clinical and radiographic evaluation of a computer-generated guiding device in bilateral sagittal...

lable at ScienceDirect

Journal of Cranio-Maxillo-Facial Surgery xxx (2013) 1e9

Contents lists avai

Journal of Cranio-Maxillo-Facial Surgery

journal homepage: www.jcmfs.com

Clinical and radiographic evaluation of a computer-generated guidingdevice in bilateral sagittal split osteotomies

Ahmed Abdel-Moniem Barakat a, Adel Abou-ElFetouh a, *, Maha Mohammed Hakam a,Hesham El-Hawary a, Khaled Mahmoud Abdel-Ghany b

a Oral and Maxillofacial Surgery Department (Prof. Ragia Mounir), Faculty of Oral & Dental Medicine, Cairo University, 11 El-Saraya Street, El-Manial,11451 Cairo, Egyptb Central Metallurgical Research and Development Institute (Dr. Khaled Abdel-Ghany), El-Tebbeen, Helwan, 11421 Cairo, Egypt

a r t i c l e i n f o

Article history:Paper received 14 May 2013Accepted 28 August 2013

Keywords:Sagittal splitComputer guidedImage-guidedCondyle positionNerve injuryRapid prototypingCAD/CAMOrthognathic surgeryComputed tomography

* Corresponding author. Tel.: þ20 111 4408950.E-mail address: [email protected]

1010-5182/$ e see front matter � 2013 European Asshttp://dx.doi.org/10.1016/j.jcms.2013.08.007

Please cite this article in press as: Abdel-Monin bilateral sagittal split osteotomies, Journa

a b s t r a c t

The bilateral sagittal split osteotomy (BSSO) is one of the main orthognathic surgery procedures used formanaging skeletal mandibular excess, deficiency or asymmetry. It is known to be a technique-sensitiveprocedure with high reported incidences of inferior alveolar nerve injury, bad splits and post-surgicalrelapse. With the increasing use of computer-assisted techniques in orthognathic surgery, the accuratetransfer of the virtual plan to the operating room is currently a subject of research. This study evaluated theefficacy of computer-generated device at maintaining the planned condylar position and minimizinginferior alveolar nerve injury during BSSO. The device was used in 6 patients who required isolatedmandibular surgery for correction of their skeletal deformities. Clinical evaluation showedgood recoveryofthe maximal incisal opening and a reproducible occlusion in 5 of the 6 patients. Radiographic evaluationshowedbetter control of the condyle position in both the vertical and anteroposterior directions than in themediolateral direction. The degree of accuracy between the planned and achieved screw positions werejudged as good to excellent in all cases. Within the limitations of this study and the small sample size, theproposed device design allowed for good transfer of the virtual surgical plan to the operating room.

� 2013 European Association for Cranio-Maxillo-Facial Surgery. Published by Elsevier Ltd. All rightsreserved.

1. Introduction

The bilateral sagittal split osteotomy (BSSO) is one of the mainorthognathic surgery procedures used for managing skeletalmandibular excess, deficiency or asymmetry. It is known to be atechnique-sensitive procedure with high reported incidences ofinferior alveolar nerve injury (Bothur and Blomqvist, 2003;Teltzrow et al., 2005; D’Agostino et al., 2010; Iannetti et al., 2013),bad splits (Teltzrow et al., 2005; Chrcanovic and Freire-Maia, 2012)and post-surgical relapse (Ow and Cheung, 2009).

Post-surgical relapse has been classified into immediate anddelayed relapse. The immediate relapse has been mainly attributedto improper seating of the mandibular condyles in the glenoidfossae whereas delayed relapse has been attributed to unstableocclusion, inadequate fixation and condylar resorption. Variousauthors have stressed avoiding condylar displacement and havesuggested intraoperative guidance of the condylar position (Ayoub

g (A. Abou-ElFetouh).

ociation for Cranio-Maxillo-Facial

iem Barakat A, et al., Clinicall of Cranio-Maxillo-Facial Su

et al., 1997; Reyneke and Ferretti, 2002; Emshoff et al., 2003; Freyet al., 2007).

Different condylar positioning devices (CPDs) have been inves-tigated for effectiveness in re-establishing the preoperativecondylar position following BSSO with contradicting outcomes asregards to both skeletal stability and temporomandibular disorders(TMD). These devices ranged from intraoperative monitoring de-vices; such as infrared diodes (Bettega et al., 1996, 2002) or ultra-sonography (Gateno et al., 1993); to various instruments that attachthe proximal condyle-bearing segment to either osseous or dentalfixed distal anchoring sites or to the maxilla (Yagami and Nagumo,1996; Harada et al., 1997; Joos, 1999). Some clinicians have inves-tigated the feasibility of intraoperative awakening of the patientbefore fixation of the proximal segment (Politi et al., 2007).

Clinical investigations have shown a high incidence (40e85%) ofneurosensory deficits associated with BSSO (MacIntosh, 1981;Nishioka et al., 1987; Westermark et al., 1998). Possible causesinclude traction on the inferior alveolar nerve (IAN) during opera-tion, direct injury to the nerve when the ramus is split or the screwholes are drilled, and compression of the bony segments on the IANas a result of rigid fixation.

Surgery. Published by Elsevier Ltd. All rights reserved.

and radiographic evaluation of a computer-generated guiding devicergery (2013), http://dx.doi.org/10.1016/j.jcms.2013.08.007

mailto:[email protected]

www.sciencedirect.com/science/journal/10105182

http://www.jcmfs.com

http://dx.doi.org/10.1016/j.jcms.2013.08.007



A. Abdel-Moniem Barakat et al. / Journal of Cranio-Maxillo-Facial Surgery xxx (2013) 1e92

One study reported a 13% increase in the incidence of subjectivesensory changes when screws were used for fixation compared toMMF after BSSOs (Bouwman et al., 1995). Other authors reportedmore abnormalities of evoked potentials and pathological neuronalchanges when bicortical screws were used compared to mono-cortical screws and miniplate fixation in monkeys (Hu et al., 2007).

Emphasizing the negative effect of nerve compression, someauthors continuously recorded orthodromic sensory nerve actionpotentials (SNAPs) of the IAN in 20 patients with mandibular ret-rognathia during BSSO (Teerijoki-Oksa et al., 2002). Despite usingpositional screws and no visible damage to the nerve duringsplitting, the SNAP disappeared in four cases during fixation andupon tightening of the screws. In accordance with the above find-ings, another study retrospectively reviewed 68 patients (136 sites)who had a BSSO for mandibular advancement and were fixed usingeither wires, bicortical lag screws or monocortical mini-plates(Nesari et al., 2005). Of the sites that received lag screws, 34%showed neurosensory affection compared to only 15% of the sitesthat received monocortical plates at 2.5 years. The authors sug-gested that compression of the nerve during fixation is one majorreason for nerve dysfunction and that with bicortical screws, injuryto the inferior alveolar nerve can occur during drilling or screwplacement.

With the increasing popularity of computer-assisted orthog-nathic surgery, great interest has been shown in developing tech-niques to transfer the virtual plan into the operating room usingeither dynamic navigation techniques (Bell, 2011) or with the aid ofvarious positioning wafers and templates (Metzger et al., 2008; Baiet al., 2010, 2012; Shehab et al., 2013). There are only two reports onthe use of tooth-borne computer-generated CPDs for BSSO (Zinseret al., 2012; Polley and Figueroa, 2013).

In this study we aimed at investigating the use of a bone-bornecomputer-generated stent for maintaining the condylar positionand minimizing IAN injury during BSSO.

2. Material and methods

2.1. Patient selection

Patients were selected from the pool of patients presenting tothe departments of orthodontics and oral and maxillofacial surgery

Fig. 1. Point-based registration of the maxillary cast dentition to artifact-fre

Please cite this article in press as: Abdel-Moniem Barakat A, et al., Clinicalin bilateral sagittal split osteotomies, Journal of Cranio-Maxillo-Facial Su

at Cairo University with a skeletal class II or III malocclusion thatrequired isolated mandibular orthognathic surgery.

Clinical examination of the patients included assessment offacial symmetry and proportions, occlusal relationship andtemporomandibular joint function.

2.2. Preoperative planning

Where necessary, patients received orthodontic preparation,occlusal appliances to alleviate TMD at least 4 weeks prior to sur-gery and/or occlusal wafers to guide the condyles in their centricrelation if it did not coincide with the maximum intercuspationposition. Maxillary andmandibular impressions were taken using apolyvinyl siloxane impression material and casts poured in extra-hard stone. The casts were then mounted on a simple hinge artic-ulator in the final desired occlusion.

Preoperative radiographic records of the patients consisted of acomputed tomographic (CT) scan of the mandible acquired using acone beam CT scanner. The image acquisition parameters were asfollows: voxel size ¼ 0.3 mm, extended field of view(FOV) ¼ 17 � 23 cm, 120 KV, 5 mA and an exposure time of 4. Themaxillary and mandibular casts; in final occlusion; were then CBCTscanned according to the following protocol: FOV ¼ 8 � 8 cm and avoxel size of 0.125mm. All images were acquired in a digital DICOMformat.

The patients’ CBCTs were imported into surgical planning soft-ware (Mimics 10.0: Materialise, Leuven, Belgium). A segmentationprocesswas then carried out to select only the bony structures out ofit and calculate a 3D model. Virtual anatomic Frankfurt and Mid-sagittal planes were constructed and the 3D model aligned accord-ingly. The artifact dentitionwas separated from the skull model andreplaced with the 3Dmodel of the scanned casts. Themaxillary castwas registered to the maxillary CT dentition using a markerlessiterative closest point (ICP) registration method with a minimum of6 artifact-free dental landmarks identified on both the 2D and 3Dviews (Fig. 1). The same was repeated for the mandibular dentition.An artifact-free composite skull/dentitionmodelwas then achieved.

The medial, vertical (anterior) and connecting cuts were thensimulated on each side of the 3D mandible to complete a virtualBSSO. The right and left inferior alveolar nerves were thensegmented and their 3D paths generated.

e dental landmarks of the CBCT in the axial, coronal and sagittal views.


Fig. 2. The pre-osteotomy template consisted of a proximal and a distal part connectedtogether by 2 arms. Two 2.1 mm wide holes were incorporated into each part of thetemplate through which positioning screw holes were to be drilled. Fig. 4. A virtual 3D image of the distal segment and teeth after they have been

registered to the mandibular cast dentition in final occlusion. Note the proximal seg-ments (in orange and violet) still maintain their pre-osteotomy positions. The proximal(green) and distal (blue) parts of the template also maintain their relationship to theircorresponding mandibular segments.

Fig. 5. The locations of the fixation screws have been chosen so as to avoid injury ofthe inferior alveolar nerve in its new simulated position as dictated by the movementof the distal segment.

A. Abdel-Moniem Barakat et al. / Journal of Cranio-Maxillo-Facial Surgery xxx (2013) 1e9 3

A bone-supported template was then designed with one partfitting on the lateral surface of the proximal segment and engagingthe anterior border of the ramus and another part fitting on thelateral surface of the distal segment. Each part had two 2.1 mmholes to accommodate for monocortical 2.0 mm screws. Both partswere connected together using two arms (Fig. 2).

The mandibular CT dentition as well as the virtual distalsegment and the distal part of the template were then registered tothe final planned occlusion; as indicated by the scanned casts;utilizing the same markerless registration technique as mentionedbefore. The virtual proximal segments were now checked for anyinterference with the new position of the distal segment. If anyinterference existed, they were virtually repositioned in such a wayto eliminate the interference while avoiding excessive torqueing ordisplacement of the condyles (Fig. 3). If on the other hand a gapexisted between the virtual distal and proximal segments, the gapwas maintained to keep the condyles in their same preoperativepositions.

At least two new 4.0 mm holes were added to the proximal partof the template, positioned so as to allow for the screws to engageboth mandibular cortices and at the same time avoid inferioralveolar nerve injury. Two new connecting arms were then createdin between the new positions of the proximal and distal parts of thetemplate (Figs. 4 and 5).

A horse-shoe shaped virtual block was designed, aligned inbetween the virtual maxillary and mandibular casts and a digital

Fig. 3. 3D view of the distal segment in its final position with some interference with the lefbeen manipulated to alleviate the interference (in green). Note that the new position of th


subtraction process carried out to create the final wafer to be usedin the surgery.

A total of 5 parts were designed, pre-osteotomy and post-osteotomy templates for each side in addition to the interocclusal

t proximal segment in its preoperative position. To the right, the proximal segment hase proximal segment maintained the same vertical and AP positions.


Fig. 7. With the wafer in place and secured using MMF, the post-osteotomy templatewas secured to the proximal and distal segments using the same pre-drilled posi-tioning screw holes. The template repositioned the proximal segment to its preoper-ative position while the gap created between the segments was maintained.


wafer. All designs were exported in a stereolithographic (STL)format to a multi-jet modelling (thermal material application withultra-violet curing) additive manufacturing machine (InVision Si2,3D Systems e Rock Hill, SC) where they were fabricated from aplastic material (VisiJet SR 200, 3D Systems e Rock Hill, SC). Thisprocess was carried out at the Rapid Prototyping Department,Central Metallurgical Research and Development Institute e Cairo,Egypt.

2.3. Surgical procedure

The surgery was carried out under general anaesthesia andthrough standard intraoral vestibular incision to expose themandibular rami.

The preoperative template was placed laterally on the mandibleand oriented accurately in place so as to be in intimate contact withthe anterior border of the ramus and the buccal surface of themandible (Fig. 6). A 1.5 mm drill was used to drill holes for thepositioning screws through the template at the predefined loca-tions through both the proximal and distal parts. The template wasthen removed. The osteotomies were then completed and themandible split into proximal and distal segments. The wafer wasthen used to guide the distal segment in the new position and themaxillary and mandibular dentition were secured using MMF. Thepost-osteotomy template was then fixed to both the distal andproximal segments using the same previously drilled positioningscrew holes and monocortical 2.0 mm screws (Fig. 7). The templatenow holds the proximal segment in the same preoperative plannedposition while the distal segment is in its new position. The prox-imal segment was then fixed to the distal segment using bicortical2.0 mm screws placed through the planned 4.0 mm holes in the

Fig. 6. After complete reflection of the masseter muscle off the lateral surface of theramus, the pre-osteotomy template was placed engaging the anterior border of theramus for secure placement.


proximal part of the template. The monocortical template-positioning screws and the template were then removed.

After both sides were fixed, the MMF was released and the oc-clusion achieved compared to the planned occlusion. The incisionswere sutured using running 3/0 Vicryl sutures.

2.4. Clinical and radiographic follow-up

Patients were followed up as regards healing, occlusion, TMJfunction and inferior alveolar neurosensory function on the 2nd,7th and 14th day postoperatively and monthly thereafter for thenext 3 months. Patients were started on passive and active exer-cises to regain normal range of mouth opening on the 7th daypostoperatively. The MIO at 3 months was recorded. Inferior alve-olar neurosensory function was checked for pain perception bypricking using a dental probe at multiple points and for light touchperception using a cotton wisp. Orthodontic treatment wasresumed at the beginning of the 3rd month postoperatively.

A cone beam CT scan was done using the same preoperativeprotocol 3e7 days following the surgery. DICOM data were im-ported into the same surgical planning software; a 3D modelcalculated using the same previous steps and the pre- and post-operative CBCTs superimposed using a point registration methodbased on the anterior cranial base and/or maxillary dentition.

The deviation between the planned and achieved condylar po-sitions was measured on the 2D images in anteroposterior (AP),mediolateral and superoinferior (vertical) directions. On the axialviews, the cut with the greatest mediolateral dimension of thecondylar head was selected. A line representing the mid-sagittalplane was drawn passing through the nasal septum and the mid-point of the clivus in the skull base. Perpendicular distances weremeasured from the constructed mid-sagittal line to the medial poleof the pre- and post-operative condyles and the difference



calculated (Fig. 8). The axial condylar long axis passing from thelateral to medial poles of the condyle was drawn for both the pre-and post-operative condyles and its angle with themid-sagittal linemeasured and the difference calculated. In cases where the pre-operative condyle position has been intentionally modified toalleviate interferences between proximal and distal segments, thecontour of the repositioned condyle was used for themeasurements.

On the sagittal view, the cut showing both the condylar headand neck was selected. The sagittal long axes of the condyles wereconstructed through a point at the top of the condyles and passingparallel to the anterior border of the condylar neck, the angle be-tween both axes was then measured.

On the same sagittal cut, vertical condyle position changes weremeasured by extending lines from the summits of pre- and post-operative condyles to and perpendicular on the previously con-structed Frankfurt horizontal plane (Fig. 9). The difference in lengthbetween both lines was calculated. For the AP change, a coronalplane has been constructed to be perpendicular to the Frankfurtand mid-sagittal planes. Parallel lines were drawn from the tops ofboth condyles to the coronal plane and the difference in theirlengths calculated.

Fig. 9. Sagittal view showing vertical distances measured from the pre- and post-operative condyles to the Frankfurt plane.

Fig. 8. Axial view at the greatest mediolateral dimension of the right mandibularcondyle showing the mid-sagittal line (in red), the preoperative condyle position(in greyscale) and the contour of the superimposed postoperative condyle posi-tion (in blue). Distances were measured as two parallel lines (pink & yellow)from the medial poles of the condyles uptill and perpendicular on the mid-sagittal line.


2.5. Statistical analysis

Statistical analysis was performed using statistical software(SPSS (Statistical Package for the Social Sciences) version 15,Echosoft Corp., U.S.A). Data were represented as mean þ standarddeviation. Kurtosis and skewness showed that some of the vari-ables were not normally distributed. The one-sample T-test hasbeen used to compare variables to a fixed test value. Non-para-metric correlation coefficients were used to show relations be-tween variables. In all tests, the result was considered statisticallysignificant if the p-value was equal to or less than 0.05.

3. Results

3.1. Clinical findings

3.1.1. PreoperativeOur study included 6 patients, their ages ranged from 19 to 26

years with a mean of 21.3 years.No interference was detected between the virtual proximal and

distal segments in any of the patients except on one side of onepatient. The resultant gaps between the virtual proximal and distalsegments at the retromolar region measured 0.08e4.47 mmwith amean of 1.68 mm. The distance between the inferior alveolar canaland the inner aspect of the buccal cortex as measured at the ret-romolar region ranged from 1.09 to 4.96 mm with a mean of2.42 mm.

3.1.2. IntraoperativeIn all 6 cases, the BSSO cuts and splitting went uneventful. After

splitting, the inferior alveolar nerve was located in the distalsegment in all 12 surgical sites, however, the neurovascular bundlewas exposed but intact in one site; with 1.09 mm distance betweenthe buccal cortex and inferior alveolar neurovascular bundle; andpartially injured in another site; with 1.20 mm distance betweenthe buccal cortex and inferior alveolar neurovascular bundle.

The computer-generated rapid prototyped occlusal wafershowed perfect fit on both the upper and lower teeth in all patientsbut one. For that patient, the occlusal wafer required minorgrinding at 2 sites on its mandibular fitting surface before a properfit could be achieved.

One of the post-osteotomy templates fractured during theprocess, that condyle was manually positioned in place and theresults of this side excluded. The resultant gaps between bothsegments were clinically evaluated to be similar to those that wereplanned virtually in all surgical sites except for one. Three or fourbicortical 2.0 mm screws were used for fixation in all surgical sites,all screws that were placed through the pre-planned 4.0 mm holesin the proximal parts of the templates were found to engage bothbuccal and lingual cortices of the rami.

After the MMFwas released, 5 of the 6 patients had an occlusionexactly as planned virtually. One patient showed 2 mm midlineshift to the right andwas placed on guiding elastics postoperatively.

3.1.3. PostoperativeThe postoperative recovery and healing phasewas uneventful in

all patients. The pre-planned occlusion was maintained in all pa-tients except for one where guiding elastics were applied asmentioned before. At 3 months postoperatively, this patientcontinued to have a midline shift of 2 mm to the right and an openbite of 1 mm in the canine/premolar area of the left side.

All patients experienced limited mouth opening in the earlypostoperative phase that improved by time by active and passivemouth opening exercises. At 3 months postoperatively, the per-centage recovery of MIO ranged between 67.5% and 93.3% with a


Table 2One-sample T-test statistical analysis for the normally distributed condylar positionchanges variables.

N Min Max Mean Std.deviation

p-value(Testvalue ¼ 0)

Lineardisplacement(mm)

Mediolateral(Rt.)

5 1.35 4.00 2.6080 1.04889 0.005*

Mediolateral(Lt.)

6 0.56 2.34 1.3500 0.71425 0.006*

Vertical (Lt.) 6 0.00 1.61 0.6200 0.55624 0.041*AP (Rt.) 5 0.00 2.10 0.8080 0.87070 0.107

Angularrotation (�)

Axial longaxis (Lt.)

6 0.93 5.83 3.1967 1.87116 0.009*

Sagittal longaxis (Lt.)

6 1.68 6.53 3.3883 1.77719 0.005*

*Significant changes, p < 0.05.

Table 3


mean of 83.7%. All patients enjoyed an acceptable MIO except forthe patient that showed amidline shift who had a relatively limitedMIO of 27 mm at 3 months. None of the patients showed any signsof TMD postoperatively in terms of pain or TMJ sounds.

Comparison of the pre- and post-operative MIOs at 3 monthsusing the Wilcoxon Signed Ranks Test, revealed a statistically sig-nificant reduction in MIO with a p-value of 0.028.

In the early postoperative period, all patients reported symp-toms of inferior alveolar nerve dysfunction in terms of lower lip/chin numbness and were irresponsive to pain and light touch. Signsof nerve recovery could be appreciated gradually in all operatedsides. At 3 months postoperatively, 10 of the 12 involved inferioralveolar nerves showed signs of complete recovery. In one of the 2sites with incomplete recovery of neurosensory function the pa-tient still felt numbness but was responsive to pain and couldlocalize light touch. In the second of these 2 sites, the patient wasnot responsive to pain and could not localize light touch at 3months postoperatively.

One-sample T-test statistical analysis of the non-normally distributed condylarposition changes variables.

N Min Max Mean Std.deviation

p-value(Testvalue ¼ 0)

Lineardisplacement(mm)

Vertical (Rt.) 5 0.63 1.09 0.7800 0.18069 0.001*AP (Lt.) 6 0.45 2.34 1.2633 0.80817 0.012*

Angularrotation (�)

Axial longaxis (Rt.)

5 0.00 2.07 0.9080 0.94840 0.099

Sagittal longaxis (Rt.)

5 0.85 3.33 2.1700 1.07917 0.011*

*Significant changes, p < 0.05.

3.2. Radiographic findings

Superimposition of the post- and pre-operative CBCTs for the 6patients (11 condyles) revealed linear and angular differences be-tween the planned and achieved condylar positions.

In the mediolateral direction, 6 condyles were positionedlaterally by 0.56e4 mm (mean 1.77 mm) while 5 were positionedmedially by 0.99e3.2 mm (mean 2.096 mm). Vertically, 8 condyleswere positioned inferiorly by 0.48e1.61 mm (mean 0.82 mm), 2were positioned superiorly by 0.25e0.78 mm (mean 0.52 mm)while 1 condyle maintained its exact preoperative position. In theAP plane, 7 condyles were positioned anteriorly by 0.45e2.34 mm(mean 1.23 mm), 2 were positioned posteriorly by 0.92e2.10 mm(mean 1.51 mm) while 2 condyles maintained their exact preop-erative positions (Table 1).

For angular measurements and along the axial long axis, 6condyles rotated laterally by 0.34e5.83�(mean 2.84�), 4 rotatedmedially by 0.34e3.37� (mean 1.68�) while 1 condylemaintained itsexact preoperative position. Along the sagittal long axis, 7 condylesrotated superiorly (anticlockwise) by 0.85e3.84� (mean 2.37�)while 4 rotated inferiorly (clockwise) by 1.68e6.53� (mean 3.64�).

Analysis of the condylar position changes showed normal dis-tribution of the data for 6 of the 10 evaluated variables as follows:mediolateral displacement (Rt.), mediolateral displacement (Lt.),vertical displacement (Lt.), AP displacement (Rt.), axial long axisrotation (Lt.) and sagittal long axis rotation (Lt.). The one-sampleT-test showed insignificant condylar position changes i.e.adequate repositioning by the CPD for the AP displacement (Rt.)with a p-value of 0.107 (Table 2).

Variables that did not show a normal distribution were: verticaldisplacement (Rt.), AP displacement (Lt.), axial long axis rotation

Table 1Descriptive data of the right and left condylar position changes in the various planes for

ML displacement (mm) Axial long axis rotation (�) Vertical displace

Rt. Lt. Rt. Lt. Rt.

1 �4 �0.56 2.07 �4.15 1.092 N/A �2.34 N/A �5.83 N/A3 2.61 �1.74 0.34 �3.75 0.684 �1.35 �0.67 0 0.93 0.765 3.2 1.80 �0.34 3.37 0.746 1.88 0.99 �1.79 �1.15 0.63

A negative sign indicates a lateral, superior or posterior displacement.A negative sign indicates a lateral or upward rotation.ML ¼ Mediolateral, AP ¼ Anteroposterior, N/A ¼ not applicable (manually positioned co


(Rt.) and sagittal long axis rotation (Rt.). The one-sample T-testconducted for these variables showed insignificant condylar posi-tion changes i.e. adequate condylar repositioning using the CPD forthe axial long axis rotation (Rt.) (Table 3).

None of the postoperative CTs showed direct injury of theinferior alveolar canal by any of the bicortical screws used for fix-ation. The degree of accuracy between the planned and achievedscrew positions as evident on the superimposed pre- and post-operative CTs were judged as good to excellent in all cases (Figs. 10and 11).

4. Discussion

Point-based rigid technique was used to register 3D dentalmodels to patients’ CT models in this study. This is the sameregistration strategy used by several other authors (Gateno et al.,2003; Nkenke et al., 2004; Uechi et al., 2006; Swennen et al.,2007, 2009a). The accuracy of such a technique was proved to bein the sub-millimetre range which is still acceptable in terms of the

each patient.

ment (mm) Sagittal long axis rotation (�) AP displacement (mm)

Lt. Rt. Lt. Rt. Lt.

�0.25 �0.85 6.53 �0.92 2.34�0.78 N/A 3.02 N/A 1.941.61 �1.39 �3.84 �2.10 0.450 3.33 1.68 0 1.610.60 �3.15 �3.50 0 0.480.48 �2.13 �1.76 1.02 0.76

ndyle due to CPD fracture).


Fig. 10. Coronal sections showing relationship of the bicortical fixation screws to the inferior alveolar nerve (in orange).

Fig. 11. 3D view of the superimposed postoperative mandible showing the relation-ship between the planned fixation screw holes incorporated into the post-osteotomytemplate (in grey) and the actual screws location (projected in yellow).


observed clinical outcome. Using more advanced techniques suchas the sequential point- and surface-based registrations (Kim et al.,2010) or voxel-based registration (Swennen et al., 2009b); ifavailable; might have added to the overall accuracy of the tech-nique. Additionally, our registration technique was a markerlessone. It has been shown by others (Nkenke et al., 2004; Kim et al.,2010) that dental landmarks such as cusp tips or central groovescan be successfully used as registration points with high precisionevenwhen some of the teeth show beam-hardening artifacts due tometallic restorations or orthodontic brackets. More recently,anatomical landmarks have also been shown to offer good regis-tration accuracy during image-guided bimaxillary surgeries (Sunet al., 2012).

All the earlier CPDs reproduced the preoperative condylar po-sition exactly as it was with no allowance for possible interferencesbetween proximal and distal segments that might occur as a resultof the direction and magnitude of the planned distal segmentmovement. This raises a question about the value of such devices infacial asymmetry cases where such interferences are likely, espe-cially when large yaw movements are planned. The use of virtualsimulation allowed us to visualize and predict such interferencesbeforehand, so wewere able to design our CPD to accommodate for


such unavoidable condylar position changes or, when such changeswere deemed excessive we could then change the whole plan andopt for different types of osteotomies. One of our patients experi-enced such interference between the osteotomized segments, dueto the 7.3� planned yaw rotation, that the proximal segment on theleft side was manipulated to alleviate it.

The results obtained for that patient emphasize the negativeeffects of condylar torque on the TMJ. Her planned 7.3� medialrotation of the left condyle in addition to a further 3.37� as a resultof insufficient repositioning by the CPD on that side, resulted in theleast percentage recovery of MIO (67.5%) among all patients. Inaddition, this was the only patient among the group that showed adeviation from the planned occlusion that required prolongedelastic guidance and orthodontic treatment for correction. Suchresults are consistent with those of others who reported 62.65%mandibular motion recovery when the condyles were repositionedin the sagittal plane only with no torque control compared to a77.58% when they were repositioned all 3 planes of space (Bettegaet al., 2002). When such condylar position modifications wereinevitable, we tried our best to minimize the condylar torque sinceit has been related by some authors to TMD and limited mouthopening (Bettega et al., 2002; Gerressen et al., 2006). Under allcircumstances no vertical or anteroposterior position changes wereundertaken. These directions of movement were shown to be acommon cause for immediate as well as late relapses due tocondylar resorption and remodelling (Ayoub et al., 1997; Reynekeand Ferretti, 2002; Frey et al., 2007).

The need to trim interferences between the split segments toensure proper condylar positioning, especially in asymmetric cases,has been emphasized; even with the concomitant use of CPDs(Harada et al., 1997). It has been shown that asymmetry cases wereparticularly accompanied by higher condylar position changes(Ueki et al., 2012). A very interesting technique to minimize suchinterferences was described by osteotomizing the distal segmentbehind the last molar thus allowing passive alignment of theproximal segment (Iwai et al., 2012). Such a technique can becombined with virtual simulation and computer-generated tem-plates to ensure accurate condylar positioning while minimizingthe need for manipulating the proximal segment to alleviateinterferences.

As previously reported, higher incidences of gaps being createdwere found when the proximal segment position was controlled inall three dimensions of space (Bettega et al., 2002). One might



argue that the gap we intentionally maintained between theproximal and distal segments; if present; could have resulted ininstability of the segments or allowed for fibrous ingrowthwith lessthan optimum bone healing, but in all six cases, the gap did notexceed 4.47 mm. Similar gaps were also left ungrafted in previousreports (Kang et al., 2010). None of our patients experienced anyhealing problems and postoperative function was resumed fromday 1 after surgery with no periods of MMF. If the virtual planninghad revealed larger gaps, we might have thought of using differentosteotomies, slightly adjusting the proximal segment to minimizethe gap or bone grafting such gaps.

Our CPD design allowed us to minimize potential inferior alve-olar nerve injury in two ways. First of all, maintaining the gap be-tween proximal and distal segments, if present, prevented possiblecompression of the nerve. This has been shown by other authors toresult in diminished sensory nerve action potential (Teerijoki-Oksaet al., 2002). The 4.0 mm screw holes incorporated into our CPDallowed us to locate the bicortical fixation screws safely away fromthe inferior alveolar bundle in its virtually simulated position.Bicortical positioning screws were shown to cause more inferioralveolar nerve dysfunctionwhen compared to monocortical screwsand plates, probably due to higher incidences of hitting the nervedirectly during drilling (Teerijoki-Oksa et al., 2002; Nesari et al.,2005).

Only 2 of our patients (2 out of 12 operated sites) reported signsand symptoms of inferior alveolar neurosensory dysfunction, one ofthem with partial recovery at 3 months postoperatively. In bothpatients, the inferior alveolar canal was located relatively close tothe buccal cortex of the ramus (1.20 & 1.09 mm as measured on CT)and the neurovascular bundles were encountered during splitting,with one partially severed. We do believe that such neurosensorydysfunction was due to direct injury of the nerves rather than dueto compression between the proximal and distal segments or injuryby the bicortical screws.

In accordance with other authors (Metzger et al., 2008; Hsuet al., 2013), we found the computer-generated occlusal wafersoffered similar, and in many cases superior, fit when compared toconventional acrylic splints. Combining the computer-generatedCPD and occlusal wafer provided us with excellent outcomes in 5of the 6 patients regarding the postoperative occlusion and itsreproducibility. The occlusion achieved was judged clinically asexactly similar to that planned and we could easily guide the teethinto the splint while taking the postoperative CBCT scan. The onlypatient that required guiding elastics postoperatively was the pa-tient where excessive torqueing of the left condyle occurred. Ourresults in this aspect are quite similar to others (Bettega et al., 2002;Renzi et al., 2003).

Most of our patients showed inferior, posterior and lateraldisplacement in addition to posterior rotation. It seems that ourCPDwasmore effective in controlling the proximal segment in boththe vertical and AP directions than in the mediolateral direction.Our results are in agreement with others who found that the Luhrdevice was unable to prevent condylar rotation or displacement(Rotskoff et al., 1991). One possible cause might be the heavy pullexerted by the lateral pterygoid muscles onto the condyles. Anotherexplanation might be that the screws showed some resiliency and,being applied at the anterior part of the ramus, they were unable toprovide adequate 3D control of the proximal segment. Extendingthe proximal part of our CPD more posteriorly or using more solidmaterials for its construction might have improved the results.

To the best of our knowledge, only a few studies report on theuse of CAD/CAM templates to maintain the condyles in placefollowing BSSO. In one of these studies (Zinser et al., 2012), the CPDdesign seemed to be more capable of reproducing the preoperativecondylar position, with less than 0.19 mm linear deviations as


measured from the coronal and Frankfurt planes; while in the otherstudy (Polley and Figueroa, 2013); no reports on the accuracy oftheir device was provided. This higher accuracy might be becausethese CPDs were teeth- rather than bone-supported, where thedetailed dental anatomy offers a more solid and reproducibleorientation of the device.

5. Conclusions

Within the limitations of this study and the small sample size,the proposed guiding device design offered very good clinicaloutcomes in terms of occlusion and postoperative TMD while itoffered less than optimal condylar repositioning as evaluatedradiographically. The CPD described showed better linear condylarrepositioning in the vertical and anteroposterior directions than inthemediolateral direction. Angular condylar position changes weremore difficult to control. It also showed good control over thelocation of bicortical fixation screws and maintained gaps betweenproximal and distal segments which resulted in a low incidence ofinferior alveolar nerve neurosensory dysfunction.

Ethical approvalThis research has been approved by the ethical committee at the

Faculty of Oral and Dental Medicine, Cairo University. Surgerieshave been carried out only after an informed consent has beensigned by all patients.

FundingNone.

Conflict of interest statementNone declared.

Acknowledgement

Special thanks goes to Dr. Amr Maher; professor of anaesthesi-ology at the Faculty of Medicine e Cairo University, for his greatefforts preparing the statistics for this piece of work.

References

Ayoub AF, Trotman CA, Stirrups DR, Wilmot JJ: Stability of bimaxillary osteotomyfollowing surgical correction of class II skeletal deformities: a two-centre study.Br J Oral Maxillofac Surg 35: 107e115, 1997

Bai S, Bo B, Bi Y, Wang B, Zhao J, Liu Y, et al: CAD/CAM surface templates as analternative to the intermediate wafer in orthognathic surgery. Oral Surg OralMed Oral Pathol Oral Radiol Endod 110: e1e7, 2010

Bai S, Shang H, Liu Y, Zhao J, Zhao Y: Computer-aided design and computer-aidedmanufacturing locating guides accompanied with prebent titanium plates inorthognathic surgery. J Oral Maxillofac Surg 70: 2419e2426, 2012

Bell RB: Computer planning and intraoperative navigation in orthognathic surgery.J Oral Maxillofac Surg 69: 592e605, 2011

Bettega G, Dessenne V, Raphaël B, Cinquin P: Computer-assisted mandibularcondyle positioning in orthognathic surgery. J Oral Maxillofac Surg 54: 553e558, 1996

Bettega G, Cinquin P, Lebeau J, Raphaël B: Computer-assisted orthognathic surgery:clinical evaluation of a mandibular condyle repositioning system. J Oral Max-illofac Surg 60: 27e34, 2002

Bothur S, Blomqvist JE: Patient perception of neurosensory deficit after sagittal splitosteotomy in the mandible. Plast Reconstr Surg 111: 373e377, 2003

Bouwman JP, Husak A, Putnam GD, Becking AG, Tuinzing DB: Screw fixationfollowing bilateral sagittal ramus osteotomy for mandibular advancement e

complications in 700 consecutive cases. Br J Oral Maxillofac Surg 33: 231e234,1995

Chrcanovic BR, Freire-Maia B: Risk factors and prevention of bad splits duringsagittal split osteotomy. Oral Maxillofac Surg 16: 19e27, 2012

D’Agostino A, Trevisiol L, Gugole F, Bondí V, Nocini PF: Complications of orthog-nathic surgery: the inferior alveolar nerve. J Craniofac Surg 21: 1189e1195, 2010

Emshoff R, Scheiderbauer A, Gerhard S, Norer B: Stability after rigid fixation ofsimultaneous maxillary impaction and mandibular advancement osteotomies.Int J Oral Maxillofac Surg 32: 137e142, 2003


http://refhub.elsevier.com/S1010-5182(13)00250-3/sref1








































Frey DR, Hatch JP, Van Sickels JE, Dolce C, Rugh JD: Alteration of the mandibularplane during sagittal split advancement: short- and long-term stability. OralSurg Oral Med Oral Pathol Oral Radiol Endod 104: 160e169, 2007

Gateno J, Miloro M, Hendler BH, Horrow M: The use of ultrasound to determine theposition of the mandibular condyle. J Oral Maxillofac Surg 51: 1081e1086, 1993

Gateno J, Xia J, Teichgraeber JF, Rosen A: A new technique for the creation of acomputerized composite skull model. J Oral Maxillofac Surg 61: 222e227, 2003

Gerressen M, Zadeh MD, Stockbrink G, Riediger D, Ghassemi A: The functional long-term results after bilateral sagittal split osteotomy (BSSO) with and without acondylar positioning device. J Oral Maxillofac Surg 64: 1624e1630, 2006

Harada K, Ono J, Okada Y, Nagura H, Enomoto S: Postoperative stability after sagittalsplit ramus osteotomy with condylar-positioning appliance and screw fixation:asymmetric versus symmetric cases. Oral Surg Oral Med Oral Pathol Oral RadiolEndod 83: 532e536, 1997

Hsu SS, Gateno J, Bell RB, Hirsch DL, Markiewicz MR, Teichgraeber JF, et al: Accuracyof a computer-aided surgical simulation protocol for orthognathic surgery: aprospective multicenter study. J Oral Maxillofac Surg 71: 128e142, 2013

Hu J, Zhao Q, Tang J, Zheng Z, Qi MC: Changes in the inferior alveolar nervefollowing sagittal split ramus osteotomy in monkeys: a comparison of mono-cortical and bicortical fixation. Br J Oral Maxillofac Surg 45: 265e271, 2007

Iannetti G, Fadda TM, Riccardi E, Mitro V, Filiaci F: Our experience in complicationsof orthognathic surgery: a retrospective study on 3236 patients. Eur Rev MedPharmacol Sci 17: 379e384, 2013

Iwai T, Maegawa J, Aoki S, Tohnai I: Ultrasonic vertical osteotomy of the distalsegment for safe elimination of interference between the proximal and distalsegments in bilateral sagittal split osteotomy for mandibular asymmetry. Br JOral Maxillofac Surg 51: e192ee194, 2013

Joos U: An adjustable bone fixation system for sagittal split ramus osteotomy:preliminary report. Br J Oral Maxillofac Surg 37: 99e103, 1999

Kang MG, Yun KI, Kim CH, Park JU: Postoperative condylar position by sagittal splitramus osteotomy with and without bone graft. J Oral Maxillofac Surg 68: 2058e2064, 2010

Kim BC, Lee CE, Park W, Kang SH, Zhengguo P, Yi CK, et al: Integration accuracy ofdigital dental models and 3-dimensional computerized tomography images bysequential point- and surface-based markerless registration. Oral Surg Oral MedOral Pathol Oral Radiol Endod 110: 370e378, 2010

MacIntosh RB: Experience with the sagittal osteotomy of the mandibular ramus: a13-year review. J Maxillofac Surg 9: 151e165, 1981

MetzgerMC, Hohlweg-Majert B, Schwarz U, TeschnerM, Hammer B, Schmelzeisen R:Manufacturing splints for orthognathic surgery using a three-dimensionalprinter. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 105: e1ee7, 2008

Nesari S, Kahnberg KE, Rasmusson L: Neurosensory function of the inferior alveolarnerve after bilateral sagittal ramus osteotomy: a retrospective study of 68 pa-tients. Int J Oral Maxillofac Surg 34: 495e498, 2005

Nishioka GJ, Zysset MK, Van Sickels JE: Neurosensory disturbance with rigid fixationof the bilateral sagittal split osteotomy. J Oral Maxillofac Surg 45: 20e26, 1987

Nkenke E, Zachow S, Benz M, Maier T, Veit K, Kramer M, et al: Fusion of computedtomography data and optical 3D images of the dentition for streak artefactcorrection in the simulation of orthognathic surgery. Dentomaxillofac Radiol33: 226e232, 2004

Ow A, Cheung LK: Skeletal stability and complications of bilateral sagittal splitosteotomies and mandibular distraction osteogenesis: an evidence-based re-view. J Oral Maxillofac Surg 67: 2344e2353, 2009


Politi M, Toro C, Costa F, Polini F, Robiony M: Intraoperative awakening of the pa-tient during orthognathic surgery: a method to prevent the condylar sag. J OralMaxillofac Surg 65: 109e114, 2007

Polley JW, Figueroa AA: Orthognathic positioning system: intraoperative system totransfer virtual surgical plan to operating field during orthognathic surgery.J Oral Maxillofac Surg 71: 911e920, 2013

Renzi G, Becelli R, Di Paolo C, Iannetti G: Indications to the use of condylar repo-sitioning devices in the surgical treatment of dental-skeletal class III. J OralMaxillofac Surg 61: 304e309, 2003

Reyneke JP, Ferretti C: Intraoperative diagnosis of condylar sag after bilateralsagittal split ramus osteotomy. Br J Oral Maxillofac Surg 40: 285e292, 2002

Rotskoff KS, Herbosa EG, Villa P: Maintenance of condyle-proximal segment posi-tion in orthognathic surgery. J Oral Maxillofac Surg 49: 2e7, 1991

Shehab MF, Barakat AA, AbdElghany K, Mostafa Y, Baur DA: A novel design ofa computer generated splint for vertical repositioning of the maxilla afterLe Fort I osteotomy. Oral Surg Oral Med Oral Pathol Oral Radiol 115: e16e25,2013

Sun Y, Luebbers HT, Agbaje JO, Schepers S, Vrielinck L, Lambrichts I, et al: Validationof anatomical landmarks-based registration for image-guided surgery: an in-vitro study. J Craniomaxillofac Surg 41: 522e526, 2013

Swennen GR, Barth EL, Eulzer C, Schutyser F: The use of a new 3D splint and doubleCT scan procedure to obtain an accurate anatomic virtual augmented model ofthe skull. Int J Oral Maxillofac Surg 36: 146e152, 2007

Swennen GR, Mommaerts MY, Abeloos J, De Clercq C, Lamoral P, Neyt N, et al:A cone-beam CT based technique to augment the 3D virtual skull model with adetailed dental surface. Int J Oral Maxillofac Surg 38: 48e57, 2009a

Swennen GR, Mollemans W, De Clercq C, Abeloos J, Lamoral P, Lippens F, et al:A cone-beam computed tomography triple scan procedure to obtain a three-dimensional augmented virtual skull model appropriate for orthognathic sur-gery planning. J Craniofac Surg 20: 297e307, 2009b

Teltzrow T, Kramer FJ, Schulze A, Baethge C, Brachvogel P: Perioperative compli-cations following sagittal split osteotomy of the mandible. J CraniomaxillofacSurg 33: 307e313, 2005

Teerijoki-Oksa T, Jääskeläinen SK, Forssell K, Forssell H, Vähätalo K, Tammisalo T,et al: Risk factors of nerve injury during mandibular sagittal split osteotomy. IntJ Oral Maxillofac Surg 31: 33e39, 2002

Uechi J, Okayama M, Shibata T, Muguruma T, Hayashi K, Endo K, et al: A novelmethod for the 3-dimensional simulation of orthognathic surgery by using amultimodal image-fusion technique. Am J Orthod Dentofacial Orthop 130:786e798, 2006

Ueki K, Moroi A, Sotobori M, Ishihara Y, Marukawa K, Yoshizawa K, et al: Changes intemporomandibular joint and ramus after sagittal split ramus osteotomy inmandibular prognathism patients with and without asymmetry.J Craniomaxillofac Surg 40: 821e827, 2012

Westermark A, Bystedt H, von Konow L: Inferior alveolar nerve function aftermandibular osteotomies. Br J Oral Maxillofac Surg 36: 425e428, 1998

Yagami K, Nagumo M: A transoral approach for three-dimensional repositioning ofthe proximal segment after mandibular sagittal split ramus osteotomy. J OralMaxillofac Surg 54: 1256e1258, 1996

Zinser MJ, Mischkowski RA, Sailer HF, Zöller JE: Computer-assisted orthognathicsurgery: feasibility study using multiple CAD/CAM surgical splints. Oral SurgOral Med Oral Pathol Oral Radiol 113: 673e687, 2012












































































































































clinical and radiographic evaluation of a computer-generated guiding device in bilateral sagittal...

Documents