The effectiveness of clear aligners in levelling the curve of Spee and related maxillofacial factors in patients with a deep overbite

Clear aligners have recently been introduced to address the increasing demand by adult patients for orthodontic appliances that are both aesthetically pleasing and comfortable.¹ Since its development in 1997, clear aligner technology has been established worldwide as an aesthetic and accepted alternative to labial fixed appliances.^2,3 However, as there are often discrepancies between the intended goals of clear aligner treatment and the observed outcomes, the clinical efficacy of this approach has remained controversial.^4,5

Since their inception, the efficacy of clear aligners has been continually evaluated. Kravitz et al.² conducted the first prospective clinical study and reported an average accuracy of 41% for tooth movement using the Invisalign appliance. More recently, several studies have measured aligner efficacy.^2,5–8 In 2020, Haouili et al.⁹ reported a mean accuracy of 50% for all tooth movements, which was a notable increase compared to previous studies. However, incisor intrusion remained a challenge which failed to improve even with G5 enhancement. Al-balaa et al.¹⁰ reported the mean precision of true anterior intrusion using Invisalign clear aligners as 51.19%, with a mean correction of 48.81%. While their results differed numerically, some reports have suggested that deep overbite malocclusions pose a challenge for clear aligner treatment.^11,12

A deep overbite is defined as an excessive vertical overlap of the mandibular incisors by the maxillary incisors in centric occlusion. A deep overbite >5 mm is a malocclusion found in 20% of children and 13% of adults.¹³ The correction of a deep overbite is a prime objective of orthodontic treatment and is related to the stability of the occlusion.¹⁴ Tooth movement involves incisor intrusion, incisor proclination, the extrusion of the lower posterior segments, or a combination of these changes. Numerous methods have been established to treat deep overbite malocclusions¹⁵ and range from clear appliances to fixed appliances with or without orthognathic surgery, including non-surgical options, such as nickel-titanium reverse curve arch wires, utility arches,¹⁶ anterior bite turbos¹⁷ and clear aligners. The effectiveness of clear aligners in correcting a deep overbite is controversial. Blundell et al.¹² were the first to assess the efficacy of the SmartTrack material and reported that ClinCheck software (Align Technology; San Jose, CA, USA) over-predicted overbite reduction in 95.3% of patients, with only 39.2% of the predicted overbite reduction realised. This finding supports that of Kravitz et al.,² who reported a 41% mean accuracy of tooth movement. It was also demonstrated that precision bite ramps did not improve the ability of the SmartTrack material to predictably open a deep bite.¹⁸ Furthermore, Burashed et al.¹⁹ reported that the actual amount of overbite reduction using clear aligners was significantly lower than planned. However, Henick et al.²⁰ claimed that both fixed appliances and the Invisalign appliance were effective in opening deep bites at the dentoalveolar and skeletal levels. Shin²¹ reported that clear aligners can maintain the overbite in patients with a normal overbite, as well as improve malocclusions in the vertical dimension (dental deep bite) within a mild-to-moderate overbite range. Khosravi et al.¹¹ concluded that clear aligners appear to manage the vertical dimension relatively well, and that the primary mechanism is incisor movement.

Levelling of the curve of Spee (COS) is a common approach for correcting a deep overbite²² and has been recommended as a primary goal of orthodontic treatment. Pandis et al.²³ investigated the effects of flat arch wires in a 0.022 inch slot edgewise appliance, without elastics nor auxiliaries, in individuals with a mild to moderate mandibular COS and found that the appliance levelled 50% of the initial COS depth. Goh et al.²⁴ investigated the ability of the Invisalign appliance to level the COS and reported a mean levelling accuracy of 35%. Unlike continuous arch wire levelling, which incorporates significant extrusion of the posterior teeth and proclination of the mandibular incisors, clear aligners only caused significant intrusion of the mandibular incisors.²⁵ Nevertheless, a pronounced COS poses a particular challenge.²⁶ Facial type, the proclination angle of the anterior teeth, and arch space are widely recognised as factors likely associated with ease of tooth movement and COS levelling. Evaluating the factors that could predict the effectiveness of COS levelling using clear aligners would allow clinicians to achieve more successful outcomes.

The present study was designed to evaluate the effectiveness of clear aligners in levelling the COS and included a statistical analysis to characterise the relationships while seeking to identify those factors of influence. The aim was to determine the effect of a range of maxillofacial characteristics to guide the treatment of deep overbite.

The study was approved by the Beijing Stomatology Hospital and the Ethics Committee of Capital Medical University (No. CMUSH-IRB-KJ-YJ-2021-29), who issued an endorsement certificate, and all patients provided informed consent to participate in the study. The study was retrospective and included 42 patients (22 females and 20 males) with a mean age of 23.74 ± 6.44 years. All patients underwent treatment with Invisalign clear aligners under the guidance of two experienced orthodontists certified in Invisalign treatment. All patients started treatment in 2019 or later, after Align Technology (San Jose, CA, USA) introduced the Smart-Track aligner material and launched their deep bite protocol.

Patients were included if (1) they underwent treatment in both arches; (2) they attended appointments, showed good compliance, and consistently wore the clear aligners; (3) the treatment plan did not include tooth extraction; (4) the treatment for the mandibular dentition involved levelling of the COS; (5) they had an established second molar occlusion; and (6) they presented with a deep overbite (above Class I). The exclusion criteria were (1) the depth of the COS < 1.5 mm, (2) if other supplementary methods were applied for COS levelling, (3) if periodontal disease or a morphological tooth anomaly was present, and (4) if self-reported temporomandibular disorder symptoms existed.

Pre-treatment (T1) and post-treatment (T2) intraoral scans were conducted by one examiner using an iTero scanner (Align Technology, 100–240V, 50/60Hz, USA), and the predicted outcomes (T3) were exported from the ClinCheck plan. The COS depth was measured using Materialise 3-matic Research 13.0 software (Materialise, Belgium). A virtual reference plane based on the midpoint of the central incisor edges (Aa) and distobuccal cusp tips (L_db and R_db) of the mandibular second molars was constructed using the plane tool. The perpendicular distances from the buccal cusp tip of an involved tooth to the constructed plane was evaluated as shown in Figure 1A, and the deepest points of the COS were measured on the right and left sides. The depth was defined as the mean value of both sides. The achieved levelling value (ΔD_a) was calculated as the COS depth (T1) subtracted from the COS depth (T2). The predicted value (ΔD_p) was calculated as the COS depth (T1) subtracted from the COS depth (T3). The percentage of levelling effectiveness was determined using the following equation²: percentage of accuracy = 100% – [(|ΔD_p – ΔD_a|/|ΔD_p|) × 100%]. This equation accounted for directionality and ensured that the percentage of accuracy would not exceed 100% for teeth that underwent movements beyond their predicted value. Crowding was defined as an arch length discrepancy, which was calculated as the difference between the sum of the mesiodistal widths of the teeth in each arch and the total recorded arch perimeter (Figure 1B).

Figure 1.

All lateral cephalograms were obtained with the mandible in natural head and postural position. Measurements were made using Dolphin Imaging software (version 11.8, Chatsworth, CA, USA), according to the reference lines shown in Figure 2A. All panoramic radiographs were obtained with the mandible in a protrusive position. Measurements of mandibular incisor parallelism were made via Digimizer software (v5.6; MedCalc Software) (Figure 2B) (See Table I for definitions). These procedures were repeated three times for each outcome parameter after an interval of two weeks. The intraclass correlation coefficient was used to assess consistency among the three measurements.

Figure 2.

The intraclass correlation coefficient was 0.801– 0.979. Please see Table II for more information. Descriptive statistics for the measurements are presented in Table III. The average reduction in the COS achieved by the clear aligners was 0.901 ± 0.455 mm. The levelling effectiveness was 71.8% ± 19.0%.

The results of the multiple linear regression analysis are presented in Table IV. The regression model explained 88.5% of the variance in levelling effectiveness (F-value = 167.209, adjusted R² = 0.885, Akaike information criterion [AIC] [E] = -146.94). Frankfort-mandibular plane angle (MP-FH)(T1) and P(T1) were considered significant at values <0.05. D(T1) and IMPA(T1) were considered significant at values <0.01. Both MP-FH(T1) and P(T1) showed a strong association; however, D(T1) and IMPA(T1) showed a stronger association in the regression model. The levelling effectiveness of clear aligners was positively associated with the pretreatment MP-FH, the incisor-mandibular plane angle (IMPA), and the depth of the COS, but negatively associated with the degree of mandibular incisor parallelism. The regression equation for the levelling efficiency of clear aligners can be written as: E = 0.008 × MP-FH(T1) + 0.005 × IMPA(T1) -0.254 × P(T1) + 0.120 × D(T1)+ εi. Stepwise regression eliminated all other variables (dental cast and lateral cephalometric measurements) that were not statistically significant and yielded a larger AIC when included in the model.

The fabrication of clear aligners involves digital setups and virtual three-dimensional tooth alignment. Unlike actual biological tooth movement, virtual movements are infinite, and the predicted outcomes may not be achievable.²⁷ The ability to evaluate the level of tooth movement is important in facilitating a deeper understanding of the biomechanics and treatment outcomes of clear aligners. While several studies have used superimposition to calculate the mean accuracy of tooth movement,^7,8 the mandibular arch lacks stable landmarks for best-fit overlap.²⁸ It is difficult to determine whether a tooth has undergone extrusion or intrusion; however, anterior intrusion and premolar extrusion ultimately result in a decreased depth of the COS. To the best of current knowledge, the present study is the first to evaluate changes in the COS and the influence of various maxillofacial factors following treatment by clear aligners.

The levelling of the COS requires complex tooth movements. To minimise the effects of different tooth movement patterns²⁹ and attachment design,³⁰ patients were only included if treatments shared the same orthodontic strategies. Initially, all cases required the same tooth movement pattern, which comprised tilting of the anterior teeth by proclination prior to intrusion. Additionally, the intrusion pattern was sequential, which indicated the respective intrusion of the incisors and then the canines. Secondly, attachments were used in the premolar region to facilitate tooth movement. Boyd³ reported that a rectangular attachment bonded to the premolar could resist the reaction force produced by intrusion of the anterior teeth, increase retention of the clear aligner, and improve the effectiveness of vertical movement of the anterior teeth. Therefore, the patients had rectangular attachments with strong retention attached in the premolar area.

The average reduction in the COS achieved by the clear aligners was 0.901 ± 0.455 mm, which was higher than the values reported by Goh et al.²⁴ (-0.29 mm) but lower than those reported by Rozzi et al.²⁵ (-2.2 mm). The levelling effectiveness was 71.8 ± 19.0%, which exceeded that reported by Goh et al. This may be related to differences in the methods employed in each case, since Rozzi et al. measured the deepest point of the mandibular arch as the COS depth. The enhanced accuracy in the present study, compared to that of Goh et al., could be due to a difference in treatment objectives. The participants in their study were ≥ 18 years of age before treatment, whereas the present study involved patients ≥ 14 years old. The COS depth decreased slightly during the adolescent dentition stage, which took place at the same time as the procedure for COS levelling. This may have resulted in a higher effectiveness of the levelling procedure in younger individuals. Another factor might be variations in attachment type, number, size, and location, or the amount and type of tooth movement. Because modifications of the Clincheck plan may be made by orthodontists, treatment outcomes may differ according to the specific modification employed.

The results of the present study suggest that the levelling effectiveness of clear aligners is influenced by the patient’s vertical skeletal pattern. Nevertheless, the MP-FH only explained 2.5% of the variance in levelling effectiveness, which is similar to the correlation between facial type and alveolar bone morphology of the mandible demonstrated previously.^31–34 Handelman et al.³¹ showed that cases with a low mandibular plane angle exhibited thicker bone lingual to the maxillary and mandibular incisors, while cases with a high mandibular angle had thinner bone labial to the mandibular incisors. Gracco et al.³² observed that the total thickness of the symphysis was greater in individuals with a short face than in those with a long face. Further, as bone density plays an important role in facilitating orthodontic tooth movement,³³ levelling effectiveness may also be influenced by differences in mandibular symphysis bone density³⁴ and mandibular remodelling time after orthodontic force application in patients with different skeletal patterns.

It was observed that levelling effectiveness was positively associated with IMPA. This finding supports the conclusion that labial tipping of the anterior teeth may facilitate COS levelling. Previous studies have reported that lip inclination affects the vertical position of the incisor edge, which can facilitate the correction of a deep overbite.^23,35,36 Pandis et al.²³ demonstrated that the sole predictor of curve flattening is the lower incisor to mandibular plane angle, with 1 mm of levelling being achieved with a 4° proclination of the mandibular incisors (without an increase in arch width).

Of the dental-related measurements investigated in the present study, the parallelism of the mandibular incisors had a significant influence on levelling effectiveness. Although previous studies^37–39 have reported that mesiodistal tooth angulations on panoramic films are inaccurate, radiographic distortion is primarily limited to the maxillary and mandibular canines. Moreover, the utility of panoramic radiographs to measure incisor parallelism has been demonstrated in a prior study.⁴⁰ Mandibular incisor parallelism reflects the axial inclination of the mandibular anterior teeth and their occupation of mesiodistal space.¹⁴ This parameter can be used to indicate discrepancies in the quantity and osseous volume in the mandibular anterior tooth region. Higher values reflect a greater degree of convergence of the mandibular anterior teeth, which in turn indicates that the anterior mesio-distal space is insufficient, and that COS levelling will be more difficult.

It was found that levelling effectiveness was also influenced by the depth of the COS before treatment, which explained 53.1% of the total variance. This further indicates that an increased COS can be altered by a greater amount during treatment, resulting in a higher levelling efficiency. Conversely, with the same amount of treatment movement per aligner, patients with a deep COS would require a greater number of aligners and experience a longer period of treatment to make their levelling more efficient.

It is generally believed that 1 mm of arch space is required for the levelling of the COS by 1 mm on each side.⁴¹ Previous studies have demonstrated that the space required for levelling the COS is often overestimated.^42,43 The results show that levelling effectiveness was not statistically significantly correlated with crowding analysis, which may be related to the inclusion of adolescent participants. Generally, adolescents still experience some growth in the length and width of the mandibular dental arch.⁴⁴ This increased arch length and width improves crowding to some degree. Therefore, the crowding analysis did not result in any significant findings. However, further studies with larger sample sizes are needed to investigate this.

The present study has several limitations. Firstly, although orthodontic treatment was provided by two orthodontists with a high level of experience and strictly adhering to treatment protocols, the risk of selection bias was difficult to avoid. Secondly, owing to the limitations of the retrospective study design, cone-beam computed tomography was not performed post-treatment. This data would have allowed further analysis of angular changes of the teeth. Nevertheless, digital models can be an acceptable and an accurate alternative for measurements (particularly COS depth) taken from dental models.^45–47 Thirdly, the age range of the included patients was between 14 and 36 years. While the depth of the COS reportedly changes slightly with age,⁴⁸ this issue could not be avoided, as orthodontic treatment is a long-term process. Marshall et al.⁴⁹ suggested that the depth of the COS decreases slightly during the adolescent dentition stage, subsequently remaining relatively stable into early adulthood. These results are similar to those observed and resulted in more effective levelling. Massaro et al.⁵⁰ reported that the COS remained stable from the age of 17 to 48 years. Future studies should account for different age groups in their analyses to obtain more accurate results, provided that sample sizes are sufficient.

Overall, the findings of the present study suggest that levelling the COS using clear aligners requires multiple considerations. Additional insight into the morphological and functional significance of COS may further improve the stability of orthodontic treatment. Li et al.⁵¹ used multilayer perception artificial neural networks to predict the outcomes of orthodontic treatment plans, and the COS was one of the most important factors determining these predictions. Future studies on the COS may consider digital model measurements and longitudinal and cross-sectional study designs comprising different age and ethnic groups.

Clear aligners effectively level the COS. Patients with a high MP-FH, a deep COS before treatment, a reduced IMPA, and splayed mandibular incisors are prime candidates for COS levelling using clear aligners.