Introduction

Miniscrew-Assisted Rapid Palatal Expansion (MARPE) has recently gained international attention as a non-surgical treatment for transverse maxillary deficiency in late adolescents and adults. Conventional Rapid Palatal Expansion (RPE) is a well-documented treatment for maxillary expansion in growing patients [1]. In late adolescents and adults, however, the skeletal effect of RPE was limited and it resulted in significant dental side-effects. Surgically-Assisted Rapid Palatal Expansion (SARPE) was, hence, considered necessary to enable separation of the midpalatal suture, but it was associated with the inherent surgical risks, higher costs, and comorbidity of a surgical procedure [2, 3]. A less invasive approach to expand the maxilla was thus envisaged.

Previous histological and radiological studies reported that the ossification of the midpalatal suture displayed a great variation and was often incomplete in late adolescents and adults, providing the basis for the development of MARPE [4]. Lee et al. were among the first to design a MARPE appliance as a tooth-bone-borne RPE-hyrax device anchored in the palate with four miniscrews through helical loops, which were soldered to the expansion screw base. The miniscrews were instrumental for transferring the expansion force directly into the palate, leading to separation of the midpalatal suture [5].

Moon et al. enhanced the MARPE appliance and introduced the Maxillary Skeletal Expander (MSE), which was positioned posteriorly, at the level of the first molars, in order to overcome the high skeletal resistance in this area, resulting in a more parallel suture expansion [6]. As the MSE was based on a prefabricated expansion screw and connectors that had to be soldered to the molar bands by the dental technician, a weakness in the metal structure was introduced that was more prone to deformation and fracture [7]. The shortcomings of the available expanders inspired the development of the Dutch Maxillary Expansion Device (D-MED), which was used in this study.

Several publications have examined the efficacy of MARPE, including a recent systematic review and meta-analysis. The findings suggested that MARPE was an efficient treatment for maxillary expansion in late adolescents and adults, but also that the vast majority of included studies was of retrospective nature and had a high risk of bias, especially regarding patient selection, and indicated the need for prospective studies of higher quality [2). Following 3D-printing, the base of an expansion screw (Forestadent, Pforzheim, Germany) is soldered on the structure. The screw is positioned parallel to the palate at the level of the upper first molars and has an expansion capacity of either 10 mm or 12 mm, selected according to the required maxillary expansion.

Fig. 2
figure 2

Miniscrew positioning with D-MED on an intra-oral scan

Full size image

Clinical procedure

All procedures were performed by one orthodontist, the first author (A.K.), with the assistance of two senior orthodontists (J.S. and R.N.) in accordance with the clinical protocol for this MARPE device. As described above, an IOS was acquired for the design and fabrication of the D-MED (see: Fig. 3). Palatal thickness was measured on the pre-treatment CBCTs to determine the appropriate length of the miniscrews (11.0 mm or 13.0 mm; diameter: 2.0 mm) in order to achieve bicortical anchorage. When placing the D-MED, it was first cemented with light cure band cement (Ultra Band-Lok®, Reliance Orthodontic Products, Itasca, IL, USA) to facilitate the insertion of the miniscrews as designed, with the screw holes functioning as a surgical guide. Following local anaesthesia infiltration, an electric screw driver (iSD900, PSM Medical Solutions, Gunningen, Germany) with 40 Ncm torque at 20 rpm turning speed was used to insert the four miniscrews in the center of the screw holes, with the fitting diameter providing a tight grip. If the measured palatal thickness was > 6.0 mm, pre-drilling through the screw holes at the insertion sites of the miniscrews was performed.

Fig. 3
figure 3

Frontal view of an intra-oral scan with D-MED at T0 (A: left) and T1 (B: right)

Full size image

Maxillary expansion was initiated immediately after insertion of the D-MED (see: Fig. 4a). The screw was activated once a day, equivalent to 0.25 mm, and the expansion was closely monitored with weekly check-ups. The expansion was measured intra-orally with a digital calliper at the expansion screw, the central diastema, and inter-molar and inter-canine widths, and periodontal pockets deeper than 3 mm around the upper central incisors were recorded. The expansion was considered successful when the occlusal aspect of the palatal cusp of the upper first molars contacted the occlusal aspect of the buccal cusp of the lower first molars, and thus, when the necessary amount of expansion was achieved. The expansion screw was then blocked with resin-based composite. According to the retention protocol, the D-MED was left in place for 3 months after the end of expansion to allow remodelling of the bone in the separated midpalatal suture (see: Fig. 4b). After 3 months, the bands and connectors were removed and the treatment could be continued with a fixed straight-wire appliance, while the four miniscrews and the expansion screw were left in place for retention (see: Fig. 4c). The patients were then seen for a retention check-up at 6 and 12 months after termination of expansion. Finally, the expansion screw and miniscrews were removed either at 12 months or, in case of a surgical orthognathic treatment, prior to surgery.

Fig. 4
figure 4

Maxillary occlusal view with D-MED at T0 (A: left), T1 (B: middle), and 3 months after expansion (C: right)

Full size image

Data collection

IOS and CBCT scans were acquired before expansion (T0) and immediately at the end of expansion (T1). The CBCT scans were made on KaVo equipment (KaVo 3D eXam, KaVo Dental, Biberach, Germany) at 120 kV and 5 mA pulse mode and the patients were scanned seated upright, with the Frankfort horizontal plane parallel to the floor. At T0, the “Extended Height” modus was used (field of view: 23 cm diameter × 17 cm height; scan time: 17.8 s; voxel size: 0.3 mm) so it was also possible to extract an orthopantomogram and lateral cephalogram from the datafile for treatment planning purposes. At T1, the “Maxilla Only” modus was used (16 cm diameter × variable height; scan time: 14.7 s; voxel size: 0.2 mm) to reduce the radiation dose.

The stereolithography files from the IOSs and the Digital Imaging and Communications in Medicine files from the CBCT scans were fused using IPS CaseDesigner® software (KLS Martin Group, Tuttlingen, Germany), creating separate fusion models for T0 and T1 per patient. The created models were imported into Maxilim® software (Medicim NV, Mechelen, Belgium) and a 3D reference frame was set up as follows: the horizontal plane was defined as a plane passing through the anterior nasal spine (ANS-point) anteriorly and the left and right notch located on the posterior maxillary border (see: Fig. 5a); the midsagittal plane was constructed as a plane perpendicular to the horizontal plane passing through the ANS-point and the middle of the line between the left and right posterior maxillary notch; the vertical (coronal) plane was constructed perpendicular to the horizontal and midsagittal planes (see: Fig. 5b). After setting up the reference frame, the dental and skeletal landmarks were identified manually on each fusion model.

Fig. 5
figure 5

Orientation of the CBCT. (A: left) CBCT orientation of the horizontal plane defined by 3 landmarks: ANS-point, maxillary posterior border notch on the right (MbR), maxillary posterior border notch on the left (MbL). (B: right) CBCT orientation with 3D reference frame

Full size image

Measurements

The measurements were taken at T0 (see: Fig. 6a) and T1 (see: Fig. 6b). The transverse measurements were measured along the X-axis, while the vertical measurements were measured along the Y-axis. The primary outcome measure was the total transverse maxillary expansion at the level of the upper first molar and first premolar, defined as the change in intermolar width (IMW) or interpremolar width (IPW) between T1 and T0, respectively. The change in intercanine width (ICW) was measured as well (see: Fig. 6c). The total transverse maxillary expansion was the sum of skeletal expansion, alveolar expansion, and dental expansion. Skeletal expansion was defined as the midpalatal suture width (MSW) measured at T1. Alveolar expansion was defined as the change in palatal alveolar width (PAW) between T1 and T0, minus the MSW. Dental expansion was calculated by subtracting the skeletal expansion and the alveolar expansion from the total transverse maxillary expansion. The nasal cavity width (NCW) and the nasal cavity expansion (T1-T0) were also measured (see: Fig. 6d). The measurements were taken in the coronal plane on the CBCT at the level of the buccal groove of the first molar or the buccal cusp tip of the first premolar. Detailed descriptions of the measurements are shown in Table 1. The success rate of the treatment was the percentage of patients achieving the required maxillary width.

Fig. 6
figure 6

The hard palate at T0 and T1, and skeletal and dentoalveolar measurements on CBCT and IOS. (A: upper left) Hard palate at T0, before midpalatal suture separation, on an axial CBCT slice. (B: upper middle) Hard palate at T1, with separated midpalatal suture, on an axial CBCT slice. (C: upper right) Measurement of IMW, IPW, and ICW on an intra-oral scan. (D: lower left) Measurement of MSW, PAW, and NCW on a coronal cross-sectional CBCT slice through the buccal groove of the M1. (E: lower middle) Measurement of TI left and right on a coronal cross-sectional CBCT slice through the buccal groove of the M1. (F: lower right) Measurement of BBT left and right on a coronal cross-sectional CBCT slice through widest portion of the mesiobuccal root of the M1

Full size image
Table 1 Definitions of the measurements used in this study
Full size table

The secondary outcome measures, assessing dental and periodontal side-effects, were tooth inclination (TI), as a measure of angular dental tip** (see: Fig. 6e), clinical crown height (CCH), as a measure of gingival recession, and buccal bone thickness (BBT), as a measure of buccal bone loss. These measurements were taken at T0 and T1 at the level of the first molar and first premolar on the left (L) and right (R) sides and mean values were used for statistical analysis. BBT at the first molar was measured at the mesiobuccal root because, given its greater proximity to the maxillary buccal alveolar cortex, it was found to be more prone to significant changes than the distobuccal root (see: Fig. 6f) [10]. Lastly, treatment duration of the expansion was recorded as the number of days from the first to the last day of activation of the MARPE appliance.

Statistical analysis

A sample size analysis resulted in a minimum sample size of 33 individuals based on an alpha of 0.05, power of 95%, and an assumed effect size of 0.66 (G*Power, version 3.1.9.7., Dusseldorf, Germany). All measurements were performed by two independent observers (A.K. and B.O.). A random selection of 50% of the measurements was repeated by both observers individually after 3 weeks. Intra-observer reliability and inter-observer reliability were assessed using the intraclass correlation coefficient (ICC) and the duplicate measurement error (DME). The pre- and post-operative measurements were explored with descriptive statistics. Normal distribution of data was confirmed using the Kolmogorov–Smirnov test, except for mean buccal bone thickness at M1 at T1 and at P1 at T0. Paired t-tests, Mann–Whitney U test, and Wilcoxon signed ranks test were used to investigate changes between T0 and T1. The level of significance was set at 0.05. All statistical analyses were performed using SPSS® Statistics version 25.0 (IBM Corp., Armonk, NY, USA).

Results

From October 2018 to September 2020, 35 patients were consecutively enrolled. One patient chose to terminate the orthodontic treatment after 2 days and this patient was excluded. Consequently, 34 patients (8 male, 26 female; mean age: 27.0 ± 9.4 years; age range: 17.1–56.0 years) were included for further analysis. The majority of patients were skeletally mature and presented with an Angle Class II/1 malocclusion with a posterior crossbite (see: Table 2). The maxillary expansion was successful in 32 patients (6 male, 26 female; mean age: 26.1 ± 8.2 years; age range: 17.1–47.5 years), resulting in a success rate of 94.1%. In two patients (2 men; 26.4 and 56.0 years old), separation of the midpalatal suture was not achieved as no diastema was observed clinically, in one due to a fracture of the expansion screw, and in the other due to substantial dental side-effects (buccal translation of the molars). The treatment time from initiation to termination of expansion with the MARPE appliance, the observation time, was on average 31.7 ± 8.0 days (range: 21–56 days). A post-hoc power analysis based on an alpha of 0.05, a sample size of 32, and an effect size of 3.86 (mean total expansion 6.56 mm and SD 1.7 mm) demonstrated that a power of 0.99 was obtained in this study.

Table 2 Basic characteristics of the total study population, as well as of the midpalatal suture separation and non-separation group
Full size table

Inter- and intra-observer reliability analysis

Intra-observer reliability was excellent for both observer 1 (A.K.) (mean ICC: 0.94 ± 0.05; mean DME: 0.48 ± 0.38) and observer 2 (B.O.) (mean ICC: 0.92 ± 0.08; mean DME: 0.51 ± 0.35). Also, an excellent inter-observer reliability was found, with a mean ICC of 0.95 ± 0.06 and DME 0.43 ± 0.33.

Maxillary expansion

The pre- and post-expansion intermolar, interpremolar, and intercanine widths are displayed in Table 3. The mean total transverse maxillary expansion at M1 and P1 was 6.56 ± 1.70 mm and 4.19 ± 1.29 mm, respectively (see: Table 4). The skeletal expansion (midpalatal suture width) at the level of M1 and P1 is shown in Tables 3 and 4. The mean skeletal expansion was 3.75 ± 1.02 mm at M1 and 3.86 ± 0.95 mm at P1. The mean anterior skeletal expansion was 3.0% (0.11 mm) larger than the posterior expansion (p = 0.29). After deducting the midpalatal suture width from the palatal alveolar width, the resulting value, defined as alveolar expansion or alveolar bone bending, was 0.68 ± 1.65 mm at M1 and 0.00 ± 1.22 mm at P1 (see: Table 5). The dental expansion, the component of expansion after deduction of the skeletal and alveolar expansion, was 2.12 ± 1.58 mm at M1 and 0.33 ± 1.16 mm at P1 (see: Table 5). All expansion results were statistically significant (p < 0.001). Based on these results, the contribution to expansion at M1 and P1 was 60.4 ± 20.1% and 92.2 ± 14.5% skeletal, 8.1 ± 27.6% and 0.0 ± 18.6% alveolar, and 31.6 ± 20.1% and 7.8 ± 17.7% dental, respectively (see: Table 5).

Table 3 Mean (± SD) dental and skeletal measurements (mm) at M1, P1, and/or C at T0 and T1
Full size table
Table 4 Mean (± SD) skeletal and dentoalveolar expansion (mm) (T1-T0) at M1, P1, and/or C
Full size table
Table 5 Mean (± SD) skeletal, alveolar, and dental component (mm; %) of total expansion at M1 and P1
Full size table

The mean nasal cavity widths at T0 and T1 at the level of M1 are shown in Table 3. The nasal cavity expansion, the mean difference between T1 and T0, was 2.07 ± 1.42 mm (p < 0.001; see: Table 4), which was 55.2% of the skeletal expansion.

Side-effects of expansion

The mean values of side-effects at T0 and T1 and their difference (T1-T0) are shown in Table 6. The mean increase in tooth inclination was 3.88 ± 3.92° (p < 0.001) at M1 and 2.29 ± 3.89° (p < 0.01) at P1. Clinical crown height increased on average 0.12 ± 0.31 mm (p < 0.05) at M1 and 0.04 ± 0.22 mm (p = 0.31) at P1. Buccal bone thickness decreased on average 0.31 ± 0.49 mm at M1 (p < 0.01) and 0.01 ± 0.45 mm at P1 (p = 0.91).

Table 6 Mean (± SD) TI (degrees), CCH (mm), and BBT (mm) at T0, T1, and T1-T0 at M1 and P1
Full size table

Discussion

Study and appliance design

The aim of the present study to provide a higher degree of evidence on the efficacy of of MARPE was supported by a number of methodological choices. In contrast to most previous studies, a prospective study design was applied. The selection bias, one of the main shortcomings of previous reports [17]. Different factors, such as tip** of the miniscrews, low palatal thickness, or bone density, could have contributed to tip**, as they could have led to increased force delivery on the anchoring molars [18, 37]. However, the mean increase of clinical crown height was not clinically relevant, as well as the decrease of buccal bone thickness, which was similar to that found in earlier MARPE reports [15,16,17], while studies on RPE presented contrasting results, with minimal changes in some [35], and a significant decrease in others [10, 36]. However, in patients with a weak periodontium, thinning of the buccal bone could be a potentially relevant side-effect, as these patients may be more prone to further periodontal deterioration [10]. This was the case in one out of two failed expansions observed in this study.

Study limitations and future perspectives

Considering study design, ideally, an RCT would provide stronger scientific evidence by comparing the effects of MARPE and SARPE. However, due to the surgical nature of SARPE, random allocation of patients to a surgery and a non-surgery group may face challenging ethical issues. In order to address concerns pertinent to the ALARA principle of radiation exposure, the CBCT at T1 had a smaller field of view. On the other hand, this created a challenge relevant to consistency between the first and the second image acquisition, which, nonetheless, was overcome by the selection of the reference frame.

Despite the positive treatment outcome of the D-MED in the immediate post-expansion period, the observation time was very short and studies regarding the long-term effects and stability following MARPE expansion would be recommended. Due to the limited observation time, it was not possible to evaluate root resorption as a potential side-effect. Furthermore, as there were only two failures, a reliable prediction on which participants might not achieve a successful expansion could not be made. This indicates the need for further research on the prognostic factors, such as palatal morphology in a prospective clinical setting. Nonetheless, it can be expected that the use of MARPE for the correction of transverse maxillary discrepancy will continue to increase.

Conclusions

This prospective clinical cohort study contributes to a higher degree of evidence on the efficacy of MARPE, which achieved a significant transverse maxillary expansion in a large and diverse group of late adolescents and adults.

The Dutch Maxillary Expansion Device demonstrated to be an effective MARPE appliance with a success rate of 94.1%. Moreover, the skeletal component of expansion was very high both at molar and at premolar levels, 60.4% and 92.2%, respectively. Buccal dental tip**, clinical crown height increase, and buccal bone thinning were observed, but could be considered clinically insignificant in patients with a healthy periodontium.

Therefore, MARPE is an effective and safe non-surgical treatment for transverse maxillary deficiency in late adolescents and adults.