Three-dimensional crown–root movement in first premolar extraction cases treated with two clear aligner systems based on different materials: a retrospective cohort study Authors: Luyao Qu, Runzhe Fang, Ruomei Li, Zitong Yang, Yixin Li, Xuehui Chen, Boxi Chen, Jian Wang, Lingjun Yuan, Lunguo Xia, Bing Fang Published: 2026 ABSTRACT AimsTo quantify and compare the three-dimensional (3D) deviations of crown and root movements produced by two clear aligner systems based on different materials—SmartTrack (ST) and MasterControl S (MCS)—in first premolar extraction cases using a 3D crown–root registration approach.MethodsThis retrospective cohort study included 29 patients (58 dental sides) treated with ST (n = 15) or MCS (n = 14). Pretreatment (T0) CBCT images were integrated with T0 and post-treatment (T1) intraoral scans and superimposed using palatal rugae as reference structures. Linear mixed-effects models (LMM) were applied to estimate between-material differences while adjusting for age, sex, and standardized planned movement magnitude. Estimates are reported with 95% confidence intervals (CI).ResultsThe two systems demonstrated no statistically significant difference in deviation in sagittal crown retraction of the maxillary central incisor (U1) (ST: 1.25 (1.30) mm; MCS: 2.13 (1.82) mm; LMM difference: − 0.81 mm; 95% CI − 1.89–0.28; P = 0.145). MCS demonstrated smaller deviation in sagittal root positioning for the second premolar (U5) (difference: 1.07 mm; 95% CI 0.21–1.93; P = 0.015) and first molar (U6) (difference: 1.04 mm; 95% CI 0.13–1.94; P = 0.025). ST showed improved control of U5 axial inclination (difference: 3.33°; 95% CI 0.27–6.39; P = 0.033) and torque (difference: 3.04°; 95% CI 0.08 to 6.00; P = 0.044). Both systems exhibited a consistent “roller-coaster effect”, with no significant difference in unplanned U1 extrusion (difference: 0.14 mm; 95% CI − 0.73–1.02; P = 0.747).ConclusionsThe biomechanical performance of clear aligner systems appears to vary depending on tooth position and movement vector. SmartTrack demonstrated better control of premolar angulation and torque, whereas MasterControl S showed closer agreement with the planned sagittal root position. Introduction The clinical scope of clear aligner therapy (CAT) has expanded substantially and increasingly includes complex malocclusions such as first premolar extraction cases [1]. These treatment protocols introduce demanding biomechanical requirements, as successful outcomes depend on the coordinated control of en masse anterior retraction, posterior anchorage preservation, and maintenance of root parallelism [2]. Despite advances in digital treatment planning, the viscoelastic behavior of aligner polymers remains a significant biomechanical challenge in extraction mechanics. Thermoplastic materials are susceptible to stress relaxation and elastic deformation, which can lead to rapid force decay during the aligner wear cycle. As a consequence, undesirable tooth movement patterns are frequently observed during space closure, including mesial tipping of posterior teeth (“tipping-in”), distal tipping of canines, and anchorage loss. These unplanned movements may compromise root parallelism and deepen the occlusal curve—a phenomenon often referred to as the “roller-coaster effect” [2–4]. The clinical effectiveness of CAT is largely influenced by the mechanical properties of the thermoplastic materials used in aligner fabrication. Contemporary aligner systems primarily employ thermoplastic polyurethane (TPU) or polyethylene terephthalate glycol (PETG), which differ in elastic modulus and dimensional stability [1, 5]. SmartTrack (ST; Invisalign) is a multi-layer TPU-based material designed to provide high elasticity and sustained force delivery [5]. In contrast, MasterControl S (MCS; Angelalign) is a PETG-based material characterized by higher initial stiffness and improved resistance to deformation [6]. These material-specific characteristics may influence the moment-to-force (M:F) ratio transmitted to the dentition, potentially affecting whether extraction space closure occurs through bodily movement or uncontrolled tipping [7, 8]. Accurate evaluation of these biomechanical effects requires three-dimensional analysis of both crown and root displacement. This approach is consistent with recent applications of three-dimensional deviation analysis in digital dentistry, which have demonstrated its reliability in quantifying discrepancies between planned and achieved structures [9].Conventional two-dimensional cephalometric analysis is insufficient to capture complex tooth movements such as torque expression or root translation [10]. Intraoral scans provide high-resolution crown morphology but lack root information. Hybrid reconstruction approaches combining CBCT-derived anatomical data with intraoral STL models have therefore been developed to generate root-incorporating 3D models [4, 10, 11]. This integrated method enables more accurate quantification of root displacement and axial changes than crown-only analyses. Nevertheless, comparative studies investigating differences between aligner systems bases on different materials in three-dimensional crown and root movements during extraction treatment remain limited. Specific objectives and hypotheses The objective of this retrospective cohort study was to quantify and compare the three-dimensional crown and root movement characteristics associated with SmartTrack (ST) and MasterControl S (MCS) aligner systems based on different materials in first premolar extraction cases using integrated CBCT–STL reconstructed models. It was hypothesized that the stiffer PETG-based material (MCS) would demonstrate smaller deviation in sagittal root positioning of posterior teeth, owing to its potential to generate a more favorable moment-to-force (M:F) ratio for root control during extraction space closure. In contrast, the more elastic TPU-based material (ST) was expected to provide improved control of premolar axial inclination and incisor torque expression due to its ability to deliver more continuous and uniform orthodontic forces. Methods Study design This study was designed as a single-center retrospective cohort study. Participants, eligibility criteria, and setting Participants were identified from clinical records at Department of Orthodontics, Ninth People’s Hospital, Shanghai Jiao Tong University School of Medicine. To minimize inter-operator variability, all patients were treated by three experienced orthodontists from the same clinical team, who followed a standardized departmental protocol for diagnosis, attachment bonding, and interproximal reduction (IPR). Treatments were performed at department of orthodontics, Ninth People’s Hospital, Shanghai Jiao Tong University School of Medicine between January 2022 and January 2025. Inclusion criteria: (1).Adult patients (≥ 16 years) with permanent dentition and Angle Class I malocclusion presenting with mild maxillary crowding (≤ 4 mm) and requiring extraction of four first premolars; (2).Treatment with SmartTrack (ST; Invisalign) or MasterControl S (MCS; Angelalign) aligners; (3).Treatment plans involving en masse anterior retraction during extraction space closure; (4).Availability of complete T0/T1 CBCT images, intraoral scans, and digital treatment plans; (5).High treatment compliance (20–22 h/day) with a standardized protocol of aligner replacement every 14 days; (6).No history of previous orthodontic treatment or orthognathic surgery. Exclusion criteria:(1) Periodontal attachment loss greater than 3 mm; (2)Systemic diseases affecting bone metabolism; (3) Craniofacial anomalies; (4)Missing teeth (excluding third molars); (5)Absence of CBCT imaging or incomplete palatal rugae in intraoral scans. Variables Exposures, predictors, potential confounders, and effect modifiers The primary exposure in this study was the type of aligner systems (SmartTrack vs MasterControl S), which differ in material composition. The magnitude of prescribed tooth movement derived from the digital setup was considered as a predictor of treatment outcomes. Potential confounders included patient-related factors such as age and sex, as well as baseline tooth position and malocclusion characteristics. Patient compliance was considered a potential effect modifier. Outcomes Primary Outcome: Sagittal movement (SM) deviation of the roots for U1, U3, U5, U6, and U7. Secondary Outcomes: Linear (VM, TD) and angular (tipping, torque, rotation) deviations. Deviation = Achieved Movement (AM)-Designed Movement (DM). Data measurement Data acquisition and preprocessing Pretreatment CBCT scans (T0) were acquired with a voxel size of 0.3 mm. Intraoral scans at T0 and T1 were obtained using the iTero Element 2 scanner. Digital treatment plans were exported from ClinCheck® (ST) and iOrtho® (MCS), including the initial (D0) and final (D1) digital models. Three-dimensional (3D) image processing and analysis were performed using Mimics Medical 2021 software. T0 and T1 intraoral scans were superimposed using the palatal rugae as stable reference structures. Digital treatment models (D0/D1) were subsequently aligned based on crown morphology (Fig. 1).Fig. 1Schematic diagrams of three-dimensional model superimposition and tooth separation protocols. A. Superimposition of pretreatment and posttreatment intraoral scan models using the palatal rugae as stable anatomical landmarks; B. Superimposition of digital treatment plan models (initial vs. final designs) incorporating root structures; C. Co-registration of pretreatment/posttreatment intraoral scan models with initial/final digital design models; D. Segmentation of individual teeth from the final design’s root-included digital model and alignment with the posttreatment intraoral scan model; E, F. Superimposition of the posttreatment intraoral scan model (integrated with root structures) with the initial and final digital design models Schematic diagrams of three-dimensional model superimposition and tooth separation protocols. A. Superimposition of pretreatment and posttreatment intraoral scan models using the palatal rugae as stable anatomical landmarks; B. Superimposition of digital treatment plan models (initial vs. final designs) incorporating root structures; C. Co-registration of pretreatment/posttreatment intraoral scan models with initial/final digital design models; D. Segmentation of individual teeth from the final design’s root-included digital model and alignment with the posttreatment intraoral scan model; E, F. Superimposition of the posttreatment intraoral scan model (integrated with root structures) with the initial and final digital design models Coordinate system A three-dimensional coordinate system was established using the T0 model as reference (Fig. 2).Fig. 2Schematic diagram for establishing the sagittal plane, coronal plane and occlusal plane Schematic diagram for establishing the sagittal plane, coronal plane and occlusal plane The occlusal plane (OP) was defined by the buccal cusps of U6 and the mesial contact point of U1. The coronal plane (CP) was constructed perpendicular to the occlusal plane through U6. The sagittal plane (SP) was defined as perpendicular to both the occlusal and coronal planes through U1. Landmark definitions: Crown: incisal edge midpoint (U1), cusp tip (U3), or central fossa (U5–U7). Root: apex (U1, U3) or furcation center (U5–U7). The long axis of each tooth was defined as the line connecting the crown and root landmarks. Linear and angular tooth movements were then quantified relative to the reference coordinate planes. Linear movements (unit: mm).Sagittal movement (SM): Distance from the crown or root landmark to the coronal plane (CP); positive values indicate displacement toward the extraction space, and negative values indicate displacement in the direction opposite to the extraction space.Transverse displacement (TD): Distance from the crown or root landmark to the sagittal plane (SP); positive values indicate buccal movement, and negative values indicate palatal movement.Vertical movement (VM): Distance from the crown or root landmark to the occlusal plane (OP); positive values indicate intrusion, and negative values indicate extrusion. Sagittal movement (SM): Distance from the crown or root landmark to the coronal plane (CP); positive values indicate displacement toward the extraction space, and negative values indicate displacement in the direction opposite to the extraction space. Sagittal movement (SM): Distance from the crown or root landmark to the coronal plane (CP); positive values indicate displacement toward the extraction space, and negative values indicate displacement in the direction opposite to the extraction space. Transverse displacement (TD): Distance from the crown or root landmark to the sagittal plane (SP); positive values indicate buccal movement, and negative values indicate palatal movement. Transverse displacement (TD): Distance from the crown or root landmark to the sagittal plane (SP); positive values indicate buccal movement, and negative values indicate palatal movement. Vertical movement (VM): Distance from the crown or root landmark to the occlusal plane (OP); positive values indicate intrusion, and negative values indicate extrusion. Vertical movement (VM): Distance from the crown or root landmark to the occlusal plane (OP); positive values indicate intrusion, and negative values indicate extrusion. Angular movements (unit: °) (Fig. 3).Axial inclination: For anterior teeth, defined as the angle between the tooth long axis and the sagittal plane (SP); for posterior teeth, defined as the angle between the tooth long axis and the coronal plane (CP). Positive values indicate distal tipping, and negative values indicate mesial tipping.Torque: For anterior teeth, defined as the angle between the tooth long axis and the coronal plane (CP); for posterior teeth, defined as the angle between the tooth long axis and the sagittal plane (SP). Positive values indicate buccal tipping, and negative values indicate lingual tipping.Rotation: For anterior teeth, defined as the angle between the mesiodistal line of the crown and the coronal plane; for posterior teeth, defined as the angle between the buccolingual cusp line of the crown and the coronal plane. Positive values indicate mesial rotation, and negative values indicate distal rotationFig. 3A. Anterior axial inclination; B. Anterior torque; C. Anterior torsion angle D. Molar torque angle; E. Molar axial inclination; G. Molar torsion angle Axial inclination: For anterior teeth, defined as the angle between the tooth long axis and the sagittal plane (SP); for posterior teeth, defined as the angle between the tooth long axis and the coronal plane (CP). Positive values indicate distal tipping, and negative values indicate mesial tipping. Axial inclination: For anterior teeth, defined as the angle between the tooth long axis and the sagittal plane (SP); for posterior teeth, defined as the angle between the tooth long axis and the coronal plane (CP). Positive values indicate distal tipping, and negative values indicate mesial tipping. Torque: For anterior teeth, defined as the angle between the tooth long axis and the coronal plane (CP); for posterior teeth, defined as the angle between the tooth long axis and the sagittal plane (SP). Positive values indicate buccal tipping, and negative values indicate lingual tipping. Torque: For anterior teeth, defined as the angle between the tooth long axis and the coronal plane (CP); for posterior teeth, defined as the angle between the tooth long axis and the sagittal plane (SP). Positive values indicate buccal tipping, and negative values indicate lingual tipping. Rotation: For anterior teeth, defined as the angle between the mesiodistal line of the crown and the coronal plane; for posterior teeth, defined as the angle between the buccolingual cusp line of the crown and the coronal plane. Positive values indicate mesial rotation, and negative values indicate distal rotation Rotation: For anterior teeth, defined as the angle between the mesiodistal line of the crown and the coronal plane; for posterior teeth, defined as the angle between the buccolingual cusp line of the crown and the coronal plane. Positive values indicate mesial rotation, and negative values indicate distal rotation A. Anterior axial inclination; B. Anterior torque; C. Anterior torsion angle D. Molar torque angle; E. Molar axial inclination; G. Molar torsion angle Sample size calculation Sample size calculation as this was a retrospective study with consecutive sampling, the sample size was determined by the number of eligible cases during the study period. Bias To minimize measurement bias, all landmark identification and measurements were performed by a single investigator who was blinded to the aligner systems group. Measurement reliability was assessed by randomly selecting 20% of the samples for repeated measurement after a two-week interval. Selection bias was minimized through consecutive case inclusion, and potential confounding factors were controlled for using linear mixed-effects models. Statistical analysis All statistical analyses were performed using Python (version 3.x) with the statsmodels library. Continuous independent variables (age and planned movement magnitude) were standardized using Z-score normalization. Linear mixed-effects models (LMM) was used to account for clustering of multiple teeth within the same individual. The model specification included: Fixed effect: aligner systems (ST vs MCS). Covariates: age, sex, tooth side, and standardized planned movement magnitude. Random effect: patient ID (random intercept). Inference: Statistical inference focused primarily on point estimates (model coefficients) and their 95% confidence intervals rather than relying solely on P values [12]. Clinical equivalence: To assess clinical relevance, equivalence margins (δ) were predefined. Clinical equivalence between systems based on different materials was considered present when the 95% confidence interval for the difference fell entirely within the following thresholds: ± 1.0 mm for crown linear displacement ± 0.5 mm for root linear displacement ± 2.0° for angular deviations ± 1.0 mm for crown linear displacement ± 1.0 mm for crown linear displacement ± 0.5 mm for root linear displacement ± 0.5 mm for root linear displacement ± 2.0° for angular deviations ± 2.0° for angular deviations These thresholds were selected based on previously reported clinically acceptable limits for orthodontic tooth movement deviation [13–15]. Results Participant flow (Fig. 4): 29 patients (58 sides) completed treatment with no dropouts.Fig. 4Flow diagram of patient selection and allocation to the two aligner systems groups Flow diagram of patient selection and allocation to the two aligner systems groups Baseline data A total of 29 patients (58 sides) were included, with a mean age of 23.90 (6.79) years (7 males, 22 females).ST Group: 15 patients (30 sides), mean age 23.67 (6.68) years (3 males, 12 females).MCS Group: 14 patients (28 sides), mean age 24.14 (6.88) years (4 males, 10 females). ST Group: 15 patients (30 sides), mean age 23.67 (6.68) years (3 males, 12 females). ST Group: 15 patients (30 sides), mean age 23.67 (6.68) years (3 males, 12 females). MCS Group: 14 patients (28 sides), mean age 24.14 (6.88) years (4 males, 10 females). MCS Group: 14 patients (28 sides), mean age 24.14 (6.88) years (4 males, 10 females). Intra-rater reliability was excellent, with values ranging from 0.90 to 0.96. Primary outcome Sagittal movement deviation For sagittal root movement, smaller deviations were observed in the MCS group compared to the ST group in posterior teeth (Tables 3, 4). Specifically, for U5 root SM, the deviation was 1.00 (1.43) mm in the ST group and − 0.17 (1.81) mm in the MCS group, with a between-group difference of 1.07 mm (95% CI 0.21 to 1.93; P = 0.015). Similarly, for U6 root SM, the deviation was 0.60 (1.46) mm in the ST group and − 0.69 (1.75) mm in the MCS group, with a between-group difference of 1.04 mm (95% CI 0.13 to 1.94; P = 0.025). For U1 crown SM (Table 1), the deviation was 1.25 (1.30) mm in the ST group and 2.13 (1.82) mm in the MCS group, with no significant between-group difference (95% CI − 1.89 to 0.28; P = 0.145).Table 1Comparison deviation of U1 between ST and MCS systems, with between-group differences estimated using linear mixed-effects models (LMM) and 95% confidence intervals (CI) Table 1Comparison deviation of U1 between ST and MCS systems, with between-group differences estimated using linear mixed-effects models (LMM) and 95% confidence intervals (CI) Comparison deviation of U1 between ST and MCS systems, with between-group differences estimated using linear mixed-effects models (LMM) and 95% confidence intervals (CI) Secondary outcomes Vertical movement In both groups, anterior teeth showed extrusion (Tables 1, 2). For U1 crown VM, the deviation was − 2.26 (1.07) mm in the ST group and − 2.72 (1.57) mm in the MCS group, with no significant between-group difference (95% CI − 0.73 to 1.02; P = 0.747).For U3 crown VM, the deviation was − 1.18 (1.20) mm in the ST group and − 1.02 (1.28) mm in the MCS group, with no significant between-group difference (95% CI− 0.67 to 0.81; P = 0.854).Table 2Comparison deviation of U3 between ST and MCS systems, with between-group differences estimated using linear mixed-effects models (LMM) and 95% confidence intervals (CI) Table 2Comparison deviation of U3 between ST and MCS systems, with between-group differences estimated using linear mixed-effects models (LMM) and 95% confidence intervals (CI) Comparison deviation of U3 between ST and MCS systems, with between-group differences estimated using linear mixed-effects models (LMM) and 95% confidence intervals (CI) For posterior teeth (Tables 3–4), intrusion was observed. For U5 crown VM, the deviation was 0.23 (0.64) mm in the ST group and 0.24 (0.80) mm in the MCS group, with no significant between-group difference (95% CI − 0.42 to 0.53; P = 0.823).For U6 crown VM, the deviation was 0.70 (1.11) mm in the ST group and 0.60 (0.80) mm in the MCS group, with no significant between-group difference (95% CI− 0.59 to 0.54; P = 0.934).Table 3Comparison deviation of U5 between ST and MCS systems, with between-group differences estimated using linear mixed-effects models (LMM) and 95% confidence intervals (CI)Table 4Comparison deviation of U6 between ST and MCS systems, with between-group differences estimated using linear mixed-effects models (LMM) and 95% confidence intervals (CI) Table 3Comparison deviation of U5 between ST and MCS systems, with between-group differences estimated using linear mixed-effects models (LMM) and 95% confidence intervals (CI) Comparison deviation of U5 between ST and MCS systems, with between-group differences estimated using linear mixed-effects models (LMM) and 95% confidence intervals (CI) Table 4Comparison deviation of U6 between ST and MCS systems, with between-group differences estimated using linear mixed-effects models (LMM) and 95% confidence intervals (CI) Comparison deviation of U6 between ST and MCS systems, with between-group differences estimated using linear mixed-effects models (LMM) and 95% confidence intervals (CI) For U7 crown VM (Table 5), the deviation was − 0.38 (0.93) mm in the ST group and − 0.07 (0.87) mm in the MCS group, with no significant between-group difference (95% CI − 0.88 to 0.26; P = 0.284).Table 5Comparison deviation of U7 between ST and MCS systems, with between-group differences estimated using linear mixed-effects models (LMM) and 95% confidence intervals (CI) Table 5Comparison deviation of U7 between ST and MCS systems, with between-group differences estimated using linear mixed-effects models (LMM) and 95% confidence intervals (CI) Comparison deviation of U7 between ST and MCS systems, with between-group differences estimated using linear mixed-effects models (LMM) and 95% confidence intervals (CI) Angular movement For U5 axial inclination (Table 3), the ST group showed smaller deviations compared to the MCS group (− 2.89 (4.17)° vs − 6.08 (5.64)°), with a between-group difference of 3.33° (95% CI 0.27 to 6.39; P = 0.033). For U6 axial inclination (Table 4), a similar pattern was observed (− 3.77 (5.43)° vs − 7.46 (5.71)°), although the difference was not statistically significant (between-group difference: 2.43; 95% CI − 0.43 to 5.30; P = 0.096). For U5 torque (Table 3), the ST group demonstrated smaller deviations (− 0.62 (5.31)° vs − 3.31 (5.17)°), with a between-group difference of 3.04° (95% CI 0.08 to 6.00; P = 0.044). For U1 torque (Table 1), both groups showed substantial lingual torque loss (− 9.77 (10.93)° vs − 11.66 (10.07)°), with no significant between-group difference (95% CI − 4.67 to 6.29; P = 0.772). Discussion Based on three-dimensional crown–root registration analysis, the present study evaluated multidimensional tooth movement characteristics associated with two clear aligner systems based on different materials in first premolar extraction cases. Unlike many previous investigations that primarily focused on non-extraction protocols or crown-level movement assessments, extraction space closure represents a substantially more complex biomechanical environment in which three-dimensional root behavior plays a critical role in treatment outcomes [16, 17]. By integrating CBCT-based crown–root reconstruction with intraoral scans and digital treatment plans, the present approach enabled detailed three-dimensional quantification of tooth movement during extraction space closure, thereby addressing the limitations of crown-only analyses reported in earlier studies. Sagittal root control vs angulation stability One of the principal findings of this study is that differences between aligner systems may be associated with variations in posterior tooth movement during extraction space closure. The MCS system (PETG-based) showed smaller deviations in sagittal root positioning of both the second premolars (U5) and first molars (U6), whereas the ST system (TPU-based) demonstrated better control of axial inclination and torque in the upper second premolar. From a biomechanical perspective, posterior tooth movement is largely governed by the moment-to-force (M:F) ratio applied to the dentition. Classical biomechanical studies have shown that insufficient M:F ratios produce uncontrolled tipping, whereas higher ratios are required for translation or controlled root movement [18]. During extraction mechanics, inadequate control of this ratio may lead to mesial tipping of posterior teeth and anchorage loss [19]. The greater sagittal root stability observed with the PETG-based aligner may therefore be associated with the higher stiffness and dimensional stability of PETG-based materials [7, 8, 20]. Compared with more elastic polymers, stiffer materials may generate higher initial forces and moments, facilitating more effective transmission of root-controlling moments and reducing unintended mesial tipping during space closure. However, as this study compared aligner systems rather than isolated materials, other system-related factors, including digital setup and treatment protocols, may also have contributed. Conversely, the ST system demonstrated better control of axial inclination and torque of the second premolar. This may partly reflect the mechanical characteristics of TPU-based materials. SmartTrack exhibits greater elasticity and shape recovery, allowing more continuous force delivery and more uniform force transmission across the crown surface. These properties may enhance the expression of programmed angular movements such as tipping and torque [7]. Previous studies have similarly reported that TPU-based polymers exhibit improved elastic recovery and reduced stress relaxation compared with conventional thermoplastic materials, which may improve the predictability of rotational and torque movements [20]. Meanwhile, U6 exhibited an inclination pattern similar to that observed in U5. This coordinated response suggests that the premolar–molar segment may function as a mechanically continuous posterior unit during extraction space closure, with tooth movements governed by a shared force system rather than independent tooth-level effects. In addition to differences in tipping magnitude, distinct movement patterns were observed between the two aligner systems. The MCS system was associated with concurrent rotational displacement of U5 and U6 together with mesial tipping, whereas the ST system predominantly produced mesial tipping of U6 with minimal rotational deviation. These differences may reflect variations in force distribution within the posterior segment, potentially influenced by both material characteristics and system-related factors. The “Roller-Coaster Effect”: a shared biomechanical limitation Both aligner systems demonstrated a similar vertical limitation during extraction space closure, characterized by a pronounced “roller-coaster effect.” Although no statistically significant difference in anterior retraction was observed between the two systems (LMM Diff: − 0.81 mm; 95% CI − 1.89 to 0.28; P = 0.145), substantial unplanned extrusion of the maxillary central incisor occurred in both groups. This vertical pattern—anterior extrusion combined with posterior intrusion—has been widely attributed to the bowing effect associated with en masse retraction. During space closure, the aligner spans the extraction site and tends to flex under loading, generating intrusive forces on premolars while producing extrusive forces on incisors [21, 22]. In addition, coverage of posterior occlusal surfaces creates a “bite-block effect,” which may further contribute to posterior intrusion and deepening of the occlusal curve [21, 23]. Because these biomechanical effects arise primarily from the geometric interaction between the aligner and the dental arch rather than from material-specific mechanical properties, the absence of significant differences between ST and MCS systems in vertical crown movement suggests that this deviation represents a shared biomechanical limitation of aligner-based extraction mechanics. Previous analyses similarly indicate that vertical control during en masse retraction remains challenging with aligners and may require optimized staging, attachment design, or auxiliary mechanics to mitigate the development of the roller-coaster effect. Torque expression Torque loss represented another shared limitation. Both aligner systems demonstrated substantial underexpression of U1 torque, consistent with previous clinical studies highlighting torque control as a persistent challenge in clear aligner therapy [22]. This limitation may partly result from the inherent “play” between the aligner and the tooth surface, which reduces effective torque expression. Furthermore, successful torque expression is strongly influenced by attachment design, power ridges, and staged biomechanics, suggesting that improvements in auxiliary design may play a greater role than material-related differences alone in addressing this limitation [24]. Clinical implications These findings provide preliminary guidance for clinical decision-making in extraction cases. The MCS system may be advantageous when posterior root parallelism and anchorage control are primary concerns, as its higher stiffness may contribute to more stable root positioning. However, clinicians should consider incorporating distal tipping overcorrection in the digital treatment plan to compensate for potential posterior tipping. Conversely, the ST system may be beneficial in patients requiring more precise control of premolar angulation and anterior torque expression. When anchorage control is critical, the early use of temporary anchorage devices (TADs) or intermaxillary elastics may help mitigate sagittal root deviations. Study limitations Several limitations of the present study should be acknowledged. First, the retrospective design and relatively small sample size of 29 patients may limit the generalizability of the findings. Nevertheless, all cases were treated by a single clinical team following a strictly standardized treatment protocol, which helped minimize potential bias related to inter-operator variability. In addition, clustering and potential confounding factors were addressed using linear mixed-effects models (LMM). Second, this study was conducted in a single clinical center. Although the standardized treatment protocol ensured consistency in clinical procedures, the findings may not fully reflect variations in treatment approaches or patient populations in other clinical settings. Future multicenter studies with larger samples are therefore warranted to further evaluate system- and material-related influences on tooth movement during extraction space closure. Generalizability The findings of this study may be generalizable to patients undergoing clear aligner therapy with premolar extraction using similar treatment systems. However, as this was a single-center retrospective study, the external validity may be limited. Variations in patient characteristics, treatment protocols, and compliance may influence the applicability of these results. Therefore, further multicenter prospective studies are warranted to confirm these findings in broader populations. Conclusion Three-dimensional crown–root movement deviation patterns in extraction cases appear to depend on tooth position and the specific movement vector involved. Differences between the ST and MCS systems were observed, with improved control of premolar axial inclination and torque in the ST system and closer agreement with the planned sagittal root position in the MCS system. The shared vertical deviations observed in both systems highlight the persistent biomechanical challenge of the bowing effect during extraction space closure. These findings suggest that clinicians may consider aligner system selection according to root parallelism requirements and incorporate targeted overcorrection or auxiliary mechanics to improve vertical and torque control.