This retrospective study screened Korean adult patients who visited the Department of Orthodontics, Yonsei University Dental Hospital (Seoul, Republic of Korea) between January 2019 and December 2021, underwent MARPE treatment, and completed their orthodontic treatment before December 2024. To reduce operator-related heterogeneity, the cohort was a priori limited to patients treated by a single faculty member (J.Y.C.) using a standardized MARPE protocol.

The specific inclusion criteria were as follows: (1) age ≥19 years at the time of treatment; (2) diagnosed with maxillary transverse deficiency, defined as a maxillomandibular transverse differential index >19.6 mm ; (3) treatment with hyrax-type MARPE using either short or long miniscrews; and (4) baseline Pittsburgh Sleep Quality Index score ≤5.

The exclusion criteria were as follows: (1) failure of MARPE (no midpalatal suture opening on periapical radiographs); (2) lack of acceptable-quality cone-beam computed tomography (CBCT) scan; (3) presence or history of craniofacial syndromes or deformities; (4) history of previous orthodontic treatment; (5) history of nasal conditions (eg, allergic rhinitis and habitual mouth breathing); and (6) history of nasal surgery (eg, septoplasty and turbinate reduction).

A total of 32 participants were included in this study. The patients were divided into 2 groups according to miniscrew length: the short miniscrew group (group S: anterior, 8 mm; posterior, 6 mm; n = 17) and the long miniscrew group (group L: anterior, 13 mm; posterior, 11 mm; n = 15) ( Fig 1 ). Ethical approval was obtained from the institutional review board of the Yonsei University Dental Hospital (CRNo: 2-2019-0037).

Study flow diagram. Initial screening and eligibility identified 43 patients with MARPE. Eleven were excluded (full medical and dental history criteria are described in the Materials and Methods). A total of 32 patients were included for skeletal and dentoalveolar analyses (group S: n = 17; group L: n = 15). For CFD, 2 additional patients were excluded because of severe structural nasal obstruction on CBCT, resulting in a CFD analysis subset of 30 (group S: n = 16; group L: n = 14). PSQI, Pittsburgh Sleep Quality Index.

Sample size was calculated using G∗Power (version 3.1; University of Düsseldorf, Germany) for repeated-measures analysis of variance (RMANOVA) with within-between interaction, assuming 80% power, α = 0.05, and a medium effect size (f = 0.25). Power analysis indicated that at least 28 participants were required; thus, a final sample of 32 participants was considered sufficient to ensure appropriate statistical power.

All patients underwent hyrax-type MARPE (Hyrax II; Dentaurum, Ispringen, Germany) connected to the palate in the para-midsagittal area with 4 miniscrews (diameter, 1.5 mm; self-drilled type; BMK, Biomaterials Korea, Seoul, Republic of Korea) and connected to the first premolar and first molar (tooth-bone-borne). After cementation, 4 miniscrews were inserted perpendicular to the center of the hooks under infiltration anesthesia. The heads of the miniscrews were then connected to the hooks using a core resin (Light-Core, BISCO Dental, Schaumburg, Ill) for additional stability ( Fig 2 , A ). In both groups, the expander was activated at 0.2 mm/day. After 2 weeks, midpalatal suture opening was confirmed using periapical radiography. The expansion continued until the palatal cusps of the maxillary first molars contacted the buccal cusps of the mandibular first molars. The expander was then sealed with a core resin and retained for 3 months. After removal, a band with a buccal tube and a lingual sheath was placed on the maxillary first molars, followed by the insertion of a transpalatal arch for retention ( Fig 2 , B ). The self-ligating brackets were then bonded, and conventional orthodontic treatments were performed. The transpalatal arch was removed during the finishing stage, after which all appliances, including the brackets and bands, were removed. All patients received fixed anterior and removable circumferential retainers.

Clinical application of MARPE and postexpansion retention: A, Intraoral view after MARPE placement; B, Transpalatal arch fabricated with a 0.032 × 0.032-in stainless steel wire, used for postexpansion retention.

CBCT scans were obtained at 3 timepoints: before orthodontic treatment (T0), immediately after MARPE removal (approximately 3 months postexpansion, T1), and at follow-up (T2). For patients who underwent orthognathic surgery with LeFort I osteotomy, T2 was defined as CBCT obtained immediately before surgery to avoid postoperative anatomic alterations, whereas in nonsurgical patients, T2 corresponded to the end of the orthodontic treatment. The CBCT system (Alphard VEGA; ASAHI Roentgen IND, Kyoto, Japan) was operated at 80 kV and 5.0 mA, with a scanning duration of 17 seconds. The imaging parameters included a 154 × 154 mm field of view and a voxel size of 0.3 mm, using the CBCT panoramic mode with a low-dose exposure setting. During each scan, the patients were instructed to remain still and avoid swallowing at the end of expiration.

At T1, miniscrew engagement was evaluated in patients whose miniscrews remained visible on CBCT (23/32; group L, n = 11; group S, n = 12). For each of the 4 miniscrews, engagement was classified as bicortical or monocortical. Bicortical engagement was defined as clear engagement of both the palatal cortex and the nasal floor cortex along the screw trajectory on axial and coronal reconstructions.

The CBCT images were imported as digital imaging and communications in medicine files. Using the fusion module in the OnDemand3D software (Cybermed, Seoul, Republic of Korea), CBCT volumes were superimposed based on the intensity of the gray levels for each voxel in the anterior cranial base. Reorientation was performed using the Frankfort horizontal plane. Twelve bilateral landmarks were identified based on a previous study ( Fig 3 , A and B ; Table I ). ^, Linear distances between the bilateral landmarks were calculated based on their 3-dimensional (3D) coordinates, using the Euclidean distance formula. Two skeletal and 4 dentoalveolar measurements were obtained at each time point, and the differences between the time points were analyzed to assess the changes. All measurements were performed by a single researcher (E.H.C.) who was blinded to the group allocation and unaware of the airway analysis results. To assess the intraexaminer reliability, 10 patients were randomly remeasured after a 2-week interval. Intraclass correlation coefficients indicated excellent reliability (range: 0.92-0.98).

Skeletal and dentoalveolar landmarks for transverse measurements: A, Alare (1, 2), processus zygomaticus (3, 4), ectocanine (5, 6), and ectomolare (7, 8); B, Furcation (9, 10) and central fossa (11, 12) of the maxillary first molar. See Table I for definitions of each landmark.

Airway function was assessed using 2 CFD-derived parameters: pressure-flow ratio (pascals per milliliter per second) and maximum velocity (meters per second). The 3D nasal airway models were constructed from CBCT data using volume-rendering software (INTAGE Volume Editor; Cybernet Systems, Tokyo, Japan) ( Fig 4 , A ). Segmentation was based on the image intensity, with the threshold set at the midpoint between the soft tissue and airway values.

Evaluation of nasal airflow parameters (pressure-flow ratio and maximum velocity) using CFD: A, Extraction and construction of the 3D nasal airway model and numerical simulation of expiratory airflow ( black ); B, Evaluation of nasal airway pressure. The pressure-flow ratio was calculated by dividing the intranasal pressure by the airflow volume; C, Evaluation of maximum velocity.

A medical image analysis system (Imagnosis VE; Imagnosis, Kobe, Japan) was used to establish a 3D coordinate system and extract the measurement data. Mesh morphing (DEP MeshWorks/Morpher; IDAJ, Kobe, Japan) was then applied to smooth the airway surface without altering patient-specific morphology. The final models were exported in a stereolithography format and analyzed using CFD software (Phoenics; CHAM Japan, Tokyo, Japan).

Of the 32 participants included in the skeletal and dentoalveolar analysis, 2 were excluded from CFD simulation because of severe structural nasal obstruction identifiable on baseline CBCT, which precluded reliable segmentation and meshing ( Fig 1 ). Accordingly, CFD-based airflow analysis was performed in 30 participants under the following simulation conditions: (1) steady airflow at 500 cm ³/s; (2) nonslip wall boundary conditions; and (3) 300 iterative runs to obtain averaged values. Airflow was modeled as entering from the choanae and exiting through both external nares. The pressure-flow ratio was calculated according to the Ohm’s law, based on the pressure difference between the choanae and external nares and the corresponding mass airflow rate ( Fig 4 , B ). ^, Maximum velocity was defined as the highest local velocity within the nasal airway domain under the same flow conditions ( Fig 4 , C ). All CFD simulations and measurements were independently conducted by a single researcher (M.N.) who was blinded to the group allocation and transverse measurement outcomes. To assess reproducibility, CFD simulations were repeated on 5 randomly selected models, with the results showing minimal variation (<5%), indicating high analytical consistency.

The primary outcome of this study was to evaluate the difference in skeletal, dentoalveolar, and nasal airflow changes between group L and group S across the time points. The secondary outcome was the identification of potential predictors of nasal airflow changes after MARPE. Two sets of correlation analyses were conducted separately within each group: (1) between changes in transverse skeletal and dentoalveolar dimensions and concurrent changes in nasal airflow parameters and (2) between baseline anatomic (transverse, horizontal, and vertical skeletal dimensions), functional (baseline nasal airflow parameters), and demographic (age) variables and subsequent airflow changes.

In addition, the pressure changes across the 3 time points were classified into distinct temporal patterns for each subject. The maximum velocity change patterns were further classified into each pressure pattern category.

A total of 32 participants (mean age: 22.3 ± 3.0 years; range: 19-30 years) were included in this study, with no significant differences in age or sex distribution between the 2 groups ( Table II ). Of the 32 subjects, 14 (43.8%) had an ANB <0°, 8 (25.0%) were within the normal range (0°-4°), and 10 (31.3%) had values >4°. For SN-GoMe, 19 subjects (59.4%) showed values >36°, 11 (34.4%) were within 28°-36°, and 2 (6.3%) had values <28°. These distributions suggest that skeletal Class III and high-angle characteristics were relatively common in this sample. No significant intergroup differences were found in the baseline skeletal characteristics. The amount of expander screw activation and the duration between each timepoint (T1-T0, T2-T1, and T2-T0) did not differ significantly between the 2 groups (all P > 0.05; Table II ).

Among CBCT-evaluable cases (n = 23), group L (n = 11) exhibited posterior bicortical engagement in 11/11 (100.0%): 3/11 (27.3%) in all 4 miniscrews and 8/11 (72.7%) posterior screws only. Group S (n = 12) showed monocortical engagement in all 4 miniscrews in 10/12 (83.3%), with posterior-only bicortical engagement in 2/12 (16.7%).

A significant time × group interaction was observed only for the inter-processus zygomaticus width ( P = 0.002; Table III ), although pairwise comparisons between the groups at each time point did not reveal significant differences.

Stay updated, free dental videos. Join our Telegram channel

Join