Stress distribution of restorations in external cervical root resorption under occlusal and traumatic loads: a finite element analysis
Article information
Abstract
Objectives
This study analyzed the stress distribution in a maxillary central incisor with external cervical resorptive defect restored with different restorative materials under normal masticatory and traumatic loading conditions using finite element analysis.
Methods
Cone-beam computed tomography of an extracted intact incisor and created resorptive models (Patel’s 3D classification-2Bd and 2Bp) in the maxillary central incisor was performed for finite element models. The 2Bd models were restored either with glass ionomer cement (GIC)/Biodentine (Septodont) or a combination of both with composite resin. 2Bp models were restored externally with a combination technique and internally with root canal treatment. The other model was external restoration with GIC and internal with fiber post. Two masticatory loads were applied at 45˚ to the palatal aspect, and two traumatic loads were applied at 90˚ to the buccal aspect. Maximum von Mises stresses were calculated, and stress distribution patterns were studied.
Results
In 2Bd models, all restorative strategies decreased stress considerably, similar to the control model under all loads. In 2Bp models, the dentin component showed maximum stress at the deepest portion of the resorptive defect, which transfers into the adjacent pulp space. In 2Bp defects, a multilayered restoration externally and root canal treatment internally provides better stress distribution compared to the placement of a fiber post.
Conclusions
Increase in load, proportionally increased von Mises stress, despite the direction or angulation of the load. Multilayered restoration is preferred for 2Bd defects, and using an internal approach of root canal treatment is suggested to restore 2Bp defects.
INTRODUCTION
External cervical resorption (ECR) is a complex process with dynamic pathogenesis occurring in the cervical area of the tooth due to clastic cell activity [1]. Tissue injuries, such as pre-cemental or periodontal damage associated with inflammation, are most likely the reasons for the initiation of ECR [2]. ECR has varied predisposing factors, with orthodontic treatment being considered the most common factor, attributing up to an incidence of 45.7% [3]. Among the other risk factors, trauma and hypoxia play a major role in the initiation and progression of resorptive lesions [4]. The incidence of ECR is most common in maxillary central incisors, accounting for up to 29.24% [5]. Clinical characteristics of ECR range from being minimally restricted to dentin in the cervical region to being invasive and extensive, involving the full length of the roots. Based on two-dimensional and three-dimensional (3D) assessments of depth and extension, various classifications are available [6,7].
The treatment strategy of ECR depends on various factors such as clinical signs and symptoms, the ability to probe the point of entry, and the presence of bone-like tissue at the point of entry. Based on these parameters, three treatment options, such as extraction, monitoring, or a therapeutic restorative approach, are suggested [8]. Restoration of ECR can be performed by an external, internal, or combined approach depending on lesion access and debridement [9]. Over ten-year follow-ups of 274 teeth, the overall survival rate of treated ECR was found to be 84.6% in the first 3 years and gradually reduced to 28.6% in 10 years. This reduction in success rate drastically reduced after 5 years, with vertical root fracture being attributed as the main reason for failure [8]. Occlusal forces (centric and eccentric) could play a major role in the biomechanics of the tooth. As eccentric forces flex the tooth and the highest stress concentration is seen in the cervical region [10], managing resorptive defects involving the cervical region of the tooth becomes challenging for clinicians. Thus, treatment strategies for ECR should focus on materials capable of resisting occlusal load and rehabilitating the tooth by compensating for the lost structural integrity.
Currently, it is observed that 61% of ECR lesions have been treated by external surgical methods using bioactive endodontic cements and glass ionomer cement (GIC) [11]. Heithersay’s class II and class III lesions were restored mostly with bioactive endodontic cements alone (41%) or along with composite resin, while class IV lesions were completely restored with bioactive endodontic cements (57%) following root canal treatment [11]. Rajawat and Kaushik [12] have observed that the stress distribution in smaller resorptive lesions extending into the dentin (1Bd) when restored with mineral trioxide aggregate was better. However, restoring larger and deeper ECR lesions (2Bd and 3Bd) with BioAggregate (Innovative BioCeramix Inc., Vancouver, BC, Canada) and Biodentine (Septodont, Saint-Maur-des-Fossés, France) distributed the stress similar to that observed in intact teeth [12]. Askerbeyli Örs and Küçükkaya Eren [13] have shown that, for resorptive lesions involving pulp with less than 90˚ of circumferential spread, restoring with Biodentine distributes stress better than GIC or composite resin. For resorptive lesions involving pulp and with more than 90˚ of circumferential spread, GIC yielded favorable results [13]. A recent finite element study by Manaktala et al. [14] concluded that for resorptive lesions involving the pulp, irrespective of the location and size of the lesion, restoring with both mineral trioxide aggregate and Biodentine exhibited similar biomechanical performance. It would be interesting to see if restoring the ECR defect with multilayered restorations and the addition of post inside root canal to check if it would distribute the stress and reinforce the tooth.
Finite element simulations of ECR lesions and their restorations under various forces, such as normal occlusal and traumatic loads, are essential. Normal occlusal forces on the anterior could vary in a range between 50 N and 370 N [15], and the traumatic load reported in the literature ranges between 300 N and 2,000 N [16–18]. Thus, the aim of the current study is to understand the stress distribution in a maxillary central incisor with an external cervical resorptive defect restored with different restorative materials under normal masticatory and impact loading conditions using finite element analysis (FEA).
METHODS
Generation of geometric finite element models
Following approval by the Institutional Ethical Committee of Meenakshi Ammal Dental College and Hospital (MADC/IECI/031/2021), five freshly extracted intact, sound, mature human maxillary right central incisors with a regular crown and root morphology were scanned using high-resolution cone-beam computed tomography (CBCT) CS 9600 3D machine (Carestream Dental LLC, Atlanta, GA, USA) at 10 × 10 micron voxel size, 120 kV and 3.20 mA and viewed using software CS 3D Imaging ver. 3.5.18 (Carestream Dental LLC).
ECR defect 2Bd was then created in the teeth using diamond points TF13 and TF15 (Mani Medical India, Pvt. Ltd., Delhi, India) with a high-speed airotor handpiece (NSK; Nakanishi Inc., Kanuma, Japan) under water coolant. The resorption cavity was extended from the mid-labial to distal to mid-palatal region circumferentially (approximately 160˚–170˚), involving the enamel, cementum, and dentin in the cervical third of the crown and root longitudinally, with a depth of 1.2 to 1.5 mm into the dentin, and verified using multiple angled intraoral periapical radiographs. One of the five simulated defects was then subjected to a second CBCT scan using the same device settings as in the first scan. The simulated 2Bd defect was then modified using the same armamentarium by extending its depth by 0.5 mm in its deepest portion to involve the pulp in order to create a 2Bp defect, which was similarly verified, validated, and scanned.
The DICOM (Digital Imaging and Communications in Medicine) files generated by the CBCT scans were converted to STL (stereolithography) files using MIMICS CS ver. 18.0 (Materialise, Leuven, Belgium), which were further utilized to generate 3D models with delineated boundaries of enamel, dentin, pulp, and cementum using Geomagic studio and control version 2014. The models were then imported into SolidWorks (Dassault Systèmes, Cedex, France) for further work. The 3D models of the control tooth, tooth with 2Bd resorptive defect, and tooth with 2Bp resorptive defect are illustrated in Figure 1A–C.
Virtually created three-dimensional (3D) model of an intact tooth (control) (A), 2Bd unrestored tooth (B), 2Bp unrestored tooth (C), representation of the 3D meshed model (D), and loading conditions (E).
Experimental models with simulation of various restorative strategies
The 2Bd models were restored using external approaches as follows: (a) completely restored with GIC (R1) (Figure 2A), (b) completely restored with Biodentine (R2) (Figure 2B), and (c) layered restorations of Biodentine and GIC subgingivally and Biodentine and composite supragingivally (R3) (Figure 2C).
Virtually created three-dimensional models of five restoration types based on material combinations. (A) R1: restoration with glass ionomer cement (GIC) alone. (B) R2: restoration with Biodentine (Septodont, Saint-Maur-des-Fossés, France) alone. (C) R3: restoration with Biodentine, GIC, and composite resin. (D) R4: internal restoration with gutta-percha and composite resin, and external restoration with Biodentine, GIC, and composite resin. (E) R5: internal restoration with fiber-reinforced composite (Everstick post, GC Corp., Tokyo, Japan), gutta-percha, and composite resin, and external restoration with GIC.
Combined therapeutic options for the 2Bp models were simulated as follows: the internal approach—endodontic treatment with incisal access was simulated corresponding to a no. 7 round bur. A conical canal preparation was simulated with an apical diameter of #50 with a 2% taper. This simulation was further restored as follows: (a) internally obturated with gutta-percha and externally restored with Biodentine and GIC subgingivally and Biodentine and resin composite supragingivally (R4) (Figure 2D), and (b) internally obturated with 5-mm gutta-percha (GP) and fiber-reinforced composite (FRC) post placed in the canal coronally and externally restored with GIC completely (R5) (Figure 2E).
Finite element meshing details
The 3D meshes were generated with 10-node tetrahedral elements and quadratic displacement shape functions with three degrees of freedom per node (Figure 1D). The average mesh size was about 0.5 mm, and the number of elements and nodes of each model is tabulated (Table 1).
Number of elements and nodes for control, unrestored, and restored models
Properties of the materials used
The generated models were transferred to Ansys software (Ansys 2022, R1 student version; Ansys, Inc., Canonsburg, PA, USA) for subsequent stress analysis. All tissues were presumed to be linearly elastic, homogeneous, and isotropic, with the bonding between the tissues considered ideal. The needed properties are presented in Table 2 [19–24].
Properties of materials used in finite element models
Boundary and loading conditions
A thickness of 0.25 mm of periodontal ligament, 2 mm each of cancellous and cortical bone were simulated around all the models, starting 1.5 mm apical to the cementoenamel junction and uniformly wrapped around the root [25,26], except for at the site of resorptive defect. Under all loading conditions, it was assumed that the simulated cortical bone was fixed.
Four loading conditions were applied: F1 and F2, masticatory load of 50 N [15,27] and 100 N [28,29], respectively, applied at 45° to the long axis of the tooth, palatally and incisal to the cingulum; F3 and F4, impact forces of 200 N and 300 N [16,17], respectively, applied to the middle third of the buccal surface of the crown at 90˚ to the long axis of the tooth (Figure 1E).
Von Mises stress evaluations were carried out using Ansys software, and the highest equivalent stresses in the entire tooth structure and resorptive areas for each model were visualized and noted from the color scale.
RESULTS
The number of elements and nodes for control, unrestored, and restored models is presented in Table 1. The overall maximum stress was always found to be in enamel at the point of force application. Von Mises stresses of control, unrestored, and restored models for 2Bd and 2Bp are presented in Tables 3 and 4, respectively.
Maximum von Mises stresses in Patel’s three-dimensional classification 2Bd models
Maximum von Mises stresses in Patel’s three-dimensional classification 2Bp models
2Bd models
In 2Bd models, on application of forces (F1, F2, F3, and F4), the highest stress was observed in experimental models with resorptive defects without any restoration. The dentin component showed maximum stress at the deepest portion of the resorptive defect. On F1 and F2 applications, the stress gets distributed to the mid root region on the buccal side and the buccal and palatal sides for F3 and F4. In the dentin component, R1, R2, and R3 have been found to reduce stresses equally and considerably similar to control under F1, F2, F3, and F4. The maximum stress concentration gets transferred to the mid root region on the buccal aspect on application of F1, F2, and to the mid root region on both buccal and palatal aspects, with more concentration on the palatal aspect with F3 and F4. The stresses exponentially increase from F1 to F2 (2 times) and from F3 to F4 (1.5 times) in all models. The stresses increase by 1.9 times from F2 to F3 in all models. With respect to dentin, the control model, unrestored model, and restored models show an increase of 5.4, 3.5, and 4.5 times, respectively. The maximum stresses of the periodontal ligament and cementum components of 2Bd unrestored were observed to be adjacent to the deepest part of the resorptive site. In cementum, von Mises stress after restoration in 2Bd models (R1 and R3) under F1, F2, F3, and F4 was closer to the range of stress in the intact tooth model. However, in R2, the stress was greater than the unrestored defect model. Within the restorations, Biodentine takes up the maximum stress concentration compared to GIC and composite. Multilayered restorations are attributed to more stress concentration compared to single restorations. The distribution of von Mises stresses in dentin and cementum in the 2Bd models is given in Figure 3.
2Bp models
In 2Bp models, on application of F1 and F2, the highest stress was observed in the control model. On application of F2, the maximum stress concentrations observed for R4 and R5 were greater than the unrestored model. On application of F3 and F4, the maximum stress was observed in the unrestored model. In the 2Bp unrestored model, the dentin component showed maximum stress at the deepest portion of the resorptive defect, which gets transferred into the pulp space adjacent to it. In the dentin component, R4 and R5 have been found to reduce stresses equally and considerably similar to control under F1, F2, F3, and F4. Also, the maximum stress concentration gets transferred to the mid root region on the buccal aspect upon application of F1 and F2, and to the mid root region on both buccal and palatal aspects, with more concentration on the palatal aspect on application of F3 and F4. The stresses increase by 1.8, 3.1, and 2.9 times from F2 to F3 in control, unrestored, and restored models, respectively. With respect to dentin, the control model, unrestored model, and restored models show an increase of 5.4, 2.6, and 3.5 times, respectively. The maximum stress of the cementum component of 2Bp unrestored was observed to be adjacent to the deepest part of the resorptive defect extending buccally on the application of F1 and F2, and palatally on the application of F3 and F4. The maximum stress of the periodontal ligament component was found to be at the apex on the application of all four loads. In 2Bp models, under F1 and F2, von Mises stress was closer to the intact tooth model, while it got reduced below it under F3 and F4. Considering the stress on restorations exclusively, R4 has more stress concentration than R5. The external part that is restored with multilayered restorations is subject to more stress concentration than single-layer restorations, as observed in 2Bd models. On the contrary, R5 has more stress concentration than R4 when the internal part is considered in particular, because FRC takes up more stress. The distribution of von Mises stresses in dentin and cementum in the 2Bd models is given in Figure 4.
DISCUSSION
The strategies in the management of ECR have evolved over the years. However, the loss of structural integrity of the tooth due to chemical, mechanical, and physical reasons could result in a massive drop in success and survival rates after 5 years [8,30–33]. In previous endodontic literature, the impact of traumatic loads on ECR has not been studied. It is observed that the incidence of recurrent trauma ranges from 8% to 45%, and patients under 9 years of age have an eightfold increased risk of re-trauma in comparison to 12-year-olds [34]. Thus, in our study, other than the two normal occlusal loads, two traumatic loads of 200 N and 300 N were chosen. Unlike normal occlusal load, a specific force cannot be attributed to traumatic clinical situations [15,27–29]. Thus, a traumatic load of 200 N was chosen as double the normal masticatory load, and 300 N was chosen based on previous literature [16,17].
Von Mises stress in dentin in the 2Bd unrestored model was 1.8 and 1.2 times more as compared to that of the respective control models under masticatory and traumatic loads. Amongst the 2Bd models, all the restorative strategies resulted in von Mises stress closer to the range of stress in the intact tooth model in the dentin component under masticatory load. It was interesting to note that after being restored, the stress values decreased below the intact tooth model under traumatic load in R1, R2, and R3. In our study, 2Bd models that were completely restored with GIC or Biodentine showed similar von Mises stresses to those of the normal dentin. This was in concurrence with the previous literature where a 2Bd defect in the buccal aspect was restored with Biodentine and Bioaggregate [12,13]. Biodentine and GIC are considered excellent dentin replacement materials due to properties such as tensile stress, modulus of elasticity, Poisson’s ratio, and resilience being similar to that of dentin [35,36]. These properties could aid in the uniform distribution of stresses in R1 and R2 in a similar pattern to that of the intact tooth. A recent systematic review has shown that bioactive endodontic cements exclusively are preferred for restoring ECR in approximately 32% of cases [11]. However, a combination of bioactive cements with composites is preferred in 42% of cases [11]. Low wear resistance and non-esthetic properties of these materials restrict the placement to the crestal bone level and layering composite over them in a clinical situation [37,38]. Thus, a combination of various materials (R3) has been considered in this study. R3 has also been shown to reduce von Mises stress similar to normal dentin and hence will be preferred over R1 and R2.
Von Mises stress in dentin in the 2Bp unrestored model was 3.7 and 1.8 times more as compared to the respective control models under masticatory and traumatic loads. It is interesting to note that the masticatory loads showed more stresses than traumatic loads, and this could possibly be because of the change in the angle of load application and the involvement of pulp. In comparison to 2Bd models, when the resorptive lesion involves the pulp, von Mises stresses increase by approximately 2 and 1.5 times under both normal and traumatic load. A previous study has compared various treatment options in multiple resorptive defects depicting dentin and pulpal involvement [13]. However, the increase in stress in pulpal models is not mentioned elaborately.
In our study, amongst the 2Bp models, internal approaches with or without FRC post resulted in von Mises stress closer to the range of stress in the intact tooth model in the dentin component under masticatory load. In the recent systematic review, it was observed that nearly 34% of ECRs were treated with root canal treatment (internal approach) and only 5% of cases were managed with a combined approach [11]. It was interesting to note that after being restored, the stress values decreased below the intact tooth model under traumatic load in R4 and R5. The use of GIC externally and GP internally has generated the lowest von Mises stress for the 2Bp models in this study, which is similar to that of a previous study [13]. This is the first study to evaluate the role of placing an FRC post and the stress distribution caused thereby in teeth with ECR. It is observed that the placement of an FRC post does not improve the stress distribution when restoring an ECR lesion. Thus, the R4 therapeutic option is preferable as it will reduce multiple interfaces and thereby the bonding difficulties inside the root canal.
In all restorative strategies, the stress gets transferred from the defect to the buccal aspect extending from the middle to the apical third in the root under 50 N and 100 N. These results are similar to the previous FEA studies on ECR [12,13]. In one study, the stress distribution in the root dentin was on the palatal aspect as the defect was on the palatal aspect [14]. In our study under 200 N and 300 N, the stress gets transferred to the cervical to the middle 3rd on the palatal aspect and the middle 3rd on the buccal aspect. This might be due to an increase in the force and the point of application (90˚ to the long axis of the tooth).
In our study, Biodentine, which has a similar Young’s modulus (22 GPa) [23] to that of dentin (18.6 GPa) [19], was found to have decreased the stress towards GIC and composite resin. Thus, future research can aim at constructing models with undermining resorptive lesions as well as increasing the thickness of Biodentine in combination with restorations.
The strengths of the current study are the non-invasive, standardized methodology possible with FEA, and that two classes of ECR being repaired with different restorative strategies, including a multilayered restoration for 2Bd models and the combination of GP with FRC post for 2Bp models under a wide range of loading conditions, were evaluated.
However, the limitations include that only a single representative model of each of the 2Bd and 2Bp classes was studied, though clinically, many such patterns of resorption can occur. Secondly, the resorptive defect was created with drills, which might not be as precise as the clinical situation. This was performed to standardize the models (2Bd and 2Bp) at baseline in all dimensions with the exception of pulpal involvement. It is also not very definite if the influence of the surrounding tissue architecture would influence the restorations in the real resorptive defects, as this cannot be replicated in this scenario. The assumption that all tissues and materials were homogenous, isotropic, and elastic, and all surfaces were considered to be ideally bonded, is an inherent disadvantage of FEA, which should be considered while clinically translating the results.
CONCLUSIONS
Within the limitations of the study, it can be concluded that restoring with a single restorative material or multiple layers decreases the stress concentration; multilayered restorations are preferred in 2Bd defects. Placement of the FRC post does not add a significant difference in stress distribution compared to GP alone, and therefore, the use of GP alone is suggested for the internal approach restoration of 2Bp defects.
Notes
CONFLICT OF INTEREST
No potential conflict of interest relevant to this article was reported.
FUNDING/SUPPORT
A research seed money grant was received as part of funding through the Meenakshi Academy of Higher Education and Research for the project. The authors would like to thank this support.
AUTHOR CONTRIBUTIONS
Conceptualization: All authors. Data curation, Resources: Ramanujam P, Karthikeyan PVA. Formal analysis, Investigation, Methodology, Visualization: Ramanujam P, Karthikeyan PVA, Suresh N, Srinivasan V, Ulaganathan S. Project administration, Supervision, Validation: Suresh N, Srinivasan V, Ulaganathan S, Natanasabapathy V. Software: Karthikeyan PVA, Ulaganathan S. Writing - original draft: Suresh N, Karthikeyan PVA. Writing - review & editing: Suresh N, Srinivasan V, Ulaganathan S, Natanasabapathy V. All authors read and approved the final manuscript.
DATA SHARING STATEMENT
The datasets are not publicly available but are available from the corresponding author upon reasonable request.
References
