Controlling incisor torque with completely customized lingual appliances

All consecutive patients treated with a CCLA (WIN; DW Lingual Systems; Bad Essen, Germany) and debonded in the time period from 1 March 2013 to 30 June 2016 in one orthodontic center (Prof. Dr. Wiechmann, Dr. Beyling and colleagues, Bad Essen, Germany) were screened for potential eligibility and were consecutively selected if they met the following inclusion criteria:

Subjects were excluded if they fulfilled one of the following exclusion criteria:

After application of the inclusion and exclusion criteria, a total of 29 Angle class II division 2 malocclusion subjects (male/female ratio 11/18; mean age at the beginning of treatment (T1): 15.6 [range 13–27] years) were included. None of the patients received skeletal anchorage or fixed functional appliances. No patient was excluded from this retrospective analysis for any other reasons such as missed appointments, lack of compliance, or missing records or informed consent to anonymized data release forms, which is occasionally seen in sample compositions of retrospective studies [12].

All subjects were treated with a CCLA (WIN, DW LingualSystems, Bad Essen; Germany) and a standardized sequence of archwires: an initial NiTi with round diameter (diameter depending on extent of crowding 0.012″, 0.014″, or 0.016″), followed by a rectangular 0.016″ × 0.022″ NiTi archwire, and a 0.016″ × 0.024″ stainless steel wire with an incorporated maxillary canine to canine extra-torque bend of 13°. One patient received an additional 0.016″ × 0.024″ stainless steel wire with an extra-torque of 21°. A TMA 0.018″ × 0.018″ archwire was used for finishing in all treatments. Space closure and Angle class II correction was achieved with power chains (Morita Energy Chain, Rocky Mountain Orthodontics, Denver, CO, USA) and class II elastics following incorporation of the stainless steel archwires in both arches.

Sagittal occlusion was measured using high-resolution, digital, intra-oral photographs (D200, with Nikkor 105 mm; Nikon, Tokyo, Japan) at bonding (T1), following leveling and aligning and prior to using class II elastics (T2), and after debonding (T3). The photographs were taken directly, without an intra-oral mirror, using cheek retractors (Fig. 1). The camera axis was as close as possible to 90° to the premolar/posterior tooth group. We used the premolars to assess the extent of sagittal occlusion. For the assessment of the sagittal occlusion, we compared the position of the upper first premolar to that of the two lower premolars. We defined the Angle class I in our study by the upper premolar being perfectly centered with the lower premolars, equivalent to a value of 0 mm. If the cusp of the upper first premolar is centered with the mesial side of the first lower premolar, it was defined as a full Angle class II occlusion. An edge-to-edge position of the respective mesial and distal sides of upper and lower first premolar was defined as a sagittal distal occlusion of ½ cusp. The true dimensions of the assessed premolar and malocclusion on the digital photographs were assessed using a calibration technique previously employed [13]: The intraoral photographs were scaled to the corresponding plaster casts by adjusting the dimensions of the upper first premolar to its corresponding dimensions taken from direct cast measurements, using a sliding caliper. In cases in which there were different extents of sagittal occlusion on either side, the side with greater Angle class II occlusion was chosen for assessment. A neutral occlusion was defined by the first maxillary premolar centered between the lower premolars and was assigned a value of 0 mm of sagittal malocclusion. Deviations from neutral sagittal canine occlusion in the posterior direction (i.e., distal or Angle class II occlusion) were assigned positive mm values. Overbite measurements were performed on corresponding plaster cast models at T1 and at the end of treatment (T3). Wax bites taken in centric relation were utilized to correctly position upper and lower plaster casts, and the maximum overlapping was marked on the lower incisors with a pencil tip of 0.35 mm thickness. Distances between the pencil mark and the incisal edge of the incisor were taken using a sliding caliper (Dentaurum, Münchner Modell, Ispringen, Germany). Likewise, maxillary spacing was measured on the plaster casts taken at T1 and T3. All assessments of sagittal occlusion, incisor inclination, and overbite were performed manually by one operator (OA), twice, and the mean values for each of the repeated measurements were used for analysis.

For assessing the potential method error for assessment of the C_ROT, measurements of 10 randomly selected patients were repeated (M1; M2) by the same examiner (OA). The mean difference of (M1 − M2) was 0.42% and no significant difference was detected (standard deviation [SD] 2.5; p = 0.54 by sign test). A small variation between a minimum of −2% and a maximum of 7% was considered acceptable compared to the differences seen between C_RES and C_ROT.

To assess the quality of incisor movements during CCLA treatment, the measurement data for upper incisor inclination to the palatal plane (U1/PP), sagittal occlusion, overbite, the extent of spacing of the maxillary dental arch and the location of the C_ROT were analyzed descriptively using mean and standard deviation (SD), as well as minimum and maximum values (min–max) at specific time points. The primary endpoint was the location of the C_ROT in relation to C_RES.

The potential difference between the center of rotation (C_ROT) and the center of resistance (referred to the fixed value C_RES = 0.36) was analyzed using a nonparametric sign test, due to the asymmetric data distribution. The sign test was also used to assess the method error between repeated measurements of M1 and M2. The change in sagittal occlusion, U1/PP and overbite between T1 and T3 was analyzed using a paired t‑test. The significance level was set to α = 5%. All statistical analyses were derived using the statistical software SAS version 9.4 (SAS, Cary, NC, USA).

With the available sample size of 29 patients and different assumed standard deviations of 20 and 30, a significance level of α = 5%, power of 90%, as well as C_RES = 36%, the following true expected effect sizes (∆ = expected value of C_ROT − C_RES) were established: ∆ = 12.5% (assuming \(SD=20\)) and ∆ = 18.7% (assuming \(SD=30\)).

At T1, the mean total maxillary spacing was 3.0 mm (SD 1.9; min–max 1–10 mm). At the end of CCLA treatment (T3), all maxillary spacings were closed, as planned in the target set-up (Tables 1 and 2).

Mean sagittal malocclusion as measured on the worse side at T2 (following leveling and aligning) was 4.6 mm distal occlusion (SD 1.2; min–max 3.5–7 mm) which had improved in all patients to a mean 0.2 mm (SD 0.3; min–max 0–1.5 mm) at T3. At T3, 21 out of 29 subjects had a perfect Angle class I occlusion (0 mm), whereas 7 patients (24.1%) showed a minor distal occlusion of 0.5 mm and one subject (3.4%) had a distal occlusion of 1.5 mm. The mean sagittal correction measured on the worse side was 4.4 mm (SD 1.2; min–max 3–7 mm; Tables 1 and 2).

All of the 29 patients had a deep bite (>2 mm) at T1. Deep bite correction was achieved in all patients: at T1, the mean overbite was 4.4 mm (SD 1.0 mm), which decreased significantly (p < 0.001) by a mean difference of 2.9 mm (SD 1.0 mm) to a mean 1.5 mm (SD 0.4 mm; Tables 1 and 2).

Treatment duration varied from 16–43 months, with a mean value of 28 months (SD 3.5 months) between CCLA bonding and debonding.

In general, there are three different ways to improve the inclination of a retroclined incisor. If a single force is applied to the crown of the tooth in an anterior direction, the crown is tipped in the buccal direction. Without the simultaneous application of a couple of force with opposite directions (a moment), the tooth will rotate with a center of rotation located inside the root. The location depends on the distance of the applied force to the center of resistance. Some of these types of tooth movements meet the definition of uncontrolled tipping and are typically achieved with a variety of removable appliances or with fixed appliances in combination with round archwires. At the end of such tooth movement, the inclination of the incisor is more positive and, as the crown of the tooth is tipped to the buccal side, space is created, leading to an increased arch length and an improved incisor inclination. However, it is worth noting that this improvement in incisor inclination is not equal to a (root) torque movement or a bodily movement. To achieve a torque movement, the combination of a force and a moment (couple of forces in opposite directions, e.g., a force shifted apically to the C_RES by a moment created by the segmented arch approach resulting in a palatal root torque) is necessary. Only this combination allows for shifting the center of rotation towards the incisal edge. To achieve movement such as a root torque (torque movement, bodily movement as torque) with a center of rotation close to or at the incisal tip [16], a buccally oriented force on the crown of the tooth has to be combined with a clockwise moment (buccal root torque), or a palatally oriented force on the crown of the tooth has to be combined with a counterclockwise moment (palatal root torque). Typically, buccal root torque is needed in the maxillary anterior segment for dentoalveolar compensation of an Angle class III malocclusion, whereas palatal root torque in the same segment is needed for most of the nonsurgical Angle class II corrections. If the inclination of a retroclined incisor is corrected by a palatal root torque, no space opening is observed, as the crowns do not move buccally. The third possibility to improve the inclination of a retroclined incisor is controlled tipping of the crown to the buccal side. In this case, the center of rotation is shifted to the apex.

Putting aside these fundamental orthodontic rules, virtually impossible results have been reported by some authors mainly during the past 7 years, regarding the alleged torqueing capabilities of some kinds of clear aligner treatment (CAT) [19‐21]. A closer look at the methods applied in these studies has revealed an ambiguous use of the term “torque movements”, since they measured the inclination of the labial surface of the incisors on digital dental arches and misinterpreted each inclination change of the labial surface as a bodily movement of torque. The orthodontic community should indeed be aware of the unreflective repetition of this kind of mistake—even in systematic review articles about CAT.

The importance of palatal root torque and/or correct incisor inclination for dental arch length and the achievement of a proper molar relationship has been the subject of many studies [22‐25]. With fixed appliances, a torque moment is created by the rectangular archwire’s action to untwist when engaged in a bracket slot [26], thereby generating a couple of forces with opposite directions or moment that change the inclination of the incisor [25, 27]. Being a free vector, this moment is subject to spatial variability and may be located on any point along the long axis of the tooth, or outside of the tooth. Based on this knowledge, Andrews developed the straight-wire appliance using angulated, pre-adjusted bracket slots [28]. It is, however, well known that torque expression is highly dependent on archwire dimensions versus slot size [29]. Commonly used conventional labial brackets are manufactured using the technique of metal injection molding. The initially molded bracket—the so-called “brown part”—is about 30% larger than the final bracket and is made of compressed metal powder. During the sintering process, this oversized bracket is heated up very close to its melting temperature and shrinks in a very uncontrollable way, resulting in high slot tolerances in both directions: positive and negative. In order to avoid situations in which archwires do not fit into undersized bracket slots (negative tolerance), the target slot size is set to a higher value, aggravating the average oversize of the final bracket slot to values of up to 20% or more [29]. The consequences of increasing torque play by shifting the manufacture of orthodontic brackets from milling or precision casting to metal injection molding (MIM) have been discussed extensively [30‐33].

In lingual orthodontic treatment, third order control is of substantial importance for achieving precision in finishing [34‐37]. In a finite element analysis, Liang et al. [11] simulated and compared retraction of incisors with labial and lingual appliances and found that it was critical to control the moment/force ratio and increase lingual root torque properly with lingual appliances. It is noteworthy that, to the best of our knowledge, this topic has been previously addressed on the evidence level of in vitro experiments and retrospective studies, but not with clinical prospective study or randomized, controlled trial designs that addressed outcome differences when comparing labial, lingual or aligner technique approaches to meet a higher level of evidence. In order to address these issues of controlling the moment/force ratio adequately, the bracket slots of the CCLAs used in this study were made by high-speed milling with manufacturing tolerances of only up to ±2 μm (<0.5%) [38]. As a consequence, the magnitude of torque expression with these appliances compared to conventional labial or lingual brackets can be easily controlled by the orthodontic specialist [37, 39‐45]. In order to account for torque play, the undersized maxillary rectangular 0.016″ × 0.024″ stainless steel archwires of the CCLA used here during space closure were provided with an incorporated canine-to-canine extra torque bend of 13°, and, in the case of one patient, 21°. The use of the 0.016″ × 0.024″ stainless steel archwires with built-in extra torque bends is crucial for maintaining the necessary moment to counteract and overcome the crown tipping forces in the palatal direction generated using class II elastics and power chains during space closure.

According to calculations made by Burstone and Pryputniewicz, the center of resistance (C_RES) of an upper incisor is located approximately at a point one-third of the distance from the alveolar crest to the apex [16]. Likewise, Graber et al. [34] reported that the center of resistance in a single-rooted tooth with a parabolic shape should be calculated by multiplying the distance from the alveolar crest to the apex by 0.33. Proffit et al. [43] took into consideration the condition of the periodontal tissues and described a tooth’s center of resistance to be slightly more apical, i.e., “at the approximate midpoint of the embedded portion of the root (i.e., about halfway between the root apex and the crest of the alveolar bone)”. Due to the young age of the subjects included in our study and the absence of periodontal diseases or bone loss, the alveolar crest was assigned to be located about 2 mm below the enamel/cement junction [16]. The alveolar crest was approximated by a line, and 2 mm was subtracted (Fig. 4).

The null hypothesis of no significant deviation between the center of rotation (C_ROT) and the center of resistance (C_RES) during space closure in Angle class II/2 subjects was rejected. While the center of resistance of the tooth was located at 36% from the total length of the tooth starting from the apex, the mean C_ROT was at 88.6% (min–max, 51–100%). This difference (mean C_ROT − C_RES, 52.6%) was found to be statistically significant (p < 0.001, Table 2).

In our study, all C_ROTs were found to be located above the incisor’s C_RES in the incisal direction, i.e., they were shifted towards the incisal edge or beyond, indicating that neither controlled nor uncontrolled incisor tipping occurred in the treated individuals.

We corrected for vertical (overbite) changes in tooth position during treatment by measuring the vertical position of the incisor before and after treatment and adjusted the center of rotation along the long axis of the tooth with the same distance. One of three incisor templates was individually chosen to account for deviating crown–root angles in Angle class II/2 subjects ([14, 15]; Fig. 2). The same template that was used for the baseline assessment of each incisor was also used for the second assessment of the respective incisor, in order to avoid within-subject changes in incisor shape and dimensions.

The calculation of the incisor’s C_RES as one factor to describe the nature of the achieved tooth movements was based on calculations made by Burstone and Pryputniewicz [16]. As the location of the C_RES is multifactorially influenced by tooth morphology and the properties and condition of the individual periodontum, slightly different descriptions of the C_RES have been provided by various authors [34, 46], and the extent to which the three-dimensional situation can be generalized has been discussed [47, 48]. However, the definition of the root centroid or C_RES used here is widely accepted in orthodontic literature and is considered to be a valid orientation to describing incisor movements.