Effect of attachment wear on maxillary canine rotation in clear aligner therapy: a three-dimensional finite element analysis

Abstract

Background

This study aimed to explore the effect of attachment wear on maxillary canine rotation in clear aligner therapy through three-dimensional finite element analysis.

Methods

Finite element models, comprising the maxillary bone, periodontal ligament, upper dentition, attachments (initial and worn), and clear aligners, were created to simulate maxillary canine rotation. Two attachment types (vertical rectangular and rotation-optimized) with five wear levels (M0, M2, M4, M6, and M8) were considered for canine rotation analysis. Tooth displacement and equivalent stress in the roots and periodontal ligaments were analyzed.

Results

With increased attachment wear, both groups showed reduced canine rotation angles and lower stress levels in roots and periodontal ligaments. The most significant reduction in the canine rotation angles and stress values for the rectangular group occurred between M2 and M4, with the efficiency of canine rotation falling below 50% from M4 onward. In contrast, the optimized group experienced the most substantial decline between M6 and M8, with the efficiency of canine rotation falling below 50% at M8.

Conclusions

Attachment wear decreased maxillary canine rotation efficiency. It is preferable to use optimized rotation attachments when correcting maxillary canine rotation. After six months of optimized attachment use, if maxillary canines still require rotation, orthodontists should closely monitor attachment wear to ensure treatment effectiveness. Immediate restoration or re-bonding of the attachments is required if it is determined to be essential.

Background

Over the past two decades, clear aligners (CAs), as removable orthodontic appliances, have been used to treat nearly 20 million patients worldwide [1]. Owing to their aesthetic appeal, comfort, and ease of maintaining oral hygiene, CAs are favored by orthodontists and patients [2, 3]. However, unlike traditional fixed orthodontic appliances, CAs, as an emerging technology, still pose challenges in regard to achieving complex tooth movement [4]. A clinical trial found Invisalign’s average accuracy for all tooth movements was just 50%, with rotation showing the lowest precision (46%) [5]. Several studies have shown that for rotational movements, the predictability is lowest for rounded shaped teeth, with the maxillary canines being the most affected teeth [5, 6]. Therefore, attachments are often required to enhance the success rate of these movements and prevent aligner off-tracking. Attachments are composite resin structures bonded to the tooth surface, enhancing aligner retention and facilitating tooth movement [7]. To achieve precise tooth movement, close-fitting between the aligners and teeth is crucial, while attachment wear may affect movement efficiency [8]. Some scholars have conducted morphological observations of attachment wear, but a thorough investigation into the patterns of wear and their effect on tooth movement is lacking [9, 10]. Given the limited research on this topic, we carried out a series of in vivo and in vitro studies to explore this issue [11, 12].

Three-dimensional (3D) finite element analysis (FEA) is an effective engineering technique used to calculate the stress and deformation of geometric entities when subjected to external forces [13]. Because of its noninvasive nature and ability to simulate diverse scenarios, FEA is now widely applied in biomechanical research within the field of orthodontics [14, 15].

Therefore, building on our preliminary research, we conducted an FE study to investigate the effect of attachment wear on maxillary canine rotation in CA therapy. This study aims to provide a theoretical basis for attachment design and adjustment, ultimately improving the efficacy of CA treatment.

Materials and methods

We recruited a healthy adult volunteer with individual normal occlusion to undergo cone beam computed tomography (CBCT) imaging using a NewTom VGi scanner (Italy). The research protocol received full ethical approval from the Research Ethics Committee of Beijing Stomatological Hospital, Capital Medical University, before data collection (CMUSH-IRB-KJ-PJ-2022–25). The informed consent was obtained from the volunteer. The inclusion criteria for the volunteer were described in our previous article [12].

Establishment of 3D finite element models

On the basis of the CBCT data, 3D models of the maxillary bone and upper dentition were reconstructed with Mimics (v21.0, Materialise, Belgium) and Geomagic Studio 2014 (3D Systems, USA). Geometric models of periodontal ligaments (PDLs), attachments, and CAs were created using Pro/E (v5.0, Parametric, USA). The PDL was set as 0.20 mm [16]. The right maxillary canine was set to have a mesial rotation of 2° along its long axis. The CA models were generated by offsetting the upper dentition and initial attachments externally by 0.75 mm [17]. Components were built as 3D FE models in Hypermesh (v14.0, Altair, USA) using unstructured 4-noded tetrahedral elements. Table 1 displays the elements and nodes for each component. The assembly models were imported into Abaqus (v6.14, Dassault Systèmes, France) (Fig. 1A). The modeling process was described in detail in the previous FE study [12].

Table 1.

Elements and nodes of the finite element models

Component	Rectangular group		Optimized group
	Elements	Nodes	Elements	Nodes
Maxillary bone	358,873	74,341	358,873	74,341
Teeth	70,856	20,656	70,856	20,656
PDL	61,621	19,995	61,621	19,995
Attachment
M0	260	114	243	109
M2	1453	492	1231	429
M4	1947	628	1667	556
M6	1705	559	440	183
M8	585	223	545	209
Clear aligner	33,682	10,682	58,887	18,225

PDL Periodontal ligament

Fig. 1.

FE models and attachment design. A All components of the FE models. B Attachment FE models across various wear conditions

Two groups of models were established using the above method. In the rectangular group, the horizontal rectangular attachment was placed on the right maxillary canine. The initial size was set as 3 × 2 × 1 mm. In the optimized group, the optimized rotation attachment was placed on the same tooth. The initial geometry and positioning of the optimized rotation attachments were planned according to Clincheck (Align Tech, USA). In the prior eight-month clinical trial, we employed intraoral scanning and 3D superimposition techniques to determine the primary locations and degrees of wear for commonly used attachments in clinical settings [11]. For both groups, five attachment wear levels were modeled based on prior clinical study results. The M0, M2, M4, M6, and M8 Conditions reflect attachment wear levels and positions in patients using CAs after 0, 2, 4, 6, and 8 months. (Fig. 1B).

The global coordinate was established on the maxillary bone, with the x-axis aligned along the sagittal plane (positive toward the distal side), the y-axis along the coronal plane (positive toward the lingual side), and the z-axis vertically oriented (positive toward the crown).

Material properties and boundary conditions

Material properties were assigned to each component (Table 2) [18–20]. The maxillary bone, teeth, PDL, attachments, and CAs were set up as continuous, homogeneous, and isotropic linear elastic materials. The superior region of the maxillary bone served as a fixed constraint. The maxillary bone-PDL, PDL-teeth, and teeth-attachments contacts were bonded [21]. Frictional contacts were set between the CAs and teeth and between the CAs and attachments, with a friction coefficient of 0.2 [22].

Table 2.

Material properties

Component	Young’s modulus (MPa)	Poisson’s ratio
Maxillary bone	1.37 × 103	0.30
Teeth	1.96 × 104	0.30
PDL	0.67	0.45
Attachment	12.5 × 103	0.36
Clear aligner	842	0.36

PDL Periodontal ligament

Data measurement and outcomes

The right maxillary canine served as the reference tooth. The measurement of the canine rotation angle is shown in Fig. 2. The initial labial midpoint of the canine is defined as point A. A plane perpendicular to the long axis of the tooth is created through point A, intersecting the long axis at point O. After applying force with CAs, the new labial midpoint of the canine is defined as A'. The measurement of the angle AOA' represents the canine rotation angle.

Fig. 2.

Measurement methods for canine rotation angles

This study examined several key metrics: (1) Patterns of tooth displacement. (2) Canine rotation angle. (3) Equivalent stress distribution and levels in roots. (4) Equivalent stress distribution and levels in PDL. To further assess how attachment wear affects canine rotation, supplementary metrics were calculated as well:

1. Rate of change in canine rotation angles, measured as [(canine rotation angle at M0 − canine rotation angle at Mn)/canine rotation angle at M0] × 100%, for n = 2, 4, 6, and 8.

2. Canine rotation efficiency, measured as (canine rotation angle at Mn/2) × 100%, for n = 2, 4, 6, and 8.

3. Rate of change in root equivalent stress, measured as [(von Mises stress values of roots at M0 − von Mises stress values of roots at Mn)/von Mises stress values of roots at M0] × 100%, for n = 2, 4, 6, and 8.

4. Rate of change in PDL equivalent stress, measured as [(von Mises stress values of PDL at M0—von Mises stress values of PDL at Mn)/von Mises stress values of PDL at M0] × 100%, for n = 2, 4, 6, and 8.

Results

Initial displacement tendencies and canine rotation angles

Figure 3 shows the distribution of dentition displacement, while Fig. 4A illustrates canine movement trends. Across all conditions, canine displacement distribution remained consistent between groups, with maximum displacement located in the cervical area near the mesial labial–lingual side of the crown and minimum displacement observed at the cusp and apical regions, showing a tendency of rotation toward the mesial along the long axis of the canine. With increasing attachment wear, the canine rotation angles decreased in both groups. In M0 and M2, the rectangular group presented greater canine rotation angles than the optimized group did; however, from M4 onward, the canine rotation angles in the rectangular group were lower than those in the optimized group were (Fig. 4B). Compared with Condition M0, the rectangular group experienced an 8.48% decrease in the canine rotation angles in M2 and a 32.35% decrease in M8. Conversely, the optimized group presented a 3.27% reduction in M2 and an 18.62% reduction in M8. Under all working conditions, the rectangular group displayed the most notable decline between M2 and M4, whereas the optimized group presented the most substantial decrease between M6 and M8 (Fig. 4C). The rotation efficiency of the rectangular group dropped below 50% starting from M4, whereas the optimized group achieved this at M8 (Fig. 4D).

Fig. 3.

Dentition displacement distribution

Fig. 4.

Canine displacement tendencies and rotation angles. A Canine displacement tendencies. B Canine rotation angles. C Rate of change in canine rotation angles. D Canine rotation efficiency

Equivalent stress in roots

In Condition M0, the rectangular group presented a stress concentration in the attachment bonding area, with a significant stress distribution throughout the root cervical and mid-root regions. The maximum stress occurred near the cervical 1/3 of the root, whereas the minimum stress occurred at the apical region. In contrast, the optimized group presented a slight stress concentration in the attachment bonding area, with considerable stress distribution in the cervical 1/3 of the roots and minimum stress in the apical region. The stress distribution patterns of the two groups were essentially the same for all conditions, with root von Mises stress decreasing progressively with increased attachment wear (Fig. 5A). The rectangular group presented higher root von Mises stress compared to the optimized group in M0 and M2, but lower stress after M4 (Fig. 5B). The rectangular group’s root equivalent stress dropped 32.61% from M0 to M2, compared to a 0.14% reduction in the optimized group. By M8, the rectangular group was reduced by 67.09% from M0, whereas the optimized group decreased by 31.34%. Notably, attachment wear affected the root equivalent stress changes in the rectangular group, with a more significant reduction between M0 and M4, whereas the optimized group displayed a more pronounced decrease between M6 and M8 (Fig. 5C).

Fig. 5.

Equivalent stress distribution and levels in roots. A Canine stress distribution. B Root equivalent stress values. C Rate of change in root equivalent stress

Equivalent stress in PDL

In Condition M0, the rectangular group experienced peak stress in the PDL near the cervical region and the lowest stress at the root apex. In contrast, the optimized group presented a relatively uniform stress distribution in the PDL, focused mainly on the cervical region and mid-root, with minimal stress present at the root apex. The stress distribution remained consistent across all conditions for both groups, with PDL von Mises stress progressively decreasing with increasing attachment wear (Fig. 6A). PDL equivalent stress alterations mirrored root equivalent stress trends for both groups. Under the M0 and M2 working conditions, the rectangular group presented higher von Mises stress values in the PDL than did the optimized group, but lower values after M4 (Fig. 6B). The rectangular group’s PDL equivalent stress dropped 31.34% from M0 to M2, compared with a 1.53% decrease in the optimized group. By M8, the rectangular group decreased by 68.20% from M0, whereas the optimized group decreased by 31.30%. Attachment wear significantly affected equivalent stress alterations within the PDL for the rectangular group, with substantial reductions observed between M0 and M4, whereas the optimized group presented significant decreases between M6 and M8 (Fig. 6C).

Fig. 6.

Equivalent stress distribution and levels in PDL. A PDL stress distribution. B PDL equivalent stress values. C Rate of change in PDL equivalent stress

Discussion

In the process of correcting rotation, rounded teeth (canines and premolars) present challenges due to their morphology, hindering the retention of CAs and increasing the risk of slippage. This difficulty particularly affects achieving effective rotational correction, with canines posing the most significant challenge. In a prospective clinical study, Kravitz et al. reported a low efficiency of 32.2% in correcting maxillary canine rotation using CAs [23]. When the required rotation angle exceeded 15°, the achievement rate for correcting rotation in maxillary canines significantly decreased. Charalampakis et al. also reported that posttreatment rotational angles for all teeth, particularly maxillary canines, deviated by an average of 3.05° from the expected values [24]. Recent research highlighted an average 3.8° difference between actual and intended canine rotations, with a loss of 0.4° observed in maxillary canines per degree of planned rotation [25]. Therefore, many scholars have shown interest in enhancing the efficiency of rotational correction in maxillary canines with CAs, suggesting that adding attachments could increase the efficiency to 63.2% [26]. This concept was supported by Stephens et al., who demonstrated an increased efficiency of 75.4% when attachments were added during mandibular canine rotation [27]. Our study revealed that under all conditions, the efficiency of canine rotation ranged from 40 to 60%, which is consistent with previous findings.

The study findings revealed consistent patterns regardless of attachment type or degree of wear. The maximum displacement was located near the cervical region on the labial-lingual side of the canine crown, whereas the minimum displacement occurred at the cusp and root apex, showing a tendency to rotate around the long axis of the tooth, aligning with Cai et al.'s results [28]. With increasing attachment wear, both groups presented decreasing trends in canine rotation angles, roots, and PDL equivalent stress values, leading to reduced rotation efficiency. Our research revealed that rectangular attachment wear greatly influenced canine rotation efficiency. Compared with the optimized rotation attachments, the unworn rectangular attachments initially presented greater rotation angles and stress values. However, as the amount of wear increased to simulate two months of aligner use, a pronounced decrease in these values occurred for canines with rectangular attachments, even falling below those with optimized rotation attachments. When wear on the rectangular attachments equaled four months of aligner use, canine rotation efficiency decreased to less than 50%. These findings indicate that, considering the impact of attachment wear on the efficiency of maxillary canine rotation, adding optimized rotation attachments may be preferable to rectangular attachments. Castroflorio et al. also showed that using optimized attachments on mandibular canines can achieve higher rotation efficiency in clinical practice [25].

Based on our research, when it comes to optimized rotation attachments, surface wear comparable to six months of aligner use is marked as a tipping point. Beyond this threshold, additional wear resulted in a noticeable decline in the canine rotation angles and the root and PDL equivalent stress values. Once attachment wear was equivalent to wearing aligners for eight months, the efficiency of canine rotation dropped below 50%. Considering how attachment wear affects the efficiency of maxillary canine rotation, these findings imply that when optimized rotation attachments remain in place for beyond six months yet maxillary canines still require corrective rotation, orthodontists ought to carefully track the condition of these attachments to maintain optimal rotation efficiency. In cases of attachment loss, it is recommended that the re-bonding process be initiated by employing an attachment template. If attachments remain, monitoring their wear is essential to make necessary clinical adjustments. Our previous clinical study found that the principal areas of wear on optimized rotation attachments occurred primarily at the surface marginal ridges [11]. During follow-up appointments, clinicians may focus on this area to track attachment wear levels. In addition to visual inspection, 3D superimposition analysis through intraoral scanning can be applied to evaluate attachment wear levels. In the event of a substantial alteration in the attachment configuration, removing the existing attachments and re-bonding them is advised. If there are no significant shape changes but wear is noticeable, the original attachments can be retained, and fluid resin can be injected through perforations on the attachment template to restore the attachment shape. In clinical settings, reducing attachment wear to enhance maxillary canine rotation efficiency involves two other key factors: attachment material selection and patient behavior. This includes using more wear-resistant resins or bio-ceramic materials for making attachments and instructing patients to avoid consuming hard foods, use soft-bristled toothbrushes, and correctly wear CAs [29]. However, owing to the limited efficiency of CAs in achieving maxillary canine rotation, a combination of interactive elastics or the implementation of interproximal reduction (IPR) between adjacent teeth and the rotated tooth is recommended [30, 31].

Currently, FEA, as an effective computer simulation method, has been widely used to explore the biomechanisms of CAs [32, 33]. This research employed CBCT DICOM files for its modeling process, ensuring high accuracy of the FEA results. However, there are still some limitations in the modeling process. First, similar to previous studies, the FE models established in this research can simulate only static conditions, providing instantaneous displacements and stresses, but lack the ability to replicate the dynamic tooth movement process [14, 34]. As a result, the study outcomes may not entirely correspond to clinical reality. Nevertheless, researchers are increasingly exploring methods to achieve dynamic simulations of FE models to better align with clinical practice. Scholars such as Yokoi et al. have attempted to use FEA to dynamically simulate the process of closing diastemas between maxillary central incisors [35]. With advancements in technology, it is believed that dynamic simulations of FE models can be realized in the future, offering more precise research results for clinical applications. Our prior study measured levels and key areas of wear for the top two attachments used in canine rotation. The models for wear attachments were derived using Boolean operations on statistical analyses from prior research, thus improving the general applicability of the findings. However, in clinical settings, diverse patient dietary habits, brushing techniques, and aligner wear frequency contribute to potential discrepancies in attachment wear compared with our clinical research findings. Consequently, FE model results may not perfectly correspond to real-world clinical conditions. Moreover, owing to patient loss to follow-up and aligner refinement, our study collected data only after eight months of aligner use. Obtaining attachment wear data over longer periods, such as after one year, could have more substantial implications for clinical practice, necessitating larger sample sizes and extended observation periods in future studies. Additionally, to simulate real-world scenarios, the resin used to make the attachments in our clinical study was the composite resin commonly used in clinical settings (Filtek Z350, 3M, USA). Therefore, the attachment material in this study was also a composite resin. If the attachment is made from different materials, its material properties and friction coefficient will consequently change, altering the wear pattern of the attachment and its effect on canine rotation. For instance, an attachment made of bio-ceramic material may exhibit superior wear resistance compared to composite resin. This would result in less attachment wear and a reduced impact on canine rotation efficiency. Consistent with the previous FE study, there is an important issue to consider in the modeling process [12]. A slight discrepancy between the aligner and its optimized attachment on the active force surface allows the aligner to engage in a compressive interaction. This compression generates an active force on the teeth. Owing to the lack of standardized guidelines and challenges in quantifying specific loading values, no additional force was applied to the active force surface of the optimized rotation attachment. As a result, there may be slight variations between the outcomes and clinical reality. Given that optimized rotation attachments can actively apply forces to assist in rotation, the efficiency of achieving canine rotation with the addition of optimized rotation attachments might be greater in clinical practice. Given all the limitations, the conclusions of this study should be interpreted with caution. To increase the validity of the research findings, further real-world cases and clinical trials should be conducted.

Conclusions

Considering the study's limitations, the following conclusions can be drawn:

1. With increasing attachment wear, the efficiency of maxillary canine rotation gradually decreases, with rectangular attachments having a significant impact on canine rotation efficiency. To correct maxillary canine rotation, incorporating the influence of attachment wear, it is preferable to use optimized rotation attachments.

2. If optimized rotation attachments remain in place for over six months and maxillary canines still need corrective rotation, orthodontists should carefully track attachment wear to ensure rotation efficiency. Immediate restoration or re-bonding of the attachments is required if it is determined to be essential.

Abbreviations

CA

Clear aligner

3D

Three-dimensional

FEA

Finite element analysis

CBCT

Cone beam computed tomography

PDL

Periodontal ligament

IPR

Interproximal reduction

Authors’ contributions

QYL: Conceptualization, Data curation, Funding acquisition, Methodology, Software, Visualization, Writing-original draft, Writing-review and editing. BWX: Methodology, Software, Visualization, Writing-review and editing. DYF: Methodology, Visualization, Writing-review and editing. KY: Conceptualization, Funding acquisition, Methodology, Supervision, Visualization, Writing-review and editing.

Funding

This study was supported by grants from the Capital’s Funds for Health Improvements and Research [Grant Number: CFH2022-2–2144] and the Young Scientist Program of Beijing Stomatological Hospital, Capital Medical University [Grant Number: YSP202405].

Data availability

The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.

Declarations

Ethics approval and consent to participate

This study was performed in line with the principles of the Declaration of Helsinki. Ethical approval was obtained from the Research Ethics Committee of Beijing Stomatology Hospital, Capital Medical University, and all participants signed informed consent prior to the study (CMUSH-IRB-KJ-PJ-2022–25).

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.