Impact of different cephalometric skeletal configurations on anatomic midface parameters in adults

The maxillary complex, which comprises the premaxillary as well as the maxillary and palatine bones, is a landmark of the midface with utmost importance for orthodontic treatment planning [1]. A profound knowledge of midface anatomy is indispensable for any treatment devoted to the post-development of the upper jaw in the transversal, sagittal, and vertical plane [2‐4]. A transversal expansion of the upper jaw is necessary for severe maxillary constrictions or crossbites, with chronological age and skeletal features contributing to an increased likelihood of transversal resistance [5, 6]. Among others, mid-palatal suture maturation is responsible for the lack of success of transversal expansion, with an increased probability of unfavourable dental side effects, such as buccal tipping, recessions, and gingival ulcers [7‐9]. The abovementioned factors determine, whether a primarily tooth-borne rapid maxillary expansion device (RME) should be additionally or solely anchored cortically. A micro-implant-assisted rapid palatal expansion (MARPE) utilizes both, the hard palate and the dentition as an anchorage, while a bone-borne distractor directly transfers the orthodontic forces to the palatal bone [10, 11].

Apart from transversal expansion, multiple orthodontics conditions call for maximum skeletal anchorage, with the hard palate and the interradicular space of the alveolar crest being preferred insertion sites for orthodontic implants [12, 13]. Orthodontic implants have experienced an increasing popularity over the last decades, improving anchorage and expanding treatment options for adult patients [14, 15]. Consequently, knowledge of the respective anatomy is crucial for the accurate positioning and stability of orthodontic implants, avoiding premature implant loss. Due to its great importance for skeletal anchorage, palatal thickness has been investigated intensively in literature [16‐20]. Concerning the hard palate, the T-zone, which describes the area immediately posterior to the palatal rugae, has been validated as a reliable location with sufficient bone thickness [21]. Overall, palatal thickness has been reported to be most extensive in the anterior part of the palatal vault, comprising median and paramedian areas of the anterior hard palate, as opposed to posterior areas of the hard palate [12]. Nevertheless, a tremendous inter-individual inhomogeneity for palatal thickness has been depicted in the literature [20, 22]. Explanations alluded to in the literature are that palatal thickness is most likely influenced by a plethora of co-factors such as maxillary body length, sex, and age [17, 18, 20].

It has been described that bone thickness of the alveolar crest is equally influenced by age, sex, and craniofacial growth patterns [23]. Notably, vertical facial growth patterns have a more significant impact than the respective sagittal relation, with the highest alveolar bone thickness found in hypodivergent individuals [24, 25].

Besides bony features of the midface, soft-tissue landmarks such as the muscle thickness of the masseter are determined by the respective viscerocranial configurations, with the thickest muscles found in brachyfacial phenotypes [26]. The present study investigates the influence of vertical skull configurations (brachyfacial, mesofacial, dolichofacial) as well as the respective skeletal Class I, II, and III relationships on predefined hard and soft tissue parameters of the midface.

We retrospectively investigated 240 patients, who had received a diagnostic CT of the skull concerning soft and hard tissue parameters of the midface. Before enrolling in this study, the institutional review board had given their positive consent (IRB Number: 22–174-Br). In a period from May 2021 to May 2022, CT datasets from 240 patients of European descent (m = 110 (54%), f = 130 (46%)), 20–79 years of age (mean age: 42 ± 15), who were examined within the scope of cranial trauma, inflammatory disease or tumor staging, were retrospectively selected from our archives. We enclosed CT datasets with high resolutions down to 0.6 mm slice thickness and a 512 pixel matrix (SOMATOM X.cite, Siemens Healthcare GmbH, Forchheim, Germany), and ultra-high resolution using a dedicated filter at the detector side to obtain 0.4 mm slice thickness combined with a maximum of 1024 pixel matrix (SOMATOM X.ceed, Siemens Healthcare GmbH, Forchheim, Germany). Ultra-high-resolution datasets were chosen preferably. Patients were only enclosed if they had a maximum of one singular tooth missing per jaw. Subjects with facial skull asymmetries, neoplasms of the skull, craniofacial malformations such as orofacial clefts, displaced teeth in the palatal region, and metabolic disorders of the bones were also excluded.

All CT data sets were evaluated with a three-dimensional post-processing console (Syngo.via VB60_A, Siemens Healthcare GmbH, Erlangen, Germany). This software offers the possibility to align the datasets in a standardized way in all three spatial planes ensuring a standardized evaluation. Each CT was aligned in the axial, coronal, and sagittal slices according to predefined reference points and planes to ensure reproducibility. The midline of the face served as a reference in the coronal slice, and the spinal plane was utilized as a reference line for the axial and sagittal slices (Fig. 1).

If the axis alignment is adjusted in one of the planes, the alignment of the other two axes is also changed since they are always perpendicular to each other.

A total of 106 variables were measured on every single CT dataset, resulting in 25.440 measurements for the entire investigation collective of 240 patients without any missing values. The measurements, which were performed within the axial slice of the three-dimensional dataset, comprised the length of the hard palate (Fig. 2B, C), ranging from the anterior to the posterior spine of the hard palate, the length of the mid-palatal suture starting at 0.5 mm dorsal to the incisive foramen, the thickness of the maxillary sinus facial wall and the width of pterygomaxillary junction (Fig. 2B).

In the coronal slice, the palatal thickness of both, the maxillary and the palatine bones, was measured in the midline of the hard palate (mid-palatal sutural region) and 0.25 cm lateral of the suture on the right and left side of the hard palate (Fig. 2D). The palatal thickness of the maxillary was measured starting 0.5 mm dorsal of the incisive foramen, with five subsequent measurements being performed in the consecutive coronal CT slices. The measurements of the palatal thickness in the palatine bones were performed accordingly, being initiated 0.5 mm dorsal to the transverse palatine suture. All measurements of palatal thickness were strictly conducted at a 90° angle to the spinal plane (Fig. 2A, D). Furthermore, the thickness of the alveolar ridge in the molar and premolar region, both on the left and right side of the face, was measured in the coronal view (Fig. 2D). With regard to soft tissue parameters, the length and thickness of the masseter muscle were measured on the left and right sides of the face (Fig. 2B).

A cephalometric analysis was carried out using a sagittal thick slice MPR reconstruction as an alternative for a lateral radiograph, which was imported into Onxy Ceph (Image Instruments, Germany), where a cephalometric analysis was carried out [27, 28]. Based on the outcome of the lateral cephalometry, employing the cephalometric angles NL-NSL, ML-NL, ML-NSL, Me-tgo-Ar and the Jarabak index, every patient was assigned to either a brachyfacial, mesofacial, and dolichofacial skull configuration [28, 29]. Furthermore, based on SNA, SNB, ANB, and Wits-values every patient was allocated to skeletal Class I (neutral), Class II (distal) or Class III (mesial) [28‐30]. The cephalometric analysis of all 240 MRP reconstructions was performed by three raters with more than three years of expertise in maxillofacial radiology. Based on a preceding statistical power analysis, 50 selected CT datasets were analysed by the abovementioned three raters, calibrated beforehand, to determine the interrater reliability, ensuring the validity of the measurements. Based on the preceding interrater reliability, the remaining 190 CT datasets were evaluated by one rater.

As bony and soft tissue landmarks are directly targeted during orthodontic treatment, a profound knowledge of midface anatomy is essential for orthodontists. In the maxilla, the T-zone has been established to safely place temporary anchorage devices (TADs), while the interradicular regions of both, the maxillary and mandibular buccal alveolar bones have proven to be equally suitable [21, 33, 34]. Since the correct placement is essential for the primary stability of TADs, the respective insertion regions are intensively studied in the literature. Overall, the literature reports on significant inter-individual differences with regard to palatal anatomy. These differences might be attributed to the fact that most studies do not record the respective viscerocranial phenotypes when evaluating the hard palate’s thickness. Therefore, we elaborated on the influence of vertical skull relations and the respective skeletal Class I, II or III configurations in great detail. While the skeletal Class I, II or III had no marked impact on palatal thickness, significant differences were observed between the three vertical subgroups, with the most significant palatal thickness found in brachyfacial skull configurations. On average, the median palatal thickness was 1 mm thicker in brachyfacial patients than in mesofacial or dolicofacial ones. Using surgical insertion guides for TADs has proven to increase placement accuracy, reducing early TAD loss. Those insertion guides might be essential in patients with mesofacial or dolicofacial skull configurations as opposed to brachyfacial patients, who present with more bone support in the anterior T- zone [35]. Palatal thickness significantly decreased from ventral to dorsal and from median regions of the palate to paramedian ones. These results agree with studies investigating TAD sites of the palatal posterior supra-alveolar bone [12, 17, 36]. In our collective of adult patients, age did not have a pronounced effect on the thickness of the hard palate. In contrast, statistically significant differences between male and female patients were observed (p < 0.01). This finding is congruent with sources stating gender-specific morphological variations, while other authors negate this [12, 17, 22, 34, 37].

Concerning maxillary body length, Class III patients proved to have a significantly shorter palate (p < 0.01) than Class I or Class II patients. With an average length of 5.313 cm, maxillary body length was reduced in Class III as opposed to Class I (5.406 cm) and Class II (5.404 cm) patients. Those results align with previous studies, which report a shorter maxillary body length in mesial configurations [1, 18, 38]. In our collective, the length of the mid-palatal suture is influenced by both, vertical skull configuration as well as skeletal Class I, II and III. Regarding vertical relations, brachyfacial patients had the longest suture, and dolicofacial patients presented with significantly shorter ones.

With reference to skeletal Class, and comparable to palatal length, the shortest mid-palatal sutures were found in Class III patients. A shorter mid-palatal suture in Class III patients might indicate less resistance to transversal expansion, compared to Class I and Class II patients. However, bone-anchored RME appliances in combination with maxillary protraction are beneficial in Class III patients for sagittal maxillary development [10, 39, 40]. The positive effect of bone-anchored RME in combination with maxillary protraction might be more critical in terms of stimulation of midface structures with distinct maxillary protraction protocols [39, 41].

Alveolar ridge thickness was also correlated with vertical skull relations, with brachyfacial patients presenting a significantly broader alveolar crest than those with a dolicofacial or mesofacial configuration, further underlining the influence of the viscerocranial phenotype on interindividual morphology. Hence, interradicular TADs might have a higher success rate in patients with brachyfacial skull configurations than those with dolicofacial or mesofacial ones. Concerning the measured soft tissue parameters, masseter thickness and length correlate with brachyfacial skull configurations, which goes along with the results of previously published data [26, 42].

The findings of this study underline that besides sex and age, vertical skull configurations and skeletal Class I, II and III, as routinely obtained on lateral cephalometric images, influence soft and hard tissue landmarks of the skull. Clinical implications that could be derived from these findings might be within the scope of TAD placement. In the present study, patients with a dolicofacial skull configuration and skeletal Class III present with a significantly shorter maxilla and mid-palatal suture, as well as reduced palatal height and alveolar ridge thickness, combined with masticatory muscles, which are reduced in length and thickness. These factors could necessitate shorter implants and lower orthodontic forces to achieve clinical treatment success. Furthermore, in knowledge of the respective morphology, conventional and cortically anchored orthodontic appliances could be tailored to individual needs.

Nevertheless, this study also presents some limitations primarily associated with the retrospective study design. While selecting the patients from the local Department of Radiology archives leads to a large sample size and sufficient power of the presented measurements, no records of an initial orthodontic treatment could be obtained. The distribution of the patient collective into the cephalometric subgroups was nonetheless conducted and should therefore be interpreted as a proof of principle design. Further prospective studies with orthodontic pre-treatment records are needed to validate these findings.