Available online at www.sciencedirect.com

British Journal of Oral and Maxillofacial Surgery 53 (2015) 446–450

Panoramic imaging is not suitable for quantitative evaluation, classification, and follow up in unilateral condylar hyperplasia J.W. Nolte a,b,∗ , L.H.E. Karssemakers b,c , D.C. Grootendorst d,e , D.B. Tuinzing b,c , A.G. Becking a,b,f a

Department of Oral and Maxillofacial Surgery, Academic Medical Center (AMC), Amsterdam, The Netherlands Academic Centre for Dentistry Amsterdam (ACTA), The Netherlands c Department of Oral and Maxillofacial Surgery/Oral Pathology, VU University Medical Center, The Netherlands d Department of Epidemiology and Biostatistics, Linnaeus Institute, Spaarne Gasthuis, Haarlem, The Netherlands e Landsteiner Institute, Medical Center Haaglanden, The Hague, The Netherlands f Department of Oral and Maxillofacial Surgery, Spaarne Gasthuis, Haarlem, The Netherlands b

Accepted 17 February 2015 Available online 19 March 2015

Abstract Patients with suspected unilateral condylar hyperplasia are often screened radiologically with a panoramic radiograph, but this is not sufficient for routine diagnosis and follow up. We have therefore made a quantitative analysis and evaluation of panoramic radiographs in a large group of patients with the condition. During the period 1994–2011, 132 patients with 113 panoramic radiographs were analysed using a validated method. There was good reproducibility between observers, but the condylar neck and head were the regions reported with least reliability. Although in most patients asymmetry of the condylar head, neck, and ramus was confirmed, the kappa coefficient as an indicator of agreement between two observers was poor (−0.040 to 0.504). Hardly any difference between sides was measured at the gonion angle, and the body appeared to be higher on the affected side in 80% of patients. Panoramic radiographs might be suitable for screening, but are not suitable for the quantitative evaluation, classification, and follow up of patients with unilateral condylar hyperplasia. © 2015 The British Association of Oral and Maxillofacial Surgeons. Published by Elsevier Ltd. All rights reserved.

Keywords: Unilateral condylar hyperplasia; Mandibular asymmetry; 2-Dimensional imaging; Panoramic radiograph

Introduction Unilateral condylar hyperplasia or hyperactivity is rare, and the aetiology is unknown. It is characterised by excessive pathological activity that resembles growth in one of the mandibular condyles, and this causes asymmetry

∗ Corresponding author at: Department of Oral and Maxillofacial Surgery, Academic Medical Center (AMC) Amsterdam, P.O. Box 22660, 1100 DD Amsterdam, The Netherlands. Tel.: +31 20 5663997. E-mail address: [email protected] (J.W. Nolte).

of the lower jaw and subsequent facial asymmetry. It is characterised clinically by various degrees of mandibular overgrowth. Obwegeser and Makek described a classification of 3 categories: hemimandibular elongation, hemimandibular hyperplasia, and a combination of these two (hybrid form).1 The expected coexisting skeletal and dental variables can include a class III malocclusion on the affected side with a deviation of the chin and midline towards – and cross-bite on – the unaffected side. These may or may not be accompanied by a malocclusion with canting of the occlusal plane and unilateral open bite on the affected side without deviation to

http://dx.doi.org/10.1016/j.bjoms.2015.02.014 0266-4356/© 2015 The British Association of Oral and Maxillofacial Surgeons. Published by Elsevier Ltd. All rights reserved.

J.W. Nolte et al. / British Journal of Oral and Maxillofacial Surgery 53 (2015) 446–450

the midline. The asymmetrical development in these patients is often compounded by functional and aesthetic problems. Traditionally, routine radiological screening comprises a panoramic or posteroanterior radiograph, or both. Obwegeser and Makek described the typical characteristics of an affected mandible using panoramic radiographs.1,2 To our knowledge, Habets et al. are the only authors who have made reproducible measurements on panoramic radiographs. They concluded that a 6% vertical difference between sides resulted in 3% asymmetry. Patients who have 3% asymmetry or less are not thought to have real asymmetry, and such a small difference is the result of technical irregularities (such as malpositioning of the head).3,4 Other authors, however, recommend caution in the use of panoramic images for the assessment of mandibular asymmetry, as they often turn out to be inaccurate.5,6 We know of no previous study that has evaluated Obwegeser’s descriptions. The aim of the present study was to analyse quantitatively the available 2-dimensional panoramic radiographs in a large group of patients with condylar hyperplasia, which we expect to be insufficient for routine diagnostic procedures and follow up of the condition. Approval was granted by the Medical Ethics Review Committee.

Patients and methods We retrospectively analysed the clinical records of all patients who were referred during the period 1994–2011 to the Department of Oral and Maxillofacial Surgery of the VU University Medical Centre Amsterdam, and the Spaarne Gasthuis Haarlem, with progressive asymmetry of the mandible. Patients were diagnosed with unilateral condylar hyperplasia if they gave a history of mandibular asymmetry, and had clinically progressive disease confirmed by bone scan (technetium labelled-phosphonate scan that showed differences in affected compared with unaffected regions of interest of more than 10%),7,8 or condylectomy. Patients with a confirmed mandibular fracture, those who had had previous mandibular (orthognathic) surgery, or those whose facial asymmetry was thought to be caused by something else, were excluded from this study. Sex, side affected, and age at presentation were recorded. All panoramic radiographs were assessed, and those that did not show or only partly showed one or two condyles were excluded. The radiographs were analysed using a validated method (Figs. 1 and 2).4 Additional variables included the angle of the gonion, which was measured as the angle between the tangential line of the ramus and the mandibular body; and the height of the body, which was measured as the line perpendicular to the tangential line of the mandibular body distal to the first molar. The condylar neck was defined as the distance between the condylar head and the perpendicular line through the lowest point of the semilunar border. The height of the ramus was defined as the total length between the most upper and lower points perpendicular to

447

Fig. 1. Panoramic radiograph showing hemimandibular elongation.

Fig. 2. Panoramic radiograph showing hemimandibular hyperplasia.

the tangential line of the mandibular ramus. Measurements were made independently by two observers who used the same calibrated Protractor (3M Unitek TM Cephalometric Protractor, Monrovia, USA). Asymmetry was calculated as (affected − healthy)/(affected + healthy) × 100%. Differences of more than 3% were considered as true asymmetry. Statistical analysis The significance of differences between variables was assessed with the help of PASW Statistics for Windows (version 18, SPSS Inc., Chicago, IL, USA). We used the chi square test for non-parametric data and Student’s t test for parametric data. Between-observer reproducibility of continuous variables was calculated after an analysis of variance. The between-observer reproducibility was subsequently computed by using the formula: [mean square between groups (MSB) − mean square within groups (MSW)/(MSB + MSW)]. Cohen’s kappa statistic (k) was calculated to assess the between-observer agreement in the classification of patients with or without asymmetry of the mandible. Both indicators of agreement can have a maximum value of 1, indicating perfect agreement.

Results During the period 1994–2011, 241 patients with asymmetry of the mandible were referred, 132 of whom were diagnosed as having unilateral condylar hyperplasia according to our criteria (Table 1). We found no association between the side affected and sex (p = 0.24).

448

J.W. Nolte et al. / British Journal of Oral and Maxillofacial Surgery 53 (2015) 446–450

Table 1 Details of 132 patients studied. Data are number (%) unless otherwise stated. Variable Sex Male Female Mean (SD) age (years) Range Side affected Left Right Bilateral

61 (46) 71 (54) 22 (8) 9–43 70 (53) 60 (46) 2 (2)

Table 2 Between-observer reproducibility (kappa) of variables of unilateral hemimandibular hyperplasia as assessed on orthopantomogram, stratified by affected and healthy condylar head and neck. Between-observer reproducibility

Condylar head Condylar neck Height of the ramus Angle of the gonion Height of the body

Affected side

Healthy side

0.51 0.78 0.88 0.98 0.97

0.65 0.76 0.96 0.93 0.95

Table 3 Prevalence of asymmetry of more than 3% between the affected and healthy sides together with agreement of asymmetry between the two observers. No (%) with asymmetrya

Observer 1

Observer 2

Kappa

Condylar head Condylar neck Height of the ramus

95 (84.1) 98 (86.7) 76 (67.3)

98 (86.7) 102 (90.3) 80 (70.8)

0.114 −0.04 0.504

a

Asymmetry was healthy)I × 100%.

calculated

as

I(affected − healthy)/(affected +

A total of 113 panoramic radiographs were available and suitable for evaluation. Table 2 shows the between-observer reproducibility of the measurements. The condylar neck, and particularly the area of the condylar head, were the least reliable regions according to reproducibility of measurements. Table 3 shows the prevalence of asymmetry of more than 3% between affected and non-affected sides according to the measurements of our two observers. Most patients had some degree of asymmetry: for more than 84% of the patients this was in the area of the condylar head and neck, and in over 67% in the height of the ramus. However, the Cohen’s kappa coefficient as indicator for agreement between the two observers is poor (−0.040 to 0.504). The angle of the gonion and the height of the body are shown in Table 4. There is hardly any difference between the Table 4 Number (%) of patients with affected side larger than the healthy side.

Angle of the gonion Height of the body

Observer 1

Observer 2

Agreement (kappa)

57 (50.9) 90 (80.4)

56 (50.0) 90 (80.4)

0.77 0.60

sides at the angle of the gonion, indicating good betweenobserver agreement, kappa coefficient 0.77). The height of the body seemed larger on the affected side in 80% of patients (kappa coefficient 0.60). Discussion Unilateral condylar hyperplasia is an intriguing condition and many questions are still to be answered. As it is so rare, we know of no large studies on incidence and clinical characteristics. Nitzan et al. published a series of 61 patients in which they concluded that it can occur at any age, and it does not stop at the end of the growth period. They found that the condition is more common among women, and the affected side is sex-dependent, with a predominance to the left side in men.9 Raijmakers et al. reviewed 10 studies that included 275 patients.10 They showed a clear predominance among women, and no association between sex and side affected. The mean age at presentation in those studies was in the early twenties. In the present retrospective study of 132 Dutch patients we found a slight preference for female sex (54%), but no significant connection between side affected and sex, and that the mean age at presentation was 22 years. According to Obwegeser and Makek a panoramic radiograph of the mandible is one of the best diagnostic media for mandibular anomalies, and it presents a fair overview of mandibular discrepancies.2 Objective measurements, however, are difficult to obtain, and several studies have evaluated the reliability of panoramic measurements. Opinions vary from useful (a simple diagnostic tool to measure mandibular length) to cautious (apparent geometric discrepancies result from incorrect positioning of the head, magnification factors, or distortion and type of camera).5,6,11–15 According to the classification, hemimandibular elongation includes an elongated hemimandible with lengthening of the horizontal and ascending ramus and of the condylar neck, and a stretched angle. In hemimandibular hyperplasia there is an increase in volume in all directions, and on 2-dimensional panoramic views this is seen as a clearly-elongated ascending ramus in all its sections, the neck of the condyle is longer, and the height of the horizontal ramus is enlarged on the affected side. The angle is rounded off or bowed.2 To be able to objectify and quantify these features, reproducible measurements are necessary, and it is clear from published reports that horizontal measurements on panoramic radiographs are unreliable.12–14 Habets et al. developed a validated method for assessment of vertical asymmetry, in which they used a left:right ratio, and specified that an outcome of more than 3% can be considered to be true asymmetry.3,4 Kjellberg et al. measured the effects of condylar lesions quantitatively on an orthopantogram and suggested that condylar ratios should be used to look for asymmetry in condylar height, as those are unaffected by positioning error, distortion, or magnification errors.13 Considering the inclusion criteria and the nature of the disease we expected to find true asymmetry on all panoramic

J.W. Nolte et al. / British Journal of Oral and Maxillofacial Surgery 53 (2015) 446–450

radiographs of the patients in this study. Using the mathematical method described by Habets et al., we found asymmetry of more than 3% the area of the condylar head and neck in 84%–90% of our patients and in the height of the ramus in only 67%–71%. Although the measurements showed quite good reproducibility between observers (> 0.88) in the height and angle of the ramus and the height of the body, the condylar region seems to be the least reliable region for linear measurements (0.51–0.78 for the condylar head and neck). This is even less on the affected side. The kappa coefficient as an indicator of agreement between observers about asymmetry was also poor (−0.040 to 0.504). There is a tendency among patients towards Class III malocclusion on the affected side. In hemimandibular elongation there is often a crossbite on the unaffected side as well. Uysal et al. evaluated vertical asymmetry of the condyle and the ramus in adolescent subjects with normal occlusion patterns, unilateral and bilateral crossbites. They found similar asymmetrical indices with no significant differences between them.16 Sezgin et al.17 studied the effects of different occlusion types on mandibular asymmetry in young people and concluded that malocclusions have an effect on the height of the condyle rather than the ramus and that an Angle Class II/1 seems to be more related to condylar asymmetry.17 Turp et al. measured 25 adult male skulls with stable occlusions and compared these with the measurements on the orthopantographs.18 The overall measurement error was too small to explain the observed asymmetry and so they concluded that the asymmetry of the condyle and ramus were part of the biological variations in humans, but it is not clear when this should be considered pathophysiological.18 As the slender type of hemimandibular elongation tends to have a stretched angle, and the hyperplastic type has a rounded or bowed angle, one would expect to find different angles in affected and healthy sides in these patients. Larheim et al. showed that measurements of the angle on panoramic images have acceptable reproducibility with angles that are almost identical to those on measurements of the skull.19 We found hardly any difference between sides in the gonial angle, with reasonably good agreement between observers (kappa coefficient 0.77). It is clear that there are more 2-dimensional radiographs suitable for evaluation of mandibular asymmetry, such as the anterioposterior view. However these have their limitations as well, such as the inaccuracy of positioning landmarks.11 These authors used panoramic views to attempt to reproduce the Obwegeser and Makek’s descriptions and to evaluate them as a follow up tool. However, they identified a major drawback, as the panoramic radiographs were not made using the same machine and no mean linear measurement could be calculated or significant differences found, so they could not evaluate the orthopantograph as a useful tool. We conclude that panoramic radiographs might be suitable for screening purposes, to give overall information on dimensions and a description of characteristics in patients with unilateral condylar hyperplasia, but that they are not

449

suitable for quantitative evaluation, classification, or follow up. A major drawback is the 2-dimensional reproduction of a 3-dimensional condition. The variable shape of the mandible in unilateral condylar hyperplasia makes it difficult to draw correct tangential lines to the ramus and the mandibular body. It is also impossible to measure volumes, whereas volumetry is an important diagnostic factor in this disease. It is necessary to be able to differentiate between elongation and hyperplasia, and to follow-up the progress of the disease or the therapeutic effects of surgery (condylectomy). Our findings are not surprising, and we have now launched a study to evaluate Obwegeser and Makek’s classification of unilateral condylar hyperplasia using 3-dimensional radiographs and quantitative techniques. Conflict of interest We have no conflict of interest. Ethics statement/confirmation of patients’ permission The protocol was approved by the Medical Ethics Committee, and as a retrospective study does not fall within the scope of the Medical Research Involving Human Subjects Act (WMO). References 1. Obwegeser HL, Makek MS. Hemimandibular hyperplasia – hemimandibular elongation. J Maxillofac Surg 1986;14:183–208. 2. Obwegeser HL. Documentation for diagnosis, treatment planning and follow up. In: Obwegeser HL, Luder H-U, editors. Mandibular growth anomalies, terminology, aetiology, diagnosis, treatment. Berlin: Springer; 2001 [chapter 4]. 3. Habets LL, Bezuur JN, van Ooij CP, et al. The orthopantomogram, an aid in diagnosis of temporomandibular joint problems. I. The factor of vertical magnification. J Oral Rehabil 1987;14:475–80. 4. Habets LLMH, Bezuur JN, Naeiji M, et al. The orthopantomogram, an aid in diagnosis of temporomandibular joint problems. II. The vertical symmetry. J Oral Rehabil 1988;15:465–71. 5. Turp JC, Vach W, Harbich K, et al. Determining mandibular condyle and ramus height with the help of an Orthopantomogram – a valid method? J Oral Rehabil 1996;23:395–400. 6. Van Elslande DC, Russett SJ, Major PW, et al. Mandibular asymmetry diagnosis with panoramic imaging. Am J Orthod Dentofacial Orthop 2008;134:183–92. 7. Saridin CP, Raijmakers PG, Tuinzing DB, et al. Comparison of planar bone scintigraphy and single photon emission computed tomography in patients suspected of having unilateral condylar hyperactivity. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 2008;106:426–32. 8. Saridin CP, Raijmakers PG, Tuinzing DB, et al. Bone scintigraphy as a diagnostic method in unilateral hyperactivity of the mandibular condyles: a review and meta-analysis of the literature. Int J Oral Maxillofac Surg 2011;40:11–7. 9. Nitzan DW, Katsnelson A, Bermanis I, et al. The clinical characteristics of condylar hyperplasia: experience with 61 patients. J Oral Maxillofac Surg 2008;66:312–8. 10. Raijmakers PG, Karssemakers LHE, Tuinzing DB. Female predominance and effect of gender on unilateral condylar hyperplasia: a review and meta-analysis. J Oral Maxillofac Surg 2012;70:e72–6.

450

J.W. Nolte et al. / British Journal of Oral and Maxillofacial Surgery 53 (2015) 446–450

11. Ongkosuwito EM, Dieleman MMJ, Kuijpers-Jagtman AM, et al. Linear mandibular measurements: comparison between orthopantomogram and lateral cephalograms. Cleft Palate Craniofac J 2009;46: 147–53. 12. Laster WS, Ludlow JB, Bailey LJ, et al. Accuracy of measurements of mandibular anatomy and prediction of asymmetry in panoramic radiographic images. Dentomaxillofac Radiol 2005;34:343–9. 13. Kjellberg H, Ekestubbe A, Kiliaridis S, et al. Condylar height on panoramic radiographs a methodologic study with a clinical application. Acta Odontol Scand 1994;52:43–50. 14. Tronje G, Welander U, McDavid WD, et al. Image distortion in rotational panoramic radiography. I. Object morphology: outer contours. Acta Radiol Diagn (Stockh) 1981;22:689–96.

15. Akcam MO, Altiok T, Ozdiler E. Panoramic radiographs: a tool for investigating skeletal pattern. Am J Orthod Dentofacial Orthop 2003;123:175–81. 16. Uysal T, Sisman Y, Kurt G, et al. Condylar and ramal vertical asymmetry in unilateral and bilateral posterior crossbite patients and a normal occlusion sample. Am J Orthod Dentofacial Orthop 2009;136:37–43. 17. Sezgin OS, Celenk P, Arici S. Mandibular asymmetry in different occlusion patterns. Angle Orthod 2007;77:803–7. 18. Turp JC, Alt KW, Vach W, et al. Mandibular condyles and rami are asymmetric structures. Cranio 1998;16:51–6. 19. Larheim TA, Svanaes DB. Reproducibility of rotational panoramic radiography: mandibular linear dimensions and angles. Am J Orthod Dentofacial Orthop 1986;90:45–51.

Panoramic imaging is not suitable for quantitative evaluation, classification, and follow up in unilateral condylar hyperplasia.

Patients with suspected unilateral condylar hyperplasia are often screened radiologically with a panoramic radiograph, but this is not sufficient for ...
568KB Sizes 2 Downloads 10 Views