lable at ScienceDirect

Journal of Cranio-Maxillo-Facial Surgery xxx (2015) 1e7

Contents lists avai

Journal of Cranio-Maxillo-Facial Surgery

journal homepage: www.jcmfs.com

3D analysis of condylar remodelling and skeletal relapse followingbilateral sagittal split advancement osteotomies

Tong Xi a, *, Ruud Schreurs a, Bram van Loon a, Martien de Koning a, Stefaan Berg�e a,Theo Hoppenreijs b, Thomas Maal a

a Department of Oral and Maxillofacial Surgery, Radboud University Nijmegen Medical Centre, Geert Grooteplein 10, 6525 GA, Nijmegen, The Netherlandsb Department of Oral and Maxillofacial Surgery, Rijnstate Hospital, Wagenerlaan 55, 6815 AD, Arnhem, The Netherlands

a r t i c l e i n f o

Article history:Paper received 21 August 2014Accepted 10 February 2015Available online xxx

Keywords:Bilateral sagittal split osteotomyCondylar remodellingCondylar resorptionCBCTRelapse3D

* Corresponding author. Department of Oral and MUniversity Nijmegen Medical Centre, P.O. Box 9101number 590, The Netherlands. Tel.: þ31 24 361 45 50

E-mail addresses: Tong.Xi@radboudumc.nl (T. Xi),nl (R. Schreurs), Bram.vanLoon@radboudumc.nl (B. varadboudumc.nl (M. de Koning), Stefaan.Berge@thoppenreijs@rijnstate.nl (T. Hoppenreijs), Thomas.Ma

http://dx.doi.org/10.1016/j.jcms.2015.02.0061010-5182/© 2015 European Association for Cranio-M

Please cite this article in press as: Xi T, eadvancement osteotomies, Journal of Cranio

a b s t r a c t

A major concern in mandibular advancement surgery using bilateral sagittal split osteotomies (BSSO) ispotential postoperative relapse. Although the role of postoperative changes in condylar morphology onskeletal relapse was reported in previous studies, no study so far has objectified the precise changes ofthe condylar volume. The aim of the present study was to quantify the postoperative volume changes ofcondyles and its role on skeletal stability following BSSO mandibular advancement surgery. A total of 56patients with mandibular hypoplasia who underwent BSSO advancement surgery were prospectivelyenrolled into the study. A cone beam computed tomography (CBCT) scan was acquired preoperatively, at1 week postoperatively and at 1 year postoperatively. After the segmentation of the facial skeleton andcondyles, three-dimensional cephalometry and condylar volume analysis were performed. The meanmandibular advancement was 4.6 mm, and the mean postoperative relapse was 0.71 mm. Of 112 con-dyles, 55% showed a postoperative decrease in condylar volume, with a mean reduction of 105 mm3 (6.1%of the original condylar volume). The magnitude of condylar remodelling (CR) was significantly corre-lated with skeletal relapse (p ¼ 0.003). Patients with a CR greater than 17% of the original condylarvolume exhibited relapse as seen in progressive condylar resorption. Female patients with a highmandibular angle who exhibited postoperative CR were particularly at risk for postoperative relapse.Gender, preoperative condylar volume, and downward displacement of pogonion at surgery wereprognostic factors for CR (r2 ¼ 21%). It could be concluded that the condylar volume can be applied as auseful 3D radiographic parameter for the diagnosis and follow-up of postoperative skeletal relapse andprogressive condylar resorption.

© 2015 European Association for Cranio-Maxillo-Facial Surgery. Published by Elsevier Ltd. All rightsreserved.

Introduction

Bilateral sagittal split osteotomy (BSSO) is the most frequentlyused orthognathic surgical procedure to correct mandibular hy-poplasia (skeletal Class II deformity). A major concern in the sur-gical advancement of the mandible is the potential postoperativerelapse (Borstlap et al., 2004; Chen et al., 2013). Postoperativeskeletal relapse is affected by bone movements at the osteotomy

axillofacial Surgery, Radboud, 6500 HB Nijmegen, Postal; fax: þ31 24 354 11 65.Ruud.Schreurs@radboudumc.n Loon), Martien.deKoning@radboudumc.nl (S. Berg�e),al@radboudumc.nl (T. Maal).

axillo-Facial Surgery. Published by

t al., 3D analysis of condyla-Maxillo-Facial Surgery (201

sites as well as by changes in the position and morphology of thecondyles. Since the amount of skeletal relapse at the osteotomysites was significantly reduced by the application of rigid fixationtechniques, the influence of the condyles on the postoperativerelapse has become more and more the focus of clinical research(Joss and Vassalli, 2009).

Previous studies reported that early postoperative skeletal relapse(<6months after surgery) is often associatedwithmalposition of oneorbothcondylesduring surgery, causingcondylar sagandsubsequentunfavourable displacement of the mandible (Mobarak et al., 2001;Joss and Vassalli, 2009; Kim et al., 2014). Late postoperative relapse(>12months after surgery), however, is often related tomorphologicchanges of the condyles, also known as progressive condylarresorption (PCR) (Hoppenreijs et al., 1999; Mobarak et al., 2001; Jossand Vassalli, 2009; Hoppenreijs et al., 2013).

Elsevier Ltd. All rights reserved.

r remodelling and skeletal relapse following bilateral sagittal split5), http://dx.doi.org/10.1016/j.jcms.2015.02.006

Table 1Definitions of the 3D cephalometric landmarks and measurements.

Landmarks Definition Bilateral

A-point (A) The point of maximum concavity in the midlineof the alveolar process of the maxilla.

C-point (C) The most caudal point of the sigmoid notch. XCondor (Con) The most posterior point of the mandibular

ramus intersecting the C-plane.X

Gonion (Go) The most caudal and most posterior point of themandibular angle.

X

Mean gonion(Gomean)

A computed landmark based on the projectionof left and right gonion on the median plane.

Menton (Men) The most inferior midpoint of the chin on theoutline of the mandibular symphysis.

Nasion (N) The midpoint of the frontonasal suture.Pogonion (Pog) The most anterior midpoint of the chin.Sella (S) The center of the hypophyseal fossa (sella

turcica).Measurement Definition BilateralGonial angle The angle between the landmarks condor,

gonion, and menton.X

Intergonialdistance

The distance between the landmarks left andright gonion.

Mandibularplane angle

The angle between the SN-line and themandibular plane.

Occlusalplane angle

The angle between the SN-line and the occlusalplane.

S-Gomean The distance between the landmarks sella andmean gonion along the vertical plane.

SNA angle The angle between the landmarks sella, nasion,and A-point projected on the median plane.

SNPog angle The angle between the landmarks sella, nasion,and pogonion projected on the median plane.

T. Xi et al. / Journal of Cranio-Maxillo-Facial Surgery xxx (2015) 1e72

In cases inwhich the alteredmechanical load, during or after thesurgical procedure, exceeds the intrinsic adaptive capability of oneor both temporomandibular joints, some morphologic post-operative changes of the condyles may occur. According to theirmagnitude and rate of progression, the morphologic changes of thecondyles can be defined as physiologic condylar remodelling (CR)or pathologic PCR (Scheerlinck et al., 1994; Arnett et al., 1996a,b;Hoppenreijs et al., 1999). Radiographic assessment is required toquantify these condylar changes.

The use of cone-beam computed tomography (CBCT) in thetreatment planning and postoperative follow-up of orthognathicpatients has provided the surgeons with a new three-dimensional(3D) imaging modality to evaluate the postoperative skeletalrelapse and postoperative morphologic changes of the condyles.Skeletal relapse, as well as positional and dimensional changes ofthe condyles, can be quantified in three dimensions (Chen et al.,2013; Kim et al., 2014). The application of a validated 3D condylarsegmentation technique allows accurate volumetric measurementsof condyles (Paniagua et al., 2011; Xi et al., 2013). The potential to beable to quantify the volumetric changes of condyles in the post-operative follow-up may help clinicians to identify CR and PCR, aswell as associated skeletal relapse.

The aims of this study were to quantify the postoperative vol-ume changes of condyles in a 3D way, to assess the postoperativeskeletal relapse, and to identify the role of condylar changes onskeletal stability following BSSO mandibular advancement surgery.

Material and methods

Patients

All patients were treated with a BSSO mandibular advancementsurgery between 2007 and 2011 at the Department of Oral andMaxillofacial Surgery in Radboud University Nijmegen MedicalCentre. The inclusion criteria were nonsyndromic mandibular hy-poplasia, minimum age of 16 years, and availability of pre- andpostoperative CBCT scans. The exclusion criteria were previousorthognathic surgery, simultaneously performed other orthog-nathic procedures (e.g., Le Fort I osteotomy, genioplasty), buccalplate fracture during surgery, and severe facial asymmetry.

Surgical technique

All surgeries were performed under general anaesthesia withnasotracheal intubation. A BSSO mandibular advancement surgeryaccording to the Hunsuck modification was performed (Hunsuck,1968). After the completion of the osteotomies, the distalsegment of the mandible was placed in the planned position usingthe prefabricated interocclusal splint and stabilized with inter-maxillary fixation. The proximal segments were gently pushedbackward and upward to seat the condyles in the glenoid fossa. Themandibular segments were fixed with two titanium miniplates(one on each side) and monocortical screws (Champy 2.0 mm, KLSMartin, Tuttlingen, Germany). After the release of intermaxillaryfixation, the final occlusionwas checked. Depending on the stabilityof the occlusion, the interocclusal splint was left in place, and tightelastics were used in the first week after surgery. Class II guidingelastics were used after the first week, in conjunction with thepostoperative orthodontic treatment.

3D cephalometry and measurements

CBCT scans were acquired for each patient on the followingoccasions: 1e4 weeks prior to surgery (Tpre), 1 week after surgery(T1wk), and 1 year after surgery (T1yr). Patients underwent scanning

Please cite this article in press as: Xi T, et al., 3D analysis of condylaadvancement osteotomies, Journal of Cranio-Maxillo-Facial Surgery (201

while seated in a natural head position, using a standard CBCTscanning protocol (field of view: 22 � 16 cm; scan time: 40 s; voxelsize 0.4 mm; i-CAT, 3D Imaging System, Imaging Sciences Inter-national Inc, Hatfield, PA, USA). CBCT data were exported in aDICOM format and rendered to 3D models of the facial skeletonusing Maxilim software (Medicim NV, Mechelen, Belgium).

A 3D cephalometric reference framewas set up for the 3Dmodelof a patient at Tpre according to the validated procedure of Swennenet al. (2006). The 3Dmodels of T1wk and T1yr were superimposed onthe Tpre 3D model using voxel-based registration on the unalteredsubvolume of the anterior cranial base, forehead, and zygomaticarches (Nada et al., 2011). 3D Cephalometric landmarks wereidentified for the 3D cephalometric hard tissue analysis (Swennenand Schutyser, 2006). Cephalometric measurements were per-formed to analyse the actual surgical displacement of mandibularsegments, and to assess the postoperative skeletal relapse (Table 1).

A semi-automated CBCT condylar segmentation protocol wasapplied to segment and render the condyles accurately (Xi et al.,2013). The 3D-rendered condyles were then loaded into the orig-inal 3D head models. A plane that goes through the C-point (themost caudal point of the sigmoid notch) and runs parallel to theFrankfurter plane (C-plane) was constructed inMaxilim. The part ofthe condylar process situated cranially to the C-plane was definedas the condylar volume, and its volume was measured at Tpre andT1yr (Fig. 1).

Statistical analysis

Statistical data analysis was carried out with the IBM SPSSsoftware programme, version 20.0 for Windows (IBM Corp.,Armonk, NY, USA). Twelve variables of cephalometric measure-ments were selected and analysed over three time intervals toquantify the surgical displacements (Tpre � T1wk) and postoperativeskeletal relapse (T1wk � T1yr). A total of 28 patients with the lowestpreoperative mandibular plane angle (mean 27.9�; SD 3.18�; range

r remodelling and skeletal relapse following bilateral sagittal split5), http://dx.doi.org/10.1016/j.jcms.2015.02.006

Fig. 1. The C-plane, defined as a plane that runs parallel to the Frankfurt plane through the C-point, is visualized in the left picture. The magnification on the right shows the C-point(yellow) and the superimposition of 3-dimensionally rendered preoperative (white) and postoperative (purple) condyles. The part of the condylar process that is located cranially tothe C-plane is defined as the condylar volume. A postoperative reduction in condylar size (volume) is clearly visible.

Table 3Changes (mm or degree) in cephalometric variables among 56 mandibularadvancement patients attributed to surgery and postoperative relapse.

Variable Surgical displacement Postoperative relapse

Mean SD Range Mean SD Range

Horizontal(mm)

A 0.16a 0.79 �2.2e2.4 �0.18a 0.77 �2.0e2.0Pog 4.59a 3.43 �4.7e14.4 �0.71a 2.04 �3.4e6.1

Vertical(mm)

A 0.11 1.54 �2.4e4.3 �0.02 1.30 �2.3e2.7Pog 3.62a 2.10 �1.2e8.7 �0.81a 1.82 �2.6e4.5Go 0.11 1.58 �5.3e3.5 �1.68a 1.59 �1.0e5.5

Transverse(mm)

C-point 1.65a 1.35 �1.7e4.9 �0.76a 0.88 �1.4e3.1Go 6.35a 3.06 �0.6e13.1 �2.91a 2.28 �1.8e8.9

Angle (�) SNA �0.01 0.83 �2.6e1.5 �0.21a 0.82 �1.4e2.3SNPog 2.81a 1.63 �0.1e8.1 �0.49a 1.06 �1.8e3.2Mandibularplane

�0.58a 2.26 �5.1e4.7 1.10a 1.72 �2.2e6.0

Occlusalplane

�0.07 2.20 �5.4e6.3 1.44a 2.12 �2.2e8.6

Gonial �0.13 2.69 �7.9e6.0 2.01a 2.42 �2.3e11.3

T. Xi et al. / Journal of Cranio-Maxillo-Facial Surgery xxx (2015) 1e7 3

18.9�e31.2�) and 28 patients with the highest preoperativemandibular plane angle (mean 38.7�; SD 5.64�; range 31.8�e53.0�)formed respectively the low-angle and high-angle group. Analysisof variance was used to test for differences in skeletal relapse andcondylar volume changes between the groups with correction forpossible confounding factors at the 5% level of significance(p � 0.05). Univariate and multivariate regression analysis wereapplied to identify the prognostic factors for postoperative changesin condylar volume.

Results

The study population consisted of 56 patients, 39 female (69.6%)and 17 male (30.5%). Their characteristics are described in Table 2.No statistically significant difference in age was found betweenmale and female (p ¼ 0.42) participants. A total of 112 splits of themandibular ramus were performed.

Skeletal movements and postoperative relapse

The mean changes in cephalometric measurements as the im-mediate result of the BSSO advancement surgery and the post-operative skeletal relapse for all patients are presented in Table 3.The mean surgical advancement at pogonion was4.59 mm ± 3.43 mm. Pogonion showed a downward movement of3.62mm± 2.10mmdirect postoperatively. A statistically significantpostoperative relapse (p < 0.05) at pogonion was found both in thehorizontal (0.71 mm) and vertical direction (0.81 mm), respectively15% and 23% of the surgical displacement. The mandibular planeangle decreased by 0.58� following surgery (p < 0.05). However, themandibular plane displayed a tendency to increase (1.1� ± 1.72�) inthe follow-up period (p < 0.05). Flaring of the proximal segmentwas measured at the mandibular angles. The intergonial distanceincreased immediately after surgery (6.35 mm ± 3.06 mm), but

Table 2Age and gender distribution within the study population.

Total population(n ¼ 56)

Male(n ¼ 17)

Female(n ¼ 39)

p-Value

Age (years) Mean (SD) 30.2 (12.5) 28.2 (14.4) 31.1 (11.7) 0.42Range 15e54 16e54 15e54

Please cite this article in press as: Xi T, et al., 3D analysis of condylaadvancement osteotomies, Journal of Cranio-Maxillo-Facial Surgery (201

decreased significantly in the first year after surgery(2.91 mm ± 2.28 mm). The gonial angle was increased by 2� in thepostoperative follow-up.

No significant correlation was found between the postoperativeskeletal relapse at pogonion, and the lingual split pattern and gender,both in the horizontal (p ¼ 0.07, p ¼ 0.30) and vertical direction(p ¼ 0.07, p ¼ 0.91). However, the high-angle patients displayedsignificantlymore skeletal relapse than low-angle patients. Themeanrelapse at pogonion in the horizontal direction was 1.24 mm in thehigh-angle group and 0.20 mm in the low-angle group (p ¼ 0.005).

Condylar volume and postoperative skeletal changes

The mean preoperative and postoperative condylar volumeswere 1728 mm3 (SD 496 mm3; range 532e2844 mm3) and

angle

a Significant at the 5% level (p< 0.05). Horizontal changes: positive value indicatesanterior displacement, negative value indicates posterior displacement. Verticalchanges: positive value indicates inferior displacement, negative value indicatessuperior displacement. Transverse changes: positive value indicates transversewidening, negative value indicates transverse narrowing. Surgical displacement:difference between Tpre and T1wk. Postoperative relapse: difference between T1wk

and T1yr. The direction of the relapse opposes the direction of surgical displacementin all cases.

r remodelling and skeletal relapse following bilateral sagittal split5), http://dx.doi.org/10.1016/j.jcms.2015.02.006

Table 4Mean postoperative skeletal changes in the CR, CRm, CR1SD and CR2SD groups.

Variables CR group (n ¼ 62) CRm group (n ¼ 22) CR1SD group (n ¼ 10) CR2SD group (n ¼ 4)

Mean (SD) p Value Mean (SD) p Value Mean (SD) p Value Mean (SD) p-Value

Relapse at Pog Horizontal (mm) 0.90 (2.00) 0.30 1.82 (2.73) 0.004* 2.61 (3.20) 0.002* 5.18 (1.59) <0.001*Vertical (mm) 0.72 (1,76) 0.53 0.76 (1.94) 0.88 �0.24 (1.43) 0.054 �0.95 (1.91) 0.047*

Intergonial distance (mm) 2.63 (2.35) 0.15 3.07 (2.60) 0.71 2.87 (2.70) 0.95 3.50 (3.23) 0.60S-Gomean (mm) 1.99 (1.73) 0.02* 2.77 (1.88) <0.001* 2.75 (1.77) 0.02* 4.15 (0.66) 0.001*SN-Pog (�) 0.64 (1.06) 0.08* 1.12 (1.40) 0.001* 1.51 (1.51) 0.001* 2.65 (0.71) <0.001*Mandibular plane (�) 1.35 (1.77) 0.06* 2.00 (2.17) 0.004* 2.80 (2.65) 0.001* 4.80 (1.75) <0.001*Occlusal plane (�) 1.75 (2.32) 0.08* 2.72 (3.01) 0.001* 2.51 (2.29) 0.094 4.28 (1.33) 0.006*Gonial angle (�) 1.82 (2.68) 0.37 2.33 (3.09) 0.49 1.52 (1.96) 0.50 1.98 (1.88) 0.98

CR group: condyles with a decrease in condylar volume >0mm3. CRm group: condyles with a decrease in condylar volume >105 mm3 (mean decrease). CR1SD group: condyleswith a decrease in condylar volume >197 mm3 (mean þ 1 standard deviation). CR2SD group: condyles with a decrease in condylar volume >289 mm3 (mean þ 2 standarddeviations). A two-way ANOVA was used to calculate the p values.*p < 0.05.

T. Xi et al. / Journal of Cranio-Maxillo-Facial Surgery xxx (2015) 1e74

1703 mm3 (SD 552 mm3; range 494e2822 mm3) respectively. Of112 condyles, 62 condyles (55%) showed a postoperative reductionin condylar volume, with a mean reduction of 105 mm3 (SD92 mm3; range 2e428 mm3). This reduction in condylar volumewas statistically significant (p < 0.001; 95% confidence interval [CI]81e127 mm3). Of a total of 56 patients, 14 patients (25%) demon-strated a postoperative horizontal relapse of more than 2 mm atpogonion. The mean decrease in condylar volume among thesepatients was 72.5 mm3, against 8.4 mm3 for the rest of the patients(p¼ 0.013). The magnitude of postoperative CR was correlated withskeletal relapse (Pearson's r ¼ 0.23, p ¼ 0.003).

The condyles that decreased in volume (CR group) were furtherdivided into three subgroups (CRm, CR1SD, and CR2SD groups) ac-cording to the magnitude of the volume reduction. The meanpostoperative skeletal relapse within each group is summerised inTable 4. The amount of the skeletal relapse increased significantlyin patients who experienced more condylar resorption. Thedecrease in condylar volume was associated with an increase inskeletal relapse at pogonion, a decrease in posterior facial height (S-Gomean) and an increase in mandibular plane angle. Postoperativeflaring at gonion and an increase in gonial angle were not associ-ated with a decrease in condylar volume. It is worth noting that inthe CR2SD group, an increase in the anterior facial height in com-bination with a decrease in posterior facial height was observed.

Condylar volume changes and gender

The mean preoperative condylar volume for female and malepatients were 1631 mm3 and 1952 mm3, respectively (p ¼ 0.001).Among female patients, 63% of condyles decreased in volumepostoperatively, against 35% among male patients. Postoperatively,the mean reduction in condylar volume was significantly higheramong female patients in comparison to male patients (mean dif-ference 81.2 mm3; SD 125.7 mm3; p ¼ 0.001). Gender did not affect

Table 5Mean postoperative skeletal relapse (millimetres and degrees) among male and female

Mean postoperativeskeletal relapse

Patients with no postoperative decrease in condylar volume

Male Female p Value

SN-Pog (�) 0.22 0.35 0.67Pog horizontal (mm) 0.39 0.56 0.77Pog vertical (mm) 0.80 1.03 0.67S-Gomean (mm) 0.87 1.57 0.061Mandibular plane (�) 0.37 0.99 0.17Occlusal plane (�) 0.94 1.14 0.70Gonial angle (�) 2.30 2.20 0.86

*p < 0.05.

Please cite this article in press as: Xi T, et al., 3D analysis of condylaadvancement osteotomies, Journal of Cranio-Maxillo-Facial Surgery (201

the amount of skeletal relapse among patients whose condylarvolume stayed stable after surgery. However, in patients whoexhibited CR after surgery, significantly more relapse at pogonionand of mandibular plane angle occurred in female patients(Table 5).

Condylar volume changes and low-angle/high-angle patients

Patients in the low-angle group had preoperatively significantlylarger condyles than in the high-angle group (mean difference of302 mm3, p ¼ 0.001). In comparison to the high-angle group, lesscondyles displayed a postoperative decrease in volume amongpatients with a lowmandibular angle (<31.5�) (Tables 6 and 7). Thehigh-angle group also exhibited more decrease in condylar volumethan the low-angle group (þ50.3 mm3, p ¼ 0.031). Among thepatients whose postoperative condylar volume was unchanged, nosignificant differences were noticed in skeletal relapse between thelow-angle and high-angle groups. However, among patients whodisplayed a postoperative decrease in condylar volume, the high-angle group had significantly more skeletal relapse in SN-Pog andat pogonion (in the horizontal direction) compared to the low-angle group (p < 0.05).

Prediction of condylar remodelling

Univariate regression analysis was applied to explain the post-operative decrease in condylar volume from a list of prognosticfactors (Table 8). The highest explained variance was 11% (preop-erative condylar volume). The age of patient at surgery, lingual splitpattern, flaring of mandibular angle, and vertical postoperativerelapse at pogonion were poor prognostic factors. Multivariateregression analysis involving all variables from Table 8 gave anexplained variance of 21% (adjusted r2), indicating that the preop-erative condylar volume, gender and downward displacement of

patients with and without a decrease in postoperative condylar volume.

Patients with a postoperative decrease in condylar volume

Male Female p Value

0.03 0.80 0.018*�0.12 1.17 0.038*0.76 0.70 0.920.87 2.29 0.007*0.15 1.67 0.005*0.50 2.08 0.028*0.80 2.10 0.12

r remodelling and skeletal relapse following bilateral sagittal split5), http://dx.doi.org/10.1016/j.jcms.2015.02.006

Table 6Number of condyles in the CR, CRm, CR1SD and CR2SD groups among patients with alow mandibular angle (<31.5�) and patients with a high mandibular angle (>31.5�).

Low-angle group High-angle group

CR group (n) 28 34CRm group (n) 7 15CR1SD group (n) 1 9CR2SD group (n) 0 4

Table 8Prognostic value of different variables in predicting condylar remodeling usingunivariate regression analyses.

Variable Univariate regression

Explained variancea P value

Gender (male/female) 8% <0.001Age (years) 0% 0.33Lingual split pattern (LSS) 0% 0.96Flaring at Go (mm) 0% 0.66Preoperative condylar volume (mm3) 11% <0.001Mandibular plane angle (�) 2% 0.10SN-Pog displacement (�) 6% 0.005SN-Pog relapse (�) 7% 0.003Pog displacement Horizontal (mm) 8% <0.001

Vertical (mm) 6% 0.005Pog relapse Horizontal (mm) 5% 0.008

Vertical (mm) 1% 0.27

a Explained variance ¼ adjusted R2.

T. Xi et al. / Journal of Cranio-Maxillo-Facial Surgery xxx (2015) 1e7 5

pogonion during surgery were the most important prognosticfactors.

Discussion

Morphological changes of the condyles are observed frequentlyfollowing the surgical advancement of the mandible (Arnett et al.,1996a,b; Hoppenreijs et al., 1999; Borstlap et al., 2004; Mottaet al., 2011; Kobayashi et al., 2012; Franco et al., 2013;Hoppenreijs et al., 2013). In contrast to the self-limiting physio-logical CR, PCR is characterized by severe morphological changes ofthe condylar shape, a reduction of condylar volume, and post-operative skeletal relapse (Chen et al., 2013; Hoppenreijs et al.,2013; Xi et al., 2013). Although many previous studies have char-acterised the extent of PCR following mandibular advancementsurgery, none of them have objectified the reduction in condylarvolume and its relationship with postoperative skeletal relapse. Tothe knowledge of the authors, the present study is the first that hasquantified the postoperative changes in condylar volume in a 3Dway and investigated the influences of these changes on post-operative skeletal relapse.

The strength of the current study is the application of CBCTscans to quantify the 3D postoperative changes of the facial skel-eton and condylar volume, and the use of a validated methodologyto measure condylar volume. The major advantage of CBCTcompared to conventional radiographs (orthopantomography andlateral cephalogrammetry) is the possibility to render 3D headmodels, allowing linear, angular, and volumetric measurements ofthe facial skeleton (Hoppenreijs et al., 1999; Plooij et al., 2009). Inthe present study, all cephalometric landmarks were identified byan experienced observer using the validated protocol described bySwennen et al. with reported intraobserver and interobserver errorof respectively 0.88 mm and 1.26 mm (Swennen et al., 2006).Cephalometric landmarks with poor reliability such as condylion,the position of which could be affected by condylar resorption,were therefore excluded in this study (de Oliveira et al., 2009).

The total measurement error for cephalometric measurementscould be regarded as the sum of errors caused by the superimpo-sition of CBCT scans and the not-fully-automatised identification of3D cephalometric landmarks. The absolute mean distance error

Table 7Mean postoperative skeletal relapse (mm and degrees) in the low-angle and high-angle gr

Mean postoperative skeletal relapse Patients with no postoperative decrease in con

Low-angle group High-angle group

SN-Pog (�) 0.16 0.46Pog horizontal (mm) 0.20 0.86Pog vertical (mm) 1.08 0.75S-Gomean (mm) 1.42 1.10Mandibular plane (�) 0.58 0.92Occlusal plane (�) 1.23 0.83Gonial angle (�) 1.83 2.78Decrease in condylar volume (mm3) 74.6 74.8

Low-angle group: mandibular plane angle <31.5� . High-angle group: mandibular plane*p < 0.05.

Please cite this article in press as: Xi T, et al., 3D analysis of condylaadvancement osteotomies, Journal of Cranio-Maxillo-Facial Surgery (201

induced by the voxel-based registration ranged from 0.05 to0.12mm, which is clinically negligible (Nada et al., 2011; Almukhtaret al., 2014). As far as the measurement error of condylar volumewas concerned, the mean intraobserver and interobserver varia-tions in the condylar volume were respectively 8.62 mm3 and6.13 mm3 (de Oliveira et al., 2009). Regarding a mean postoperativecondylar resorption of 105 mm3, it was unlikely that this relativelysmall measurement error in condylar volume would have influ-enced the results significantly.

In the literature, large surgical advancements, female gender,and a high mandibular plane angle are associated with a higherincidence of PCR (Scheerlinck et al., 1994; Arnett et al., 1996a,b;Hoppenreijs et al., 1999; Borstlap et al., 2004; Kobayashi et al.,2012). The mean mandibular advancement of 4.6 mm for the 56patients is comparable to previous studies on BSSO advancementwith rigid fixation (Mobarak et al., 2001; Eggensperger et al., 2006;Joss and Thuer, 2008; Nada et al., 2011; den Besten et al., 2013;Hoppenreijs et al., 2013). A mean postoperative relapse of0.71mm (15.5% of the surgical advancement) is in the middle rangeof the previously reported figures. No relationship was observedbetween the amount of advancement and skeletal relapse asdemonstrated in previous studies (Borstlap et al., 2004; Joss andThuer, 2008; Joss and Vassalli, 2009), probably due to the rela-tively mild mandibular advancement. Patients with a clinical sig-nificant relapse of more than 2 mm experienced significantly moredecrease in condylar volume than those with a relapse less than2 mm. This finding underlined a correlation between postoperativeskeletal relapse and a decrease in postoperative condylar volume.

A reduction in the postoperative condylar volume may beclassified as functional CR or as dysfunctional and pathological PCR(Joss and Vassalli, 2009). Although the cephalometric changes

oups among patients with and without a decrease in postoperative condylar volume.

dylar volume Patients with a postoperative decrease in condylar volume

p Value Low-angle group High-angle group p Value

0.32 0.29 0.93 0.017*0.27 0.17 1.50 0.009*0.55 0.98 0.50 0.290.40 1.66 2.26 0.180.47 1.03 1.62 0.200.44 1.15 2.24 0.0640.11 2.44 1.32 0.100.99 76.8 127.1 0.031*

angle >31.5� .

r remodelling and skeletal relapse following bilateral sagittal split5), http://dx.doi.org/10.1016/j.jcms.2015.02.006

T. Xi et al. / Journal of Cranio-Maxillo-Facial Surgery xxx (2015) 1e76

associated with CR and PRC were evaluated in many previousstudies (Hoppenreijs et al., 1999; Swennen and Schutyser, 2006;Joss and Vassalli, 2009; Xi et al., 2013), it is unclear to whatextent the condyle had to resorb until it could be categorised asPCR. The descriptive diagnose PCR in previous studies is based onsignificant reduction in volume of the condyle, reduction of themandibular ramus height and clinical relapse (Hoppenreijs et al.,1999). Data from Table 4 showed that when the reduction incondylar volume exceeded 17% of the preoperative condylar vol-ume, a significant relapse was noticed in the horizontal as well asvertical direction. Only in this small group did the anterior facialheight increase, indicating the onset of an anterior open bite,whereas the posterior facial height decreased, suggesting areduction in the condylar and/or ramus height, both of which wereindicative for PCR (Swennen and Schutyser, 2006; Joss and Vassalli,2009; Nada et al., 2011; Kobayashi et al., 2012). A postoperative CRofmore than 17% of the original condylar volume could therefore beindicative for PCR. However, the number of patients with PCR in thecurrent study might be too small to allow the establishment of areliable new radiological parameter to distinguish PCR from CR interms of condylar volume.

The incidence of PCR in the present study is 3.6%. This finding isin line with the incidence of 4% reported by Borstlap et al. in theirprospective, multi-center study involving patients who had un-dergone BSSO mandibular advancement (Borstlap et al., 2004). Itneeds to be noted that although most of the PCR were expected toinitiate during the first year after surgery, late PCR is possible afterthe first year (Arnett et al., 1996a,b; Hoppenreijs et al., 2013).Therefore, future prospective clinical studies on PCR with a largerpopulation and longer follow-up period are recommended.

A strong female predilection for condylar resorption was foundin several studies (Arnett et al., 1996a,b; Hoppenreijs et al., 1999;Gunson et al., 2012). There is evidence suggesting that the sexhormone 17b-estradiol has a regulatory effect on bone metabolismin the temporomandibular joint (Hajati et al., 2009; Gunson et al.,2012). The results of this study confirmed the female preponder-ance for postoperative condylar resorption following orthognathicsurgery. A higher percentage of female patients exhibited CR thanmale patients, and the reduction in condylar volume is significantlylarger among female patients. In addition, the preoperativecondylar volume of female patients was found to be significantlylower than that of male patients, as reported by Saccucci et al.(2012), which underlined an unfavourable baseline for female pa-tients even before surgery. The question now arises how thepostoperative relapse was influenced by gender. Although no sta-tistically significant differences were found between female andmale concerning skeletal relapse, Table 5 shows that the post-operative relapse was significantly higher in female patients, onlyamong those who exhibited postoperative CR. So it is unlikely thatthe female gender was the causal factor for postoperative relapse,but in combination with postoperative CR, it has certainly influ-enced the postoperative skeletal stability negatively.

The influence of the mandibular plane angle on postoperativerelapse has been reported in several studies (Hoppenreijs et al.,1999; Mobarak et al., 2001; Borstlap et al., 2004; Eggenspergeret al., 2006). The results from the present study also favoured thelow-angle group, who showed significant less postoperativerelapse than the high-angle group (0.20 mm vs. 1.24 mm). Afterexamining the data more closely in Table 6, it was evident that thepostoperative relapse differed between the high-angle and low-angle patients only among those who experienced a post-operative reduction in condylar volume. The mandibular angle didnot seem to affect relapse in patients with a stable condyle. Becausethe high-angle patients had smaller condyles than low-angle pa-tients, and also because the preoperative condylar volume was a

Please cite this article in press as: Xi T, et al., 3D analysis of condylaadvancement osteotomies, Journal of Cranio-Maxillo-Facial Surgery (201

prognostic factor for postoperative condylar resorption, the high-angle patients were more susceptible to CR and thus mightdisplay more skeletal relapse. This interesting finding suggests thata significant reduction in condylar volume could be a causal factorfor postoperative skeletal relapse, whereas the mandibular planeangle is merely an effect modification factor among those whoexhibit postoperative relapse.

Arnett et al. suggested that the torque on the proximal segmentsand subsequent compression on the condyles induced by theadvancement of the distal segment and rigid fixation created thepossibility for late condylar resorption and skeletal relapse (Arnettet al., 1996a,b). Although the intergonial distance was increased by6.35mm immediately after surgery, in the same range as reported byBecktor et al. (2002, 2008), the postoperative condylar volume,however, was not influenced by flaring at the mandibular angles.Franco et al. demonstrated, in their 3D study, that the greatest boneremodelling in the condylar areas occurred at the anterior and su-perior surfaces following BSSO mandibular advancement (Francoet al., 2013). Because the flaring of mandibular angles was observedin the transverse direction, it had probably induced only limitedlateral compression of the condyles, causing functional remodellingat the site of compression andboneappositionat siteswith a reducedloading, without changing the net condylar volume. The cephalo-metric measurements had also shown that flaring is mainly a tem-porary effect, with 50% of the original flaring disappearing by1 yearafter surgery. Therefore, it is unlikely that flaringwill induce PCR andthe subsequent skeletal instability as suggested by some earlierstudies (Arnett et al., 1996a,b; Becktor et al., 2002, 2008).

Conclusion

CBCT-based evaluation of changes in postoperative volumefollowing BSSO advancement surgery has demonstrated a positivecorrelation between postoperative skeletal relapse and a decreasein condylar volume. Female patients with a high mandibular anglewho exhibited a postoperative reduction in condylar volume wereparticularly at risk for postoperative relapse. Gender, preoperativecondylar volume, and magnitude of downward displacement ofpogonion were prognostic factors for postoperative CR. In cases inwhich the postoperative CR exceeded 17% of the preoperativecondylar volume, clinical signs of PCR were observed. Therefore,the condylar volume can be regarded as a useful radiographicparameter for the diagnosis and even prediction of postoperativeskeletal relapse and PCR in clinical practice.

Acknowledgements

This study was supported by a Mozaiek grant (017.005.006)from the Netherlands Organisation for Scientific Research, theNetherlands (NWO).

References

Almukhtar A, Ju X, Khambay B, McDonald J, Ayoub A: Comparison of the accuracy ofvoxel based registration and surface based registration for 3D assessment ofsurgical change following orthognathic surgery. PLoS One 9: e93402, 2014

Arnett GW, Milam SB, Gottesman L: Progressive mandibular retrusiondidiopathiccondylar resorption. Part I. Am J Orthod Dentofacial Orthop 110: 8e15, 1996a

Arnett GW, Milam SB, Gottesman L: Progressive mandibular retrusion-idiopathiccondylar resorption. Part II. Am J Orthod Dentofacial Orthop 110: 117e127,1996b

Becktor JP, Rebellato J, Becktor KB, Isaksson S, Vickers PD, Keller EE: Transversedisplacement of the proximal segment after bilateral sagittal osteotomy. J OralMaxillofac Surg 60: 395e403, 2002

Becktor JP, Rebellato J, Sollenius O, Vedtofte P, Isaksson S: Transverse displacementof the proximal segment after bilateral sagittal osteotomy: a comparison of lagscrew fixation versus miniplates with monocortical screw technique. J OralMaxillofac Surg 66: 104e111, 2008

r remodelling and skeletal relapse following bilateral sagittal split5), http://dx.doi.org/10.1016/j.jcms.2015.02.006

T. Xi et al. / Journal of Cranio-Maxillo-Facial Surgery xxx (2015) 1e7 7

Borstlap WA, Stoelinga PJ, Hoppenreijs TJ, van't Hof MA: Stabilisation of sagittal splitadvancement osteotomies with miniplates: a prospective, multicentre studywith two-year follow-up. Part IIIdcondylar remodelling and resorption. Int JOral Maxillofac Surg 33: 649e655, 2004

Chen S, Lei J, Wang X, Fu KY, Farzad P, Yi B: Short- and long-term changes ofcondylar position after bilateral sagittal split ramus osteotomy for mandibularadvancement in combination with Le Fort I osteotomy evaluated by cone-beamcomputed tomography. J Oral Maxillofac Surg 71: 1956e1966, 2013

de Oliveira AE, Cevidanes LH, Phillips C, Motta A, Burke B, Tyndall D: Observerreliability of three-dimensional cephalometric landmark identification on cone-beam computerized tomography. Oral Surg Oral Med Oral Pathol Oral RadiolEndod 107: 256e265, 2009

den Besten CA, Mensink G, van Merkesteyn JP: Skeletal stability after mandibularadvancement in bilateral sagittal split osteotomies during adolescence.J Craniomaxillofac Surg 41: e78ee82, 2013

Eggensperger N, Smolka K, Luder J, Iizuka T: Short- and long-term skeletal relapseafter mandibular advancement surgery. Int J Oral Maxillofac Surg 35: 36e42,2006

Franco AA, Cevidanes LH, Phillips C, Rossouw PE, Turvey TA, Carvalho Fde A, et al:Long-term 3-dimensional stability of mandibular advancement surgery. J OralMaxillofac Surg 71: 1588e1597, 2013

Gunson MJ, Arnett GW, Milam SB: Pathophysiology and pharmacologic control ofosseous mandibular condylar resorption. J Oral Maxillofac Surg 70: 1918e1934,2012

Hajati AK, Alstergren P, Nasstrom K, Bratt J, Kopp S: Endogenous glutamate in as-sociation with inflammatory and hormonal factors modulates bone tissueresorption of the temporomandibular joint in patients with early rheumatoidarthritis. J Oral Maxillofac Surg 67: 1895e1903, 2009

Hoppenreijs TJ, Maal T, Xi T: Evaluation of condylar resorption before and afterorthognathic surgery. Semin Orthod 19: 106e115, 2013

Hoppenreijs TJ, Stoelinga PJ, Grace KL, Robben CM: Long-term evaluation of patientswith progressive condylar resorption following orthognathic surgery. Int J OralMaxillofac Surg 28: 411e418, 1999

Hunsuck EE: A modified intraoral sagittal splitting technic for correction ofmandibular prognathism. J Oral Surg 26: 250e253, 1968

Joss CU, Thuer UW: Stability of the hard and soft tissue profile after mandibularadvancement in sagittal split osteotomies: a longitudinal and long-term follow-up study. Eur J Orthod 30: 16e23, 2008

Joss CU, Vassalli IM: Stability after bilateral sagittal split osteotomy advancementsurgery with rigid internal fixation: a systematic review. J Oral Maxillofac Surg67: 301e313, 2009

Please cite this article in press as: Xi T, et al., 3D analysis of condylaadvancement osteotomies, Journal of Cranio-Maxillo-Facial Surgery (201

Kim YJ, Lee Y, Chun YS, Kang N, Kim SJ, Kim M: Condylar positional changes up to 12months after bimaxillary surgery for skeletal class III malocclusions. J OralMaxillofac Surg 72: 145e156, 2014

Kobayashi T, Izumi N, Kojima T, Sakagami N, Saito I, Saito C: Progressive condylarresorption after mandibular advancement. Br J Oral Maxillofac Surg 50:176e180, 2012

Mobarak KA, Espeland L, Krogstad O, Lyberg T: Mandibular advancement surgery inhigh-angle and low-angle class II patients: different long-term skeletal re-sponses. Am J Orthod Dentofacial Orthop 119: 368e381, 2001

Motta AT, Cevidanes LH, Carvalho FA, Almeida MA, Phillips C: Three-dimensionalregional displacements after mandibular advancement surgery: one year offollow-up. J Oral Maxillofac Surg 69: 1447e1457, 2011

Nada RM, Maal TJJ, Breuning KH, Berge SJ, Mostafa YA, Kuijpers-Jagtman AM: Ac-curacy and reproducibility of voxel based superimposition of cone beamcomputed tomography models on the anterior cranial base and the zygomaticarches. PLoS One 6, 2011

Paniagua B, Cevidanes L, Walker D, Zhu H, Guo R, Styner M: Clinical application ofSPHARM-PDM to quantify temporomandibular joint osteoarthritis. ComputMed Imaging Graph 35: 345e352, 2011

Plooij JM, Naphausen MT, Maal TJ, Xi T, Rangel FA, Swennnen G, et al: 3D evaluationof the lingual fracture line after a bilateral sagittal split osteotomy of themandible. Int J Oral Maxillofac Surg 38: 1244e1249, 2009

Saccucci M, D'Attilio M, Rodolfino D, Festa F, Polimeni A, Tecco S: Condylar volumeand condylar area in class I, class II and class III young adult subjects. Head FaceMed 8: 34, 2012

Scheerlinck JP, Stoelinga PJ, Blijdorp PA, Brouns JJ, Nijs ML: Sagittal split advance-ment osteotomies stabilized with miniplates. A 2-5-year follow-up. Int J OralMaxillofac Surg 23: 127e131, 1994

Swennen GR, Schutyser F: Three-dimensional cephalometry: spiral multi-slice vscone-beam computed tomography. Am J Orthod Dentofacial Orthop 130:410e416, 2006

Swennen GR, Schutyser F, Barth EL, De Groeve P, De Mey A: A new method of 3-Dcephalometry. Part I: the anatomic Cartesian 3-D reference system. J CraniofacSurg 17: 314e325, 2006

Xi T, van Loon B, Fudalej P, Berge S, Swennen G, Maal T: Validation of a novel semi-automated method for three-dimensional surface rendering of condyles usingcone beam computed tomography data. Int J Oral Maxillofac Surg 42:1023e1029, 2013

r remodelling and skeletal relapse following bilateral sagittal split5), http://dx.doi.org/10.1016/j.jcms.2015.02.006