ‘V’-shaped alveolar bone loss patterns, respectively. On the other hand, the dual-cure composite resin cement (DCRC) showed lower resistance values of 629.52 ± 52.43 N, 567.70 ± 45.29 N, and 633.58 ± 37.15 N for the same bone defect patterns (
Table 2). Finally, the cyanoacrylate rebonded teeth samples had the lowest fracture resistance values across all bone defect patterns.
The step bone defect group (
Table 2) showed the lowest tooth fracture resistance in both the DCRC (567.70 N) and cyanoacrylate (484.52 N) groups, while it was a “V”-shaped bone defect in the RMGI group (663.40 N). The two-way analysis of variance results in
Table 3 demonstrated significant differences in fracture resistance based on the bonding agent type (
p = 0.001) and bone loss pattern (
p = 0.000). Moreover, a significant interaction between the bonding agent and bone defect pattern was also observed, with F = 157,554.22,
p = 0.000. A significant difference (
Table 4) in fracture load between different bonding agents was observed across all groups (
p = 0.000).A pairwise comparison (
Table 5) of the various bone loss patterns in relation to mean fracture load revealed a significant difference, except for ‘V’- and step-shaped bone defects (
p = 0.114).
Table 6 shows the fracture modes for different groups. In all groups, both favorable and unfavorable fracture modes were observed. However, teeth samples rebonded with cyanoacrylate exhibited predominantly unfavorable failures in angular and step-shaped bony defects. Conversely, 80% of the control group samples showed favorable failures. Within the experimental groups, teeth samples bonded with RMGI in angular and step-shaped bone defects mostly displayed favorable failure modes.
4. Discussion
Vertical root fracture is a frequent reason for tooth extraction. As a result, dentists frequently encounter VRF, and conservative management options like rebonding the fractured segments are worthy of exploring for their long-term prognosis. While previous researchers have explored agents like MTA and Bio dentine, this in vitro study assessed routinely used luting agents in clinical practice like RMGI, DCRC, and cyanoacrylate agents for regaining fracture resistance. The study also considered the routine clinical presentation of VRF teeth with alveolar bone loss patterns and their impact on fracture resistance. The results of the study indicated that the fracture resistance of rebonded teeth significantly varied between different bonding agents. Additionally, the various bone loss patterns had impacted the fracture resistance of rebonded teeth, leading to the rejection of the null hypothesis that bonding agents and bone loss patterns do not alter the fracture resistance of rebonded VRF teeth.
The premolar teeth without VRF (control group) demonstrated the highest resistance to fracture when compared to all other experimental groups (see
Table 2). Previous research [
33] has confirmed that intact teeth have higher fracture resistance than root canal-treated teeth with different core materials. The fracture resistance of the control group was measured to be 1203.66 ± 54.06 N, which is similar to the results reported by Al-Ibraheemi ZA et al. [
34] and Kaur B et al. [
35]. The decrease in strength can be attributed to the uneven distribution of stress caused by the endodontic post, which does not effectively balance the load along weakened root walls due to root canal preparation [
36]. Additionally, variations in the physical properties of bonding agents and stress concentration at bonded interfaces can also increase the occurrence of fractures in dental restorations involving composites. This happens when distinct materials with varying properties and diverse elastic moduli converge at the adhesive interface layer, which is considered a weaker area prone to debonding. This is acknowledged as a primary cause of restorative failure.
The use of a dental post with high stiffness can increase stress on the root canal walls, which in turn can raise the risk of a fracture near the cement–enamel junction [
37]. This is especially true when forces are applied at an oblique angle, causing tension in the cement layer from the crown to the root–post interface [
38]. Recent research using 3D finite element methods and fatigue analysis in teeth restored with fiber-reinforced composite posts has shown that the failure of coronal composite fillings results in higher dentin stress at the cavity floor, potentially leading to vertical root fractures [
39]. Stress concentrations have been identified at cement–root dentin junctions, which can ultimately lead to root fractures [
40]. According to Maravić et al.’s [
41] findings, von Mises stresses are highest at the base of post-preparation cavities rather than in tooth cervical regions. While posts absorb some stress, they also transfer it to tooth bases like wedges [
42]. Researchers [
43] note that the density and alignment of dentinal tubules play a critical role in crack initiation; tubule density is notably higher in middle and cervical radicular dentin thirds compared to apical sections. The brittle peritubular dentin leads to cracks that initiate more easily from this area, spreading into the intertubular matrix. Consequently, the cervical third is more susceptible to fractures and cracks than the apical bulk.
The lower fracture resistance of VRF rebonded teeth compared to intact control teeth could be due to the presence of additional interfaces for bonding agents between fractured segments (
Table 2). Differences in the modulus of the elasticity of interface materials involving a composite core, FRC post, and bonding agents lead to non-homogeneous stress distributions and reduced fracture resistance [
44]. Teeth with VRF that were bonded with DCRC showed significantly higher fracture resistance compared to the experimental groups that were cemented with cyanoacrylate. Adhesive resin cement proves to be highly effective in establishing superior bond strength and creating a dentin–post–core monobloc system that evenly distributes the applied forces along the root, thus efficiently absorbing excessive loads [
45]. In contrast, cyanoacrylates are weaker and more brittle than resin-based materials, exhibiting unfavorable fatigue and compressive characteristics when compared to resin-based luting cements [
46].
The resin-modified glass ionomer luting cement combines the best qualities of composite resin and conventional glass ionomer luting cement. It is highly biocompatible, has a strong bonding strength, releases fluoride, and has lower solubility, making it a popular choice in clinical practice. VRF teeth bonded with RMGI luting cement showed better fracture resistance in most experimental groups (
Table 2). Marchi et al. [
47] conducted a study on the fracture resistance of weakened roots and found that resin-modified glass ionomer (RMGI) cement was more effective than resin cement. When comparing DCRC with RMGI, the lower fracture resistance of DCRC may be attributed to variations in the amount of load-bearing filler particles present in these materials. A greater number of particles typically results in higher material resistance [
48]. The observed fracture resistance in the RMGI group can be credited to the combination of the basic acid–base curing reaction and an additional light-initiated curing process inherent in the material. This unique dual-curing approach could have enhanced their bonding capabilities and mechanical properties [
49]. Research has shown that resin-modified glass ionomer cement has a lower elastic modulus (35.2–36.8 MPa) compared to dual-cure self-etch resin cement [
50]. However, the elastic modulus of the former is more similar to human dentin, resulting in less stress concentration at the dentin–cement junction, which could be the reason why teeth with resin-modified glass ionomer luting cement tend to perform better. The dual-cure self-adhesive resin cement contains biphenyl-containing dimethacrylate. Its acidic nature makes it easier to infiltrate the dental surface, thereby enhancing micromechanical retention. Additionally, the functional monomers also promote better bond strength and long-term stability. The adhesive resin cement can create a dentin–post–core monobloc system, which helps distribute the applied forces along the root, absorbing excessive loads [
46]. However, the high modulus of elasticity (70.86 MPa) exhibited by this cement was a crucial factor that contributed to the occurrence of minimal elastic deformations and subsequently lower fracture resistance values compared to RMGI [
51].
Shorter root lengths within the alveolar bone can amplify the force exerted on both the alveolar bone and the post tip, increasing the risk of root fracture [
52]. Reinhardt RA [
53] states that a decline in alveolar bone levels can significantly increase stress, concentrating it around the small dentin area surrounding the post apex and making it more prone to fracture, even under lower loads. A study by Ona M et al. [
27] suggests that reduced bone height led to the highest tensile stress at the lingual cervical region, potentially resulting in damage to the periodontal ligament and altered stress distribution. Naumann et al. [
54] studied maxillary central incisors restored with fiber posts and all-ceramic crowns at various levels of bone loss, confirming that teeth with greater bone loss are more prone to fracturing. Moreover, the combination of the tapered anatomy of the root canal and the parallel design of an intracanal post exacerbates stress in the apical region, leading to fracture [
55]. Diana HH et al. [
56] in their study on 3D-FEA noted a significant increase in stress concentration in the apical third of radicular dentin during oblique force application, especially in the area opposite the loading point, attributed to the rigidity of the metal–ceramic crown and tooth bending. The study conducted by Kyogoku K et al. [
57] showed that the width of V-shaped defect did not affect the fracture load of rebonded teeth. However, the presence of ferrule increased the root fracture load. However, few studies [
58] also indicated that there is no significant difference in the fracture resistance of teeth restored with a post and crown, regardless of the level of bone loss.
The study found that teeth samples with DCRC and cyanoacrylate cement consistently showed irreparable fractures (
Table 6). Additionally, teeth samples with larger bone defects, such as V-shaped and step bone loss, exhibited a higher incidence of unfavorable fractures. The findings by Komada et al. [
59] and Ni et al. [
60] confirmed that samples with bone loss are more prone to experiencing unfavorable root fractures than those with favorable bone levels. People with normal occlusion typically have a maximum molar bite force ranging from 465 to 522 N. The results indicate that teeth samples re-bonded with RMGI and DCRC exhibit higher fracture resistance than the expected masticatory force for vertical root fractures. Therefore, these luting cements could be considered an alternative and conservative treatment option for managing VRF-affected teeth.
The study limitations include the fact that the use of nickel–chrome metal coping to restore the teeth samples during testing may have influenced the higher fracture load. Secondly, embedding experimental samples in auto-polymerizing acrylic resin did not account for the cushioning effect of periodontal ligaments and its influence on stress distribution in the results. The investigation did not include the chewing simulator to replicate the aging and assess the impact of masticatory stress on the fracture resistance of rebonded VRF teeth. Further studies are needed to assess how different endodontic posts made from various materials and shapes impact the fracture resistance of rebonded teeth with vertical root fractures in the presence of various bone defects. Further investigations also recommended to evaluate the impact of diverse resin cements and restorations using different types of all-ceramic materials on the fracture resistance of VRF teeth. Additionally, clinical studies are necessary to evaluate the long-term performance of rebonded VRF teeth with different luting agents and bone defects in oral environments.