Cone-beam computed tomography evaluation of alveolar bone and root changes after clear aligner therapy with different extraction protocols: Balancing tissue loss, tooth control, and treatment alternatives

Abstract

Objective

To evaluate changes in alveolar bone and tooth root dimensions in anterior teeth of patients with different tooth extraction types undergoing clear aligner therapy (CAT) and to provide reliable information for preventing tissue loss and providing tooth control in severe cases through a large-scale sample analysis of the clinical outcomes of CAT.

Methods

We selected 281 patients (186 non-extraction [NE], 59 with two-premolar extraction [TPE] in both the maxilla and mandible, and 36 with TPE in the maxilla and one lower-incisor extraction [OLIE] in the mandible) from the records of recent three years. Quantitative changes in the dentoalveolar apparatus were analyzed using pre- (T1) and post-treatment (T2) cone-beam computed tomography. The measured parameters included the alveolar bone height and thickness, and root length in the anterior teeth in different types of tooth extraction.

Results

Alveolar bone height loss was common in all groups after CAT. Compared to patients with NE, patients with TPE showed a higher risk of lingual bone dehiscence and torque loss (P < 0.05), whereas those with OLIE showed a higher risk of open gingival embrasures (P < 0.05). A more severe alveolar bone loss was observed in the mandibular anterior teeth than in the maxillary anterior teeth (P < 0.05).

Conclusions

Different tooth extraction types can lead to different degrees of bone loss in the direction of tooth movement, and orthodontists should adopt more cautious measures for mandibular anterior teeth. Despite numerous experimental studies for improving techniques and designs in CAT, tooth control and complication prevention in extraction cases remain challenging for orthodontists.

INTRODUCTION

Clear aligner therapy (CAT) is a rapidly evolving domain in modern orthodontics, which improves patient acceptance of the appliances and simplifies oral hygiene procedures.¹ Clear aligners (CAs) were initially used for mild to moderate malocclusion with highly predictable treatment efficacy,² but are now increasingly used in more complex cases, such as presurgical orthodontic and extraction cases.^3,4

In contemporary orthodontics, extraction is common, and approximately 30% of patients require extractions of premolars to manage malocclusion.⁵ However, 60% of practitioners rarely combine CAT with premolar extractions.^6,7 Moreover, owing to the its recent development and the limited number of patients treated using CAT, large-scale studies on CAT in different types of extractions are scarce, especially in lower incisor extractions. This may be because CAs have an advantage in distalizing molars; therefore, the possibility of choosing non-extraction (NE) treatment in borderline cases increases with CAT.⁸ Therefore, alternate treatment for extraction in CAT warrants further exploration. Most previous studies and our preliminary research^9,10 focused on comparing fixed orthodontic appliances and CAs in patients undergoing the same type of extractions; however, no comparative studies have included different extraction types in CAT.

Different extraction types correspond to different degrees of bone remodeling, and bone loss after orthodontic treatment may be a side effect of orthodontic extraction in specific types of malocclusion.^11,12 The morphology of the alveolar bone is typically quantified as bone height and thickness. Meanwhile, orthodontically induced inflammatory root resorption (OIIRR) is common in CAT.¹³ Additionally, cone-beam computed tomography (CBCT) overcomes the limitations of panoramic and periapical films; it provides accurate information in the coronal and sagittal planes. An overall understanding of bone remodeling and OIIRR assessed using CBCT in different types of extraction cases helps in the selection of appropriate treatment alternatives and prevention of any iatrogenic periodontal sequelae, such as bone dehiscence and fenestration.

This retrospective study aimed to conduct a large-scale sample analysis in adult patients undergoing CAT and evaluate the presence of alveolar bone changes and root resorption in anterior teeth with different extraction protocols, to provide clinical references for case selection and risk assessment in CAT.

MATERIALS AND METHODS

Records of healthy adult patients for 3 years after undergoing CAT (Invisalign^®, Align Technology, Tempe, AZ, USA) in the Department of Orthodontics, Stomatological Hospital of the Air Force Military Medical University, Xi’an, China, were reviewed in this study. Maxillary or mandibular anterior teeth were selected as the research objects. The inclusion criteria were as follows: Patients with (1) no periodontal conditions before or during orthodontic treatment; (2) age ≥ 18 years; and (3) clear CBCT scans at pre-treatment (T1) and post-treatment (T2). The exclusion criteria were as follows: Patients with (1) maxillary or mandibular anterior teeth that were previously restored or had undergone root canal treatment; (2) orthodontic treatment history; (3) interproximal enamel reduction during orthodontic treatment; and (4) history of trauma or surgery in the anterior region before or during treatment. Data pertaining to 281 patients (186 with NE, 59 with two-premolar extraction [TPE] in both maxilla and mandible, and 36 with TPE in the maxilla and one lower-incisor extraction [OLIE] in the mandible) were selected from the original records. In the maxillary or mandibular arch analyzed separately, NE, TPE or OLIE groups were used to represent different extraction protocols for each individual arch, respectively. The data were selected with a similar sex ratio of patients. The demographic and pre-treatment characteristics of the patients are shown in Table 1. The study protocol was approved by the Ethics Committee of the Stomatological Hospital of the Air Force Military Medical University (IRB - REV - 2022130). Consent for participation was not required.

Table 1.

Demographic and pre-treatment characteristics of the sample

Variable	Maxillary		P value	Mandibular			P value
	NE (n = 186)	TPE (n = 95)		NE (n = 186)	TPE (n = 59)	OLIE (n = 36)
Age (yr)	26.5 ± 2.1	25.4 ± 3.6	0.379	26.5 ± 2.1	25.5 ± 3.5	25.3 ± 3.6	0.435
Treatment duration (mo)	23.5 ± 10.9	25.1 ± 8.3	0.153	23.5 ± 10.9	25.3 ± 8.3	24.6 ± 6.8	0.132
ANB (°)	2.5 ± 3.2	3.7 ± 2.9	0.027*	2.5 ± 3.2a	3.9 ± 2.9b	3.2 ± 2.6b	0.036*
Overjet (mm)	1.4 ± 2.4	2.4 ± 3.9	0.197	1.4 ± 2.4	2.5 ± 3.9	2.0 ± 3.2	0.369
Anterior teeth crowding (mm)	1.8 ± 2.7	2.3 ± 1.5	0.100	1.6 ± 3.2a	2.4 ± 1.9a	3.3 ± 2.3b	0.032*
U1-SN/L1-MP (°)	106.3 ± 5.8	109.3 ± 4.9	0.049*	93.7 ± 3.4a	96.2 ± 4.0b	96.7 ± 3.7b	0.043*

The superscript letters ‘a’ and ‘b’ indicate a significant difference between groups.

NE, non-extraction; TPE, two-premolar extraction; OLIE, one lower-incisor extraction.

*P < 0.05.

CBCT (KaVo 3D eXam, KaVo Dental GmbH, Biberach, Germany; 120 kV, 5 mA, voxel size of 0.3 mm, and 9.6-second scan time) data, obtained at T1 and T2, were uploaded in the Digital Imaging and Communications in Medicine format to Dolphin Imaging version 11.9 (Chatsworth, CA, USA) for storage and interpretation. Reconstruction was performed such that the long axis of each tooth or root was perpendicular to the axial slices. Image slices were automatically reconstructed perpendicular to the axial slices. This results in optimal visualization of the alveolar crest (AC) relative to the cementoenamel junction (CEJ) in axial, coronal, and sagittal views, as previously mentioned.¹⁴ Using the axial view landmarks as a guide, two reference lines were placed: one between the CEJs on the labial and palatal/lingual surfaces and one between the CEJs on the mesial and distal surfaces. Parallel to these two lines, additional lines were placed at the AC on the labial, palatal/lingual, mesial, and distal surfaces.¹¹ Bone height was defined as the vertical distance from the reference line (CEJ) to the AC on the labial (LAAC-CEJ), lingual (LIAC-CEJ), mesial (MAC-CEJ), and distal (DAC-CEJ) surfaces (Supplementary Figure 1). The thickness of the labial alveolar bone thickness at the apex (LAABT-apex) and 3, 6, and 9 mm (LAABT-3, LAABT-6, LAABT-9, respectively) from the reference line on the buccal CEJ. Similarly, thickness of the palatal/lingual alveolar bone was measured (LIABT-apex, lingual alveolar bone thickness 3/6/9 mm from the CEJ [LIABT-3, LIABT-6, and LIABT-9]; Supplementary Figure 2). Root length was measured in the sagittal view, and distance between the interproximal contact point and AC (ICP-AC) was measured in the coronal view (Supplementary Figure 3). Δ indicates the change between post-treatment and pre-treatment values (Δ = T2−T1). Patient scans were assessed and measurements were obtained twice over a two-month period by a single author (YZ).

Data obtained from ClinCheck^® (Invisalign^®, Align Technology) and clinical records of 59 patients with TPE in both the maxilla and mandible revealed the following: first, all cases involved programmed torque compensation in the incisors, with an average value of + 9°; second, 59.2% cases involved power ridges, with root-lingual torque values exceeding 3°; and third, despite the absence of standardized protocols for the sequence of anterior retraction, the general clinical practice involved initial alignment and leveling, followed by adjustment of the overbite. Retraction was initiated once space was evident between the lateral incisors and canines. Lastly, since optimized attachments intended specifically for torque control were not introduced for the CA system used in this study (Invisalign^®), none were applied in any of the included cases.

Statistical analysis

The inter-rater reliability of the measurements was assessed using the Intraclass Correlation Coefficient, which was 0.91. The normality of the data was evaluated using the Shapiro–Wilk test. For normally distributed data, t tests and one-way analysis of variance were employed to identify statistically significant differences, whereas the Mann–Whitney U test and Kruskal–Wallis H test were used for non-normally distributed data. Post-hoc analyses were conducted using Tukey’s HSD test (for parametric data) or Dunn’s test with Bonferroni adjustment (for non-parametric data) to control the family-wise error rate during multiple pairwise comparisons. Statistical analyses were conducted using SPSS software version 24.0 (IBM Corp., Armonk, NY, USA), and figures were created with GraphPad Prism, version 8.0.2 (GraphPad Software, San Diego, CA, USA). All statistical tests were two-tailed, and P < 0.05 was set as the threshold for statistical significance.

RESULTS

A descriptive comparison of the maxillary anterior teeth in different tooth extraction types is presented in Table 2. Regarding the alveolar bone height in patients with NE and TPE, the LIAC-CEJ at T2 was 1.48 ± 1.74 mm and 2.83 ± 1.48 mm for central incisors, 1.76 ± 2.13 mm and 3.08 ± 2.44 mm for lateral incisors, and 1.99 ± 1.22 mm and 2.87 ± 1.64 mm for canines, respectively (P < 0.01), while the same trend was seen in the ΔLIAC-CEJ among central incisors, later incisors and canines (P < 0.05). The LIAC-CEJ at T2 in patients with TPE were > 2 mm. Regarding changes in alveolar bone thickness, a significantly smaller change in LAABT were measured in patients with NE compared with TPE (P < 0.05; Table 2): the significant NE values were ΔLAABT-3 (0.03 ± 0.36 mm) and ΔLAABT-6 (–0.01 ± 0.45 mm) for central incisors and ΔLAABT-3 (–0.07 ± 0.31 mm), ΔLAABT-9 (–0.02 ± 0.43 mm), and ΔLAABT-apex (0.00 ± 0.80 mm) for canines. On the other hand, a significantly larger NE change in LIABT values compared to TPE were exhibited in central incisors (ΔLIABT-3 [–0.11 ± 0.67 mm] and ΔLIABT-6 [0.01 ± 1.15 mm]; P < 0.05). The horizontal alveolar bone thickness at the apex (ΔLABT-apex; 1.65 ± 4.75 mm) for canines in patients with TPE was significantly greater than that in patients with NE (P < 0.05). After treatment, the ICP-AC in patients with TPE showed a statistically significant increasing trend compared to that in patients with NE (ΔICP-AC [0.85 ± 1.37 mm] and ICP-AC at T2 [5.82 ± 1.14 mm] for central incisors and ΔICP-AC [1.15 ± 1.02 mm] for lateral incisors; P < 0.05). Overall, in the labial-lingual direction, the height and thickness of the labial alveolar bone in patients with TPE showed an increasing trend compared to that in patients with NE, whereas the height and thickness of the lingual alveolar bone showed a decreasing trend. In the mesial-distal direction, ICP-AC after treatment showed an increasing trend in patients with TPE compared to that in patients with NE.

Table 2.

Tooth and alveolar bone morphology in maxillary anterior teeth of patients with different tooth extraction types

Variable	Central incisor				Lateral incisor				Canine
	NE (n = 186)	TPE (n = 95)	P value		NE (n = 186)	TPE (n = 95)	P value		NE (n = 186)	TPE (n = 95)	P value
ΔMAC-CEJ	0.17 ± 0.71	0.32 ± 0.82	0.302		0.20 ± 0.51	0.38 ± 0.65	0.097		0.11 ± 0.60	0.17 ± 0.73	0.658
ΔDAC-CEJ	0.00 ± 0.63	0.06 ± 0.52	0.594		0.13 ± 0.57	0.19 ± 0.54	0.606		0.05 ± 0.55	0.03 ± 1.09	0.899
ΔLAAC-CEJ	–0.04 ± 1.29	–0.07 ± 0.71	0.895		0.32 ± 1.73	0.09 ± 1.64	0.455		0.68 ± 3.36	0.06 ± 1.00	0.230
ΔLIAC-CEJ	0.09 ± 0.71	0.87 ± 0.77	0.007**		0.43 ± 1.13	1.63 ± 1.66	0.006**		0.43 ± 1.10	0.97 ± 1.37	0.038*
MAC-CEJ at T2	1.93 ± 0.88	1.90 ± 0.71	0.799		2.18 ± 0.62	1.99 ± 0.68	0.093		1.94 ± 0.74	1.87 ± 0.64	0.598
DAC-CEJ at T2	1.92 ± 0.71	1.93 ± 0.56	0.920		2.31 ± 0.63	2.21 ± 0.56	0.341		2.11 ± 0.76	1.99 ± 0.84	0.398
LAAC-CEJ at T2	2.12 ± 1.62	1.82 ± 0.78	0.234		2.66 ± 1.61	2.58 ± 1.25	0.758		3.02 ± 3.29	2.49 ± 0.98	0.154
LIAC-CEJ at T2	1.48 ± 1.74	2.83 ± 1.48	0.001**		1.76 ± 2.13	3.08 ± 2.44	0.001**		1.99 ± 1.22	2.87 ± 1.64	0.002**
ΔLAABT-3	0.03 ± 0.36	0.26 ± 0.57	0.017*		0.00 ± 0.37	–0.04 ± 0.59	0.683		–0.07 ± 0.31	0.08 ± 0.35	0.016*
ΔLAABT-6	–0.01 ± 0.45	0.24 ± 0.52	0.004**		0.10 ± 0.34	0.14 ± 0.48	0.600		–0.05 ± 0.38	0.05 ± 0.57	0.287
ΔLAABT-9	0.04 ± 0.63	0.27 ± 0.73	0.079		0.11 ± 0.49	0.27 ± 0.99	0.326		–0.02 ± 0.43	0.20 ± 0.60	0.043*
ΔLAABT-apex	0.16 ± 0.88	0.36 ± 1.08	0.232		0.07 ± 0.88	0.38 ± 1.02	0.092		0.00 ± 0.80	0.33 ± 1.02	0.047*
ΔLIABT-3	–0.11 ± 0.67	–0.45 ± 0.52	0.004**		–0.16 ± 0.52	–0.32 ± 0.64	0.121		–0.15 ± 0.61	–0.30 ± 0.68	0.197
ΔLIABT-6	0.01 ± 1.15	–0.54 ± 1.01	0.008**		–0.19 ± 0.83	–0.25 ± 1.47	0.812		–0.19 ± 0.88	–0.44 ± 0.90	0.131
ΔLIABT-9	–0.09 ± 1.94	–0.76 ± 2.11	0.086		–0.23 ± 1.28	–0.15 ± 2.03	0.830		–0.37 ± 1.20	–0.72 ± 2.63	0.295
ΔLABT-3	–0.03 ± 0.50	–0.12 ± 0.70	0.422		–0.10 ± 0.55	–0.28 ± 0.64	0.108		–0.31 ± 1.38	0.09 ± 1.61	0.144
ΔLABT-6	0.02 ± 0.77	–0.17 ± 0.83	0.193		–0.05 ± 0.68	–0.22 ± 1.29	0.413		–0.38 ± 1.50	–0.01 ± 1.88	0.217
ΔLABT-9	0.13 ± 1.15	–0.35 ± 1.95	0.093		–0.12 ± 0.90	–0.15 ± 2.09	0.940		0.19 ± 6.55	0.12 ± 2.25	0.947
ΔLABT-apex	0.19 ± 1.12	0.00 ± 1.56	0.422		–0.08 ± 0.97	0.33 ± 2.08	0.226		–0.49 ± 5.22	1.65 ± 4.75	0.016*
ΔRL	–0.34 ± 0.72	–0.59 ± 0.86	0.073		–0.02 ± 0.82	–0.34 ± 1.07	0.088		–0.03 ± 1.36	–0.36 ± 1.09	0.170
ΔICP-AC	–0.20 ± 0.96	0.85 ± 1.37	0.003**		0.37 ± 1.20	1.15 ± 1.02	0.001**		0.18 ± 1.08	0.61 ± 0.97	0.080
ICP-AC at T2	5.13 ± 1.13	5.82 ± 1.14	0.022*		4.69 ± 0.95	4.80 ± 0.97	0.562		4.53 ± 1.18	4.80 ± 0.78	0.239

Unit: mm.

T1, pre-treatment; T2, post-treatment; Δ, T2-T1; CEJ, cementoenamel junction; AC, alveolar crest; MAC/DAC/LAAC/LIAC-CEJ, distance from mesial/distal/labial/lingual AC to the CEJ; LAABT/LIABT-3/6/9, labial/lingual alveolar bone thickness 3/6/9 mm from the CEJ; LAABT-apex, labial alveolar bone thickness at the apex; LABT-3/6/9, horizontal alveolar bone thickness 3/6/9 mm from the CEJ; LABT-apex, horizontal alveolar bone thickness at the apex; RL, root length; ICP-AC, distance between the interproximal contact point and AC; NE, non-extraction; TPE, two-premolar extraction.

*P < 0.05, **P < 0.01.

A descriptive comparison of the mandibular anterior teeth in different tooth extraction types is presented in Table 3. As for the alveolar bone height, the ΔLIAC-CEJ for central incisors was 0.26 ± 1.38 mm, 1.07 ± 1.57 mm, and –0.02 ± 1.58 mm and for lateral incisors was –0.01 ± 1.50 mm, 0.93 ± 1.64 mm, and 0.99 ± 2.09 mm, respectively in the three groups (all P < 0.05). The LIAC-CEJ at T2 for central incisors was 2.70 ± 1.57 mm, 3.62 ± 1.96 mm, and 3.17 ± 2.18 mm, and for lateral incisors was 2.44 ± 1.26 mm, 3.12 ± 1.50 mm, and 3.23 ± 2.02 mm, respectively (all P < 0.05). Similar to that in the maxilla, the change in LAABT in patients with TPE was greater than that in patients with NE: the significant TPE values were ΔLAABT-6 (0.19 ± 0.49 mm) for central incisors and ΔLAABT-6 (0.33 ± 0.59 mm) and ΔLAABT-9 (0.67 ± 0.80 mm) for canines (P < 0.05). In contrast, the change in LIABT in patients with TPE was smaller than that in patients with NE: the significant TPE values were ΔLIABT-6 (–0.50 ± 0.72 mm) for lateral incisors and ΔLIABT-3 (–0.45 ± 0.75 mm) and ΔLIABT-6 (–0.68 ± 0.79 mm) for canines (P < 0.05). Differences were also observed between NE and TPE groups in ΔLABT for mandibular canines (P < 0.05): ΔLABT-3 (0.02 ± 0.58 mm and –0.41 ± 0.77 mm, respectively) and ΔLABT-6 (–0.03 ± 0.71 mm and –0.43 ± 0.89 mm, respectively). In terms of mesiodistal alveolar bone height, ICP-AC at T2 in the NE group (5.12 ± 1.40 mm for central incisors, 5.12 ± 0.12 mm for lateral incisors, and 4.58 ± 1.06 mm for canines) was lower than that in the other two groups (P < 0.05). Interestingly, the highest DAC-CEJ at T2 was observed in the OLIE group: 3.04 ± 0.71 mm for central incisors and 2.20 ± 0.85 mm for canines. The lowest ΔMAC-CEJ was observed in the NE group: 0.14 ± 0.60 mm for central incisors. Importantly, root resorption of canines in patients with NE (0.08 ± 1.02 mm) was less than that in patients with TPE (–0.51 ± 0.65 mm) and OLIE (–0.64 ± 1.18 mm) (P < 0.001). In short, the index values in the mandibular labial-lingual direction of TPE and NE showed a trend similar to that in the maxilla. In the mesial-distal direction, patients with NE showed the smallest ICP-AC after treatment compared to patients with TPE and OLIE, while patients with OLIE showed the highest DAC-CEJ at T2 compared to patients with NE and TPE.

Table 3.

Tooth and alveolar bone morphology in mandibular anterior teeth of patients with different tooth extraction types

Variable	Central incisor					Lateral incisor					Canine
	NE (n = 186)	TPE (n = 59)	OLIE (n = 36)	P value		NE (n = 186)	TPE (n = 59)	OLIE (n = 36)	P value		NE (n = 186)	TPE (n = 59)	OLIE (n = 36)	P value
ΔMAC-CEJ	0.14 ± 0.60a	0.52 ± 0.80b	0.84 ± 0.37b	0.019*		0.10 ± 0.53	0.11 ± 0.67	0.15 ± 0.27	0.960		0.11 ± 0.65	0.15 ± 1.00	0.20 ± 0.66	0.069
ΔDAC-CEJ	0.15 ± 0.56	0.40 ± 0.82	–0.02 ± 0.83	0.234		0.20 ± 0.65	–0.02 ± 0.93	0.05 ± 0.47	0.364		–0.01 ± 0.53	0.01 ± 0.53	0.08 ± 0.50	0.085
ΔLAAC-CEJ	0.98 ± 2.32	0.76 ± 1.70	0.79 ± 2.99	0.881		0.73 ± 2.25	0.95 ± 2.38	0.48 ± 0.91	0.810		0.41 ± 2.83	0.29 ± 3.32	0.37 ± 1.30	0.976
ΔLIAC-CEJ	0.26 ± 1.38b	1.07 ± 1.57a	–0.02 ± 1.58b	0.043*		–0.01 ± 1.50a	0.93 ± 1.64b	0.99 ± 2.09b	0.009**		0.39 ± 1.19	0.80 ± 1.13	0.77 ± 1.61	0.191
MAC-CEJ at T2	2.34 ± 0.79	2.44 ± 0.83	3.15 ± 1.40	0.097		2.27 ± 0.70	2.31 ± 0.89	2.57 ± 0.87	0.403		2.08 ± 0.73	2.21 ± 0.72	2.45 ± 0.95	0.139
DAC-CEJ at T2	2.34 ± 0.75a	2.46 ± 0.98a	3.04 ± 0.71b	0.007**		2.12 ± 0.79	2.16 ± 0.88	2.39 ± 0.87	0.538		1.73 ± 0.61a	1.89 ± 0.69ab	2.20 ± 0.85b	0.015*
LAAC-CEJ at T2	2.99 ± 2.44	2.97 ± 1.57	3.68 ± 2.73	0.537		3.08 ± 2.81	3.87 ± 2.58	2.68 ± 1.89	0.321		3.65 ± 2.92	3.47 ± 2.42	2.91 ± 1.79	0.063
LIAC-CEJ at T2	2.70 ± 1.57a	3.62 ± 1.96b	3.17 ± 2.18ab	0.043*		2.44 ± 1.26a	3.12 ± 1.50b	3.23 ± 2.02b	0.032*		2.06 ± 1.34	2.61 ± 1.14	2.73 ± 1.65	0.054
ΔLAABT-3	–0.14 ± 0.34	–0.20 ± 0.40	–0.20 ± 0.52	0.702		–0.18 ± 0.41	–0.11 ± 0.35	–0.02 ± 0.23	0.328		–0.06 ± 0.31	–0.04 ± 0.33	–0.20 ± 0.26	0.132
ΔLAABT-6	–0.08 ± 0.38a	0.19 ± 0.49b	–0.19 ± 0.61a	0.044*		–0.03 ± 0.36	–0.16 ± 0.46	–0.05 ± 0.23	0.147		0.02 ± 0.33a	0.33 ± 0.59b	–0.08 ± 0.75a	0.029*
ΔLAABT-9	–0.19 ± 0.91	0.07 ± 0.83	0.25 ± 0.93	0.124		0.00 ± 0.70	0.30 ± 0.97	0.18 ± 0.60	0.218		0.22 ± 0.67a	0.67 ± 0.80b	0.31 ± 0.58ab	0.009**
ΔLAABT-apex	–0.08 ± 0.98	0.18 ± 1.09	0.31 ± 0.92	0.233		0.38 ± 0.95	0.13 ± 2.00	0.26 ± 1.19	0.804		0.75 ± 1.13	0.70 ± 1.48	0.27 ± 1.46	0.282
ΔLIABT-3	0.08 ± 0.44	–0.16 ± 0.47	0.06 ± 0.62	0.063		0.17 ± 0.54	–0.12 ± 0.59	–0.04 ± 0.51	0.050		0.02 ± 0.47a	–0.45 ± 0.75b	–0.11 ± 0.50ab	0.011*
ΔLIABT-6	0.11 ± 0.67	–0.15 ± 0.53	–0.08 ± 0.75	0.150		0.11 ± 0.67a	–0.50 ± 0.72b	–0.10 ± 0.65ab	0.001**		–0.14 ± 0.70a	–0.68 ± 0.79b	–0.38 ± 0.61ab	0.002**
ΔLIABT-9	–0.01 ± 1.02	0.06 ± 0.63	–0.15 ± 1.11	0.797		–0.20 ± 0.87	–0.42 ± 0.84	–0.04 ± 0.39	0.189		–0.33 ± 0.74	–0.71 ± 0.79	–0.47 ± 0.65	0.053
ΔLABT-3	0.02 ± 0.42	–0.17 ± 0.49	0.03 ± 0.46	0.128		0.08 ± 0.58	–0.25 ± 0.71	0.02 ± 0.45	0.060		0.02 ± 0.58a	–0.41 ± 0.77b	–0.26 ± 0.58ab	0.004**
ΔLABT-6	0.10 ± 0.53	0.06 ± 0.44	–0.04 ± 0.63	0.304		0.12 ± 0.57a	–0.46 ± 0.88b	–0.17 ± 0.63ab	0.010*		–0.03 ± 0.71a	–0.43 ± 0.89b	–0.31 ± 0.88ab	0.035*
ΔLABT-9	–0.05 ± 0.95	0.05 ± 0.55	0.24 ± 0.56	0.455		–0.05 ± 0.95	–0.26 ± 0.83	–0.05 ± 0.56	0.581		–0.03 ± 0.73	–0.14 ± 0.88	–0.17 ± 0.66	0.655
ΔLABT-apex	–0.09 ± 1.06	0.03 ± 0.61	0.19 ± 0.77	0.529		0.08 ± 0.57	–0.13 ± 1.38	0.04 ± 0.97	0.766		0.18 ± 0.78	0.25 ± 0.92	–0.28 ± 1.04	0.073
ΔRL	–0.33 ± 0.59	–0.37 ± 0.56	–0.35 ± 0.77	0.952		–0.16 ± 0.81	–0.46 ± 1.07	–0.50 ± 0.63	0.183		0.08 ± 1.02a	–0.51 ± 0.65b	–0.64 ± 1.18b	0.001**
ΔICP-AC	0.47 ± 1.50	1.43 ± 1.43	1.64 ± 1.76	0.054		0.93 ± 1.30a	2.13 ± 1.53b	1.44 ± 1.60ab	0.001**		0.39 ± 1.20	1.14 ± 1.92	0.93 ± 1.02	0.054
ICP-AC at T2	5.12 ± 1.40a	5.95 ± 1.44b	6.66 ± 1.25b	0.024*		5.12 ± 0.12a	5.68 ± 0.34b	6.25 ± 0.50b	< 0.001**		4.58 ± 1.06a	5.12 ± 0.84b	5.77 ± 0.68b	< 0.001**

Unit: mm.

The superscript letters ‘a’ and ‘b’ indicate a significant difference between groups.

T1, pre-treatment; T2, post-treatment; Δ, T2-T1; CEJ, cementoenamel junction; AC, alveolar crest; MAC/DAC/LAAC/LIAC-CEJ, distance from mesial/distal/labial/lingual AC to the CEJ; LAABT/LIABT-3/6/9, labial/lingual alveolar bone thickness 3/6/9 mm from the CEJ; LAABT-apex, labial alveolar bone thickness at the apex; LABT-3/6/9, horizontal alveolar bone thickness 3/6/9 mm from the CEJ; LABT-apex, horizontal alveolar bone thickness at the apex; RL, root length; ICP-AC, distance between the interproximal contact point and AC; NE, non-extraction; TPE, two-premolar extraction; OLIE, one lower-incisor extraction.

*P < 0.05, **P < 0.01.

Figure 1 illustrates the changes in LIAC-CEJ in the maxillary and mandibular anterior teeth. Compared to patients with extraction, patients with NE exhibited minimal ΔLIAC-CEJ, and LIAC-CEJ at T2 in maxillary anterior teeth did not exceed the 2 mm threshold. However, in patients with TPE, LIAC-CEJ at T2 in both the maxilla and mandible exceeded 2 mm. The loss of LIAC-CEJ in the mandibular anterior teeth is generally greater than that in the maxillary anterior teeth. In patients with OLIE, even with minimal sagittal movement of the mandibular anterior teeth, considerable LIAC-CEJ loss was observed after treatment. Figure 2 shows the changes in ICP-AC in the maxillary and mandibular anterior teeth. Compared to patients with extraction, patients with NE exhibited minimal ΔICP-AC and changes in ICP-AC at T2, while those with OLIE showed the highest ICP-AC at T2.

Figure 1.

Error bar plot of mean LIAC-CEJ at T2 and ΔLIAC-CEJ in maxillary and mandibular anterior teeth in patients with different tooth extraction types. A, ΔLIAC-CEJ in maxillary anterior teeth in patients with different tooth extraction type. B, ΔLIAC-CEJ in mandibular anterior teeth in patients with different tooth extraction type. C, LIAC-CEJ at T2 in maxillary anterior teeth in patients with different tooth extraction type. D, LIAC-CEJ at T2 in mandibular anterior teeth in patients with different tooth extraction type.

LIAC-CEJ, distance from lingual alveolar crest to cementoenamel junction; ΔLIAC-CEJ, change in LIAC-CEJ between T2 and T1; T1, pre-treatment; T2, post-treatment; NE, non-extraction; TPE, two-premolar extraction; OLIE, one lower-incisor extraction.

*P < 0.05.

Figure 2.

Error bar plot of mean ICP-AC a T2 and ΔICP-AC in maxillary and mandibular anterior teeth in patients with different tooth extraction type. A, ΔICP-AC in maxillary anterior teeth in patients with different tooth extraction type. B, ΔICP-AC in mandibular anterior teeth in patients with different tooth extraction type. C, ICP-AC at T2 in maxillary anterior teeth in patients with different tooth extraction type. D, ICP-AC at T2 in mandibular anterior teeth in patients with different tooth extraction type.

ICP-AC, distance between the interproximal contact point and AC; ΔICP-AC, change in ICP-AC between T2 and T1; T1, pre-treatment; T2, post-treatment; NE, non-extraction; TPE, two-premolar extraction; OLIE, one lower-incisor extraction.

*P < 0.05.

DISCUSSION

Compared to fixed appliances, CAs exhibit distinct biomechanical properties, with treatment outcomes being significantly reliant on patient adherence. Their applications have expanded to include more complex cases involving various extraction strategies. Nevertheless, comparative studies evaluating different extraction types in adults undergoing CAT are lacking. Unlike previous studies focusing on fixed orthodontics,^11,15,16 we conducted a large-scale analysis of CAT in patients with different extraction types, including relatively rare OLIE cases. We identified extraction-specific bone remodeling patterns and risks of tissue loss, thereby enhancing the generalizability of the findings.

The ΔMAC-CEJ, ΔDAC-CEJ, ΔLAAC-CEJ, and ΔLIAC-CEJ increased in nearly all groups after treatment (Tables 2 and 3). Loss of alveolar bone height or thickness is common after orthodontic treatment.¹⁷ There was a greater loss of lingual bone height in patients with TPE (LIAC-CEJ at T2 > 2 mm) than in patients with NE. Meanwhile, the lingual bone thickness decreased in patients with TPE than in those with NE (Tables 2 and 3, Figure 1). Indeed, previous studies have found that in patients with TPE, alveolar bone height may decrease after anterior teeth retraction, but lingual thickness decrease was the main problem,^18-20 which was significantly correlated with the amount of anterior tooth retraction, particularly in the cervical region.^21-23 Bone or soft tissue augmentations are mostly performed on the labial or buccal side, and are not easy to conduct on the lingual side;²⁴ meanwhile, lingual bone dehiscence is not easily detected and treated but poses a significant dental risk in patients. Additionally, our results showed that, although less than that in patients with extractions, patients with NE showed generalized bone loss after CAT (Tables 2 and 3). Excessive bodily advancement or proclination of the anterior teeth to overly pursue NE treatments in borderline cases could harm periodontal tissues, rendering the periodontal tissue less resistant to the external environment, especially in the presence of specific triggering factors such as traumatic toothbrushing and plaque-induced inflammation.²⁵ Orthodontists should pay particular attention to the critical remodeling of the alveolar bone during orthodontic tooth movement.

Different extraction types correspond to varying degrees of tissue remodeling; thus, bone loss varies significantly across extraction types.^23,26,27 Except for lingual bone height, bone height in the coronal plane (ΔMAC-CEJ for mandibular central incisors and DAC-CEJ at T2 for mandibular central incisors and canines) showed significant differences between extraction and NE cases, causing bone loss. In addition, root resorption of the anterior teeth did not differ between the NE and extraction cases (Table 3). These data should be interpreted with caution. Tissue loss may be a side effect of extraction for orthodontic treatment for specific types of malocclusion. The difference in bone loss was observed in the direction of tooth movement because the alveolar bone under compression undergoes more resorption than under tension.^28,29 Mesiodistal bone loss may occur because extraction cases involve more mesiodistal movement of the anterior teeth than NE cases (such as midline correction and tooth rotation adjustments), with OLIE cases being the most notable. Alveolar bone morphology is a limiting factor in tooth movement. Therefore, tissue loss should not be solely attributed to the choice of treatment alternatives; instead, patient's individual susceptibility and iatrogenic factors during treatment (such as adverse forces and excessive tooth movement) may be key contributors.

The opposite trend was found for bone thicknesses at 3 mm and 9 mm from the CEJ in the mandibular incisors in patients with TPE (Tables 2 and 3). The least predictable movement with CAs was torqueing, especially in the mandibular incisors^30,31 and the central incisors showed greater lingual crown torque than predicted in patients with TPE.³² Further, even the presence of power ridges, two-weekly wearing protocols, and intrusion or labial torque overcorrection cannot compensate for the lingual crown torque effect during the retraction of anterior teeth.^31,33,34 Additionally, in the cases recruited in our study, CBCT data were not integrated in the digital treatment planning or aligner setup process. With our results, CBCT integration into ClinCheck^® is crucial, as the torque expression with CBCT is always higher than that without CBCT.³⁵ Second, the planning of retraction should be divided into alternating stages of overcorrection and space closure,³⁶ which can prevent or promptly correct torque loss throughout treatment. Third, a greater number of prescribed aligners can increase the accuracy of torque expression;³¹ therefore, multiple refinement protocols of the CAT can be helpful.

If the ICP-AC distance is < 5 mm, open gingival embrasures (OGEs) are rarely evident; when the ICP-AC distance ≥ 6 mm, the occurrence of OGE increases to 44%; and when the ICP-AC distance ≥ 7, the occurrence rate increases to ≥ 73%.³⁷ Our results showed that extraction could be associated with an increase in ICP-AC, while the patients with OLIEs showed the highest ICP-AC at T2 (> 6 mm) (Tables 2 and 3, Figure 2). ICP-AC distance in maxillary incisors with OGEs (5.51 mm) was greater than that in mandibular incisors (5.85 mm),³⁸ which is consistent with our results (Tables 2 and 3, Figure 2). In addition to hard bony tissues, soft tissues may play an important role in OGE formation. OGEs are common after mandibular incisor extraction,²⁷ and OLIEs may result in more crowded and overlapping incisors than NEs and TPEs. The gingival papillae in the interproximal spaces of overlapping incisors are poorly developed even before orthodontic treatment and appear narrow and elongated. As the teeth become aligned, the width of the interproximal space increases, causing the gingival fibers in these areas to stretch, decreasing the height of the gingival papilla.^39-41 While considering the limitations of gingival remodeling and compression/stretching in the direction of tooth movement, a compromise in the final ideal position of teeth can be considered, or a step-by-step approach can be selected to achieve the ultimate design goal. Therefore, the benefits of achieving the ideal tooth position should be considered in light of potential risks of OGE.¹⁰

We also found an increased ΔLABT-apex for the canines in patients with TPE (Tables 2 and 3). This reflected the widening of the alveolar bone morphology, as the canines are distalized to the region of the first premolars in patients with TPE. However, the ΔLABT-3 and ΔLABT-6 of mandibular canines decreased significantly in patients with TPE than in those with NE (Table 3), possibly because of more rotation and crowding of canines before treatment. Because of the oval cross-section of the root, after derotation of the tooth, an increased bone width was required to maintain the normal bone tissue on the labial-palatal/lingual sides.⁴²

More importantly, after therapy, the mandibular anterior teeth experienced more severe bone loss than the maxillary anterior teeth, which specifically manifested as follows (Tables 2 and 3): first, bone height after treatment at every site in each group was > 2 mm in the mandibular incisors; second, in the TPE group, mandibular incisors were more prone to lose torque control and exhibited a risk of tissue loss after distalization; and third, ICP-AC at T2 in the mandible was greater than that in the maxilla. This difference may be due to the thin anatomical structure of the mandible.⁴³ Previous studies have suggested that the virtual biomechanical stimulation and smart design for each tooth during different movement stages should be personalized, which is challenging in the short term.⁴⁴ However, the mechanical properties of the aligner material, thickness or shape of the aligner membrane, and amount of movement should be differentiated based on the anatomical differences and biomechanics between the maxilla and mandible.

This study had several limitations. First, longitudinal follow-up data are required to evaluate the stability of alveolar bone remodeling and its clinical implications. Second, owing to the limitations of retrospective data collection, we were unable to obtain more information on tooth movement for further analysis, which may have confounded the interpretation of post-treatment changes. Third, the current conclusions should be validated in future studies with larger multicenter samples and more rigorous statistical methods, such as mixed-effects models or machine learning approaches, to control baseline heterogeneity and improve causal inferences.

CONCLUSIONS

Within the limitations of this study, it was concluded that:

• Alveolar bone loss is common after CAT and different tooth extraction types may be associated with different degrees of bone loss in the direction of tooth movement.

• Despite numerous studies and applications aimed at preventing torque loss and lingual bone dehiscence during anterior tooth retraction in patients with TPE, the clinical outcomes in large sample sizes have been suboptimal. This remains a long-term challenge for orthodontists and requires further exploration.

• Extraction could be associated with an increase in ICP-AC, while OLIEs showed the most severe loss of bone height among all groups, which may increase the risk of OGE.

• There was more severe tissue loss around the mandibular anterior teeth after treatment than around the maxillary anterior teeth. Different design concepts, tooth movements, and material properties should be applied to the maxilla and mandible.

SUPPLEMENTARY MATERIAL

Supplementary data is available at https://doi.org/10.4041/kjod25.095