The short- and long-term changes of upper airway and alar in nongrowing patients treated with Mini-Implant Assisted Rapid Palatal Expansion (MARPE): a systematic review and meta-analysis

This meta-analysis was conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) statement guidelines. The Prospero registration number for the study protocol is CRD42023406225.

As shown in Table 1, the PICOS principle was used to construct the inclusion and exclusion criteria.

The PubMed, Embase, Scopus, Web of Science, and Cochrane Library electronic databases were searched up to 2 August, 2023 without date or language restrictions. Unpublished studies were eligible for inclusion. The reference lists of previous systematic reviews and meta-analyses were also manually searched. Search strategy was formulated and the details are shown in Appendix Table 1.

Duplicate documents were removed with Endnote X9. Based on the title and abstract information of the articles, the two reviewers (LCY and WKX) independently screened out the articles that met the criteria, and then reviewed the full texts of the potentially eligible articles to select those that ultimately met the inclusion criteria. Disagreements between reviewers were resolved by consensus with a third reviewer (LY).

Data extraction was independently performed by two reviewers (JCM/ZY) according to full text inclusion/exclusion criteria. When there were insufficient data in the articles, we contacted the authors by e-mail for additional information. Any disagreements between the two reviewers were resolved by discussion with a third reviewer (FCH). The following data were extracted from the included studies: sample size, sex, mean age, type of study, appliance type, quantity, length and diameter of the mini-implants, activation protocol, activation duration and measurement time.

Due to different follow-up times in the included studies, we pooled them into three time points: before treatment (T0), within 3 months after expansion (T1), and more than 3 months after expansion (T2). Changes in the main outcomes over time periods T1-T0, T2-T1 and T2-T0 were extracted and quantified as the mean difference and 95% confidence intervals. The main outcomes were divided into 3 parts:

Two reviewers (ZYY/ZQW) independently assessed the quality of 24 articles. Any disagreements were resolved by discussion with the third reviewer (LY). The criteria were modified from the method reported by Jing Huang on the basis of the CONSORT statement [15]. The detailed criteria for risk of bias assessment are mainly divided into the following 6 aspects with 17 articles:

Each item is scored as 1 if it meets the criteria and 0 if it does not. In total, the maximum sum was 17 points; scores of ≥ 15, scores of < 15 and ≥ 12 and scores of < 12 were considered to represent high, moderate, and low quality, respectively.

Summary measures and synthesis of results were independently performed by two reviewers (LCY/WKX). Stata MP 17.0 was used for data analysis. The mean difference and 95% confidence interval (CI) were used for continuous data. I² statistics were used to test the statistical heterogeneity with a significance level of α = 0.05. If the I² value was greater than 50%, there was high heterogeneity and a random effects model was adopted. Then further explore the reasons for high heterogeneity in the results, including analyzing study quality and conducting subgroup analysis based on different sample characteristics and treatment methods, such as patient age, gender, degree of palatal suture ossification and staging, palatal bone thickness, length of dental implants, and presence of dental anchorage etc. [20‐24]. When the I² value was lower than 50%, a fixed effects model was used. Based on the main outcomes, forest plots were made for visual analysis. Egger tests were used to quantitatively assess publication bias with a significance level of α = 0.05, and funnel plots were used for visual evaluation. Sensitivity analyses were performed to assess the robustness of the main outcomes [25]. T1-T0, T2-T1 and T2-T0 were used for subgroup analysis.

As shown in Fig. 1, the PRISMA flow diagram shows the study selection process. The initial results from the 5 databases were as follows: PubMed, 563; Embase, 189; Scopus, 1112; Web of Science, 285; and Cochrane Library, 84. A total of 2234 studies were included, and 1450 were obtained after removing 784 duplicates. After screening titles and abstracts, 1418 articles were excluded and 32 studies were included for full text analysis. One study [26] was obtained by manually searching the reference list of the study by Kapetanović et al. [27]. A total of 24 studies were included after examining the full texts, 22 of which were included for quantitative analysis.

Main characteristics of the included studies in this systematic review are shown in Table S1. Twelve studies were of the retrospective one-group pretest–posttest design [23, 28‐38]. Eight studies were prospective [19‐22, 39‐42] and four studies were retrospectively compared with other arch expansion methods [18, 26, 43, 44]. The 24 studies included a total of 599 patients, 240 women and 320 men. Three studies did not report information on gender [35, 40, 44]. The average age of the patients was over 15 years old. For the treatment protocol, 4 min-screws in the palate are used, but the diameter, length and type of appliance are different. The activation protocol was 1 or 2 turns (0.13–0.25 mm per turn) per day or every other day. The definition of successful expansion varied slightly across the included studies, but it was commonly considered adequate when the palatal cusps of the maxillary first molars touched the buccal cusps of the mandibular first molars.

The results of the bias assessment of the 24 included studies are shown in Table S2. Three studies were of low quality due to insufficient sample sizes and no description of sex characteristics [35, 40, 44]. Only one high-quality study was conducted [22]. Most of the remaining studies were retrospective, so the quality was moderate. Six studies conducted posttreatment follow-up [19, 28, 30, 31, 33, 41]. Seven studies used assessor blinding [18, 20, 22, 23, 30, 42, 43].

Measurement results of nasal cavity, upper airway and alar changes are shown in Table S3. Forest plots of the measurement results in T1-T0, T2-T1, and T2-T0 are shown in Figures S1, S2, S3, S4, S5, S6, S7, S8, S9, S10 and S11. The vertical line of the diagram indicates the point estimate of the means and the horizontal line shows the 95% CI. The pooled mean estimate with 95% CI of the nasal cavity, upper airway and alar changes are shown in Table S4.

Heterogeneity tests showed that the p values of the volume of the nasopharynx, palatopharynx, glossopharynx, oropharynx and hypopharynx were not statistically significant (p > 0.05), so a fixed effects model was used (I²<50%). For other main outcomes, a random effects model was used. Owing to the constraints imposed by the number of studies encompassed, further subgroup analyses is restricted solely to the width of the nasal cavity. Among the included studies, the divergence in key attributes resides in mini screws length and the activation protocols. As exemplified in Figures S12, S13 and S14, the outcomes of the subgroup analysis evinced statistical significance with a P-value <0.05, thereby intimating that the length of mini screws and the activation protocol exert an influence on MARPE treatment. It is discerned that longer mini-screws and a twice-daily activation regimen may engender more pronounced expansion.

Publication bias was assessed for indicators greater than 3 of the included studies. As shown in Figures S15, S16, S17, S18, S19, S20, S21, S22, S23, S24, S25, S26, S27, S28, S29, S30, S31 and S32 and Table S5, visual and quantitative assessments were performed using funnel plots and Egger tests respectively. The results revealed the non-significant p value of the parameters (P > 0.05), no significant publication bias was considered. The sensitivity analysis results are shown in Table S6 and Figures S33, S34, S35, S36, S37, S38, S39, S40, S41, S42, S43, S44, S45, S46, S47, S48, S49, S50, S51, S52, S53, S54 and S55. We found that the pooled mean estimates with 95% CIs of nasal cavity volume, oropharyngeal volume, hypopharyngeal volume and total volume were strongly influenced by the included studies and may have heterogeneity. The results for other parameters were more robust and were less affected by the results of the included studies.

The objective of this meta-analysis was to assess short- and long-term changes in the upper airway and alar width in nongrowing patients older than 15 years of age who received MARPE. A total of 24 studies met the eligibility criteria and 22 studies were included in a quantitative analysis.

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 cm³, 95% CI: 0.74, 1.38) at T1-T0. Nasal cavity volume (WMD: 1.88 cm³, 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 cm³, 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.

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 cm³, 95% CI: -0.25, 1.27) and glossopharyngeal volume (WMD: 0.34 cm³, 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 cm³, 95% CI: -3.62, 2.65) did not change significantly at T1-T0, but it (WMD: 0.92 cm³, 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 cm³; 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 cm³, 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.

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 I² > 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.

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.