Changes in nasal cavity width, nasal floor width and nasal cavity volume
According to the results of the present study, nasal cavity width (WMD: 2.10 mm, 95% CI: 1.73, 2.47) and nasal floor width (WMD: 2.43 mm, 95% CI: 1.87, 2.99) increased at T1-T0. Although nasal cavity width (WMD: -0.29 mm, 95% CI: -0.44, -0.13) relapsed to some extent at T2-T1, nasal cavity width (WMD: 2.81 mm, 95% CI: 1.09, 4.53) and nasal floor width (WMD: 2.37 mm, 95% CI: 1.67, 3.07) increased at T2-T0. Lim et al. also reported that nasal floor width (mean: -0.64 mm, 95% CI: -0.93, -0.35) decreased at T2-T1 [
32].
Only one study reported an increase in nasal cavity volume (mean: 1.06 cm
3, 95% CI: 0.74, 1.38) at T1-T0. Nasal cavity volume (WMD: 1.88 cm
3, 95% CI: 1.05, 2.72) also increased at T2-T0. Inconsistent with the results of a decrease in nasal cavity width (WMD: -0.29 mm, 95% CI: -0.44, -0.13) at T2-T1, Kim et al. reported an increase in nasal cavity volume (mean: 0.65 cm
3, 95% CI: 0.22, 1.08) over this time period [
31].
MARPE is generally maintained for 3 months after treatment and corresponding soft tissue modifications after skeletal tissue changes take more time, which may be the cause of width recurrence from T2-T1. Cameron et al. reported an increase in nasal cavity width can effectively enlarge the nasal cavity volume to improve respiratory function, and it was maintained after 8 years of follow-up [
45]. Arqub et al. systematically found that increased nasal cavity width after MARPE treatment can reduce respiratory resistance [
46]. There was a high level of heterogeneity for width changes, partly due to different activation protocols, with three studies [
30,
31,
36] activating one turn per day, two studies[
37,
39] activating one turn every other day, and the remaining four studies activating two turns (0.25 mm per turn) per day. Nonetheless, it must be noted that there is currently limited research on the efficacy of MARPE treatment in nongrowing patients. Earlier investigations indicate that both rapid and slow expansion protocols yield comparable overall effects [
47,
48]. Furthermore, the subgroup analysis further underscores that the length of mini screws can affect the changes in nasal cavity width after MARPE treatment. This observation aligns with the findings of Choi's study, which suggest that the longer implants may increase the amount of skeletal expansion due to their bio-cortical bone anchorage [
22].
The sensitivity analysis revealed that the findings were robust and unaffected by the included studies, so the increase in nasal cavity width and nasal floor width after MARPE treatment was significant. However, the results of nasal cavity volume in T2-T0 were strongly influenced by the included studies and the results were not robust. Therefore, conclusions should be drawn with caution regarding the increase in nasal cavity volume after MARPE treatment. Higher quality RCTs are needed to investigate the impact of MARPE on nasal cavity volume.
Changes in upper airway volume
In this study, nasopharyngeal volume (WMD: 0.75 cm3, 95% CI: 0.44, 1.06) increased at T2-T0. There was no significant heterogeneity, I2 = 32.5% (P > 0.05). Sensitivity analysis revealed that the results were stable, so the effect of MARPE treatment on nasopharyngeal volume was statistically significant.
Shetty et al. reported that there was no significant change in palatopharyngeal and glossopharyngeal volume immediately after expansion [
35]. The results of our systematic analysis suggested that palatopharyngeal volume (WMD: 0.51 cm
3, 95% CI: -0.25, 1.27) and glossopharyngeal volume (WMD: 0.34 cm
3, 95% CI: -0.26, 0.94) remained unchanged at T2-T0. We anatomically defined the oropharynx as being divided into the palatopharynx and glossopharynx. Similar to the palatopharyngeal and glossopharyngeal volume changes at T2-T0, oropharyngeal volume (WMD: -0.49 cm
3, 95% CI: -3.62, 2.65) did not change significantly at T1-T0, but it (WMD: 0.92 cm
3, 95% CI: 0.50, 1.33) increased at T2-T0. Kim et al. reported an increase in oropharyngeal volume at T2-T1 [
31]. These inconsistent results may be due to differences in the definition of anatomical boundaries, the activation protocol and the thickness of the implant through the cortex used. Kim et al. [
31] used the choanae and the third cervical vertebrae as the boundary, while the other 5 studies [
20,
32,
35,
36,
38] used the posterior nasal spine (PNS) and epiglottis as the boundary. Tang et al. [
36] activated 1 turn(0.2 mm) per day, and other studies[
31,
32,
35,
38] activated 2 turns(0.25 mm) a day. The different mandibular positions due to upright and supine positions when taking CBCT may be responsible for the contradictory results. Tang et al. [
36] placed the patient in the supine position when performing CBCT, Aneris, Kim and Yi et al. [
20,
31,
38] placed the patient in the upright position. There are studies reporting that the use of different appliance materials can affect the effectiveness of orthodontic treatment such as self-curing plastics or cast metals, molar anchorage [
49,
50].
In addition, the sensitivity analysis showed that the oropharyngeal volume was greatly influenced by the included studies. Therefore, the results were unstable and due to the small number of included articles and varying quality, it is cautionary to draw conclusions that oropharyngeal volume was increased at T2-T0 despite the results of meta-analysis suggesting an increase. Previous meta-analyses have shown that nasal cavity volume increased and oropharyngeal volume remained unchanged after MARPE, but they included pediatric patients, so the confounding factors of growth and development cannot be ruled out [
51].
Two articles described no significant changes in hypopharyngeal volume (WMD: -0.90 cm
3; 95% CI: -1.86, 0.06) and the difference was not statistically significant (
P > 0.05) [
32,
36]. This is consistent with previous studies that showed no significant change in the volume of the inferior section of the upper airway after MARPE treatment [
52]. Total volume (WMD: 1.67 cm
3, 95% CI: 0.68, 2.66) increased at T2-T0.
Li et al. reported that MARPE can produce more transverse skeletal expansion, relieve maxillary transverse deficiency and improve upper airway ventilation [
32]. Tang et al. found that respiratory resistance, speed, minimum shear force and other respiratory functions of adult patients improved after MARPE treatment, which mainly relied on their anatomical changes [
36]. Studies have found that improved airflow characteristics in patients with sleep apnoea syndrome after MARPE treatment are significantly associated with improvements in polysomnography results, suggesting that MARPE is a viable treatment option [
2]. However, Arqub et al. reported that there was no correlation between upper airway changes and airway ventilation in a systematic analysis [
46]. Ronchi et al. found that mandibular setback surgery results a statistically significant posterior airway space reduction in the medium- and long-term follow-up. But, no direct correlation was identified with OSAS risk [
53]. Due to the small number of included articles, more high-quality RCTs are needed to explore the relationship between upper airway changes and respiratory function.
Alar width changes
Based on the results of this study, alar width (WMD: 1.47 mm; 95% CI: 0.40, 2.55) and alar base width (WMD: 1.62 mm; 95% CI: 1.11, 2.13) increased in T1-T0. The difference was statistically significant (
P < 0.05) and the changes in both were consistent with the corresponding skeletal nasal cavity width changes. However, heterogeneity analyses revealed that I
2 > 80%, and sensitivity analyses showed robust results, possibly due to inconsistent measurements in the included studies. Jesus et al. [
44] and Shetty et al. [
35] used CBCT three-dimensional (3D) reconstruction, Krijt et al. and Lee et al. used 3D stereophotogrammetry [
19,
39]. An et al. used 2D frontal photos to find that the alar width also increased [
18]. 3D measurement of facial soft tissues is a new direction that can avoid the superimposition and image distortion observed with the 2D radiography technique [
28]. Staller et al. found that there are some differences between 2D photos and 3D measurements. However, this error is clinically acceptable [
54]. Lavorgna also reported that there were no significant differences emerged in the measurements made with 3D stereophotogrammetry and photogrammetry [
55].
From the perspective of clinical methodology, the activation protocol was as follows: Jesus et al. (0.5 mm/d) > Krijt et al. (0.25 mm/d) > Lee et al. (0.2 mm/d), which may be one of the reasons for the high heterogeneity of alar width. Therefore, orthodontists should pay attention to the effect of activation strategies on alar changes. This may affect the aesthetics of the nose. Brito et al. believed that the nasal framework basically determines nasal morphology, and nasal morphology changes when the skeletal nasal cavity changes [
56]. In patients with a depression in the middle of the face, the anterior movement of the maxilla can improve the facial shape, but the increase in alar width and alar base width may affect the aesthetics of the nose, resulting in a collapsed nose and a humped nose [
15]. Abedini et al. used 3DMD to analyse facial soft tissue and reported significant forwards sagittal and lateral asymmetry changes in the paranasal, upper lip, both cheeks and with greater changes in the cheek area. Those changes remained stable after 1 year [
28]. An et al. found that the maxilla A point was significantly forwarded by 1.3 mm on average, and the length of the nose and upper lip increased, but this increase was not statistically significant [
18]. Shetty et al. also found an increase in the H-angle after MARPE treatment, which may be related to the forwards shift of the A point [
35]. Almaqrami et al. reported that the amount of maxillary forwards movement was small (0.88°), which might not be clinically significant and the mandible rotated downwards and backwards. However, Nguyen et al. found more significant lateral changes in paranasal and cheek areas [
34].
The goal of orthodontic treatment is not only the alignment of teeth and a stable occlusal relationship, but the beauty and coordination of the face is also one of the goals pursued by orthodontists and patients. The coordination of the nose, lips and chin is an important parameter index to evaluate the results of orthodontic treatment. The effect of MARPE on the soft tissues has rarely been studied, which could be due to the assumption that it is expected to be minute, or overshadowed by other growth changes and therefore hard to evaluate. However, with the introduction of MARPE, more nongrowing patients can be expanded skeletally, which calls for a more in-depth study on the actual effects of expansion on soft tissues.
Strengths and limitations
A large number of studies have focused on skeletal and dental changes after MARPE treatment, with limited research on the upper airway and facial soft tissue. Patients treated with MARPE are usually in the advanced stages of growth, therefore the use of traditional expanders such as Haas Hyrax is not indicated. Compared with the meta-analysis of Li et al., we did not include studies with case groups under 15 years of age and excluded growth as a confounding factor [
51]. This meta-analysis was conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement guidelines. We strictly defined the inclusion and exclusion criteria in accordance with the PICOS principles. We used follow-up time as a subgroup and performed detailed heterogeneity tests, publication bias, and sensitivity analyses.
The quality of the literature included in this review was mostly moderate and most of them were retrospective one-group pretest–posttest study designs with a small sample size (< 25). It is difficult to conduct high-quality RCTs due to clinical ethical issues. Heterogeneity was high, but the number of studies did not allow for relevant subgroup analyses and sensitivity tests. More studies need to be included for analysis due to the potential confounding factors of race, sex, appliance type, diameter and length of implant nails, etc. The anatomical boundaries and measurement methods of the included studies were also not completely consistent. Most of the follow-up included in this study was within one year, and it was difficult to assess longer-term changes after MARPE treatment, so longer follow-up studies are encouraged.