Available online at www.sciencedirect.com
British Journal of Oral and Maxillofacial Surgery 52 (2014) 251–257
Computer-assisted orthognathic surgery for correction of facial asymmetry: results of a randomised controlled clinical trial Giacomo De Riu a,∗ , Silvio Mario Meloni a , Alessandro Baj b , Andrea Corda a , Damiano Soma a , Antonio Tullio a a b
Maxillofacial Surgery Unit, University Hospital of Sassari, Italy Maxillofacial Surgery Unit, Ospedale Maggiore Policlinico, Milan, Italy
Accepted 13 December 2013 Available online 10 January 2014
Abstract In this randomised controlled clinical trial, 2 homogeneous groups of patients with facial asymmetry (n = 10 in each) were treated by either classic or computer-assisted orthognathic corrective surgery. Differences between the 2 groups in the alignment of the lower interincisal point (p = 0.03), mandibular sagittal plane (p = 0.01), and centring of the dental midlines (p = 0.03) were significant, with the digital planning group being more accurate. © 2013 The British Association of Oral and Maxillofacial Surgeons. Published by Elsevier Ltd. All rights reserved. Keywords: Computer-assisted orthognathic surgery; Facial asymmetry; Orthognathic planning
Introduction The correct planning of successful correction of facial asymmetry can be complex. In clinical practice it is often difficult to individualise the midline facial landmarks on a standard frontal cephalogram because of the superimposition of different cranial structures. Midlines can therefore be adopted arbitrarily to assess the asymmetry of the face, which can lead to imprecise evaluation. Classical surgical simulation using plaster casts on semiindividual articulators is also imprecise. The alignment of the midlines on the articulator is hindered by many errors inherent in the planning of the classic model operation.1–6
∗ Corresponding author at: Unità Operativa di Chirurgia Maxillo-Facciale, AOU Sassari, Viale San Pietro 43/b, 07100 Sassari, Italy. Tel.: +39 079228216; fax: +39 079229002. E-mail addresses: [email protected], [email protected] (G. De Riu).
Planning the correction of asymmetry is more difficult than planning the correction of sagittal deformities because of the complex management of different movements of the plaster casts (roll, yaw, and lateral translation) that affect the position and inclination of the midlines and occlusal planes. Correct model planning of symmetrical maxillomandibular repositioning is also difficult because of the inaccuracy of linear and angular measurements, usually made between pencil marks on the rough surface of the plaster casts. Surgical planning of the correction of facial asymmetry is therefore challenging, but the difficulties lie mainly in the out-of-date and imprecise tools that we use during the diagnostic and simulation phases, which result in the incorporation of various errors in the intermediate splint. In many cases, despite meticulous planning, residual asymmetry is common. We wondered if digital planning would give more accurate results.7,8 Several currently available programmes permit virtual operations with digital control of the 3-dimensional movements of the maxilla, and the computer-aided design
0266-4356/$ – see front matter © 2013 The British Association of Oral and Maxillofacial Surgeons. Published by Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.bjoms.2013.12.010
252
G. De Riu et al. / British Journal of Oral and Maxillofacial Surgery 52 (2014) 251–257
and machining (CAD/CAM) of intermediate splints.9 These programmes can overcome the technical problems of management of the spatial positions of the jaws with reference to the soft tissues and craniofacial midlines. The purpose of this randomised, controlled, clinical study was to evaluate the most accurate procedure for orthognathic correction of facial asymmetry. We compared two different methods of surgical planning: classic and computerassisted.10 The specific aims of the study were to measure and compare the rates of alignment and reduction of cant of the dental and facial midlines between the 2 groups.
Patients, material, and methods Design of the study We designed a randomised, controlled trial (ClinicalTrials.gov id. NCT01879969) that was conducted according to the Helsinki Declaration of 1975, and approved by the Research Committee (IRB) of the Department of Surgical, Microsurgical and Medical Science, University of Sassari, Italy. We studied 20 patients who presented to the Maxillofacial Surgery Unit, University Hospital of Sassari, for evaluation and management of facial asymmetry or lateral deviation of the mandible. The randomisation codes were created (Excel, Microsoft, Redmond, WA, USA), by combining a sequence of randomised non-consecutive numbers to match the 2 procedures (classic or digital planning). The codes were assigned by an independent operator who was not involved in the trial, and were placed in envelopes. Data were then collected in sheets (Excel). All participants provided written informed consent before enrollment, after which they were treated by orthognathic repositioning of the maxilla and mandible with or without genioplasty. All operations were planned and done by the same surgeon between November 2010 and December 2012. The inclusion criteria were: cant of the occlusal plane of more than 3◦ or midline discrepancies of more than 2.5 mm, or both; the presence of all central incisors; radiographs and plaster casts taken before and after the operation (classic group), or cranial cone-beam computed tomographic (CT) images (digital group), together with digital photographs taken before and after the operation. Patients were excluded if they had had previous trauma to the hard or soft facial tissues, functional deviation of the mandible, or their records were incomplete (cone-beam CT, radiographs, casts, or digital photographs). Variables The primary predictive variable was method of planning treatment. Each eligible patient was randomly selected to have bimaxillary surgery planned with a classic, standard
procedure or using computer-assisted orthognathic surgery. The primary outcome variables were linear and angular measures that defined the alignment of the facial midlines, or reduction of maxillary and mandibular cant in the 2 groups. Technique of classic planning The deformity for each patient was diagnosed after careful analysis of frontal facial aesthetics,11–13 compared with the cephalometric findings.14 The plaster casts were mounted on an average value, semi-individual articulator (SAM II, SAM Präzisionstechnik GmbH, Munich, Germany). The vertical and mediolateral relations of the teeth were recorded by drawing vertical reference lines on the upper model and horizontal reference lines at a calculated distance. With these reference points, the precise movements of the maxillary segments were calculated in the 3 planes of space with model surgery.15 The models were repositioned based on data obtained with the surgical Visual Treatment Objective.16 Surgical simulation allows the manufacture of the resin occlusal intermediate splint that records the occlusal relations after the upper model has been moved and the mandible is still in the preoperative position, which serves as a guide during the operation. In some cases (such as those with class II, short face, or asymmetric vertical deficiency of the face) we prefer to move the mandible first, and inverted planning is used. Technique of digital planning We used cone-beam CT data (KaVo, Biberach, Germany) and the Maxilim software (Medicim, Nobel Biocare Group). The protocol for acquiring CT images involved a first scan of the patient in centric occlusion with relaxed lips, a second scan at low resolution and low-dose biting of a double tray, and a third scan of that tray to record the occlusion at high resolution.10 The total radiation dose to the patient was ∼100 Sv. Once the CT images had been acquired on to the software, a computer-assisted procedure allowed the development of a virtual 3-dimensional model of the hard and soft tissues of the face, with a better-defined window of the teeth and intercuspation (augmented model). A complete set of 2-dimensional photographs of the face captured with any digital camera was used to add information about the texture of the skin. This process created the patient’s face mask. The relations between the soft and hard tissues of the face were then shown, setting the transparency of the skin texture, and permitting the measurement of the maxillary vertical dimension compared with the length of the upper lip, or definition of deviation of the dental midline from the philtrum (Fig. 1). Next, we made a 3-dimensional cephalometric analysis. Lateral and frontal cephalograms were developed from the CT data and connected to the surface of the hard and soft tissues of the virtual model. Once all the landmarks had
G. De Riu et al. / British Journal of Oral and Maxillofacial Surgery 52 (2014) 251–257
253
Fig. 1. Three-dimensional facial planning showing the relations between the hard and soft tissues.
been identified, the angular and linear measurements were calculated automatically. Planning of treatment included virtual osteotomies on the maxilla, mandible, and chin. The operator chose whether to move the maxilla or the mandible first. Once the maxilla (or mandible) had been placed in the desired position, a special tool17 was used to move the opposite jaw and to define the best occlusion for the patient through a 3-dimensional view of the dental arches and an occlusogram. The dental arches were guided to occlusion with virtual rubber bands that allowed for the control of intercuspation. The occlusogram also quantified the effective area of contact between the teeth, and identified areas where grinding of the cusp would be necessary.10,18 In the case of a multisegmented maxilla, the standard virtual procedure was edited to individualise every piece of bone with at least 3 landmarks. This process enabled us to fix the multipiece arch in the right occlusion. Once the final occlusion had been decided, the mandible (or maxilla in inverted planning) was moved. Finally, the chin was operated on if necessary, although no surgical guide was provided. The soft tissues were remodelled automatically: the algorithm for soft tissue to bone movement simulates the aesthetic changes, which helps to predict the outcome. Once we were satisfied with the planned movements, an output file was saved and uploaded to the server for the manufacture of a CAD/CAM surgical splint, which arrived in a few days by mail.
facial midline, and between the maxillary sagittal plane, the mandibular sagittal plane and the ideal midsagittal plane. The distance between the upper and lower interincisal points was also evaluated. An ideal midline and ideal midsagittal plane were considered to be the vertical line and the vertical plane that passed by the apophysis crista galli, sella, and nasion for the hard tissues, and passed midway between the left and right internal canthi and the middle of the lip for the soft tissues. In the classic group, measurements were made before and after operation on posteroanterior cephalograms and digital photographs. In the digital group the measurements were obtained by comparing the Swennen modified 3-dimensional cephalometry taken before and after the operation.19 Analysis of data The data were recorded on a spreadsheet (Excel; Microsoft Corporation, Redmond, WA, USA), and the significances of differences between the groups were assessed using Student’s test with the help of Stata software (version 9.1, StataCorp, TX, USA). Postoperative clinical and radiographic improvement in the alignment of anatomical points for each subject and the mean improvement for each group were calculated. Probabilities of less than 0.05 were accepted as significant, and all data are presented as mean (SD) with 95% CI.
Results Collection of data In both groups measurements were made to define linear and angular discrepancies between upper and lower interincisal points, skeletal menton, soft tissue menton, and the ideal
The classic group comprised 7 women and 3 men aged from 21 to 54 years old). An acrylic intermediate splints was constructed. The digital group consisted of 7 men and 3 women aged 24–47 years old) and their intermediate surgical
254
G. De Riu et al. / British Journal of Oral and Maxillofacial Surgery 52 (2014) 251–257
Fig. 4. Three-dimensional planning.
Fig. 2. Facial asymmetry after left hypercondyly: preoperative photograph.
splints were generated using a computer-assisted procedure (Figs. 2–5). Six patients in the classic group and 8 in the digital group had genioplasties. The data shown in Table 1 summarise the measurements for different anatomical landmarks. The table shows the distances before and after operation of the considered midline landmarks and sagittal planes from the ideal facial midline and midsagittal plane for each group. “Positive” and “negative” angular values refer to clockwise and counter-clockwise rotations, respectively. Fig. 6 shows the surgical movements that depended directly on the intermediate splint and the surgical movements that were influenced only by planning, but are actually defined during operation.
Fig. 3. Preoperative 3-dimensional computed tomographic scan.
Differences between the 2 groups in percentage alignment rates of the lower interincisal point, mandibular sagittal plane, and centring of the dental midlines were significant, and the digital planning group showed greater accuracy. The postoperative alignment of the upper interincisal point, maxillary sagittal plane, and skeletal menton with the ideal midline suggest the greater accuracy of digital planning, although the differences were not significant. There was no difference in the alignment of the soft tissues of the chin between groups.
Discussion The purpose of this prospective randomised controlled study was to assess whether computer-assisted orthognathic planning of operations can improve the accuracy of correction
Fig. 5. Postoperative photograph.
G. De Riu et al. / British Journal of Oral and Maxillofacial Surgery 52 (2014) 251–257
255
Table 1 Comparison of rates of alignment: classic compared with computer-assisted planning. Data are mean (SD) and 95% CI. Assessment of postoperative deviation/asymmetry
Percentage rate of alignment (classic)
Percentage rate of alignment (digital)
p-Value
Distance from the upper interincisal point to the facial midline
76.25 (34.59) 54.81–97.69 50.82 (43.65) 23.76–77.88 58.50 (41.10) 33.03–83.97 79.67 (33.13) 59.14–100.20 79.67 (32.09) 59.78–99.56 58.83 (33.43) 38.11–79.55 42.74 (32.04) 22.88–62.60
88.10 (27.97) 70.77–105.43 88.17 (20.91) 75.21–101.13 92.59 (13.98) 83.93–101.25 85.77 (23.51) 71.20–100.34 76.67 (39.44) 52.23–101.11 71.61 (31.79) 51.90–91.32 80.18 (29.48) 61.91–98.45
0.41
Distance from the lower interincisal point to the facial midline Distance between interincisal points Distance from skeletal menton to the facial midline Distance from soft tissue menton to the facial midline Distance from the maxillary sagittal plane to the facial midsagittal plane Distance from the mandibular sagittal plane to the facial midsagittal plane
of facial asymmetry, and we showed generally better results in the computer-assisted (digital) group. Alignment of the dental midlines, alignment of the mandibular midline to the facial midline, and alignment of the mandibular sagittal plane to the mid-sagittal plane were significantly more precise in this group, compared with classic planning. We think that the favourable results are explained by several factors. Classic planning can incorporate errors at several time points. Firstly, during transfer of the models to the articulator with the facial bow as a result of articulator characteristics, variability in the patient’s anatomy, or because of a lack of cooperation.20 Secondly, while recording the wax bite in centric relation as a result of poor cooperation or technical error.1 Thirdly, while drawing the vertical and horizontal reference lines on the mounted models by hand
0.03 0.03 0.64 1.00 0.39 0.01
using 2-dimensional instruments (ruler and callipers) on 3dimensional plaster casts that do not represent the patient’s bone structure or the osteotomy lines used at operation.20 Finally, while repositioning the models, because of a lack of references for the soft tissues and difficulty measuring the movement precisely on the irregular surface of the model. Digital planning definitely allows less room for errors.21 Most arise during acquisition of CT data and are usually the result of poor cooperation. For example, the first scan must be taken in centric occlusion with relaxed lips; if the patient fails to follow these instructions the software cannot read the small spaces between the teeth and impression efficiently and “noise” and inaccuracies can result. Secondly, surgical simulation on plaster models on the articulator allows some risk of inaccuracy.2–4 Only the
Fig. 6. Rates of alignment depending directly on the intermediate splint (columns 1–4) and influenced by planning but defined during operation (columns 5–7). Red columns = computer-assisted planning and blue columns = classic planning.
256
G. De Riu et al. / British Journal of Oral and Maxillofacial Surgery 52 (2014) 251–257
patient’s teeth are shown in 3 dimensions in these models, and none of the segments of the facial skeleton that the surgeon will modify and mobilise intraoperatively.20,22 Traditional simulation provides no immediate display of the entire facial structure, which is essential for correcting facial asymmetry. This display is possible only through the abstract integration of cephalometric and clinical data, which requires clinical experience. With digital planning all necessary records are provided simultaneously through the integration of all the patient’s data into a single 3-dimensional image. The maxilla and mandible, with their midline landmarks and sagittal planes, are easily aligned to the facial midplane and consequently achievement of facial symmetry is more immediate, intuitive, and reliable. Cephalometric data are updated automatically when the maxilla and mandible are repositioned, which allows realtime 3-dimensional control of the planned movements. The semitransparent texture of the surrounding soft tissues shows the relations with the asymmetrical structures of the facial bones, and helps to achieve a more symmetrical result from an aesthetic perspective. This is much better than in classic simulation, where symmetry is achieved by aligning twisted pencil marks made on the rough surfaces of plaster casts, with the only sagittal reference represented by the vertical pin of the articulator and with no soft tissue references at all. Thirdly, a precise intermediate splint is necessary in bimaxillary surgery for asymmetry, during which complex movements are made in the 3 planes.20,23 Different authors have evaluated the accuracy of computer-assisted orthognathic surgery,20,24 and concluded that it is a reliable method of designing and creating precise surgical guides. These conclusions were made by surgeons experienced in classic model operations who faced the problems of virtual simulation for the first cases. Finally, simulation on plaster models does not allow the simulation of genioplasty, as the chin is not indicated on the models. In contrast, all anatomical structures of the facial skeleton are reproduced in virtual surgical simulation, which can simulate repositioning of the chin. However, neither method enables the realisation of a surgical guide for genioplasty, although some authors have already reported computer-assisted repositioning of the chin (prototyped surgical guides), with successful results.24,25 In conclusion, despite our relatively small sample, our results encourage the adoption of computer-assisted orthognathic operations for facial asymmetry. This procedure, compared with the classic method of planning, allowed better control and accuracy in repositioning and alignment of the maxilla and the mandible.
Conflict of interest Conflict of interest statement: all authors disclose any financial and personal relationships with other people or
organisations that could inappropriately influence (bias) their work.
Ethics statement This work was conducted according to the Helsinki Declaration of 1975 and approved by the Research Committee (IRB) of the Department of Surgical, Microsurgical and Medical Science, University of Sassari. We also confirm that patient permission has been obtained to publish the figures, along with the informed consent.
References 1. Bamber MA, Firouzal R, Harris M, Linney A. A comparative study of two arbitrary face-bow transfer systems for orthognathic surgery planning. Int J Oral Maxillofac Surg 1996;25:339–43. 2. Bowley JF, Bowman HC. Evaluation of variables associated with the transverse horizontal axis. J Prosthet Dent 1992;68: 537–41. 3. Bowley JF, Cipra DL, Herman PF. Evaluation of the accuracy of cast reorientation to a surveyor by prosthodontic residents. J Prosthet Dent 1992;68:294–8. 4. Bowley JF, Michaels GC, Lai TW, Lin PP. Reliability of a facebow transfer procedure. J Prosthet Dent 1992;67:491–8. 5. Gateno J, Forrest KK, Camp B. A comparison of 3 methods of facebow transfer recording: implications for orthognathic surgery. J Oral Maxillofac Surg 2001;59:635–41. 6. Sharifi A, Jones R, Ayoub A, et al. How accurate is model planning for orthognathic surgery? Int J Oral Maxillofac Surg 2008;37:1089–93. 7. De Riu G, Meloni SM, Pisano M, Massarelli O, Tullio A. Computed tomography-guided implant surgery for dental rehabilitation in mandible reconstructed with a fibular free flap: description of the technique. Br J Oral Maxillofac Surg 2012;50:30–5. 8. Meloni SM, De Riu G, Pisano M, Massarelli O, Tullio A. Computer assisted dental rehabilitation in free flaps reconstructed jaws: one year follow-up of a prospective clinical study. Br J Oral Maxillofac Surg 2012;50:726–31. 9. Bell RB. Computer planning and intraoperative navigation in orthognathic surgery. J Oral Maxillofac Surg 2011;69:592–605. 10. Swennen GR, Mollemans W, De Clercq C, et al. A cone-beam computed tomography triple scan procedure to obtain a three-dimensional augmented virtual skull model appropriate for orthognathic surgery planning. J Craniofac Surg 2009;20:297–307. 11. Farkas LG, Munro IR. Anthropometric facial proportions in medicine. Springfield: Charles C. Thomas; 1987. 12. Arnett GW, Bergman RT. Facial keys to orthodontic diagnosis and treatment planning. Part I. Am J Orthod Dentofacial Orthop 1993;103:299–312. 13. Arnett GW, Bergman RT. Facial keys to orthodontic diagnosis and treatment planning—Part II. Am J Orthod Dentofacial Orthop 1993;103:395–411. 14. Epker BN, Stella JP, editors. Dentofacial deformities: integrated orthodontic and surgical correction, vol. 4. St. Louis: Mosby; 1999. 15. Anwar M, Harris M. Model surgery for orthognathic planning. Br J Oral Maxillofac Surg 1990;28:393–7. 16. Proffit WR, White Jr RP, Sarver DM. Contemporary treatment of dentofacial deformities. St Louis: Mosby; 2003. 17. Nadjmi N, Mollemans W, Daelemans A, Van Hemelen G, Schutyser F, Bergé S. Virtual occlusion in planning orthognathic surgical procedures. Int J Oral Maxillofac Surg 2010;39:457–62.
G. De Riu et al. / British Journal of Oral and Maxillofacial Surgery 52 (2014) 251–257 18. Swennen GR, Mollemans W, Schutyser F. Three-dimensional treatment planning of orthognathic surgery in the era of virtual imaging. J Oral Maxillofac Surg 2009;67:2080–92 [Erratum in J Oral Maxillofac Surg 2009;67:2703]. 19. Swennen GR, Schutyser F. Three-dimensional cephalometry: spiral multi-slice vs cone-beam computed tomography. Am J Orthod Dentofacial Orthop 2006;130:410–6. 20. Aboul-Hosn Centenero S, Hernández-Alfaro F. 3D planning in orthognathic surgery: CAD/CAM surgical splints and prediction of the soft and hard tissues results – our experience in 16 cases. J Craniomaxillofac Surg 2012;40:162–8. 21. Donatsky O, Bjørn-Jørgensen J, Hermund NU, Nielsen H, HolmqvistLarsen M, Nerder PH. Immediate postoperative outcome of orthognathic surgical planning, and prediction of positional changes in hard and soft
22.
23. 24.
25.
257
tissue, independently of the extent and direction of the surgical corrections required. Br J Oral Maxillofac Surg 2011;49:386–91. Olszewski R, Reychler H. Limitations of orthognathic model surgery: theoretical and practical implications. Rev Stomatol Chir Maxillofac 2004;105:165–9 [in French]. Ellis III E. Bimaxillary surgery using an intermediate splint to position the maxilla. J Oral Maxillofac Surg 1999;57:53–6. Hsu SS, Gateno J, Bell RB, et al. Accuracy of a computer-aided surgical simulation protocol for orthognathic surgery: a prospective multicenter study. J Oral Maxillofac Surg 2013;71:128–42. Olszewski R, Tranduy K, Reychler H. Innovative procedure for computerassisted genioplasty: three-dimensional cephalometry, rapid-prototyping model and surgical splint. Int J Oral Maxillofac Surg 2010;39: 721–4.
British Journal of Oral and Maxillofacial Surgery 52 (2014) 251–257
Computer-assisted orthognathic surgery for correction of facial asymmetry: results of a randomised controlled clinical trial Giacomo De Riu a,∗ , Silvio Mario Meloni a , Alessandro Baj b , Andrea Corda a , Damiano Soma a , Antonio Tullio a a b
Maxillofacial Surgery Unit, University Hospital of Sassari, Italy Maxillofacial Surgery Unit, Ospedale Maggiore Policlinico, Milan, Italy
Accepted 13 December 2013 Available online 10 January 2014
Abstract In this randomised controlled clinical trial, 2 homogeneous groups of patients with facial asymmetry (n = 10 in each) were treated by either classic or computer-assisted orthognathic corrective surgery. Differences between the 2 groups in the alignment of the lower interincisal point (p = 0.03), mandibular sagittal plane (p = 0.01), and centring of the dental midlines (p = 0.03) were significant, with the digital planning group being more accurate. © 2013 The British Association of Oral and Maxillofacial Surgeons. Published by Elsevier Ltd. All rights reserved. Keywords: Computer-assisted orthognathic surgery; Facial asymmetry; Orthognathic planning
Introduction The correct planning of successful correction of facial asymmetry can be complex. In clinical practice it is often difficult to individualise the midline facial landmarks on a standard frontal cephalogram because of the superimposition of different cranial structures. Midlines can therefore be adopted arbitrarily to assess the asymmetry of the face, which can lead to imprecise evaluation. Classical surgical simulation using plaster casts on semiindividual articulators is also imprecise. The alignment of the midlines on the articulator is hindered by many errors inherent in the planning of the classic model operation.1–6
∗ Corresponding author at: Unità Operativa di Chirurgia Maxillo-Facciale, AOU Sassari, Viale San Pietro 43/b, 07100 Sassari, Italy. Tel.: +39 079228216; fax: +39 079229002. E-mail addresses: [email protected], [email protected] (G. De Riu).
Planning the correction of asymmetry is more difficult than planning the correction of sagittal deformities because of the complex management of different movements of the plaster casts (roll, yaw, and lateral translation) that affect the position and inclination of the midlines and occlusal planes. Correct model planning of symmetrical maxillomandibular repositioning is also difficult because of the inaccuracy of linear and angular measurements, usually made between pencil marks on the rough surface of the plaster casts. Surgical planning of the correction of facial asymmetry is therefore challenging, but the difficulties lie mainly in the out-of-date and imprecise tools that we use during the diagnostic and simulation phases, which result in the incorporation of various errors in the intermediate splint. In many cases, despite meticulous planning, residual asymmetry is common. We wondered if digital planning would give more accurate results.7,8 Several currently available programmes permit virtual operations with digital control of the 3-dimensional movements of the maxilla, and the computer-aided design
0266-4356/$ – see front matter © 2013 The British Association of Oral and Maxillofacial Surgeons. Published by Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.bjoms.2013.12.010
252
G. De Riu et al. / British Journal of Oral and Maxillofacial Surgery 52 (2014) 251–257
and machining (CAD/CAM) of intermediate splints.9 These programmes can overcome the technical problems of management of the spatial positions of the jaws with reference to the soft tissues and craniofacial midlines. The purpose of this randomised, controlled, clinical study was to evaluate the most accurate procedure for orthognathic correction of facial asymmetry. We compared two different methods of surgical planning: classic and computerassisted.10 The specific aims of the study were to measure and compare the rates of alignment and reduction of cant of the dental and facial midlines between the 2 groups.
Patients, material, and methods Design of the study We designed a randomised, controlled trial (ClinicalTrials.gov id. NCT01879969) that was conducted according to the Helsinki Declaration of 1975, and approved by the Research Committee (IRB) of the Department of Surgical, Microsurgical and Medical Science, University of Sassari, Italy. We studied 20 patients who presented to the Maxillofacial Surgery Unit, University Hospital of Sassari, for evaluation and management of facial asymmetry or lateral deviation of the mandible. The randomisation codes were created (Excel, Microsoft, Redmond, WA, USA), by combining a sequence of randomised non-consecutive numbers to match the 2 procedures (classic or digital planning). The codes were assigned by an independent operator who was not involved in the trial, and were placed in envelopes. Data were then collected in sheets (Excel). All participants provided written informed consent before enrollment, after which they were treated by orthognathic repositioning of the maxilla and mandible with or without genioplasty. All operations were planned and done by the same surgeon between November 2010 and December 2012. The inclusion criteria were: cant of the occlusal plane of more than 3◦ or midline discrepancies of more than 2.5 mm, or both; the presence of all central incisors; radiographs and plaster casts taken before and after the operation (classic group), or cranial cone-beam computed tomographic (CT) images (digital group), together with digital photographs taken before and after the operation. Patients were excluded if they had had previous trauma to the hard or soft facial tissues, functional deviation of the mandible, or their records were incomplete (cone-beam CT, radiographs, casts, or digital photographs). Variables The primary predictive variable was method of planning treatment. Each eligible patient was randomly selected to have bimaxillary surgery planned with a classic, standard
procedure or using computer-assisted orthognathic surgery. The primary outcome variables were linear and angular measures that defined the alignment of the facial midlines, or reduction of maxillary and mandibular cant in the 2 groups. Technique of classic planning The deformity for each patient was diagnosed after careful analysis of frontal facial aesthetics,11–13 compared with the cephalometric findings.14 The plaster casts were mounted on an average value, semi-individual articulator (SAM II, SAM Präzisionstechnik GmbH, Munich, Germany). The vertical and mediolateral relations of the teeth were recorded by drawing vertical reference lines on the upper model and horizontal reference lines at a calculated distance. With these reference points, the precise movements of the maxillary segments were calculated in the 3 planes of space with model surgery.15 The models were repositioned based on data obtained with the surgical Visual Treatment Objective.16 Surgical simulation allows the manufacture of the resin occlusal intermediate splint that records the occlusal relations after the upper model has been moved and the mandible is still in the preoperative position, which serves as a guide during the operation. In some cases (such as those with class II, short face, or asymmetric vertical deficiency of the face) we prefer to move the mandible first, and inverted planning is used. Technique of digital planning We used cone-beam CT data (KaVo, Biberach, Germany) and the Maxilim software (Medicim, Nobel Biocare Group). The protocol for acquiring CT images involved a first scan of the patient in centric occlusion with relaxed lips, a second scan at low resolution and low-dose biting of a double tray, and a third scan of that tray to record the occlusion at high resolution.10 The total radiation dose to the patient was ∼100 Sv. Once the CT images had been acquired on to the software, a computer-assisted procedure allowed the development of a virtual 3-dimensional model of the hard and soft tissues of the face, with a better-defined window of the teeth and intercuspation (augmented model). A complete set of 2-dimensional photographs of the face captured with any digital camera was used to add information about the texture of the skin. This process created the patient’s face mask. The relations between the soft and hard tissues of the face were then shown, setting the transparency of the skin texture, and permitting the measurement of the maxillary vertical dimension compared with the length of the upper lip, or definition of deviation of the dental midline from the philtrum (Fig. 1). Next, we made a 3-dimensional cephalometric analysis. Lateral and frontal cephalograms were developed from the CT data and connected to the surface of the hard and soft tissues of the virtual model. Once all the landmarks had
G. De Riu et al. / British Journal of Oral and Maxillofacial Surgery 52 (2014) 251–257
253
Fig. 1. Three-dimensional facial planning showing the relations between the hard and soft tissues.
been identified, the angular and linear measurements were calculated automatically. Planning of treatment included virtual osteotomies on the maxilla, mandible, and chin. The operator chose whether to move the maxilla or the mandible first. Once the maxilla (or mandible) had been placed in the desired position, a special tool17 was used to move the opposite jaw and to define the best occlusion for the patient through a 3-dimensional view of the dental arches and an occlusogram. The dental arches were guided to occlusion with virtual rubber bands that allowed for the control of intercuspation. The occlusogram also quantified the effective area of contact between the teeth, and identified areas where grinding of the cusp would be necessary.10,18 In the case of a multisegmented maxilla, the standard virtual procedure was edited to individualise every piece of bone with at least 3 landmarks. This process enabled us to fix the multipiece arch in the right occlusion. Once the final occlusion had been decided, the mandible (or maxilla in inverted planning) was moved. Finally, the chin was operated on if necessary, although no surgical guide was provided. The soft tissues were remodelled automatically: the algorithm for soft tissue to bone movement simulates the aesthetic changes, which helps to predict the outcome. Once we were satisfied with the planned movements, an output file was saved and uploaded to the server for the manufacture of a CAD/CAM surgical splint, which arrived in a few days by mail.
facial midline, and between the maxillary sagittal plane, the mandibular sagittal plane and the ideal midsagittal plane. The distance between the upper and lower interincisal points was also evaluated. An ideal midline and ideal midsagittal plane were considered to be the vertical line and the vertical plane that passed by the apophysis crista galli, sella, and nasion for the hard tissues, and passed midway between the left and right internal canthi and the middle of the lip for the soft tissues. In the classic group, measurements were made before and after operation on posteroanterior cephalograms and digital photographs. In the digital group the measurements were obtained by comparing the Swennen modified 3-dimensional cephalometry taken before and after the operation.19 Analysis of data The data were recorded on a spreadsheet (Excel; Microsoft Corporation, Redmond, WA, USA), and the significances of differences between the groups were assessed using Student’s test with the help of Stata software (version 9.1, StataCorp, TX, USA). Postoperative clinical and radiographic improvement in the alignment of anatomical points for each subject and the mean improvement for each group were calculated. Probabilities of less than 0.05 were accepted as significant, and all data are presented as mean (SD) with 95% CI.
Results Collection of data In both groups measurements were made to define linear and angular discrepancies between upper and lower interincisal points, skeletal menton, soft tissue menton, and the ideal
The classic group comprised 7 women and 3 men aged from 21 to 54 years old). An acrylic intermediate splints was constructed. The digital group consisted of 7 men and 3 women aged 24–47 years old) and their intermediate surgical
254
G. De Riu et al. / British Journal of Oral and Maxillofacial Surgery 52 (2014) 251–257
Fig. 4. Three-dimensional planning.
Fig. 2. Facial asymmetry after left hypercondyly: preoperative photograph.
splints were generated using a computer-assisted procedure (Figs. 2–5). Six patients in the classic group and 8 in the digital group had genioplasties. The data shown in Table 1 summarise the measurements for different anatomical landmarks. The table shows the distances before and after operation of the considered midline landmarks and sagittal planes from the ideal facial midline and midsagittal plane for each group. “Positive” and “negative” angular values refer to clockwise and counter-clockwise rotations, respectively. Fig. 6 shows the surgical movements that depended directly on the intermediate splint and the surgical movements that were influenced only by planning, but are actually defined during operation.
Fig. 3. Preoperative 3-dimensional computed tomographic scan.
Differences between the 2 groups in percentage alignment rates of the lower interincisal point, mandibular sagittal plane, and centring of the dental midlines were significant, and the digital planning group showed greater accuracy. The postoperative alignment of the upper interincisal point, maxillary sagittal plane, and skeletal menton with the ideal midline suggest the greater accuracy of digital planning, although the differences were not significant. There was no difference in the alignment of the soft tissues of the chin between groups.
Discussion The purpose of this prospective randomised controlled study was to assess whether computer-assisted orthognathic planning of operations can improve the accuracy of correction
Fig. 5. Postoperative photograph.
G. De Riu et al. / British Journal of Oral and Maxillofacial Surgery 52 (2014) 251–257
255
Table 1 Comparison of rates of alignment: classic compared with computer-assisted planning. Data are mean (SD) and 95% CI. Assessment of postoperative deviation/asymmetry
Percentage rate of alignment (classic)
Percentage rate of alignment (digital)
p-Value
Distance from the upper interincisal point to the facial midline
76.25 (34.59) 54.81–97.69 50.82 (43.65) 23.76–77.88 58.50 (41.10) 33.03–83.97 79.67 (33.13) 59.14–100.20 79.67 (32.09) 59.78–99.56 58.83 (33.43) 38.11–79.55 42.74 (32.04) 22.88–62.60
88.10 (27.97) 70.77–105.43 88.17 (20.91) 75.21–101.13 92.59 (13.98) 83.93–101.25 85.77 (23.51) 71.20–100.34 76.67 (39.44) 52.23–101.11 71.61 (31.79) 51.90–91.32 80.18 (29.48) 61.91–98.45
0.41
Distance from the lower interincisal point to the facial midline Distance between interincisal points Distance from skeletal menton to the facial midline Distance from soft tissue menton to the facial midline Distance from the maxillary sagittal plane to the facial midsagittal plane Distance from the mandibular sagittal plane to the facial midsagittal plane
of facial asymmetry, and we showed generally better results in the computer-assisted (digital) group. Alignment of the dental midlines, alignment of the mandibular midline to the facial midline, and alignment of the mandibular sagittal plane to the mid-sagittal plane were significantly more precise in this group, compared with classic planning. We think that the favourable results are explained by several factors. Classic planning can incorporate errors at several time points. Firstly, during transfer of the models to the articulator with the facial bow as a result of articulator characteristics, variability in the patient’s anatomy, or because of a lack of cooperation.20 Secondly, while recording the wax bite in centric relation as a result of poor cooperation or technical error.1 Thirdly, while drawing the vertical and horizontal reference lines on the mounted models by hand
0.03 0.03 0.64 1.00 0.39 0.01
using 2-dimensional instruments (ruler and callipers) on 3dimensional plaster casts that do not represent the patient’s bone structure or the osteotomy lines used at operation.20 Finally, while repositioning the models, because of a lack of references for the soft tissues and difficulty measuring the movement precisely on the irregular surface of the model. Digital planning definitely allows less room for errors.21 Most arise during acquisition of CT data and are usually the result of poor cooperation. For example, the first scan must be taken in centric occlusion with relaxed lips; if the patient fails to follow these instructions the software cannot read the small spaces between the teeth and impression efficiently and “noise” and inaccuracies can result. Secondly, surgical simulation on plaster models on the articulator allows some risk of inaccuracy.2–4 Only the
Fig. 6. Rates of alignment depending directly on the intermediate splint (columns 1–4) and influenced by planning but defined during operation (columns 5–7). Red columns = computer-assisted planning and blue columns = classic planning.
256
G. De Riu et al. / British Journal of Oral and Maxillofacial Surgery 52 (2014) 251–257
patient’s teeth are shown in 3 dimensions in these models, and none of the segments of the facial skeleton that the surgeon will modify and mobilise intraoperatively.20,22 Traditional simulation provides no immediate display of the entire facial structure, which is essential for correcting facial asymmetry. This display is possible only through the abstract integration of cephalometric and clinical data, which requires clinical experience. With digital planning all necessary records are provided simultaneously through the integration of all the patient’s data into a single 3-dimensional image. The maxilla and mandible, with their midline landmarks and sagittal planes, are easily aligned to the facial midplane and consequently achievement of facial symmetry is more immediate, intuitive, and reliable. Cephalometric data are updated automatically when the maxilla and mandible are repositioned, which allows realtime 3-dimensional control of the planned movements. The semitransparent texture of the surrounding soft tissues shows the relations with the asymmetrical structures of the facial bones, and helps to achieve a more symmetrical result from an aesthetic perspective. This is much better than in classic simulation, where symmetry is achieved by aligning twisted pencil marks made on the rough surfaces of plaster casts, with the only sagittal reference represented by the vertical pin of the articulator and with no soft tissue references at all. Thirdly, a precise intermediate splint is necessary in bimaxillary surgery for asymmetry, during which complex movements are made in the 3 planes.20,23 Different authors have evaluated the accuracy of computer-assisted orthognathic surgery,20,24 and concluded that it is a reliable method of designing and creating precise surgical guides. These conclusions were made by surgeons experienced in classic model operations who faced the problems of virtual simulation for the first cases. Finally, simulation on plaster models does not allow the simulation of genioplasty, as the chin is not indicated on the models. In contrast, all anatomical structures of the facial skeleton are reproduced in virtual surgical simulation, which can simulate repositioning of the chin. However, neither method enables the realisation of a surgical guide for genioplasty, although some authors have already reported computer-assisted repositioning of the chin (prototyped surgical guides), with successful results.24,25 In conclusion, despite our relatively small sample, our results encourage the adoption of computer-assisted orthognathic operations for facial asymmetry. This procedure, compared with the classic method of planning, allowed better control and accuracy in repositioning and alignment of the maxilla and the mandible.
Conflict of interest Conflict of interest statement: all authors disclose any financial and personal relationships with other people or
organisations that could inappropriately influence (bias) their work.
Ethics statement This work was conducted according to the Helsinki Declaration of 1975 and approved by the Research Committee (IRB) of the Department of Surgical, Microsurgical and Medical Science, University of Sassari. We also confirm that patient permission has been obtained to publish the figures, along with the informed consent.
References 1. Bamber MA, Firouzal R, Harris M, Linney A. A comparative study of two arbitrary face-bow transfer systems for orthognathic surgery planning. Int J Oral Maxillofac Surg 1996;25:339–43. 2. Bowley JF, Bowman HC. Evaluation of variables associated with the transverse horizontal axis. J Prosthet Dent 1992;68: 537–41. 3. Bowley JF, Cipra DL, Herman PF. Evaluation of the accuracy of cast reorientation to a surveyor by prosthodontic residents. J Prosthet Dent 1992;68:294–8. 4. Bowley JF, Michaels GC, Lai TW, Lin PP. Reliability of a facebow transfer procedure. J Prosthet Dent 1992;67:491–8. 5. Gateno J, Forrest KK, Camp B. A comparison of 3 methods of facebow transfer recording: implications for orthognathic surgery. J Oral Maxillofac Surg 2001;59:635–41. 6. Sharifi A, Jones R, Ayoub A, et al. How accurate is model planning for orthognathic surgery? Int J Oral Maxillofac Surg 2008;37:1089–93. 7. De Riu G, Meloni SM, Pisano M, Massarelli O, Tullio A. Computed tomography-guided implant surgery for dental rehabilitation in mandible reconstructed with a fibular free flap: description of the technique. Br J Oral Maxillofac Surg 2012;50:30–5. 8. Meloni SM, De Riu G, Pisano M, Massarelli O, Tullio A. Computer assisted dental rehabilitation in free flaps reconstructed jaws: one year follow-up of a prospective clinical study. Br J Oral Maxillofac Surg 2012;50:726–31. 9. Bell RB. Computer planning and intraoperative navigation in orthognathic surgery. J Oral Maxillofac Surg 2011;69:592–605. 10. Swennen GR, Mollemans W, De Clercq C, et al. A cone-beam computed tomography triple scan procedure to obtain a three-dimensional augmented virtual skull model appropriate for orthognathic surgery planning. J Craniofac Surg 2009;20:297–307. 11. Farkas LG, Munro IR. Anthropometric facial proportions in medicine. Springfield: Charles C. Thomas; 1987. 12. Arnett GW, Bergman RT. Facial keys to orthodontic diagnosis and treatment planning. Part I. Am J Orthod Dentofacial Orthop 1993;103:299–312. 13. Arnett GW, Bergman RT. Facial keys to orthodontic diagnosis and treatment planning—Part II. Am J Orthod Dentofacial Orthop 1993;103:395–411. 14. Epker BN, Stella JP, editors. Dentofacial deformities: integrated orthodontic and surgical correction, vol. 4. St. Louis: Mosby; 1999. 15. Anwar M, Harris M. Model surgery for orthognathic planning. Br J Oral Maxillofac Surg 1990;28:393–7. 16. Proffit WR, White Jr RP, Sarver DM. Contemporary treatment of dentofacial deformities. St Louis: Mosby; 2003. 17. Nadjmi N, Mollemans W, Daelemans A, Van Hemelen G, Schutyser F, Bergé S. Virtual occlusion in planning orthognathic surgical procedures. Int J Oral Maxillofac Surg 2010;39:457–62.
G. De Riu et al. / British Journal of Oral and Maxillofacial Surgery 52 (2014) 251–257 18. Swennen GR, Mollemans W, Schutyser F. Three-dimensional treatment planning of orthognathic surgery in the era of virtual imaging. J Oral Maxillofac Surg 2009;67:2080–92 [Erratum in J Oral Maxillofac Surg 2009;67:2703]. 19. Swennen GR, Schutyser F. Three-dimensional cephalometry: spiral multi-slice vs cone-beam computed tomography. Am J Orthod Dentofacial Orthop 2006;130:410–6. 20. Aboul-Hosn Centenero S, Hernández-Alfaro F. 3D planning in orthognathic surgery: CAD/CAM surgical splints and prediction of the soft and hard tissues results – our experience in 16 cases. J Craniomaxillofac Surg 2012;40:162–8. 21. Donatsky O, Bjørn-Jørgensen J, Hermund NU, Nielsen H, HolmqvistLarsen M, Nerder PH. Immediate postoperative outcome of orthognathic surgical planning, and prediction of positional changes in hard and soft
22.
23. 24.
25.
257
tissue, independently of the extent and direction of the surgical corrections required. Br J Oral Maxillofac Surg 2011;49:386–91. Olszewski R, Reychler H. Limitations of orthognathic model surgery: theoretical and practical implications. Rev Stomatol Chir Maxillofac 2004;105:165–9 [in French]. Ellis III E. Bimaxillary surgery using an intermediate splint to position the maxilla. J Oral Maxillofac Surg 1999;57:53–6. Hsu SS, Gateno J, Bell RB, et al. Accuracy of a computer-aided surgical simulation protocol for orthognathic surgery: a prospective multicenter study. J Oral Maxillofac Surg 2013;71:128–42. Olszewski R, Tranduy K, Reychler H. Innovative procedure for computerassisted genioplasty: three-dimensional cephalometry, rapid-prototyping model and surgical splint. Int J Oral Maxillofac Surg 2010;39: 721–4.
