Suggestion of a new standard in measuring the mandible via MRI and an overview of reference values in young women

Adult idiopathic condylar resorption (AICR) is a temporomandibular joint (TMJ) disease also known as cheerleaders’ syndrome or progressive condylar resorption. It mostly affects young women aged 15–35 years [1, 2], and other surveys propose a range of 11–45 years [3]. The term cheerleader’s syndrome refers to the perception that it mostly affects thin young women taking part in sports [4]. Based on a survey by Handelman and Greene, it is estimated that orthodontists see 1 case of idiopathic condylar resorption in 5000 patients [5].

AICR is often associated with a retruded lower jaw, a class II occlusational relationship, and an anterior open bite [4], while it remains unclear if these changes are caused by changes of the condyle or if they are to be perceived as one of many risk factors. Patients might also report progressive changes of their facial features that are accompanied by pain or cracking of the TMJ [4], headaches, malocclusion, or a decreased range of motion of the jaw [1]. According to Wolford and Cardenas, about 25% of patients with AICR experience no symptoms of TMJ disease at all [6].

Some authors use the term AICR in a way that is not clearly defined and include osteoarthritic changes of the TMJ into their definition [7, 8]. Contrary to that, some authors particularly exclude cases with osteoarthritis, previous craniomaxillofacial surgery, or trauma of the TMJ, to define AICR as truly idiopathic [4, 9, 10]. While some authors want to exclude patients that have undergone orthognathic surgery from the diagnosis, others propose that orthognathic surgery might be a triggering event in the development of AICR and therefore might be part of a pathomechanism leading to the disease. One might also argue that currently AICR is likely under-diagnosed and patients have to undergo a long journey to receive their diagnosis, on which journey they may have received orthognathic surgery anyway, so excluding subjects that have had orthognathic surgery could distort scientific results. On top of that, Wolford supposes that the relationship between orthognathic surgery and AICR might only be coincidental but not causally determined in any direction [4], which further highlights the complexity of the topic and how little we know to this date. It also emphasizes inconsistencies in the definition of AICR, which complicate scientific research and clinical diagnosis, and which are based on gaps in our knowledge regarding AICR. Sansare et al. mention in their systematic review that in research, cases of AICR should be well defined and not mixed up with other diseases [11]. However, this is problematic since the diagnostic process is further complicated as a result of a lack of known laboratory tests that are specific for AICR [4].

As Sansare et al. proposed in 2015, further investigation on the presentation of the TMJ in 3D imaging procedures is needed [11]. It is particularly important to note that the progression of the disease, estimated to be 1.5 mm/year, might be underestimated by only looking at 2D images of the mandible [11]. Therefore, 3D imaging, e.g., via MRI or CT scan, is an important improvement to assess the development and progression of AICR, and condylar resorption in general.

As the variety of radiographic findings in patients with AICR is substantial in the literature and includes both bones and soft tissues [11], we consider MRI to have great potential to assess the disease, as both bones and soft tissues can be observed, and the 3D morphology of the condyle can be visualized.

We used datasets of the LIFE-Adult-Study [12], a cohort study initiated by the University of Leipzig, which include MR images of the head suitable for the segmentation of 3D models, and a wide variety of background information to assess possible correlations to the morphology of the mandible.

The primary aim of this study was to establish reference values for measurements in segmented models of an MRI of the mandible in young women aged 15–40, i.e., the population mostly affected by AICR [1, 2], in order to gain a base line for further research, and to establish MRI measurements as an efficient tool for mandible assessment. As the definition of AICR is inconsistent throughout research and the pathogenesis of AICR is only poorly understood, we chose to include data of a large healthy cohort and put a special focus during evaluation on participants whose mandibles show measurements consistent with AICR (i.e., a low condylar volume or a small SNB angle). We assessed SNB angle, condylar volume, the volume of the processus muscularis, ramus length, depth of the antegonial notch, and length and width of the condyle. To the authors’ knowledge, there are no studies available yet that evaluate these measurements of the mandible as presented in MRI.

The second aim was to compare MRI measurements to other parameters surveyed by LIFE, such as blood count, vitamin-D-levels, dietary habits, or anthropometric data, to evaluate if there are any connections to be made between the morphology of the mandible and another parameter, thus to gain new hints on the origins of condylar resorption, and as a baseline for laboratory tests that might be interesting candidates to help the diagnostic process.

Changes in SNB angle and condylar volume as well as changes in height of the mandibular ramus are parameters that are associated with condylar resorption [11]. Therefore, we included these measurements into our study. As the length and width of the condyle (2D parameters in the axial view) are linked to the condylar volume and can be assessed without segmenting a 3D model of the mandible, and might therefore be helpful in a clinical routine setting, we included these parameters as well.

Of 158 participants, 33 had a class II SNB angle of ≤ 78°, 59 had a class I SNB angle of 78–82°, and 67 had a class III SNB angle of ≥ 82°. The minimal SNB was 59.9°, the maximum 89.1°.

We defined a normal condylar volume as between the first and third quartile; according to this, a “small” condylar volume was below the first quartile (1311.25 mm³ on the right, 1387.47 mm³ on the left), and a “large” volume above the third quartile (1855.37 mm³ on the right, 1873.65 mm³ on the left). The minimal volume measured on the right side was at 656.99 mm³, on the left side at 769.48 mm³. The maximal volume measured was 3234.58 mm³ on the right side and 2957.35 mm³ on the left. Our measurements of the condylar volume are shown in Table 5.

Condylar volumes of both sides (verage right condylar volume: 1584.21mm³, SD 415.77, average left condylar volume: 1641.67 mm³, SD 389.30) and the SNB angle (average 81.1° SD 4.05) were nearly normally distributed.

Condylar volume on the left and right side showed a correlation of 0.77 (Fig. 2), which is surprisingly low given the bilateral symmetry of the mandible.

The average length of the right condyle was 9.08 mm (median 8.85 mm; SD 1.54 mm) with a minimum of 5.7 mm and a maximum of 19.5 mm. On the left condyle, the average length was 9.11 mm (median 8.9 mm; SD 1.30 mm) with a minimum of 5.9 mm and a maximum of 13.5 mm. For both sides, the average length was 9.09 mm with a median of 8.9 mm and a standard deviation of 1.42 mm.

The average width of the right condyle was 17.74 mm (median 17.85 mm; SD 2.16 mm) with a minimum of 12.2 mm and a maximum of 22.7 mm. On the left, the average width of the condyle was 18.03 mm (median 18.1 mm; SD 1.99 mm) with a minimum of 12.4 mm and a maximum of 23.1 mm. The average for both sides combined was 17.88 mm with a median of 18 mm and a standard deviation of 2.08 mm.

We assessed an average mandibular ramus length on the right side of 53.54 mm (median 53.5 mm; SD 4.19 mm) with a minimum of 38.1 mm and a maximum of 63.7 mm. On the left, the mean ramus length was 54.22 mm (median 54.2 mm; SD 4.17 mm) with a minimum of 42.1 mm and a maximum of 65.8 mm. On both sides taken together, the mean ramus length was 53.88 mm with a median of 53.85 mm and a standard deviation of 4.19 mm.

To our knowledge, there are no referential values for antegonial notching and the volume of the muscular processus, neither for CT imaging nor for MRI. We therefore attempted to extract these from the dataset. We measured a mean volume of the muscular processus on the right side of 210.21 mm³ (median 179.57 mm³; SD 139.34 mm³) with a minimum of 10.73 mm³ and a maximum of 645.68 mm³. On the left, the average volume was 212.44 mm³ (median 188.47 mm³; SD 138.70 mm³) with a minimum of 6.27 mm³ and a maximum of 949.46 mm³. For both sides, the average volume of the muscular processus was 211.32 mm³ (median 188.47 mm³; SD 138.80 mm³). However, the segmentation of the muscular processus was limited by the presentation in MRI, as the structure of the processus is very thin and likely not to be captured in the slices properly. On the right side 3 and on the left side 4 processi musculari could not be segmented. Also, the manual correction of the outline during segmentation was hampered as the differentiation between bone and the soft tissue around it was vague. Therefore, we expect our measurements of the muscular processus to be more of an estimate.

The average depth of the right antegonial notch was 2.02 mm (median 2.15 mm; SD 0.99 mm) with a minimum of 0 mm and a maximum of 5.7 mm. On the left, the average was 1.68 mm (median 1.8; SD 0.98 mm) with a minimum of 0 mm and a maximum of 6 mm. The average of both sides taken together was 1.85 mm with a median of 2 mm and a standard deviation of 0.99 mm.

To the authors’ knowledge, none of these parameters has been assessed in MR imaging of the healthy jaw to this date. Ramus length, condylar length and width, and condylar volume have been assessed previously, but only in CT scans [14‐17], so the measurements given above represent new referential values for measurements of the mandible in MRI. For a summary and overview of all resulting measurements, see Table 6.

We next analyzed if SNB angle and condylar volume are correlated, as both small SNB angles as well as low condylar volume are used to assess the presence of AICR. Of note, there was no significant correlation between SNB angle and condylar volumes. The Pearson correlation between the left condylar volume and SNB was 0.075, while the correlation between the right condylar volume and SNB was 0.153. For comparison, the Pearson correlation between the condylar volumes of both sides of the mandible was 0.78 (Fig. 2).

To take into account other general physiological parameters that might influence jaw physiognomy, we assessed also correlations between SNB angle or condylar volume and height or weight of the participants as presented in Table 7. The SNB angle correlated with a factor of 0.19 to the height and 0.14 to the weight, so there is a slightly stronger correlation to height. The condylar volumes on the right side correlated with a factor of 0.18 with the height and a slightly higher factor of 0.22 with the weight. On the left, the correlation with height was 0.19 and with weight was 0.24, so condylar volumes seem to correlate slightly more with the weight of the participants.

Nevertheless, the overall correlations were only weak, so the height and weight of the participants do not seem to be the defining parameters for condylar volume or SNB angle.

We were not able to find any relevant correlations between the SNB angle or condylar volumes and other parameters, like eating habits, physical activity, or various laboratory parameters, likely because our sample of participants was too small to detect more subtle effects (for a list of all parameters analyzed, see Table 1).

We hypothesized that there could be a correlation between sex hormones and small condylar volumes or SNB, as AICR often occurs in young women, suggesting that changes in the hormonal balance might play a role in pathogenesis. Unfortunately, the laboratory parameters for sex hormones were not documented in relation to anamnesis of the menstrual cycle, so we chose not to use this data, as this does not permit the reconstruction of the cycle phase and therefore the interpretation of the values measured is prohibited.

We also took a closer look at parameters indicating bone metabolism like calcium, parathormone, 25-OH-vitamin D, and alkaline phosphatase, as condylar resorption affects the bone and might be accompanied by altered bone metabolism. We looked at markers of inflammation as well, e.g., hs-CRP or IL-6, as these might hint at inflammatory causes of condylar resorption. While no correlations could be observed, it should be mentioned that such a correlation in a pathological situation cannot be ruled out as the data set comprises participants from the general population.

Upon closer inspection of the three groups of small, average, and large SNBs or condylar volumes, we made an interesting observation: Three of our participants reported to be diagnosed with tinnitus in their medical anamnesis, and all three had a low condylar volume under 1300 mm³ on both sides. Of note, Alsabban et al. mention the occurrence of tinnitus in association with temporomandibular disorder [1]. Further research to validate this observation might be warranted.

Of the 10 participants that had elevated platelets, most seemed to have a rather small condylar volume and/or SNB angle. A total of 4 out of 10 had a small condylar volume on both sides and 3 out of those 4 had a small SNB angle also.

Our results give an overview on the measurements of the mandible of women aged 19–40 years in MR imaging. As MRI is part of standard diagnostics in any TMJ disease to examine soft tissues like the articular discus, while CT scans are additionally used to assess the morphology of the bone, the possibility to assess the bony structures of the TMJ in the MRI as well enables a more effective use of resources, sparing the CT scan, and imposes less radiation on the patient. These considerations make our referential data on MRI-segmented mandibles an important prerequisite to improve patient care while also sparing resources.

Unfortunately, we did not detect any correlations of SNB and condylar volume to laboratory parameters or conspicuous issues regarding lifestyle, medical anamnesis, or oral health, which could be contributed to the relatively small sample size and the fact that participants were recruited from the general population, with only very few, if any, participants likely suffering from AICR.

We did not find a significant correlation between SNB angle and condylar volume, which is particularly interesting, as the SNB angle is often used as a surrogate parameter to assess condylar resorption in clinical practice. It might therefore be warranted to repeat our measurements with people affected by AICR to elucidate if SNB angles in the presence of AICR indeed correlate with the loss of condylar volume, which is the defining element of condylar resorption.

When correlating height and weight of the participants to SNB and the condylar volumes, we only found weak correlations, while SNB correlated more strongly with height and the condylar volumes showed stronger correlations with the weight of the participants, even though the differences are minor.

As we reported above, three of our participants mentioned having a tinnitus and all three of them had low condylar volume on both sides of the mandible. As tinnitus might be associated with temporomandibular disorder [1], we thought this is an interesting observation that could be followed up in future studies.

Ten participants showed elevated platelets in the complete blood count, and we noted that of those 10 participants, 5 had a small SNB angle and tended to have rather small condylar volumes, as can be seen in Table 7. As an elevated number of platelets is associated with a higher risk of a thrombus, it might be conceivable that inhibition of the blood flow in the condylar region caused by elevated platelets could lead to reduced bone metabolism and smaller condylar volumes. A possible connection to AICR could be explored in future studies.