Factors influencing complete midpalatal suture separation in adult patients undergoing 3D-guided piezocorticotomy-assisted MARPE: A case-control study

Abstract

Objectives

To evaluate demographic, cephalometric, and craniometric biomarker (frontozygomatic separation) predictors of full versus partial midpalatal suture (MPS) separation following 3D-guided piezocorticotomy-assisted MARPE, and characterize separation morphologies.

Methods

A retrospective case-control study analyzed 108 adults (70 full separation controls; 38 partial separation cases). Separation patterns were categorized (Types 1–5). Multivariable logistic regression assessed biological sex, age, cephalometrics ($\text{ANB}$, $\text{SN}‐\text{ANS}‐\text{PNS}$), PNS-PPW width, and frontozygomatic separation. Model discrimination was evaluated via Area Under the ROC Curve ($\text{AUC}$). Linear combinations tested sex-specific age interactions ($<30$, $30-39$, $40-58$ years).

Results

Five distinct MPS separation morphologies were identified. Baseline demographic, cephalometric, and post-expansion parameters showed no group differences ($p>0.05$). Multivariable regression ($N=108$, $\text{LR}{\chi }^2\left(6\right)=26.61$, $p=0.0002$, Pseudo $R^2=0.1900$) demonstrated that baseline $\text{ANB}$ ($p=0.850$) and $\text{SN}‐\text{ANS}‐\text{PNS}$ ($p=0.519$) did not predict separation. Frontozygomatic separation was a significant positive predictor ($\text{OR}=2.81$, $95\%\text{CI}:\left[1.10,7.19\right]$, $p=0.031$). Conversely, male sex ($\text{OR}=0.14$, $p<0.001$) and advancing age ($\text{OR}=0.93$, $p=0.014$) significantly reduced separation likelihood. Adding the biomarker increased classification accuracy by $2.7\%$ ($\text{AUC}0.755\to 0.782$; ${\chi }^2=1.21$, $p=0.271$). Stratified analysis revealed age-related declines operated predominantly in males ($n=44$), with males aged $40-58$ demonstrating an $87.3\%$ odds reduction ($\text{OR}=0.127$, $p=0.024$), whereas females ($n=64$) maintained stable odds ($p>0.05$).

Conclusions

MPS separation manifests in five morphological types. Full separation is uninfluenced by baseline cephalometrics but constrained by male sex and advancing age (precipitously after age 40 in males). Frontozygomatic separation serves as a strong structural predictor.

1. Introduction

The success rate of midpalatal suture (MPS) expansion with Mini screw- Assisted Rapid Palatal Expansion (MARPE) in adults is controversial. Some studies report¹ it to be related to age of the patient (94.1% of success by the age 24 years old in adults and up to 76% in older patients), bone thickness, and not related to the stage of skeletal maturation. Study by Shin et al.² supported the age to be a confounder to the success of MPS separation, while found that the length of the MPS and its thickness also impact the outcomes of MPS separation.

The length of the midplatal suture presents a composition of two sutures: the intermaxillary suture, one connecting palatal processes of the paired maxillary bones, and the interpalatine suture, connecting horizontal processes of two palatine bones. The length of the hard palate measured from the incisive foramen to point PNS (Posterior Nasal Spine) in adolescents ranged between 31.5 and 39.9 mm.³ Thickness of the hard palate measured in the same study was divided into the thickness of the anterior palate (7.40-13.20 mm), intermediate (2.5-9.6 mm), and posterior palate (1.7-7.9 mm). In the CBCT study of 260 patients 18-45 years old Li and coauthors⁴ found significant difference between three skeletal Classes in the length of ANS-PNS distance (length of the hard palate). Thus, the ANS-PNS was significantly longer in Skeletal Class I patients (49.0 ± 3.27 mm) and the shortest in Skeletal Class III (47.7 ± 3.57 mm).

The study by Angelieri and coauth.^5,6 on MPS maturation outlines stages D and E as those directly related to the maturation of the palatine bone of the MPS. Study by Oliveira reiterated the idea of age-related success rate of the MPS separation with MARPE defining the success rate in patients with stage D MPS separation at the level of 62.5% and stage E at 58.3%. Salmoria and coauth.⁷ measured the width of interincisal diastema in patients with stage D and E MPS maturation and found that those with stage D exhibited wider diastema compared to patients with stage E maturation. A retrospective study by Bas and colleagues used maturation stage classification to determine the suitability for the midfacial expansion. Authors found that at the age of 16 yo only 41.7% of participants remained in stage C of MPS maturation. Thus, age and MPS maturation stage remain the main limiting factors for the success of MPS separation in adults.

While the Midfacial expansion in adolescents and young adults has higher success rates, analogous treatment in adults requires procedures that do not rely on age, suture maturation, MPS thickness, or other non-modifiable factors. Surgically assisted rapid palatal expansion (SARPE) has been one of such procedures, which was recently accompanied by 3D-GMP MARPE.⁸ The latter presents as a minimally invasive high-precision low-pain intervention in conjunction with MARPE of varying designs. The recent study on the effects of 3D-GMP MARPE in a mixed cohort of adult patients⁹ showed a favorable forward maxillary base movement immediately following intervention. The implications of this finding are suggestive of association between the increase in PNS-posterior pharyngeal wall distance (PNS-PPW) and maxillary base position. The PNS-PPW was shown to be smaller in subjects with OSA compared to non-OSA adults.¹⁰

The goal of the current study was to evaluate the effects of complete MPS separation following 3D-GMP MARPE compared to the incomplete (not involving palatine bones). The null hypothesis of this study was that full separation of the midpalatal suture in adult patients undergoing 3D-guided midpalatal piezocorticotomy-assisted MARPE treatment exhibits equal odds across both genders and is not associated with demographic or anatomical variables. Conversely, the alternative hypothesis posited that the odds of achieving full midpalatal suture separation are significantly influenced by sexual dimorphism, chronological age, and specific regional craniometric biomarkers.

2. Methods

Study design. This study presents a case-control study with the total number of 108 participants treated at the private practice DrKoval Orthodontics, Boca Raton, FL, USA. The manuscript is presented following the STROBE guidelines for observational studies. Ethical approval was received from Solutions Institutional Review Board (Solutions IRB) with the Ref # 0846/01-13-2026. The study was conducted according to the Declaration of Helsinki.

Setting. Participants were retrospectively recruited after the completion of the active expansion stage. The recruitment took place from January to May 2026. Data was retrospectively analyzed after initial screening for the pattern of MPS separation (partial or complete). Written informed consent for the publication of anonymized data was obtained from all participants prior to enrollment.

Participants. The eligibility criteria for entering the study included successfully completed Midfacial expansion with 3D-GMP technique with visible diastema between maxillary central incisors of more or equal to 3 mm. The age of participants was 19 years old and older. Participants were allocated to controls (n = 70) once the MPS separation followed complete pattern, and to cases (n = 38) when the MPS separation followed partial pattern with no separation of the interpalatine suture. A 1.84:1 control-to-case ratio was utilized to maximize statistical power given the limited availability of participants exhibiting the partial MPS separation pattern. Inclusion criteria for the study: age 19 years or older; presence of signs and symptoms of Sleep-Disordered Breathing (SDB) and/or Obstructive Sleep Apnea (OSA) evaluated via Epworth Sleepiness Scale (ESS ≥ 10 points) and STOP-BANG (≥2 points) questionnaires; Skeletal Class I, II, or III relationship (based on ANB angle); maxillary divergence evaluated via SN-ANS-PNS angle (SN-ANS-PNS < 11° indicating no downward maxillary base rotation; SN-ANS-PNS ≥ 11° indicating downward maxillary rotation); reduced posterior nasal width (<30 mm); reduced upper airway cross-sectional area (<400 mm²) at the level of the palatal plane; and moderate-to-severe nasal obstruction, defined as a Nasal Obstruction Subjective Evaluation (NOSE) questionnaire score of 30 or greater. Exclusion criteria included: history of bimaxillary surgery, history of Le Fort I or SARPE, history of palatopharyngoplasty, history of connective tissue disease, history of autoimmune disease, history of cancer and/or chemotherapy (Fig. 1).

Fig. 1.

Study recruitment and grouping flowchart.

Variables. Fullsep (complete MPS separation), the dependent variable, is primary outcome variable, representing the status of anatomical separation. It is coded as 1 for Full Separation (Controls) and 0 for all patterns of Partial Separation (Cases). Frontozygomasep (Frontozygomatic separation, Binary): The primary anatomical biomarker of interest, measuring the frontozygomatic separation (0 indicates no bilateral separation, 1 – bilateral separation). Male (Binary): A demographic covariate representing biological sex, coded as 1 for male and 0 for female. Age (Continuous): A demographic covariate representing the participant's chronological age in years. PNS-PPW (Posterior Nasal Spine to Posterior Pharyngeal Wall distance, Continuous): An anatomical covariate representing the distance from the point PNS (Posterior Nasal Spine) to the posterior pharyngeal wall (Fig. 2).

Fig. 2.

Fronto-maxiallry suture separation and PNS-Posterior Pharyngeal Wall measurement: 1 – Posterior Nasla Spine (PNS), 2 – posterior pharyngeal wall (PPW), 3 – distance from PNS to PPW, 4 – upper airway volume reconstruction, 5 – cervical spine, 6 – hard palate, 7 – soft palate, 8 – zygomatic process of the frontal bone, 9 – frontozygomatic suture, 1- - frontal process of the zygomatic bone, 11- MARPE, 12- nasal base separation (midpalatal suture separation).

Data sources. Measurements. Participants were classified into distinct morphological categories based on the observed patterns of midpalatal suture separation. Anatomically, the midpalatal suture comprises four structural segments formed by two paired craniofacial bones: the palatine processes of the maxillary bones anteriorly, and the two distinct palatine bones posteriorly (Fig. 3). Five distinct separation types were identified and defined as follows:

• •Type 1 (Full Separation): Complete parallel separation extending from the anterior nasal spine (ANS) to the posterior nasal spine (PNS), involving both the palatine processes of the maxillary bones and the horizontal plates of the palatine bones.
• •Type 2 (Middle Separation): Separation restricted to the middle third of the midpalatal suture, presenting with minimal to no separation in the corresponding anterior and posterior thirds.
• •Type 3 (Anterior Separation): Separation isolated to the anterior two-thirds of the midpalatal suture, with no involvement or separation observed in the region of the palatine bones.
• •Type 4 (Anterior/Unilateral Transverse Separation): Separation involving the anterior two-thirds of the midpalatal suture, accompanied by unilateral separation of the transverse palatine suture.
• •Type 5 (Anterior/Bilateral Transverse Separation): Separation involving the anterior two-thirds of the midpalatal suture accompanied by bilateral separation of the transverse palatine suture, notably presenting with no separation of the midpalatal suture between the palatine bones (Fig. 4, Table 4).

Fig. 3.

Architecture of the Midpalatal Suture (MPS): 1- MPS, 2 – transverse suture, 3 – incisive foramen, 4- PNS, 5 – ANS.

Fig. 4.

Midpalatal Suture (MPS) Separation patterns: Pattern 1 – complete parallel separation, Pattern 2 - separation of the middle third of the MPS, Pattern 3 - separation of the anterior two thirds of the MPS, Pattern 4 – “half-Y″ separation involving one side of the transverse suture with anterior two thirds of the MPS, Pattern 5 - “Y”-shaped separation involving anterior two thirds of the MPS and both sides of the transverse suture.

Table 4.

Midpalatal suture (MPS) splitting patterns.

Classification	Midpalatal Suture (MPS) Status	Transverse Suture Status	Visual Characteristic
Type 1	Complete separation from the Anterior Nasal Spine (ANS) to the Posterior Nasal Spine (PNS).	No involvement (Remains intact/fused).	Two continuous, fully separated parallel lines.
Type 2	Partial separation: Middle third is split; the anterior and posterior thirds remain fused.	No involvement.	Isolated central separation window.
Type 3	Partial separation: Two anterior thirds are split; the posterior third remains fused.	No involvement.	Front-to-back fading separation.
Type 4	Partial separation: Two anterior thirds are split; the posterior third remains fused.	Asymmetric involvement: One side of the transverse suture splits; the other remains fused.	L-shaped or asymmetric branch pattern.
Type 5	Partial separation: Two anterior thirds are split; the posterior third remains fused.	Symmetric involvement: Both sides of the transverse suture split completely.	Reverse Y-shape pattern.

Study size. A formal sample size justification was evaluated based on the events per variable (EPV) criterion for multivariable logistic regression models to ensure stable parameter estimation. With a total sample size of N = 108 consisting of 70 controls (full separation) and 38 cases (partial separation), the limiting sample size is dictated by the smaller outcome group (n = 38). For the four independent predictor variables included in the full model (Male, Age, PNSPPW, and Frontozygomasep), this yields an EPV ratio of 9.5 (38/4). This ratio satisfies the established methodological standard of 5 to 10 EPV required to mitigate severe coefficients estimation bias, ensuring that the maximum likelihood estimations, Wald z-statistics, and corresponding odds ratios remain stable and robust.

Bias. To minimize misclassification bias, the outcome variable was simplified into a definitive binary classification distinguishing exclusively between full separation (Type 1) and partial separation (all remaining types). Potential information bias was further controlled by utilizing a single, trained observer for all landmark-based metrics, eliminating inter-observer variability, though an inherent risk of intra-observer measurement drift remains. Methodological limits regarding selection bias are present due to the fixed case-control sample (N = 108), which may restrict the generalizability of the findings to a broader population.

Statistical methods. Descriptive statistics were calculated to characterize the sample layout of 108 participants, comprising 70 controls with full separation and 38 cases with partial separation. All anatomical and demographic data were collected by the same observer to maintain measurement consistency across the sample. Multivariable logistic regression was then employed to evaluate the independent predictive value of the continuous anatomical biomarker, Frontozygomasep, while adjusting for demographic covariates (Male and Age) and the metric PNS-PPW. The overall fit of the model was assessed using the likelihood ratio chi-square test and Pseudo R² values, with individual variable significance determined using odds ratios (OR) and 95% confidence intervals. Finally, the practical diagnostic utility of the biomarker was evaluated by comparing the Area Under the ROC Curve (AUC) of the full model against a baseline demographic-only model using a post-estimation chi-squared contrast.

To evaluate sex-specific age effects across the complete cohort ($n=108$), a secondary full-factorial logistic regression model was specified incorporating an interaction term between biological sex and categorized age groups ($\text{Male}\times \text{Age}\text{Category}$). This model evaluated whether the odds of full separation across age brackets ($<30$, $30-39$, and $40-58$ years) differed significantly between male and female subjects.

A logistic regression model was run to all participants (n = 108) to estimate odds ratios (ORs) and 95% confidence intervals (CIs) for the likelihood of experiencing Full Separation. Statistical significance was set a priori at alpha = 0.05.

3. Results

Participants. Participants were allocated in a 1.84:1 ratio (controls to cases) to optimize statistical efficiency and enhance the power of the study to detect significant differences between the two expansion patterns. Statistical analysis was performed using multivariable logistic regression to adjust for primary demographic confounders by including four predictive variables within the model. Given the study's sample size (N = 108), the inclusion of this number of predictors introduces a potential risk of model overfitting, which was monitored via post-estimation diagnostics. To evaluate whether this parameter density compromised classification generalizability, the discriminatory power of the full model was strictly validated against a baseline demographic model using Area Under the ROC Curve (AUC) comparisons. A non-significant change in the resulting AUC metrics was established a priori as the threshold indicating that individual predictor significance should be interpreted with caution due to sample constraints.

Baseline demographic, craniometric, and expansion-related parameters are summarized in Table 1. Pre-intervention baseline characteristics showed no statistically significant differences between cases ($n=38$) and controls ($n=70$) with respect to age ($37.32\pm 8.46$ vs. $35.49\pm 8.58$ years, $p=0.29$), sagittal skeletal relationship ($\text{ANB}$: $3.52\pm {2.82}^{\circ }$ vs. $3.58\pm {2.43}^{\circ }$, $p=0.91$), or vertical maxillary divergence ($\text{SN}‐\text{ANS}‐\text{PNS}$: $7.64\pm {3.94}^{\circ }$ vs. $8.72\pm {3.62}^{\circ }$, $p=0.16$). Furthermore, longitudinal comparisons demonstrated comparable pre- and post-expansion measurements across both groups for interpterygoid width (cases: $28.25\pm 2.55\text{mm}$ to $30.24\pm 2.80\text{mm}$; controls: $28.14\pm 2.25\text{mm}$ to $29.96\pm 2.77\text{mm}$; $p=0.63$), posterior nasal base width (cases: $28.78\pm 3.23\text{mm}$ to $31.23\pm 3.34\text{mm}$; controls: $28.13\pm 2.12\text{mm}$ to $31.06\pm 2.86\text{mm}$; $p=0.78$), and upper airway cross-sectional area at ANS-PNS (cases: $391.76\pm 140.96{\text{mm}}^2$ to $431.21\pm 128.07{\text{mm}}^2$; controls: $392.54\pm 143.31{\text{mm}}^2$ to $433.46\pm 136.25{\text{mm}}^2$; $p=0.93$).

Table 1.

Baseline demographic and craniometric characteristics of the study population.

Parameter	Patient group		p-value
	Cases (n = 38)	Controls (n = 70)
Demographic characteristics
Age, years (n = 108)	$36.13\pm 8.54$
Age, years	37.32 ± 8.46	35.49 ± 8.58	0.29
Male gender,%	49%	46%	0.67
Craniometric characteristics
ANB before intervention, degrees	3.52 ± 2.82	3.58 ± 2.43	0.91
SNA before intervention, degrees	80.92 ± 3.76	80.56 ± 4.04	0.66
Amount of Anterior MPS separation, mm	5.67 ± 2.28	5.58 ± 1.46	0.77
Difference in Maxillary plane rotation, degrees	0.29 ± 3.91	1.67 ± 2.88	0.85
Difference in PNSPPW, mm	0.14 ± 2.93	1.03 ± 2.55	0.10
SN-ANS-PNS before intervention, degrees	7.64 ± 3.94	8.72 ± 3.62	0.16
Interpterygoid width, mm
Before intervention	28.25 ± 2.55	28.14 ± 2.25	0.16
After intervention	30.24 ± 2.80	29.96 ± 2.77	0.63
Posterior nasal base width, mm
Before Intervention	28.78 ± 3.23	28.13 ± 2.12	0.21
After Intervention	31.23 ± 3.34	31.06 ± 2.86	0.78
Upper airway cross-section at the level of ANS-PNS, mm2
Before intervention	391.76 ± 140.96	392.54 ± 143.31	0.98
After intervention	431.21 ± 128.07	433.46 ± 136.25	0.93

3.1. Main results

To evaluate whether sagittal skeletal classification ($\text{ANB}$) and vertical maxillary divergence ($\text{SN}‐\text{ANS}‐\text{PNS}$) independently influenced outcomes, both continuous cephalometric parameters were incorporated into the multivariable logistic regression model alongside biological sex, continuous age, $\text{PNS}‐\text{PPW}$, and $\text{Frontozygomasep}$ ($N=108$; Model Likelihood Ratio ${\chi }^2\left(6\right)=26.61$, $p=0.0002$, Pseudo $R^2=0.1900$). Adjusting for all covariates, neither the sagittal skeletal pattern ($\text{ANB}$: $\text{OR}=1.02$, $95\%\text{CI}:\left[0.85,1.22\right]$, $p=0.850$) nor the vertical growth pattern ($\text{SN}‐\text{ANS}‐\text{PNS}$: $\text{OR}=1.04$, $95\%\text{CI}:\left[0.92,1.18\right]$, $p=0.519$) demonstrated a statistically significant association with the likelihood of complete separation. In contrast, male sex ($\text{OR}=0.14$, $95\%\text{CI}:\left[0.05,0.39\right]$, $p<0.001$), advancing chronological age ($\text{OR}=0.93$, $95\%\text{CI}:\left[0.87,0.98\right]$, $p=0.014$), and the $\text{Frontozygomasep}$ biomarker ($\text{OR}=2.81$, $95\%\text{CI}:\left[1.10,7.19\right]$, $p=0.031$) remained significant independent predictors of full separation status.

To further evaluate the clinical/diagnostic utility of the Frontozygomatic separation biomarker, the full model's classification performance was compared against a baseline demographic-only model (consisting of Male, Age, and PNS-PPW; AUC = 0.755, 95% CI [0.656, 0.854]). While the addition of Frontozygomatic separation increased the absolute classification accuracy by 2.7% (moving the AUC from 0.755 to 0.782), a post-estimation chi-squared comparison revealed that this incremental increase in discriminatory power was not statistically significant (chi² = 1.21, P = 0.271). (Fig. 5)

Fig. 5.

Receiver Operating Characteristic (ROC) curves comparing model classification performance. The blue curve represents the baseline demographic model (Model 1: Male, Age, and PNSPPWB) with an Area Under the Curve (AUC) of 0.7553. The red curve represents the full model (Model 2) following the inclusion of the anatomical biomarker Frontozygomasep01, yielding a marginal absolute increase in AUC to 0.7823. The diagonal black line denotes the reference line for chance classification (AUC = 0.50). The frequent intersection and close spatial proximity of the two curves visually reflect the post-estimation analysis, confirming that the addition of the biomarker did not yield a statistically significant improvement in overall diagnostic discrimination (P = 0.271).

Secondary analysis. To evaluate whether the impact of age differed across biological sexes, a secondary full-factorial logistic regression model incorporating an interaction term ($\text{Male}\times \text{Age}\text{Category}$) was performed across the entire cohort (n = 108; Model Likelihood Ratio ${\chi }^2\left(5\right)=18.70$, $p=0.0022$, Pseudo $R^2=0.1335$) (Table 2). Within the female cohort, advancing chronological age showed no statistically significant association with the likelihood of full separation; relative to females under 30 years old, female participants aged 30–39 years ($\text{OR}=0.71$, $95\%\text{CI}:\left[0.07,7.16\right]$, $p=0.775$) and those aged 40–58 years ($\text{OR}=0.67$, $95\%\text{CI}:\left[0.06,6.87\right]$, $p=0.733$) demonstrated stable baseline odds of full separation. In contrast, male subjects demonstrated a pronounced, age-dependent decline in separation likelihood. Compared to younger males ($<30$ years), the odds of full separation decreased moderately in males aged 30–39 years ($\text{OR}=0.64$) and fell sharply by 87% in older males aged 40–58 years ($\text{OR}=0.13$). Although individual interaction coefficients did not achieve independent statistical significance due to subgroup sample constraints ($\text{Male}\times \text{Age}40–58:p=0.270$), the composite model demonstrates that age-related biological constraints on full separation operate predominantly within male subjects. (Table 3).

Table 2.

Odds ratios from multivariable logistic regression for factors associated with Full midpalatal separation.

Predictor Variable	Odds Ratio (OR)	Standard Error	z-statistic	p-value
Male (1 = Male, 0 = Female)	0.134	0.071	−3.81	<0.001*
Age, years	0.927	0.028	−2.50	0.013*
SN-ANS-PNS (before intervention), degrees	1.041	0.062	0.65	0.519
ANB (before intervention), degrees	1.017	0.092	0.19	0.85
PNSPPW, mm	0.884	0.060	−1.80	0.071
Frontozygomaticsep (1 = yes, 0 = no)	2.857	1.366	2.20	0.028*
Constant	928.95	2094.44	3.03	0.02

Note: * Indicates statistically significant values (P < 0.05), Model fit: n = 108, p = 0.0022*, Pseudo R² = 0.1900.

Table 3.

Logistic regression predictors of full separation among all participants (n = 108).

Sex	Age group, years	OR (Odds Ratio)	Interpretation
Female	<30 (n = 6)	1.00 (Ref.)	Baseline
Female	30-39 (n = 32)	0.71	Minor, non-significant change ($p=0.775$)
Female	40-58 (n = 26)	0.67	Minor, non-significant change ($p=0.733$)
Male	<30 (n = 18)	1.00 (Ref.)	Baseline within males
Male	30-39 (n = 14)	0.64	Moderate decline in separation odds (p = 0.53)
Male	40-58 (n = 12)	0.13	87% reduction in separation odds (p = 0.024*)

Note: * indicates statistical significance. Model Overall Fit: Likelihood-Ratio chi² = 18.70; p = 0.0022; Pseudo R² = 0.14.

4. Discussion

Key results. The primary objective of this study was to identify the biological and structural factors that dictate the success of achieving complete midpalatal suture separation during 3D-guided piezocorticotomy-assisted MARPE in adults. While the surgical technique has recently been enhanced by extending the surgical excision posteriorly toward the posterior nasal spine (PNS) to facilitate complete separation, our regression model demonstrates that success is still heavily governed by individual baseline anatomy and demographics. Specifically, the binary craniometric biomarker Frontozygomatic separation emerged as a significant positive predictor of full separation (OR = 2.86, P = 0.028), indicating that the structural morphology and resistance of the peripheral zygomatic buttress zone play a vital role in how expansion forces are distributed across the midface.

Conversely, advancing chronological Age (OR = 0.927) and Male gender (OR = 0.134) acted as rigid biological constraints that significantly diminished the probability of complete separation. The distinct vulnerability of male patients to incomplete separation (an 87% reduction in odds) is likely a function of the increased cortical bone thickness, higher mineral density, and complex sutural interdigitation characteristic of male craniofacial architecture.

From a surgical and mechanical standpoint, these demographic and biomarker constraints manifest directly in the posterior third of the hard palate. Despite meticulous digital planning and depth measurements derived from preoperative CBCT scans, the physical execution of deep cortical cuts is frequently obstructed by individual structural variations of the nasal septum. A key anatomical explanation for this phenomenon is the dense attachment of the vomer and its tight structural relationship to the horizontal plates of the palatine bones, which can generate excessive mechanical resistance against a through-and-through midline division.

When encountering this deep posterior resistance, the expansion forces generated by the MARPE appliance fail to divide the midpalatal suture cleanly between the paired palatine bones. Instead, the stresses are redirected laterally along paths of lesser structural resistance, forcing the palate into the alternative separation scenarios described in our classification as Types 4 and 5. In these patterns, separation is restricted to the anterior two-thirds of the midpalatal suture and terminates by naturally separating into either a unilateral transverse palatine suture separation (Type 4) or a bilateral transverse palatine suture separation (Type 5).

Our secondary interaction analysis provides crucial nuance to the relationship between demographic factors and suture separation. While initial subgroup observations suggested a general age-related decline, expanding the age-categorized model across the entire cohort revealed that this biological constraint is predominantly sex-specific. Advancing age significantly diminishes the probability of complete separation in male subjects—with older males (ages 40–58) experiencing an 87% drop in odds compared to their younger counterparts—whereas female subjects maintain relatively stable odds of separation regardless of age group ($p>0.70$). This stark divergence suggests that age-related structural or sutural changes, such as increased interdigitation or midpalatal suture obliteration, may progress at a markedly accelerated rate in males. Consequently, while male sex and advancing age jointly represent key biological constraints, clinical risk stratification must account for their interaction: age serves as a primary limiting factor specifically within male patient populations rather than as a uniform barrier across all subjects.

Angelieri⁵ classification was earlier suggested to define stages of MPS maturation and the likely timing for each stage of maturation. Stages D and E were shown to take place later in life, following the pubertal growth spurt. These findings suggest later maturation of the anterior maxilla compared to the posterior third of the MPS. While the morphological determinants of the maturation staging derived from the CBCTs demonstrate certain patterns of interdigitation with aging, the separation patterns in our study suggest facilitated separation of the anterior thirds of the MPS. The plausible explanation could be, as mentioned earlier, the underlying anatomical relationships between the maxillary crest, vomer, and pterygoid processes of the sphenoid bone.

A recent study of Surgically-Assisted Rapid Palatal Expansion (SARPE)-assisted MPS separation patterns suggested three main classes of MPS separation: Complete (from ANS to PNS), Partial (partial separation of the MPS not involving the posterior part), and Partial with no PNS involvement.¹¹ These findings are in agreement with the classification of the current study and are further enhanced with additional details on the palatine bone region separation with defined patterns of the anterior two-thirds separation. These may deviate from the SARPE-related patterns due to differences in surgical intervention and techniques.

Walter¹² and co-authors recently reported 11 cases with unwanted peri-maxillary fractures in patients treated with MARPE and MASPE activation protocols. One of the cases described included a paramedian fracture of the palatine bone. The cases described were treated with several appliance designs; the MARPE protocol was executed at 2–3 turns a day with varying screw-size designs. While this publication provides a great resource on the scope of potential complications, the MARPE protocols described are no longer in use.

Korbmacher¹³ examined human MPS specimens and concluded that different age groups exhibit dissimilar bone density patterns. Specifically, the middle age group demonstrated higher bone density compared to younger and older age groups. This study might give additional insight into age-related MPS separation patterns.

Thi Hong Thuy¹⁴ and co-authors studied bone regeneration patterns following MARPE and concluded that full MPS regeneration occurred within 12 months. The authors also noted that smaller amounts of MPS separation are associated with faster periods of MPS regeneration. These findings may suggest that, based on the suture separation pattern, the surface area of the hard palate increases across multiple planes with various patterns of MPS separation.

Limitations. Several limitations of the present study warrant consideration, beginning with its retrospective case-control design and modest sample size (N = 108), which may introduce selection bias and limit the generalizability of these findings. Furthermore, while the four-variable regression model satisfies standard methodological criteria, the parameter density carries an inherent risk of model overfitting, explaining why the individual significance of the biomarker did not translate into a statistically significant improvement in overall AUC. Finally, the analysis could not account for residual confounding from unmeasured variables necessitating future prospective validation in larger, multi-center cohorts.

Interpretation. This study demonstrates that successful full midpalatal suture separation in adults undergoing 3D-Guided piezocorticotomy-assisted MARPE is primarily dictated by baseline patient demographics and peripheral craniometric architecture rather than the surgical intervention alone. Specifically, younger age, female gender, and favorable zygomatic buttress morphology significantly increase the odds of complete separation. When these favorable factors are absent, surgical forces are structurally obstructed by the rigid posterior attachment of the vomer, causing the expansion vectors to fail at the palatine bones and redirect laterally into partial transverse skeletal failure patterns (Types 4 and 5).

Generalizability. Due to the retrospective, single-center design and modest sample size (N = 108), the generalizability of these findings to the broader adult orthodontic population remains limited. The predictive performance of the model is inherently optimized for this specific cohort's baseline morphology and may vary significantly across diverse ethnic ancestries, varying vertical facial biotypes, or differing surgical protocols. Consequently, these structural and demographic trends require prospective validation in larger, multi-center heterogeneous populations before they can be widely adopted as universal diagnostic frameworks.

5. Conclusions

• 1.Pre-intervention sagittal ($\text{ANB}$) and vertical ($\text{SN}‐\text{ANS}‐\text{PNS}$) skeletal patterns did not independently influence the likelihood of achieving complete midpalatal suture separation.
• 2.Male sex and advancing chronological age were significant independent negative predictors of full separation. Conversely, the presence of the $\text{Frontozygomatic}$ separation biomarker significantly increased the odds of separation success.
• 3.While incorporating the $\text{Frontozygomatic}$ biomarker improved absolute classification accuracy by $2.7\%$, this incremental gain in discriminatory power over a baseline demographic model was not statistically significant.
• 4.Age-related biological constraints on complete separation operate predominantly within male subjects, who exhibited up to an $87\%$ reduction in success odds with advancing age. Female subjects, by contrast, maintained stable odds of separation across all age brackets.
• 5.Midpalatal suture separation exhibited five distinct morphological patterns ranging from complete parallel split (Type 1) to localized middle (Type 2) and anterior-restricted openings with or without unilateral/bilateral transverse palatine suture involvement (Types 3–5).

Ethics approval and consent to participate

The study protocol was reviewed and approved by the Solutions Institutional Review Board (Solutions IRB) (Approval No. 0846/01-13/2026). Written informed consent was obtained from all participants.

Consent for publication

Written informed consent for the publication of anonymized data was obtained from all participants prior to enrollment.

Clinical trial number

Not applicable.

Compliance with ethical standards

Statement of Human Rights: All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki Declaration and its later amendments or comparable ethical standards. Informed consent was obtained from all individual participants included in the study. Ethical approval was received from Solutions Institutional Review Board (Solutions IRB) with the Ref # 0846/01-13-2026. Written informed consent for the publication of anonymized data was obtained from all participants prior to enrollment.

Statement of Animal Rights: This article does not contain any studies with animal subjects performed by any of the authors.

Funding

Not applicable.

Declaration of competing interests

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Abbreviations

The following abbreviations are used in this manuscript:

MARPE

Miniscrew-Assisted Rapid Palatal Expansion

SARPE

Surgically Assisted Rapid Palatal Expansion

RPE

Rapid Palatal Expander

3D-GMP

Three-Dimensional-Guided Midpalatal Piezocorticotomy