INTRODUCTION

Lingual orthodontic treatment has become increasingly popular in recent years due to patient esthetic desire in a changing, appearance-driven society. However, due to anatomic variability of the lingual surfaces among patients, the lingual system requires a precise setup. Indeed, invisibility of the appliance is not the only relevant feature; the real clinical objective is to achieve good facial harmony and perfect occlusion combined with smile esthetics and patient satisfaction. Lingual orthodontics achieves high biomechanical efficiency through three-dimensional tooth control. Additionally, some cases require a fixed appliance, since inferior results would be obtained with aligners and, in some cases, the lingual system is even more biomechanically advantageous than the traditional buccal technique.^1,2 However, Ata-Ali et al.³ found no difference in the posttreatment peer assessment rating scores between patients treated using lingual and labial appliances.

Knowledge of the appliance limitations allows calibration to apply overcorrections during the setup procedure to achieve the best results for patients. Previous authors evaluated accuracy of different customized lingual systems with bent archwires, focusing on the differences between setup and final results, showing high accuracy, and less than 5° for the angular values.^4,5 Those studies, however, did not measure the ratio between achieved and planned movement for the lingual appliance.

The wide range of anatomic variability of the lingual surfaces among teeth requires indirect bonding to achieve high-quality treatment with lingual appliances. Therefore, the setup accuracy has assumed a fundamental role. For these reasons, the aim of this study was to analyze the accuracy of planned tooth movements of torque, tip, rotations, and transverse width changes with the lingual straight wire technique. Comparison between the setup and actual final result, as well as measuring the initial models, allows analysis of the discrepancy between planned and achieved movements in relation to the initial tooth positions.

MATERIALS AND METHODS

The study design was reviewed and approved by the Ethics Committee of the Postgraduate School of Orthodontics, Ferrara University (approval number:11/2019). Sample size was calculated in the validation study of the measurement method used. To detect an effect size of 0.6 for the average tip and torque angles, with an alpha of 0.05 and a desired power of 0.80, at least 24 dental casts were required in the sample.⁶ In the present study, 40 patients were analyzed. They were consecutively treated in a private practice using a lingual straight-wire appliance with a horizontal wire insertion slot (STb, Ormco, Glendora, Calif), by the same operator (EA) over a 4-year period.

The inclusion criteria for this case series study were: patients with a full permanent dentition (except for third molars), Class I or mild Class II Angle classification, nonextraction cases. The exclusion criteria were: patients with prosthetic rehabilitation, sucking habits, craniofacial syndromes, cysts, cleft lip or palate, multiple or advanced carious lesions, patients who needed corrective jaw surgery, and patients with incomplete or imperfect records.

A panoramic radiograph, lateral cephalograms, photographs, and dental casts were obtained for all subjects. All subjects were treated in both arches with a lingual straight-wire technique with STb brackets and a manual setup with single jigs, using the Komori KommonBase system.⁷

Treatment followed the same standard archwire sequence: an initial upper and lower 0.013-inch Copper Ni-Ti followed by a 0.018- × 0.018-inch Copper Ni-Ti and a 0.0175- × 0.0175-in. TMA (Ormco) for the finishing phase. No extractions were needed and no auxiliaries were used during the finishing phase, except for Class II elastics in the last treatment phase.

Before treatment (T0), setup (T1), and after treatment (T2) maxillary and mandibular dental casts were taken for each patient. A total of 240 dental casts were acquired, and each was assigned a number to protect patient identity and to obscure the casts from the operator who digitized them. An intraoral 3D scanner (Carestream CS 3600, Carestream Dental, Rochester, NY) was used to scan all study casts to perform the three-dimensional (3D) dental analysis. All of the models were saved as STL files.

On each virtual model, the same operator (PA) digitized a total of 74 points (total: 17,760 points), using VAM software (Vectra; Canfield Scientific, Fairfield, NJ, USA), following the methodology established by Ghislanzoni et al.⁶ VAM software is currently widely used for attaining accurate model analysis.^6,8–13 A second expert operator double-checked the exact sequencing and placement of points (EA). The virtual models at the three time points were superimposed using a best-fit process on the digitized points, and the coordinates were exported as a collection of x, y, and z coordinates.

To calculate angular measurements, the models at T1 were utilized as a reference to build a common “Andrews-like” reference plane.¹⁴ The flat occlusal plane at T1 was created in the setup as the optimal result, since the occlusal plane at T0 could eventually affect some torque measurements. The results of linear and angular measurements were examined. The angular values (tip, torque, and rotation values) were analyzed using the reference plane, while the cast superimpositions were generated using the best-fit alignment among the corresponding points on the digital casts (T0-T1-T2).

As reported by Andrews, torque was measured as the labiolingual inclination of the clinical crown's facial axis with respect to the reference plane (Figure 1a,b), and tip was quantified as the mesiodistal inclination of the clinical crown's facial axis with respect to the reference plane (Figure 1c).¹⁴ The sign of the angular measurement corresponded to the traditional bracket prescriptions: positive values indicated buccal crown inclination, whereas negative values indicated lingual crown inclination.

View Figure

• Download as Powerpoint
• Download Figure

The angular variation of the mesiodistal line (segment of the mesio and distal point) with respect to the reference plane was used to quantify rotations as described by Andrews (Figure 1d).¹⁴

Transverse arch widths at the canines, first and second premolars, and first and second molars were measured using the transverse intercusp distance (Figure 2).

View Figure

• Download as Powerpoint
• Download Figure

The average differences between time points were calculated for each tooth and type of planned movement. The following formula was used to quantify the accuracy of each movement for each tooth with respect to the setup (accuracy is the % of movement achieved compared to the movement planned):

Thus, an index of the accuracy of each movement was obtained. The closer that each value was to 1, the more precise the dental movement that was achieved (100% of the planned movement in the setup). The accuracy values presented represent the average for each individual accuracy value obtained with the formula (ratio between real [T2–T0] and programmed [T1–T0] movement) and not the average of the averages between real mean movement (average [T2–T0]) and programmed (T1–T0) movement.

RESULTS

The study sample consisted of 40 Caucasian subjects (28 female and 12 male) with a mean age of 23.9 years (SD: 9.3) and an average treatment time of 23.3 months (SD: 5.8). The method error, calculated from repeated measurements of digitized casts, was 0.6° and 0.2 mm for the angular and linear measurements, respectively.

Torque

A general increase of torque values was recorded in the setup (T1) and in the final result (T2), except for the maxillary molars (Tables 1 and 2). The mean accuracy for torque actually obtained (vs planned) was between 85% and 92% for incisors, canines, and premolars, whereas the accuracy for molars was between 52% (maxillary second molars) and 81% (mandibular first molars).

Table 1. Average Differences in Torque Values Between Initial Models (T0), Setup (T1), and Final Models (T2) and Accuracy (Torque Change Obtained/Torque Change Planned) Values

View Fullscreen

Table 2. Statistical Significance for Torque Value Differences

View Fullscreen

Tip

The tip movement mean accuracy was between 84% and 94% for incisors, canines, premolars, and first molars, whereas the second molars showed slightly lower values (Tables 3 and 4).

Table 3. Average Differences in Tip Values Between Initial Models (T0), Setup (T1), and Final Models (T2) and Accuracy (Tip Change Obtained/Tip Change Planned) Values

View Fullscreen

Table 4. Statistical Significance for Tip Value Differences

View Fullscreen

Rotations

The rotational movement mean accuracy was between 85% and 93% for incisors, canines, and premolars, while the molars showed values between 57% and 87% (Tables 5 and 6).

Table 5. Average Differences in Rotation Values Between Initial Models (T0), Setup (T1), and Final Models (T2) and Accuracy (Rotation Obtained/Rotation Planned) Values

View Fullscreen

Table 6. Statistical Significance for Rotation Value Differences

View Fullscreen

Transverse widths

A general increase of transverse widths was measured in the setup (T1–T0; mean values between 1.4 mm and 2.5 mm) and in the final models (T2–T0; mean values between 0.4 mm and 1.4 mm), except for the second molars, which showed reduced intercusp transverse width during treatment. The transverse width mean accuracy was between 70% and 90% for canines, between 46% and 70% for premolars and first molars, and 20% for the mandibular second molars, whereas the maxillary second molars showed negative mean values of accuracy (Tables 7 and 8; Figure 3).

Table 7. Average Differences in Transverse Width Values Between Initial Models (T0), Setup (T1), and Final Models (T2) and Accuracy (Width Change Obtained/Width Change Planned) Values

View Fullscreen

Table 8. Statistical Significance for Transverse Width Differences

View Fullscreen

View Figure

• Download as Powerpoint
• Download Figure

DISCUSSION

In lingual orthodontics, unlike the buccal technique, the setup is a mandatory step due to the anatomic variability of the lingual surfaces among patients. The setup is an important diagnostic and therapeutic tool. Knowledge of the limitations of the appliance used will allow clinicians to plan the treatment objectives to get closer to the ideal result. Previous literature documents that indirect bonding shows high positional accuracy in the mesiodistal, buccolingual, and occlusogingival dimensions; therefore, the movement planned in the setup is completely transferred to the teeth.¹⁵

The present study analyzed the limitations of the lingual technique, but also aimed to quantify the overcorrections in the setup required for the correction of torque, tip, rotation, and transverse width. The accuracy percentages (amount of movement achieved compared to movement planned) were the average of each accuracy value, expressed as the ratio between achieved (T2–T0) and planned (T1–T0) movement.

The results of this study confirmed the high accuracy rate of the setup and lingual systems, excluding, however, the second molars, which are more difficult to control. In particular, the final torque values approached the setup, for an average percentage greater than 85% of the programmed movements. However, the average accuracy for the molars was between 52% and 81%. These differences were attributable to the fact that the archwire is less rigid in the posterior areas. Also, the posterior teeth are subjected to chewing forces and their root structure allows less control.

On the other hand, tip and rotation values denoted the high precision of this system, achieving between 72% and 94%, and 85% and 93%, of planned movement up to the second premolars and between 57% and 87% in the molar region, the latter being influenced by the use of elastic chains during lingual treatment. Grauer and Proffit,⁴ despite having analyzed another lingual system on another sample and using different software, obtained similar values to those of the current study, concluding that the discrepancies between setup and final models were minimal, less than 1 mm for linear discrepancies and 4° for rotational discrepancies, except for the second molars, which exhibited greater discrepancies. However, the limitation of Grauer and Proffit's study⁴ was that only the discrepancies between the setup and final models were analyzed, while the present study also considered the initial positions of the teeth. In the current study, a percentage of the movement obtained was calculated between the starting situation and the ideal programmed one. Pauls et al. published two similar scientific articles in which the discrepancies between setup and final models were analyzed. In those two studies, two different lingual systems were considered, both with robotically bent wires. In both studies, although through different methodologies to those reported in the current work, values of high accuracy were obtained, with less than 1 mm in linear discrepancies and 5° for rotations, and slightly greater differences for the second molars (the arch terminal elements).^5,16

Using the same measurement methods as in the current study, Lombardo et al.¹¹ presented accuracy values for treatment with aligners. They reported accuracy for mesiodistal tip of 83%, buccal lingual tip of 73%, and rotations of 67%. Despite the differences in the sample, it is possible to highlight a greater accuracy in the movements obtained with a lingual appliance compared to that achieved with aligners. This trend is confirmed by previous literature¹ in which the scientific evidence does not support the use of aligners as a treatment alternative with the same effectiveness as fixed appliances.

As previously reported by Ata-Ali et al.,³ it may be concluded that predictable results, comparable to those shown for the lingual appliance in the current study, can be obtained using a buccal fixed orthodontic appliance. However, due to the lack of scientific data and the limited use of setups in buccal therapy, comparative analyses cannot be performed using the percent accuracy formula used in this study.

Regarding transverse expansion with the lingual appliance, it was shown in the current study that a programmed increase in transverse width was only partially realized. The accuracy of transverse expansion decreased in the terminal elements. Negative mean accuracy values suggested that the maxillary second molars actually moved in the opposite direction compared to the programmed movement in the setup. This was probably due to several factors: negative torque prescription, transverse bowing effect increased by the use of elastic chains, and effect of chewing forces on the posterior elements. Grauer and Proffit's study⁴ was in agreement with these results; they concluded that the planned expansion programmed in the setup did not fully occur. For transverse expansion in aligner therapy, there are no scientific studies describing the percentage of accuracy. However, it has been shown that transverse bodily expansion with aligners is not predictable and overestimated in the setup, often resulting in coronal tipping movement.^1,3,12 An analogous situation occurs with the use of the buccal appliance in which expansion of the interarch widths is obtained through the use of wide arches. With buccal appliances, there is also a component of buccal inclination of the posterior elements.^10,13 For all these reasons, in a case that requires transverse expansion of the arches, planning a preliminary expansion phase or overcorrection of the transverse dimension in the finishing archwires is recommended.

A limitation of this study was the absence of a control group. In a previous study, a control group with similar characteristics showed no changes in transverse and torque measures. For this reason, it was assumed that the lack of a control group would not influence the interpretation of results.

CONCLUSIONS

This study measured the accuracy of the planned tooth movements of torque, tip, rotation, and transverse width values with a lingual straight wire system. Comparison between the setup and final result allowed analysis of the differences between planned and achieved movements relative to the initial tooth positions:

• The movements planned to obtain ideal torque, tip, and rotations that actually occurred, demonstrating a high accuracy rate of the lingual system.

• The most posterior teeth (the second molars) showed less accuracy of achieved vs planned movements than the other dental elements.

• Expansion of the arches programmed in the setup occurred only partially as assessed in the final models, with greater accuracy for canines and premolars and less accuracy for the posterior elements. Therefore, when expansion is necessary, a separate expansion phase prior to lingual fixed therapy, or an overcorrection of the transverse dimensions of the archwires, is required.