INTRODUCTION

Previous studies have aimed to elucidate the possible relationship between upper airway (UA) morphology and growth of the craniofacial complex because of the important role that respiratory function has in orthodontic diagnosis and treatment planning.^1–3 As early as 1907, Angle stated that Class II, division 1 malocclusions were characterized by narrow upper pharyngeal airways and mouth breathing with dentofacial alterations such as narrow upper arch, upper incisor protrusion, and abnormal lip function.⁴ According to Moss⁵ functional matrix theory, a deviation in the respiratory or masticatory pattern could disturb normal craniofacial development of an individual, assuming soft tissue might affect hard tissue morphology. However, there has been controversy regarding this subject. Some authors could not establish an association between narrow airways and dentofacial morphology,^3,6 whereas others found evidence to support that UA influences growth and development of the craniofacial complex.¹

Among all craniofacial disharmonies, skeletal Class II is the most frequently associated condition with narrowing of the UA.^7–9 There seems to be a consensus that maxillary protrusion increases UA length,^10,11 and mandibular retrusion is related to its constriction.^7,8,12 Vertical craniofacial disharmonies are also related to UA morphology and respiratory function.^13,14

It is hard to establish a cause–effect relationship between UA morphology and a specific facial pattern because of their innate multifactorial origin.¹⁵ Previous studies analyzed this subject on lateral cephalometric radiographs that, despite providing relevant data, presented limitations as a result of anatomical variations in the three-dimensional (3D) complex that constitutes the upper airway, such as highly constricted transverse areas.^16,17 With the evolution of 3D diagnosis, this study aimed to evaluate, on cone-beam computed tomography (CBCT) scans and image reconstructions, the volume and morphology of nasopharynx (NP), oropharynx (OP), and hypopharynx (HP) in different sagittal and vertical skeletal patterns of Class II malocclusions when compared with Class I patients.

MATERIALS AND METHODS

This study was approved by the research ethics committee of the Institute of Collective Health Studies of the Universidade Federal do Rio de Janeiro in Brazil. All patients signed a consent form prior to orthodontic treatment, allowing the use of their orthodontic records for this study. Sample size was calculated based on the average standard deviation of a previous study.⁸ According to Pandis,¹⁸ to detect differences of 2500 mm³ in oropharyngeal volume, with a test power of 0.80 (α = 0.05), at least 19 cone-beam computed tomography (CBCT) scans would be necessary in each group.

The sample was composed of 197 CBCT scans selected from a total of 1446 that comprised the initial records of patients starting treatment in the Orthodontic Clinics of the Postgraduate Program in the Universidade Federal do Rio de Janeiro in Brazil. The following inclusion criteria were used: age range between 13 and 23 for both genders, no previous orthodontic therapy, skeletal Class I or Class II malocclusions (0 ≤ ANB ≤ 4; ANB > 4, respectively), good state of general health, CBCT scans that included the fourth cervical vertebrae, Digital Imaging and Communications in Medicine (DICOM) files, and Craniocervical angle (CCA) ranging from 90° to 110° during CBCT acquisition. Exclusion criteria were restricted to the presence of syndromes, Class I bimaxillary protrusion cases, and scans with artifacts.

The CBCT scans were performed on one device (KODAK 9500 Cone Beam 3D System; Carestream Health, Rochester, N.Y.) according to a standard protocol (90 kV, 10 mA, Field of View (FOV) of 18.4 × 20.6 cm, 0.3 mm voxel size, and 24-second scan). During scanning, the patients were instructed to be in maximum intercuspation and to position the tongue against the palate and not to swallow. The Frankfurt horizontal plane (FHP) was kept parallel to the ground and perpendicular to the sagittal plane.

DICOM files were exported for 3D reconstruction using Dolphin Imaging software version 11.7 (Dolphin Imaging & Management Solutions, Chatsworth, Calif), and measurements were made by an experienced operator. The head was positioned virtually in space according to the axial, coronal, and sagittal reference planes.¹⁹ Linear and angular measurements were taken on the two-dimensional lateral cephalometric radiographs created from the CBCT scans in the software to characterize skeletal patterns (Figure 1).

View Figure

• Download as Powerpoint
• Download Figure

The patients were divided into one of the following three groups according to SNA and SNB angles: Class I (CI), Class II maxillary protrusion (CII MaxP), and Class II mandibular retrusion (CII MandR). Each of these groups was reallocated into one of the following three subgroups according to the vertical pattern of mandibular growth based on the FMA angle: hypodivergent (Hpo), normal, and hyperdivergent (Hpr).The sample was therefore divided into nine groups (Table 1).

Table 1 Sample Division According to the Sagittal Position of the Maxilla and Mandible to the Cranial Base (SNA and SNB Angles) and the Vertical Pattern of Facial Growth (FMA Angle)^a

View Fullscreen

Total volume and minimal axial area were determined for NP, OP, and HP. Total height values were obtained at the OP and HP. Anatomical limits for UA analysis were used as proposed by Guijarro-Martínez and Swennen.^19,20 The anterior limit was set by the intersection of the coronal plane perpendicular to FHP, passing through the Posterior Nasal Spine (PNS). The soft tissue contour of the posterior pharyngeal wall established the posterior limit, defined by the intersection of the coronal plane perpendicular to FHP, passing through point C2up (midpoint between the upper and most posterior points of the second cervical vertebrae). The soft tissue contour of the lateral pharyngeal walls established the lateral limits, defined by the sagittal plane intersecting perpendicular to FHP, touching the lateral walls of the maxillary sinuses.

At NP, the upper limit corresponded to the soft tissue contour of the upper pharyngeal wall, defined by a transverse plane parallel to the FHP passing through the clivus base, and its lower limit was set by a plane parallel to the FHP intersecting the PNS and extended to the posterior pharyngeal wall. At OP, the upper limit was established as the lower limit of NP, and the lower limit was defined by a plane parallel to the FHP intersecting point C3la (lower most anterior point of the third cervical vertebrae). At HP, the upper limit was set as the lower limit of OP, and the lower limit was established by a plane parallel to the FHP joining the base of epiglottis to point C4la (lower most anterior point of the fourth cervical vertebrae; Figure 2).

View Figure

• Download as Powerpoint
• Download Figure

RESULTS

Intraclass correlation coefficient results for intrarater reliability were above 0.93 for all analyzed variables, thus certifying operator calibration. Sample normality was confirmed separately for each variable, considering distinct groups. Descriptive statistics and intergroup comparison analysis for all variables are presented in Tables 2 and 3. All sample groups were compatible for participant age without significant differences. Statistical significant differences were found between groups regarding ANB, FMA, SNA, and SNB angles, supporting the inclusion criteria employed (Table 2).

Table 2 Descriptive Statistics and Intergroup Comparison Considering Age and Cephalometric Pattern of All Participants^a

View Fullscreen

Table 2 Extended

View Fullscreen

Table 3 Descriptive Statistics and Intergroup Comparison of Upper Airway Analysis for All Study Groups^a

View Fullscreen

Table 3 Extended

View Fullscreen

For sagittal skeletal base length, the maxilla (Co-A) displayed statistically significant lower average values in CI Hpr when compared with groups CII MaxP Hpr, CII MaxP normal, CII MaxP Hpo, and CII MandR Hpo. Variable Co-Gn, representative of mandibular length, was not statistically different among the groups. The CCA angle was significantly greater in CII MandR Hpr when compared with CI Hpr, CI Hpo, and CII MaxP Hpo (Table 2).

Regarding UA analysis, all NP, OP, and HP variables were equivalent in intergroup comparisons (Table 3).

The correlation test results are shown in Table 4. Age had a significant positive correlation with NP, OP, and HP volumes. The variables FMA, ANB, and SNA were not significantly correlated to NP, OP, and HP volumes. SNB angle was significantly correlated only with OP volume. Similar significant correlations were observed between CCA and OP and HP volumes. The variables related to NP, OP, and HP morphology (minimal axial area and total height) were significantly positively correlated with their respective volumes (Table 4).

Table 4 Pearson Correlation Coefficient for NP, OP, and HP Volume Compared to Age, Cephalometric Variables, and Upper Airway Morphological Variables^a

View Fullscreen

Linear regression graphs are shown in Figure 3 and depict positive correlations between NP, OP, and HP with maxillary and mandibular length. A significant difference was found in males regarding NP volume and Co-Gn as well as when OP and HP volumes were related to Co-A and Co-Gn in males and in total gender averages.

View Figure

• Download as Powerpoint
• Download Figure

The comparisons of average and standard error values of NP, OP, and HP volumes between males and females are shown in Figure 4. The males displayed significantly greater OP and HP volumes when compared with females.

View Figure

• Download as Powerpoint
• Download Figure

DISCUSSION

The age range established for this research made it possible to evaluate the relationship between growth and UA morphology in individuals with different dentofacial patterns. Previous retrospective 3D studies showed an increase in UA length and volume between the ages of 8 and 18^21,22 as well as increasing airway area and length until the age of 20.²³ Similar data were found in the present study, demonstrating positive and statistically significant correlations with an increase in NP, OP, and HP volumes between the ages of 13 and 23. This result is supported by the homogeneous distribution in all groups when it came to age. As age increases and soft tissue around the UA develops, there is clearly an actual increase in UA dimensions.

The majority of authors evaluated UA using linear measurements taken on lateral cephalometric radiographs.^1,3,7,13,14 As further research evolved, 3D studies on UA analysis were performed, but without considering it as a complete structure.^2,8 Considerable obstacles still need to be surpassed to make 3D UA analysis more precise, among which is the need for better anatomical boundary definition, which has been highly variable in the literature.^7,9,20 Due to its importance, the anatomical limits in the current study were used as recently published.¹⁹ By standardizing the anatomical limits used, more accurate comparisons among studies may be achieved.

Class I and Class II skeletal malocclusions were stratified on the sagittal norm in the present study, but did not show significant differences in UA morphology according to the skeletal patterns analyzed. The results confirm those of previous studies, all showing no significant association between UA dimensions and different sagittal skeletal patterns.^13,24 However, the literature reports that UA morphology can be influenced by different sagittal malocclusions, and most authors found smaller dimensions in OP and NP in skeletal Class II patterns.^2,7,9 An important difference between the current study and that by El and Palomo⁸ was that the CCA angle was used as an inclusion criterion for the sample, decreasing the effect that head posture has on the UA dimensions during a CBCT scan.^9,25 This became quite clear due to the positive significant correlations identified in this research between CCA and OP and HP volumes, which demonstrated that even if CCA was restricted to a standard range,²⁵ it was capable of significantly influencing UA dimensions.

The sagittal relationship between the maxilla and mandible is usually determined by ANB angle. However, it may be influenced by maxillary rotation, increased vertical distance between points A and B as well as between points N and B, length of the anterior cranial base, and sagittal position of point N.²⁶ In the present study, no significant correlations were found between ANB angle and NP, OP, or HP volumes, indicating that such influential factors should be carefully considered in future studies. Indriksone and Jakobsone³ did not detect significant correlations between ANB and NP or OP volume, which supports the current findings. Nevertheless, the positive significant correlation between SNB and OP volume found in the current study was consistent with previous studies.^3,7

The positive statistically significant correlations between maxillary and mandibular lengths and OP and HP volumes are consistent with other studies that identified a significant correlation between mandibular length and OP volume.^8,12 From these results, it is possible to infer the hypothesis that the actual length of the maxilla and mandible are more determinant of the volume and shape of the UA than is their sagittal position with regard to the cranial base, represented by the SNA and SNB angles. A meta-analysis reported a clinically significant association of mandibular length with an accurate diagnosis of obstructive sleep apnea.²⁷ Associated with these findings, none of the groups from the current study showed differences in Co-Gn, which could explain in part the low variation of OP and HP volumes of the evaluated groups. Future studies should present results less dependent on the SNA, SNB, and ANB angles.

Regarding the possible influence of vertical craniofacial patterns in UA, Celikoglu et al.²⁸ found significant differences in UA volume, observing higher values in Hpo and lower values in Hpr patients. The current data support the theory that this issue itself does not cause significant modifications in its morphology, consistent with other research.^2,3

The high positive correlation between NP, OP, and HP volumes and the respective minimal axial areas is well supported by previous studies.^7,8,24 This is a relevant clinical finding because volume itself cannot provide information about UA shape, but from the current results, it can be inferred that the most constricted area for air to pass through the UA seems to be directly influenced by volume and effective jaw length, as previously discussed.

Larger oropharyngeal and hypopharyngeal volumes were observed in the males in this study and not the females, consistent with previous studies.^2,8,21 Specifics related to the development of males and females within this age range may have contributed to these findings.

The present results highlighted that actual jaw length plays a more determining role on UA volume and morphology. It is a more important risk factor to be considered in an obstructive sleep apnea diagnosis than considering vertical and sagittal skeletal malocclusions separately. Therefore, additional research adopting similar anatomical limits and methodology for UA analysis are recommended to support these findings.

CONCLUSIONS

Regarding skeletal Class II malocclusion patterns, the following conclusions can be drawn:

• Different vertical and sagittal facial patterns showed no significant differences in volume and morphology of the NP, OP, and HP.

• Significant correlations were shown between jaw length and OP and HP volumes in males and in total gender averages.

• Males had larger OP and HP volumes when compared with females.

• Age was significantly correlated with NP, OP, and HP volumes in the sample studied.