The midpalatal suture becomes more tortuous and interdigitated with increasing age.1 Rapid palatal expansion can yield acceptable outcomes when performed before the end of the adolescent growth spurt. The dentoalveolar side effects of this treatment are more obvious among adults.2-4 In mature patients, surgically assisted rapid palatal expansion and microimplant-assisted rapid palatal expansion (MARPE) constitute alternatives for skeletal expansion.5-9

The maxillary skeletal expander (MSE) is a type of MARPE appliance. The jackscrew plate houses 4 microimplants (MIs) that are bicortically engaged in the posterior region of the palate. This feature contributes to the effectiveness of the appliance in opening the midpalatal and circummaxillary sutures and promoting a superior and posterior expansion.10-14

However, the posterior palatal bone is relatively thin, and MI displacement during expansion may be inevitable when the resistance is high.15,16 The midpalatal suture, zygomatic buttress, and pterygomaxillary suture are the 3 main resistance structures that play an important role during skeletal expansion.12-14 Therefore, the anatomy of the palatal bone and circummaxillary structures must be properly evaluated through cone-beam computed tomography (CBCT) prior to determining the exact position of the MSE. The aforementioned structures provide posterior resistance, and an expansion force must be applied to the posterior palate, where the bone is relatively thin. Because the palatal bone comprises dense cortical bone immediately lateral to the midpalatal suture, this site has been determined to be the most stable site in the maxilla for MIs.16 Cortical bone thickness and density closer to the midpalatal suture are higher than those in the middle and lateral areas of the posterior palate. Palatal bone thickness and density also vary in the anterior and posterior regions. Bone thickness is higher in the anterior areas than in the middle and posterior areas.16,17 However, the MSE jackscrew plate should be positioned at the level of the zygomatic buttress, as this is one of the main resistance structures to consider during expansion.12,14

The finite element method has been used in orthodontics to evaluate the stress, strain, and force distribution of different appliances delivered to the craniofacial structures.18,19 Recent studies have assessed the effectiveness of monocortical versus bicortical MI anchorage in MSE by evaluating the stress distribution and displacement.11,20 However, this method represents a simulation of clinical situations using virtual 3-dimensional skull models. Meanwhile, CBCT allows the study of the actual pattern of movement of maxillofacial bones, dentoalveolar structures, and MIs by any type of expansion device in 3 dimensions, with minimum image distortion and radiation dosage.7,8,21,22

Previous studies have investigated the stability of MIs of different lengths and diameters.18,23 However, studies on the displacement of MIs during expansion are lacking. This study aimed to analyze the MI displacement pattern of MSE using CBCT.

This retrospective study was approved by the institutional review board (IRB number 17-000567) of the University of California, Los Angeles (UCLA). Informed consent was waived owing to the retrospective nature of the study. We included 39 participants (12 males and 27 females) who were successfully treated with the MSE II (Biomaterials Korea, Seoul, Korea) appliance and had no craniofacial anomaly or history of orthodontic treatment. The mean age of the participants was 18.2 ± 4.2 years with a maturation stage of cervical stage 4 (CS4) or higher. All patients were diagnosed with a maxillary transverse deficit on the basis of the maxillomandibular bone width discrepancy.12 One clinician supervised the treatment for all patients at the Section of Orthodontics, UCLA School of Dentistry. Post-expansion records were obtained after successfully completing the active expansion phase and before bracket bonding. Cone-beam computed tomography scans were obtained before treatment (T0) and within 3 weeks of completion of expansion (T1). A CBCT scanner (5G; NewTom, Verona, Italy) with an 18 × 16 cm field of view, 14-bit grayscale, and a standard voxel size of 0.3 mm was used for scanning of all participants.

The maxillary sagittal plane, which passed through the anterior nasal spine (ANS), posterior nasal spine (PNS), and nasion, was established on the T0 scan.12 The following steps were taken to determine the anteroposterior position of the jackscrew and assess the bone thickness for the 4 MIs (Figure 1). First, at the level of the upper dentition in the axial view, the green line (representing a coronal cut) was displaced posteriorly until the zygomatic buttress could be viewed at its maximum expression in the coronal view. This new anteroposterior level of the green line on the axial view represents the anteroposterior center of the jackscrew. The orange line in the axial and coronal views represents the sagittal axis of the jackscrew and matches the midpalatal suture. Because the holes for the 4 MIs were located in a combination of 5 mm anterior/posterior and 3 mm right/left coordinates, the orange and green lines were moved accordingly to obtain the bone thickness at the level of the MI positions. The bone thickness associated with these 4 sites helped determine the minimum required length of the MIs to be bicortically engaged into the palate (Figure 1). The landmarks were then translated on the dental casts for MSE fabrication.

Figure 1. Cone-beam computed tomography landmarks for MSE fabrication. A, The green line on the axial view is at the level of the greater extension of the zygomatic buttress (seen in the coronal view). B, The orange line on the axial view corresponds to the midpalatal suture. Planned MSE jackscrew position on the axial and sagittal views. C, Bone thickness measurements on the coronal and sagittal cuts at the level of insertion of the right anterior microimplant.
MSE, maxillary skeletal expander.

The MSE II appliance (Figure 2) consists of a jackscrew framework that houses 4 palatal MIs with a 1.8 mm diameter and 11–13 mm length. In addition, 2 supporting arms extending from the jackscrew are welded to the molar bands. The rate of expansion was 4 to 6 turns per day (0.133 mm per turn, 0.5–0.8 mm per day) until a significant diastema of 2–3 mm was achieved. Subsequently, the rate was changed to 2 turns (≈ 0.267 mm) per day until the maxillary skeletal width matched or was greater than the mandibular width. The maxillary skeletal width represented the distance between the right and left most concave points on the maxillary vestibule above the mesiobuccal cusps of the first molars. The mandibular width was defined as the distance between the right and left buccal surfaces over the furcation of the first molars.12 The MSE was maintained in place with no additional activation for at least 6 months. To assess the effects induced purely by MSE, T1 scans were obtained immediately after expansion and before the patient received any other orthodontic appliance.

Figure 2. MSE II device (Biomaterials Korea, Seoul, Korea). One MSE II expansion turn is equivalent to 0.133 mm of activation of the jackscrew. One revolution is equivalent to 6 turns or activation of the jackscrew by 0.8 mm. MSE II-12 indicates that the expansion size is 12 mm, which is equivalent to 90 turns.
MSE, maxillary skeletal expander.

OnDemand3D (Cybermed, Daejeon, Korea) software was used to superimpose the T0 and T1 CBCT images based on the anatomical structures of the entire anterior cranial base through voxel gray-scale pattern automated matching. After superimposing the CBCT data sets, the maxillary sagittal plane was identified on the T0 scan.12 Then, the extent of the suture opening at ANS and PNS was measured on the T1 axial view, and the ratio of the distance at PNS to that at ANS was calculated to determine the parallelism of the suture split. Subsequently, the anterior MIs and the horizontal line connecting the anterior MIs were identified on the T1 axial view, and the anterior-coronal-MI section was determined on the coronal view (Figure 3).

Figure 3. Anterior-coronal-microimplant section identified on the coronal view.

Angular measurements were performed. The inter-MI angle (IMIA) was determined by drawing a line through the long axes of both right and left MIs. The 2 lines were projected superiorly until they intersected, and the corresponding angle was recorded. During expansion, the jackscrew plate often bends slightly. The plate angle (PA) was measured at the point of convergence of the 2 hemisections of the jackscrew plate. The MI-to-hemiplate angle (MIPA) was measured by connecting the long axis of the right or left MI to the respective hemisection of the jackscrew plate.

Linear measurements were also recorded. The inter-MI neck distance (IMIND) was defined as the distance between the right and left central parts of the MI neck embedded in the jackscrew plate. The inter-MI apical distance (IMIAD) was recorded as the distance from the right to the left MI at the apical level. In addition, the ratio between the IMIND and IMIAD was obtained for both anterior and posterior MIs. The bone thickness supporting the MIs was calculated as the palatal thickness at the MI site (PTMI). Finally, the suture opening at the anterior and posterior MIs (SOMI) was measured (Figure 4). Once all measurements for the anterior MIs were obtained, the posterior coronal MI section was determined, and the measurement process was repeated.

Figure 4. Parameters evaluated in the study.
IMIA, inter-microimplant angle; IMIAD, inter-microimplant apical distance; SOMI, suture opening at microimplant level; PTMI, palatal thickness at microimplant level; PA, plate angle; MIPA, microimplant to plate angle; IMIND, inter-microimplant neck distance.

Statistical analysis

On the basis of the findings of a previous study,12 the minimum sample size required to reveal significant changes after MSE was 14 patients. This calculation was determined considering a power of 0.85, an alpha value of 0.05, and a mean difference of 1.0 ± 1.0 mm for the lateral displacement of the zygomaticomaxillary complex after expansion. Therefore, a sample size of 39 patients was sufficient to determine the significance. All parameters were measured for 10 randomly selected patients by 2 raters to assess the reliability of the method. The measurements were repeated after 4 weeks by the same operators to determine the reliability. Descriptive statistical analyses were performed. Distribution tests were also conducted. Independent t and Mann–Whitney U tests were used to compute the P value and determine the difference between the anterior and posterior MI displacement patterns. Additionally, independent t tests were performed to assess the differences between the male and female participants. Finally, the Pearson correlation coefficient was established to identify any correlation between the extent of the suture opening at the ANS and PNS, MI displacement pattern, and PTMI.

The average amount of activation of the MSE expansion jackscrew was 9.2 ± 1.6 mm. The average suture opening at the ANS was 5.0 ± 2.1 mm and that at the PNS was 4.9 ± 2.5 mm. The mean PNS-to-ANS ratio was 0.99 ± 0.41. There were no significant differences between the right and left sides in terms of anterior MIPA (P = 0.758), posterior MIPA (P = 0.572), anterior PTMI (P = 0.973), and posterior PTMI (P = 0.503) (Table 1).

Table 1

Right versus left microimplant to plate angle and palatal thickness at the microimplant level.

Measurement	Right (n = 39)			Left (n = 39)		P value
	Mean	SD		Mean	SD
Anterior MIPA (°)	81.93	5.20		82.30	5.50	0.758
Posterior MIPA (°)	84.34	5.13		83.66	5.54	0.572
Anterior PTMI level (mm)	3.57	1.45		3.56	1.50	0.973
Posterior PTMI level (mm)	3.01	1.60		2.76	1.61	0.503

SD, standard deviation; MI, microimplant; MIPA, MI-to-hemiplate angle; PTMI, palatal thickness at the MI site..

Considering the right and left values together, the posterior MIPA was significantly greater than the anterior MIPA (P = 0.029), and the anterior PTMI was significantly greater than the posterior PTMI (P = 0.006). There was no statistical difference between the anterior and posterior IMIA (P = 0.076), PA (P = 0.552), SOMI (P = 0.695), IMIND (P = 0.157), or IMIND to IMIAD ratio (P = 0.089). However, the posterior IMIAD was significantly greater than the anterior IMIAD (P = 0.034) (Table 2).

Table 2

Linear and angular measurements of the microimplants.

Measurement	n	Anterior			Posterior		P value
		Mean	SD		Mean	SD
MI angular measurements (°)
IMIA	39	18.55	9.84		14.46	10.20	0.076
PA	39	177.08	2.82		177.46	2.60	0.552
MIPA	78	82.12	5.35		84.01	5.32	0.029*
MI linear measurements (mm)
SOMI level	39	5.01	1.90		4.84	1.94	0.695
IMIND	39	12.96	1.92		13.60	2.04	0.157
IMIAD	39	9.99	2.37		11.22	2.68	0.034*
IMIAD ratio	39	0.77	0.13		0.82	0.14	0.089
PTMI level	78	3.57	1.47		2.89	1.60	0.006**

SD, standard deviation; MI, microimplant; IMIA, inter-MI angle; PA, plate angle; MIPA, MI-to-hemiplate angle; SOMI, suture opening at the anterior and posterior MI; IMIND, inter-MI neck distance; IMIAD, inter-MI apical distance; PTMI, palatal thickness at the MI site..

*P < 0.05, **P < 0.01..

The male and female participants had similar age distributions (P = 0.995). The PTMI was significantly higher among the male participants in both anterior (P < 0.001) and posterior regions (P = 0.031) (Table 3). No significant difference between the male and female participants was observed for the other MI displacement pattern variables (P > 0.05).

Table 3

Palatal thickness at the microimplant level in the male and female participants.

Measurement	Male (n = 12)			Female (n = 27)		P value
	Mean	SD		Mean	SD
Age (yr)	18.08	4.81		18.07	4.34	0.995
Anterior PTMI level (mm)	4.53	1.25		3.14	1.36	< 0.001***
Posterior PTMI level (mm)	3.47	1.50		2.62	1.58	0.031*

MI, microimplant; SD, standard deviation; PTMI, palatal thickness at the MI site..

*P < 0.05, ***P < 0.001..

Participant age had a strong positive correlation of 0.670 with the anterior IMIA (P < 0.01) and weak positive correlation of 0.372 with the posterior IMIA (P < 0.05). A low negative correlation of –0.347 was observed between the posterior PTMI and posterior MIPA (P < 0.01). A moderate negative correlation of –0.418 was found between the posterior IMIA and PNS-to-ANS opening ratio (P = 0.008). The intraclass correlation coefficient was 0.93, indicative of high reliability.

The MSE is a type of MARPE appliance that helps correct transverse maxillary deficiency by opening the midpalatal and circummaxillary sutures.12-14,19,24,25 The MSE enables a more superior and posterior maxillary expansion owing to the posterior and bicortical placement of MIs in the palatal bone. The MIs in the MSE are the palatal bone anchors for the jackscrew. The length of the MIs chosen for MSE treatment plays an important role. Generally, MIs of length 11–13 mm are required to achieve bicortical engagement. The bicortical engagement of the 4 MIs provides a foundation for the MSE system to overcome the posterior resistance against expansion from the zygomatic buttresses and pterygopalatine sutures. The posterior placement of the MSE disarticulates the pterygopalatine suture, allowing for more parallel expansion.13 Therefore, the anteroposterior position of the jackscrew plays a critical role. The posterior site often coincides with the sagittal location of the upper first molars. This study presented the steps to achieve the recommended anteroposterior position of the jackscrew at the level of the maximum expression of the zygomaticomaxillary buttress on the CBCT scan. On the basis of this reference, the 4 MIs were placed 3 mm lateral to the maxillary sagittal plane (coincident with the midpalatal suture) and 5 mm anterior and posterior to the coronal center of the jackscrew. This is the first study to investigate the different displacement patterns of the anterior and posterior MIs during expansion with MSE.

The palatal bone is thicker anteriorly and tapers toward the PNS.16,17 Interestingly, the palatine processes of the maxilla are connected at their posterior end with thin horizontal plates of the palatine bone, forming the entire hard palate. Therefore, different displacement patterns between the anterior and posterior MIs are anticipated after MI-supported rapid maxillary expansion with MSE. In fact, some MSE cases clinically display wider expansion at the posterior MI position (Figure 5).

Figure 5. A, MSE expansion showing a more parallel displacement of the anterior and posterior microimplants. B, MSE expansion displaying a wider displacement pattern for the posterior microimplants.
MSE, maxillary skeletal expander.

In the coronal plane, the jackscrew plate had bent slightly by 2.5–3º, with no difference between the anterior and posterior segments. The jackscrew is initially rigid to support the forces required to achieve a midpalatal suture split. Once the expansion begins, the rigidity of the anterior and posterior guiding bars reduces, and the central screw loosens. The entire plate bends as the zygomaticomaxillary complex rotates.12,14 Although the IMIA did not differ significantly between the anterior (18.5 ± 9.8º) and posterior (14.4 ± 10.2º) MIs (P > 0.05), the MIPA was significantly different at the anterior and posterior sites. The concomitant rotational movement of the zygomaticomaxillary complex12,14 contributed to the angular changes, increasing the IMIA for both anterior and posterior MIs. In contrast, the bending of the jackscrew plate may have produced underestimated angular changes in the MIPA. The results suggest that the thinner bone in the posterior area allows for more translatory displacement, and the thicker bone in the anterior area produces more tipping displacement of the MI during MSE treatment.

There was no significant difference in the palatal bone thickness at the site of the MIs between the right and left sides in both anterior and posterior regions. Consequently, the difference in the MIPA on the right and left sides was not significant for either the anterior or posterior palates. Therefore, the right and left sides were combined for the anterior or posterior values. The angle between the MIs and jackscrew plate was significantly lower for the anterior MIs than for the posterior MIs. Although there was no significant difference in the IMIND between the anterior and posterior regions, the posterior IMIAD was significantly greater (P < 0.05). The above findings clearly demonstrated a translatory displacement pattern of the posterior MIs and a tipping displacement pattern of anterior MIs. The anterior PTMI was greater than the posterior PTMI (P < 0.01). The differential thickness of the palatal bone may account for the different displacement patterns between the anterior and posterior MIs. In mature patients, a cumulative force is required to achieve an adequate level of expansion to overcome the resistance from anatomical structures. This continuous cumulative force produces a mounting force against the implants and surrounding structures until the circummaxillary sutures disarticulate. When the anchor bone is thin, as in the posterior palate, the mounting force against the MIs may not be countered by the weak anchor bone, and the MIs could cut through the bicortical layers, producing translatory displacement. In the anterior region, the anchor bone can withstand this force more effectively and prevent the MIs from cutting through the bicortical layers. Because the expansion force is generated from the inferior aspect of the palatal bone, the effects on the palatal cortical bone are greater and produce a tipping displacement pattern. This concept supports the consequent loss of parallelism between both anterior and posterior MIs. The lower anterior MIPA and IMIAD may be associated with the difference in bone thickness between the anterior and posterior regions. Pearson correlation tests revealed a significantly negative correlation between the posterior PTMI and MIPA (–0.347; P < 0.01), demonstrating that thicker bone produces more tipping displacement, and thinner bone produces more translatory displacement.

An important factor that can affect the parallelism of MIs is their insertion depth. Bicortical MI engagement provides significantly greater resistance to screw deflection and greater stability than monocortical MI engagement in the maxilla and mandible.10,26 The bicortical engagement ensures more support and retention area for the MIs. However, this stability is also related to the palatal bone thickness at the site of implantation. Bicortical engagement in a bone of 3.5 mm thickness provide greater stability than a bone of 1.2 mm thickness. This investigation presented the mean values of palatal bone thickness corresponding to the ideal sites for MI placement in 39 participants successfully treated with MSE. The mean palatal thickness for anterior MIs was 3.57 ± 1.47 mm and that for posterior MIs was 2.89 ± 1.60 mm. The palatal thickness at the anterior MI site was significantly greater among the male participants (4.53 ± 1.25 mm) than among female participants (3.14 ± 1.36 mm) (P < 0.001). Similarly, the palatal thickness at the posterior MI site was significantly greater for the male participants (3.47 ± 1.50 mm) than for the female participants (2.62 ± 1.58 mm) (P < 0.05). These values are in accordance with previous findings on maxillary bone topography; the anterior bony segments are thicker than the posterior segments, and male participants present with thicker bones than female.15-17 However, the presented values are related to the more posterior position of the MSE, in comparison to other MARPE, and to the specific site of the MI placement with bicortical engagement.

As expected, there was no significant difference in the suture opening at the anterior and posterior MI sites. The different displacement patterns between the anterior and posterior MIPA, IMIAD, and PTMI were not correlated with the extent of suture split parallelism determined by the extent of suture opening at the ANS or PNS. This indicates that the observed displacement pattern was produced prior to the disarticulation of the sutures. Once the initial split is obtained with MSE treatment, the subsequent movement involves parallel expansion, as described in previous studies.13,25 Clinicians should not be concerned about the differential displacement pattern of MIs once expansion has occurred as it would not affect the anteroposterior parallelism of the expansion. The observed displacements are of interest when the split does not occur, and any negative consequences of the treatment must be managed.

Participant age had a strong positive correlation (0.670) with anterior IMIA (P < 0.01) and weak positive correlation (0.372) with posterior IMIA (P < 0.05). This could be attributed to the reduced patency of the sutures with age, necessitating a greater orthopedic force, which induces more displacement of the MIs. In addition, there was a moderate negative correlation (–0.418) between the posterior IMIA and PNS-to-ANS ratio (P = 0.008). A greater posterior IMIA may indicate less parallelism of the suture opening at the ANS and PNS. However, the mean PNS to ANS ratio was 0.99 ± 0.41, demonstrating that this correlation is not clinically significant.

Limitations of this study are related to its retrospective nature. Further studies on the displacement pattern of MIs with a larger sample with sex, race, and age differences are needed. As all participants in the present study were successfully treated with MSE, it would be interesting to compare their findings with those of patients with failed MSE expansion treatment.

In the coronal plane, the posterior MIs underwent predominantly translatory displacement, while anterior MIs underwent predominantly tipping displacement. The palatal thickness at the anterior MI site was significantly greater than that at the posterior MI site, inducing the above patterns of displacement. Despite the differences in displacement patterns, MSE produced parallel expansion at the ANS and PNS and the anterior and posterior MI sites. The palatal thickness was significantly higher among the male participants than among the female participates at both anterior and posterior MIs. The jackscrew plate slightly bends during MSE expansion.

Conceptualization: NP, WM. Data curation: NP, AG. Formal analysis: NP, AG, RDM, OC. Investigation: NP, RDM, OC, JB. Methodology: NP, OC, TD. Project administration: NP, WM. Resources: NP, MG, FS, KB. Software: NP, MG, FS, KB. Supervision: NP, WM. Validation: NP, OC, TD. Visualization: NP, AG, JB. Writing–original draft: NP, AG, OC. Writing–review & editing: NP, WM.

No potential conflict of interest relevant to this article was reported.

A video presentation of this article is available at https://youtu.be/OdibDngJguo or www.e-kjo.org.