Journal of Cranio-Maxillo-Facial Surgery 51 (2023) 280e287

Contents lists available at ScienceDirect

Journal of Cranio-Maxillo-Facial Surgery

journal homepage: www.jcmfs.com

Accuracy of virtual surgical planning in bimaxillary orthognathic

surgery with mandible first sequence: A retrospective study
Lorenzo Trevisiol a, Massimo Bersani a, *, Alejandro Martinez Garza b, Esteban Alvarado c,
G. William Arnett d, Antonio D'Agostino a
a
Department of Surgical Sciences, Dentistry, Gynaecology and Paediatrics, University of Verona, Verona, Italy
b
n, Mexico
Private Practice, San Pedro Garza García, Nuevo Leo
c
Private Practice, San Jos

e, Costa Rica
d
Private Practice, Santa Barbara, CA, USA

a r t i c l e i n f o a b s t r a c t

Article history: The aim of this study was to verify treatment accuracy using virtual surgical planning (VSP) with a
Paper received 31 August 2022 mandible-first sequence and strict surgical protocol to determine what surgical and methodological
Received in revised form factors might influence outcomes.
27 February 2023
VSP transfer accuracy was evaluated retrospectively through a modified method involving voxel-
Accepted 23 May 2023
based superimposition in patients who had undergone bimaxillary surgery with a mandible-first
Available online 24 May 2023
sequence to correct dentoskeletal deformities. Data analysis showed that the movements planned and
Handling Editor: Prof. Emeka Nkenke those executed were substantially equivalent (p < 0.01), with the exception of mandibular and maxillary
sagittal movements that were 0.72 ± 0.90 mm and 1.41 ± 1.04 mm smaller, respectively, than planned.
Keywords: This study showed that a mandible-first sequence is accurate for transferring virtual surgical planning
Orthognathic surgery intraoperatively. There are several factors involved in the proper transfer of virtual planning beyond the
Virtual surgical planning software, such as surgical technique and sequencing. Inaccurate sagittal movements and maxillary
Accuracy repositioning seem to depend mainly on surgical factors.
Voxel-based superimposition
© 2023 The Authors. Published by Elsevier Ltd on behalf of European Association for Cranio-Maxillo-
Facial Surgery. This is an open access article under the CC BY-NC-ND license (http://creativecommons.
org/licenses/by-nc-nd/4.0/).

1. Introduction sequencing could affect the outcomes accuracy-wise, although

both sequences seem to produce similar outcomes when properly
The use of digital technology in orthognathic surgery can be planned and executed (Borikanphanitphaisan et al., 2021; Perez
considered the most important advancement in the field of and Ellis, 2016). For this reason, a mandible-first sequence has
maxillofacial surgery since the advent of rigid fixation, and interest been a topic of interest in the literature in recent years, given its
in it is reflected in the numerous articles published on the subject. supposed advantages in large maxillary advancements and
Use of digital technology has determined an improvement in counter-clockwise rotation. Furthermore, various studies on VSP
diagnostic accuracy, planning and traditional treatment outcomes accuracy, with either surgical sequence, have reported limitations
from both functional and aesthetic points of view (Lin, 2015). in interpreting their results due to weaknesses in the study design
Besides these advantages, a more in-depth review of the liter- and/or to the methods used to evaluate results. Other weaknesses
ature reveals that it is still not completely clear whether VSP is have involved limited information regarding planning and surgical
better than traditional surgical protocols as far as sagittal plane stages (Alkhayer et al., 2020; Gaber et al., 2017).
accuracy and patient satisfaction are concerned (Chen et al., 2021). The aim of this study was to verify the accuracy of sugery per-
The literature does not specify which surgical factors besides VSP formed with digital planning using a mandible-first sequence and
may influence final outcomes. For instance, it is known that surgical strict surgical protocol, and to determine which surgical and
methodological factors may affect outcomes.

* Corresponding author. Policlinico G. B. Rossi, piazzale L. Scuro, 37134, Verona,

Italy.
E-mail address: [email protected] (M. Bersani).

https://doi.org/10.1016/j.jcms.2023.05.015
1010-5182/© 2023 The Authors. Published by Elsevier Ltd on behalf of European Association for Cranio-Maxillo-Facial Surgery. This is an open access article under the CC BY-
NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
L. Trevisiol, M. Bersani, A. Martinez Garza et al. Journal of Cranio-Maxillo-Facial Surgery 51 (2023) 280e287

2. Materials and methods an extraoral marker (self-tapping screw) positioned on the glabella.
Finally, the osteotomized maxilla was stabilized using titanium
We designed a retrospective, observational study in patients mini-plates fixed to maxillary buttresses, and microplates, if
who had undergone bimaxillary surgery with VSP to correct Class I, necessary, at interdental osteotomy sites. Once again, the inter-
Class II, Class III, open bite and maxillo-mandibular asymmetry maxillary fixation was removed, and the correct maxillary position
between September 2016 and March 2020. Due to the retrospective was verified by checking correspondence to the planned final oc-
nature of this study, it was granted an exemption in writing by the clusion when manipulated in centric relation. If the maxilla did not
University of Verona IRB. Evaluation was done by comparing the fit precisely, the intermaxillary fixation was restablished and the
surgical movements executed, measured by comparing pre- and maxilla was replated. No rigid post-operative maxillo-mandibular
post-surgery cone-beam computed tomography), and those plan- fixation was used.
ned virtually in accordance with Gaber et al.‘s recommendations
(Gaber et al., 2017). 2.1.1. Surgical movement computation
We included only subjects who had undergone bimaxillary Of the many methods described in the literature to compare
surgery with a mandible-first sequence and virtual surgical plan- planned surgical movements and final dentoskeletal outcomes
ning, and only if pre-operative CBCT of the maxillofacial region in (Marlie re et al., 2019; Mazzoni et al., 2015; Tucker et al., 2010), we
centric relation and post-operative CBCT of the same region per- endeavoured to follow Gaber et al.‘s recommendations to develop
formed in maximum intercuspation and without elastic bands 3 our own modified protocol involving voxel-based superimposition
weeks after surgery were available. on the skull base of pre- and post-operative CBCT volumes to
Patients with idiopathic condylar resorption were excluded measure surgical movements and to compare them with those
from the study. planned in an effort to obtain clear, reliable, repeatable, and, most
of all, clinically interpretable results (Gaber et al., 2017). Following
2.1. Planning and surgery this protocol, we imported pre- and post-operative DICOM data
into 3D Slicer 4.10.2 software (Fedorov et al., 2012).
Pre-operative CBCT (NewTom VGI EVO, Cefla, Verona, Italy) (FOV Pre-operative orientation was transferred from the planning
24
19 cm) images were taken 15e20 days before surgery. The software to 3D Slicer via orientation matrixes. Voxel-based regis-
scans were carried out with the patient standing and with the head tration on the cranial base of the post-operative scan on the ori-
in its natural position. Patients were instructed not to swallow and ented pre-operative scan was then carried out. In each patient,
to keep the lips in a relaxed position; occlusion was guided by a wax dental landmarks were identified three-dimensionally on both
bite in a centric relation position. The need for maxillary segmental volumes in order to measure the entity of surgical movements. The
surgery was verified on stone models of dental arches. The models fiducials were positioned directly onto the radiographical slices at
were then also scanned in occlusion with the same centric wax and the mesiovestibular cusp tips of the upper and lower second mo-
in final occlusion. lars, on the mesialincisal angle of the upper and lower right central
Virtual planning was then carried out using a specifically incisors, and on the upper and lower canine cusps.
developed software (Nemo FAB, Nemotec, Legane
s, Madrid, Spain). Although the repeatibility of three-dimensional landmark
Pre-operative volume was oriented according to the natural head positioning has been previously verified in literature with a mean
position, as it had been established clinically at pre-surgical eval- error of 0.2 mm (Titiz et al., 2012), for our study it was decided to
uation and registered in the planning software (Solow and Tallgren, identify only dental landmarks, as they are easier to place, and to
1971). Surgical movements were planned following Arnett et al.‘s limit their number in an effort to reduce intrinsic errors (de Oliveira
criteria, and then an intermediate splint was designed and printed et al., 2009; Lisboa et al., 2015; Stokbro et al., 2016) (Fig. 2).
(Arnett and Bergman, 1993a, 1993b). A final splint was not used in In order to further evaluate the reproducibility of surgical
any of our cases (Fig. 1). movement computation, two separate doctors performed both the
The same clinicians (L.T. and D.A.) who did the virtual planning orientations and registrations of the volumes, then identified the
performed the surgery. Surgical technique involved a mandible- fiducial points in the cases considered in order to evaluate the
first sequence with the execution of a BSSO according to Epker repeatibility of this identification/positioning. X-Y-Z coordinates for
(1977) and subsequent positioning of the mandibular distal each landmark were exported onto a spreadsheet (Microsoft Excel,
segment into its final position by means of an intermediate splint Microsoft Corp., Redmond, WA, USA), and the surgical movements
and rigid intermaxillary fixation with steel wires. The condyles with reference to positioned landmarks were measured by coor-
were properly repositioned in the fossae with bivectoral seating dinate difference. The jaw movements considered were as follows.
(Arnett et al., 1992). Any premature bone contact that could inter-
fere with correct proximal segments positioning during plating was - sagittal, vertical and lateral (midline) movements at the incisor
eliminated to allow a better bone-to-bone fit while maintaining the level (Fig. 3)
condyle in the correct position with a bivectorial maneuver. The - canine cant modifications
osteotomy was then passively stabilized with 2 titanium mini - vertical and lateral (yaw) movements at the molar level
plates per side (Arnett et al., 1992, 2022a, 2022b). Once the osteo- - sagittal rotational (pitch) modifications of the occlusal plane
synthesis was complete, the intermaxillary fixation was removed, computed with extrinsic Euler angles.
and the correct mandibular position was verified by checking cor-
respondence to the intermediate splint when manipulated in The same measurements were then calculated automatically
centric relation. If the mandible did not fit precisely in the inter- and extrapolated from the virtual planning software for each case
mediate splint, the intermaxillary fixation was re-established with and compared with the data obtained by superimposing the scans.
the splint in place and the mandible replated. The upper jaw was
then mobilized, depending on the case, by one-piece or multi- 2.2. Statistical analysis
segment Le Fort I (Arnett et al., 2022a, 2022b; Kim et al., 2011)
and repositioned using the lower arch as a guide to determine the STATA 13 (College Station, TX, USA) was used for statistical
final occlusion and position. The anterior facial height was the only analysis. Repeatability of surgical movement measurements was
value controlled intraoperatively through measurements involving verified by determining the inter-operator inter-observer
281
L. Trevisiol, M. Bersani, A. Martinez Garza et al. Journal of Cranio-Maxillo-Facial Surgery 51 (2023) 280e287

Fig. 1. Virtual surgical simulation.

variations and interclass correlations (ICC) coefficient. Because surgical movement measured are presented in Table 4. None of the
repeatability of landmark positioning has been verified in the measured movements showed mean inter-operator discrepancies
literature (de Oliveira et al., 2009; Lisboa et al., 2015; Stokbro et al., above 0.2 mm or 0.3 ; this was observed by the high total ICC
2016; Titiz et al., 2012), the value was instead calculated for each (0.994).
surgical movement (since more than one landmark is needed to Most of the surgical mandibular and maxillary movements were
compute them), as well as for the total of 900 movements evaluated substiantially similar to those planned (Tables 2 and 3). The dis-
by each operator according to a mean-rating, absolute-agreement, crepancies measured were all statistically negligible with the
2-way, mixed-effects model ICC (Koo and Li, 2016; Sabour, 2020). exception of mandibular and maxillary advancements, which were
A comparison between planned and surgical movements was 0.72 ± 0.90 mm (p ¼ 0.10) and 1.41 ± 1.04 mm (p ¼ 0.06) smaller,
done using the two one-sided test (TOST) equivalence test for respectively, compared with those planned (Fig. 4). No correlation
paired values to compare measurements in the two groups with was found between surgical discrepancies and the amplitude and
intervals set at ±2 mm. direction of the movements (p ¼ 0.53, p ¼ 0.17, p ¼ 0.87, and
Statistical significance was set at 0.05; a lower p-value indicates p ¼ 0.41 for the sagittal, vertical, and lateral movements and
significant equivalence between planned and measured surgical occlusal plane rotation, respectively.
movements. The Pearson correlation was used to measure the de- A slightly greater mean correspondence with planned sagittal
gree of association between specific variables. movements was found for mandibular repositioning compared
with maxillary movements.
3. Results
4. Discussion
A cohort of 83 patients were evaluated. Of these, 33 had
incomplete records or poor CBCT scans; 50 were eligible (aged Virtual surgical planning is currently widely used in orthog-
26 ± 7 years); 14 patients had undergone single-piece Le Fort I, nathic surgery, as it allows surgeons to overcome certain intrinsic
whereas 36 had undergone three-piece Le Fort I (Table 1). Only 2 limitations of traditional methods thanks to the greater amount of
patients were Angle Class 1 pre-operatively, 18 were Class 2 and 30 information available. However, its ability to achieve better clinical
Class 3. Of the population considered, only two patients had un- outcomes compared with traditional planning is still unclear (Chen
dergone a slight clockwise rotation of the maxillomandibular et al., 2021). Although several variables may affect VSP study re-
complex with maxillary vertical lengthening; the remaining pa- sults, some of them may not have been taken into appropriate
tients had all undergone counter-clockwise rotation with maxillary consideration: for instance, surgical variables and clearly the
impactation of different magnitude. The planned mean mandibular method used to evaluate VSP accuracy and reproducibility. Based
and maxillary advancements were of 5.75 ± 4.95 mm and on these considerations, our study an entailed strict surgical pro-
5.62 ± 1.64 mm, respectively, whereas occlusal plane rotations in tocol and analyzed various methods of accuracy evaluation out-
the sagittal projection were of 7.80 ± 3.96 and 4.06 ± 3.90 , lined in the literature.
respectively (Table 2 and 3). Ideally, in terms of surgical movement measurements, voxel-
Repeatability of hard-tissue displacement measurement, veri- based registration of the osteotomized segments performed on
fied with ICC, was confirmed by very similar results in the mea- pre- and post-operative images is able to measure, with almost
surements taken by two separate clinicians on manually positioned absolute accuracy, surgical movements produced in the absence of
dental landmarks. The average inter-operator discrepancy, confi- operator dependency (Bazina et al., 2018; Ghoneima et al., 2017;
dence interval, and inter-operator interclass coefficient for each Haas Junior et al., 2019; Luebbers et al., 2008). However, the
282
L. Trevisiol, M. Bersani, A. Martinez Garza et al. Journal of Cranio-Maxillo-Facial Surgery 51 (2023) 280e287

Fig. 2. Visualization of the surgical movements evaluation method. A: Pre-operative cone-beam computed tomography (CBCT) acquisition. B: Post-operative CBCT acquisition. C:
Pre-operative CBCT orientation following clinical orientation parameters. DeE: Fiducial points localization on the pre-operative (D) and post-operative volume (E). F: Voxel-based
registration of the post-operative cranial base on the pre-operative cranial base. GeE: Comparison between virtual surgical planning movements and actual surgical movements
calculated on fiducial points.

Fig. 3. Example of sagittal and vertical movement computation at the lower incisor.

283
L. Trevisiol, M. Bersani, A. Martinez Garza et al. Journal of Cranio-Maxillo-Facial Surgery 51 (2023) 280e287

Table 1 Furthermore, the components of a linear traslation of a moving

Study population details. point with 6 degrees of freedom differ depending on the location
Age (y) Mean ¼ 26 (18e46) considered (Fig. 5).
Sex F ¼ 37 (74%)
Another common assessment method compares pre-operative
M ¼ 13 (26%) surfaces generated by CBCT/CT volumes with those generated
post-operatively. (Marlie re et al., 2019; Mazzoni et al., 2015; Tucker
Pre-operative class Class I ¼ 2 (4%)
Class II ¼ 18 (36%) et al., 2010). Besides radiolucency scatterings, iterative closest point
Class III ¼ 30 (60%) (ICP) algorithms can be used to calculate mean deviations in three
Le Fort I Single piece ¼ 14 (28%) spatial dimensions; they cannot, however, be applied to any point
Three pieces ¼ 36 (72%) of the mesh.
Another possible reason for inaccuracy is the use of landmarks
identified on 3-dimensional radiographic volumes with positioning
literature reports limitations in the evaluation of collected data and errors ranging from 0.02 mm to 2.47 mm in all directions,
clinical interpretation. Evaluation of rotation matrices applied to depending on the method used to identify the landmark (Titiz et al.,
points such as centroids (Otranto de Britto Teixeira et al., 2020; 2012). For this reason, in the current study a landmark-based
Stokbro et al., 2016), single incisor landmarks (Sun et al., 2013), or evaluation with multiple albeit a limited number of easily identi-
pure translational and rotational matrices (Baan et al., 2016; fiable dental fiducial points was used. This method allowed us to
Liebregts et al., 2017) is not open to easy clinical interpretation. reduce systematic error during landmark identification on CBCT
volumes (Gaber et al., 2017; Lisboa et al., 2015) and to overcome

Table 2
Mandibular maxillary post-operative versus planning discrepancies and TOST p-values.

Mandibular mean movements Mandibular discrepancies (mm)

(mm)

Planned Outcome Mean discrepancy St. deviation TOST p-value

Sagittal movements
Advancement of R. incisor 5.75 5.04 0.72 0.90 0.10

Lateral movements
Midline correction 0.12 0.54 0.42 1.11 <0.0001
Lateral correction of R. 7 0.19 0.49 0.30 1.08 <0.0001
Lateral correction of L. 7 0.14 0.41 0.27 1.04 <0.0001

Vertical movements
Vertical correction of R. incisor 5.57 4.71 0.86 1.19 <0.01
Canine CANT 0.20 0.01 0.18 0.77 <0.0001
Vertical correction of R. 7 1.16 0.76 0.40 1.20 0.00
Vertical correction of L. 7 0.86 0.63 0.22 1.16 0.00

Rotational movements
Occlusal plane rotation 7.80 6.86 0.93 1.69 <0.0001

“R. 7 ” and “L. 7 ” indicate the right second molar and left second molar, respectively. A negative sagittal discrepancy means that the mandible is positioned posteriorly when
compared to the planning. A positive lateral discrepancy means that the mandible is positioned to the right when compared to the planning. A negative vertical discrepancy
value means that the mandible is positioned cranially when compared to the planning. A negative occlusal plane rotation discrepancy means that the mandible is rotated
clockwise when compared to the planning.

Table 3
Maxillary post-op versus planning discrepancies and TOST p-values.

Maxillary mean movements Maxillary discrepancies (mm)

(mm)

Planned Outcome Mean discrepancy SD TOST p-value

Sagittal movements
Advancement of R. incisor 5.62 4.21 1.41 1.04 0.06

Lateral movements
Midline correction 0.18 0.41 0.60 1.22 <0.0001
Lateral correction of R. 7 1.29 0.87 0.42 1.71 <0.0001
Lateral correction of L. 7 1.42 0.32 1.10 1.67 <0.01

Vertical movements
Vertical correction of R. incisor 2.10 1.75 0.35 1.19 <0.0001
Canine CANT 0.09 0.14 0.23 0.94 <0.0001
Vertical correction of R. 7 0.56 0.18 0.74 1.47 <0.001
Vertical correction of L. 7 0.64 0.04 0.68 1.54 <0.01

Rotational movements
Occlusal plane rotation 4.06 2.68 1.38 2.79 0.00
 
“R. 7 ” and “L. 7 ” indicate the right second molar and left second molar, respectively. A negative sagittal discrepancy means the maxilla is positioned posteriorly when
compared to the planning. A positive lateral discrepancy means the maxilla is positioned to the right when compared to the planning. A negative vertical discrepancy value
means the maxilla is positioned cranially when compared to the planning. A negative occlusal plane rotation discrepancy means the maxilla is rotated clockwise when
compared to the planning.

284
L. Trevisiol, M. Bersani, A. Martinez Garza et al. Journal of Cranio-Maxillo-Facial Surgery 51 (2023) 280e287

Table 4
Method reliability verification: Inter-observer variations and interclass correlations (ICC) for the evaluated movements.

Mandibular inter-operator ICCs Maxillary inter-operator ICCs

Mean discrepancy 95% Confidence interval ICC Mean discrepancy 95% Confidence interval ICC

Sagittal discrepancies
Anteroposterior correction of R. incisor 0.155 (-0.6584 - 0.9679) 0.998 0.139 (-1.0025 - 0.7253) 0.989

Lateral discrepancies
Midline correction 0.041 (-0.8109 - 0.7324) 0.991 0.014 (-0.7271 - 0.7553) 0.958
Lateral correction of R. 7 0.092 (-1.3229 - 1.1395) 0.948 0.063 (-0.9482 - 0.8216) 0.978
Lateral correction of L. 7 0.090 (-1.3208 - 1.1399) 0.948 0.014 (-1.1389 - 1.1119) 0.971

Vertical discrepancies
Vertical correction of R. incisor 0.195 (-1.3655 - 0.9747) 0.983 0.060 (-0.8555 - 0.9762) 0.980
Canine CANT 0.084 (-1.1262 - 0.9577) 0.915 0.013 (-0.7711 - 0.7970) 0.969
Vertical correction of R. 7 0.092 (-0.9206 - 0.7364) 0.989 0.059 (-0.5154 - 0.3969) 0.993
Vertical correction of L. 7 0.091 (-1.1929 - 1.0119) 0.983 0.044 (-0.5502 - 0.4622) 0.995

Rotational discrepancies
Occlusal plane rotation 0.279 (-2.0852 - 1.5280) 0.983 0.228 (-1.3775 - 1.8335) 0.985

Total 0.090 (-1.2510 - 1.0712) 0.995 0.000 (-0.9456 - 0.9449) 0.992

Md þ Mx grand total 0.045 (-1.1070 - 1.0168) 0.994 (0.993e0.995)

“R. 7 ” and “L. 7 ” indicate the right second molar and left second molar, respectively.

Fig. 4. Discrepancy distributions between virtual surgical planning and surgical results.

other method limitations such as difficult clinical interpretation of establishing a coordinate system, is essential to avoid linear mea-
results. surement errors (Ruellas et al., 2016).
Another critical point concerns the protocol used to acquire These measures, together with a blind assessment of move-
CBCT images. Many studies did not clearly specify this step in their ments carried out and the number of subjects, helped to reduce
method descriptions (Antonini et al., 2020; Chin et al., 2017; Falter potential sources of bias in data collection, and allowed us to obtain
re et al., 2019). Our study's post-
et al., 2013; Hsu et al., 2013; Marlie a data sample referring to surgical movements that were as
surgical CBCT images were taken in maximum intercuspation objective, reliable and accurate as possible. This was confirmed by
without elastics approximately 3 weeks after surgery to permit the high ICC for all measurements (0.994) (95% CI ¼ 0.993e0.995).
intra-articular edema reabsorption. Pre- and post-operative vol- Regarding the comparison between planned and actually per-
umes were voxel-based superimposed and oriented according to formed movements, our results show that the intraoperative
natural head position values determined clinically at the time at transfer of VSP with a mandible-first sequence is highly accurate
which presurgical data were collected. Volume orientation, that is, and reproducible in all space dimensions. No discrepancies be-
tween planned movements and those executed were statistically

285
L. Trevisiol, M. Bersani, A. Martinez Garza et al. Journal of Cranio-Maxillo-Facial Surgery 51 (2023) 280e287

Fig. 5. Note how, in the case of large counterclockwise mandibular rotation (CCWr), it is evident how the incisor translates with pure vertical movement more than the molar.

significant with the exception of sagittal movements at the incisor could be incorrect identification of the virtual axis of the condylar
level. In any case, the mean discrepancy appeared to be lower than rotation upon which planning depends (Antonini et al., 2020; Chin
2 mm, the threshold considered to be clinically relevant in other et al., 2017). This could cause increasing inaccuracy of sagittal
studies (Chin et al., 2017; Donatsky et al., 1997; Hsu et al., 2013; Tng movements as occlusal plane rotation increases. Correlations found
et al., 1994; Zhang et al., 2016). It is important to emphasize that the in the current study validate the accuracy of virtual condylar hinge
accuracy of VSP does not actually depend on the planning pro- axis for a mandible-first sequence, because the entity of the
cedure alone. There are a number of other factors that may have a mandibular distal segment's rotation and the sagittal and vertical
bearing on results: the method used to collect pre-operative data discrepancy at the lower incisor are not correlated significantly,
(as stated above), surgical sequence (Liebregts et al., 2017), fixation showing values of p ¼ 0.27 and p ¼ 0.15, respectively.
methods (Van Sickels and Richardson, 1996), operator expertise To better investigate causes of errors, a larger cohort should be
(Antonini et al., 2020), and the intraoperative management of the analyzed and different study designs tried. This could help to
condylar position (Ellis, 1994). For example, incorrect repositioning identify which of the numerous variables considered might be the
of the condyle within the fossa during osteosynthesis, hypotheti- cause of error.
cally due to factors such as improper technique, poor bone segment
passivation, or plates and screws, can also lead to a final position of 5. Conclusions
the mandible that is different from that planned, regardless of the
quality of the VSP process (Antonini et al., 2020; Baan et al., 2016; The results of this study confirm the possibility of high-accuracy
McMillen, 1972; Ritto et al., 2018; Sharifi et al., 2008). Maxillary intraoperative VSP transfers with a mandible-first sequence
impaction by osteotomy different from that planned could also lead regardless of preoperative malocclusion and potential maxillary
to discrepancies (Antonini et al., 2020; Otranto de Britto Teixeira segmentation. The assessment method guarantees reliability of
et al., 2020; Sun et al., 2013). The roles that these variables play results and reduces the risk of inaccuracies linked to the analysis
seem to be underestimated or at least inadequately evaluated in itself.
most studies on this topic. The only statistically significant discrepancy was in the sagittal
Recurrence of one of the problems listed above, which can occur direction, and this seems to be mainly due to surgical factors rather
in patients regardless of the type and amount of movement carried than virtual planning.
out, could determine a systematic difference between the planned In conclusion, VSP accuracy is not solely dependent on the
and the final outcome. software used or the digital planning approach itself; sequencing
The current study seems to confirm this hypothesis only where and surgical techinique can have a significant influence as well.
sagittal movements were concerned, which were approximately
1 mm less than those planned. These differences were not signifi- Funding
cantly correlated with movement amplitude and direction. They
seem to be due mainly to surgical factors rather than sequencing; This research did not receive any specific grant from funding
no relationship was found even in cases with movements consid- agencies in the public, commercial, or not-for-profit sectors.
ered unfavourable for mandible-first sequencing (CW rotation of
the occlusal planes, vertical impaction of the maxilla). Furthermore, Declaration of competing interest
these results are in accordance with mandible-first literature
(Bobek et al., 2015; De Riu et al., 2018; Liebregts et al., 2017). Greater The authors declare that they have no conflicts of interest in
accuracy in mandible repositioning compared with the maxilla can regard to this work.
be explained by the fact that only an intermediate bite with no final
one and a mandible-first sequence were used during the proced- References
ures outlined in this study. This means that only mandibular
repositioning is controlled intraoperatively by VSP. Subsequently,
, J., Lippold, C., Segatto, E., 2020. Accuracy of virtual planning in
Alkhayer, A., Piffko
orthognathic surgery: a systematic review. Head Face Med. 16, 34. https://
any discrepancy in a mandible's distal segment positioning would doi.org/10.1186/s13005-020-00250-2.
be amplified in the maxilla (Stokbro et al., 2016). In addition, the Antonini, F., Borba, A.M., Pagnoncelli, R.M., Han, M., Markiewicz, M.R., Miloro, M.,
upper arch's final position is determined not by a final splint but by 2020. Does a learning curve exist for accuracy in three-dimensional planning
for maxillary positioning in bimaxillary orthognathic surgery? Int. J. Oral
using the lower arch as a template. A further reason for inaccuracy
Maxillofac. Surg. 49, 787e793. https://doi.org/10.1016/j.ijom.2019.10.005.

286
L. Trevisiol, M. Bersani, A. Martinez Garza et al. Journal of Cranio-Maxillo-Facial Surgery 51 (2023) 280e287

Arnett, G.W., Bergman, R.T., 1993a. Facial keys to orthodontic diagnosis and treat- jaw surgery: Le Fort I osteotomy and mandibular setback surgery with rigid
ment planning. Part I. Am. J. Orthod. Dentofacial Orthop. 103, 299e312. https:// fixation. Oral Surg. Oral Med. Oral Pathol. Oral Radiol. Endod. 111, 681e687.
doi.org/10.1016/0889-5406(93)70010-L. https://doi.org/10.1016/j.tripleo.2010.08.001.
Arnett, G.W., Bergman, R.T., 1993b. Facial keys to orthodontic diagnosis and treat- Koo, T.K., Li, M.Y., 2016. A guideline of selecting and reporting intraclass correlation
ment planning. Part II. Am. J. Orthod. Dentofacial Orthop. 103, 395e411. https:// coefficients for reliability research. J Chiropr Med 15, 155e163. https://doi.org/
doi.org/10.1016/s0889-5406(05)81791-3. 10.1016/j.jcm.2016.02.012.
Arnett, G.W., D’Agostino, A., Grendene, E., McLaughlin, R.P., Trevisiol, L., 2022a. Liebregts, J., Baan, F., de Koning, M., Ongkosuwito, E., Berge
, S., Maal, T., Xi, T., 2017.
Combined orthodontic and surgical open bite correction: principles for success. Achievability of 3D planned bimaxillary osteotomies: maxilla-first versus
Part 2. Angle Orthod. 92, 431e445. https://doi.org/10.2319/123121-959.1. mandible-first surgery. Sci. Rep. 7, 9314. https://doi.org/10.1038/s41598-017-
Arnett, G.W., Tamborello, J.A., Rathbone, J.A., 1992. Temporomandibular joint ram- 09488-4.
ifications of orthognathic surgery. In: Bell, W.H. (Ed.), Modern Practice in Lin, H.-H.L.L.-J., 2015. Three-dimensional computer-assisted surgical simulation and
Orthognathic and Reconstructive Surgery. WB Saunders Co, Philadelphia. intraoperative navigation in orthognathic surgery: a literature review.
Arnett, G.W., Trevisiol, L., Grendene, E., McLaughlin, R.P., D’Agostino, A., 2022b. J. Formos. Med. Assoc. 114 (4), 300e307.
Combined orthodontic and surgical open bite correction. Angle Orthod. https:// Lisboa, C. de O., Masterson, D., da Motta, A.F.J., Motta, A.T., 2015. Reliability and
doi.org/10.2319/101921-779.1. reproducibility of three-dimensional cephalometric landmarks using CBCT: a
Baan, F., Liebregts, J., Xi, T., Schreurs, R., de Koning, M., Berge
, S., Maal, T., 2016. systematic review. J. Appl. Oral Sci. 23, 112e119. https://doi.org/10.1590/1678-
A new 3D tool for assessing the accuracy of bimaxillary surgery: the OrthoG- 775720140336.
nathicAnalyser. PLoS One 11, e0149625. https://doi.org/10.1371/ Luebbers, H.-T., Messmer, P., Obwegeser, J.A., Zwahlen, R.A., Kikinis, R., Graetz, K.W.,
journal.pone.0149625. Matthews, F., 2008. Comparison of different registration methods for surgical
Bazina, M., Cevidanes, L., Ruellas, A., Valiathan, M., Quereshy, F., Syed, A., Wu, R., navigation in cranio-maxillofacial surgery. J. Cranio-Maxillo-Fac. Surg. 36,
Palomo, J.M., 2018. Precision and reliability of Dolphin 3-dimensional voxel- 109e116. https://doi.org/10.1016/j.jcms.2007.09.002.
based superimposition. Am. J. Orthod. Dentofacial Orthop. 153, 599e606. Marlie re, D. -a.-A., Deme
trio, M.-S., Schmitt, A.-R.-M., Lovisi, C.-B., Asprino, L.,
https://doi.org/10.1016/j.ajodo.2017.07.025. Chaves-Netto, H.-D.-M., 2019. Accuracy between virtual surgical planning and
Bobek, S., Farrell, Brian, Choi, C., Farrell, Bart, Weimer, K., Tucker, M., 2015. Virtual actual outcomes in orthognathic surgery by iterative closest point algorithm
surgical planning for orthognathic surgery using digital data transfer and an and color maps: a retrospective cohort study. Med. Oral Patol. Oral Cir. Bucal 24,
intraoral fiducial marker: the Charlotte method. J. Oral Maxillofac. Surg. 73, e243ee253. https://doi.org/10.4317/medoral.22724.
1143e1158. https://doi.org/10.1016/j.joms.2014.12.008. Mazzoni, S., Bianchi, A., Schiariti, G., Badiali, G., Marchetti, C., 2015. Computer-aided
Borikanphanitphaisan, T., Lin, C.-H., Chen, Y.-A., Ko, E.W.-C., 2021. Accuracy of design and computer-aided manufacturing cutting guides and customized ti-
mandible-first versus maxilla-first approach and of thick versus thin splints for tanium plates are useful in upper maxilla waferless repositioning. J. Oral
skeletal position after two-jaw orthognathic surgery. Plast. Reconstr. Surg. 147, Maxillofac. Surg. 73, 701e707. https://doi.org/10.1016/j.joms.2014.10.028.
421e431. https://doi.org/10.1097/PRS.0000000000007536. McMillen, L.B., 1972. Border movements of the human mandible. J. Prosthet. Dent
Chen, Z., Mo, S., Fan, X., You, Y., Ye, G., Zhou, N., 2021. A meta-analysis and sys- 27, 524e532. https://doi.org/10.1016/0022-3913(72)90265-x.
tematic review comparing the effectiveness of traditional and virtual surgical Otranto de Britto Teixeira, A., Almeida, M.A. de O., Almeida, R.C. da C., Maue
s, C.P.,
planning for orthognathic surgery: based on randomized clinical trials. J. Oral Pimentel, T., Ribeiro, D.P.B., Medeiros, P.J. de, Quint~ ao, C.C.A., Carvalho, F. de A.R.,
Maxillofac. Surg. 79, 471.e1e471.e19. https://doi.org/10.1016/ 2020. Three-dimensional accuracy of virtual planning in orthognathic surgery.
j.joms.2020.09.005. Am. J. Orthod. Dentofacial Orthop. 158, 674e683. https://doi.org/10.1016/
Chin, S.-J., Wilde, F., Neuhaus, M., Schramm, A., Gellrich, N.-C., Rana, M., 2017. Ac- j.ajodo.2019.09.023.
curacy of virtual surgical planning of orthognathic surgery with aid of CAD/CAM Perez, D., Ellis, E., 2016. Implications of sequencing in simultaneous maxillary and
fabricated surgical splintea novel 3D analyzing algorithm. J. Cranio-Maxillo-Fac. mandibular orthognathic surgery. Atlas Oral Maxillofac Surg Clin North Am 24,
Surg. 45, 1962e1970. https://doi.org/10.1016/j.jcms.2017.07.016. 45e53. https://doi.org/10.1016/j.cxom.2015.10.004.
de Oliveira, A.E.F., Cevidanes, L.H.S., Phillips, C., Motta, A., Burke, B., Tyndall, D., Ritto, F.G., Schmitt, A.R.M., Pimentel, T., Canellas, J.V., Medeiros, P.J., 2018. Com-
2009. Observer reliability of three-dimensional cephalometric landmark iden- parison of the accuracy of maxillary position between conventional model
tification on cone-beam computerized tomography. Oral Surg. Oral Med. Oral surgery and virtual surgical planning. Int. J. Oral Maxillofac. Surg. 47, 160e166.
Pathol. Oral Radiol. Endod. 107, 256e265. https://doi.org/10.1016/ https://doi.org/10.1016/j.ijom.2017.08.012.
j.tripleo.2008.05.039. Ruellas, A.C. de O., Tonello, C., Gomes, L.R., Yatabe, M.S., Macron, L., Lopinto, J.,
De Riu, G., Virdis, P.I., Meloni, S.M., Lumbau, A., Vaira, L.A., 2018. Accuracy of Goncalves, J.R., Garib Carreira, D.G., Alonso, N., Souki, B.Q., Coqueiro, R. da S.,
computer-assisted orthognathic surgery. J. Cranio-Maxillo-Fac. Surg. 46, Cevidanes, L.H.S., 2016. Common 3-dimensional coordinate system for assess-
293e298. https://doi.org/10.1016/j.jcms.2017.11.023. ment of directional changes. Am. J. Orthod. Dentofacial Orthop. 149, 645e656.
Donatsky, O., Bjørn-Jørgensen, J., Holmqvist-Larsen, M., Hillerup, S., 1997. https://doi.org/10.1016/j.ajodo.2015.10.021.
Computerized cephalometric evaluation of orthognathic surgical precision and Sabour, S., 2020. 3D virtual surgical planning for maxillary positioning and orien-
stability in relation to maxillary superior repositioning combined with tation in orthognathic surgery: methodological issues on accuracy and agree-
mandibular advancement or setback. J. Oral Maxillofac. Surg. 55, 1071e1079. ment. Orthod. Craniofac. Res. https://doi.org/10.1111/ocr.12438.
https://doi.org/10.1016/s0278-2391(97)90282-2 discussion 1079-1080. Sharifi, A., Jones, R., Ayoub, A., Moos, K., Walker, F., Khambay, B., McHugh, S., 2008.
Ellis, E., 1994. Condylar positioning devices for orthognathic surgery: are they How accurate is model planning for orthognathic surgery? Int. J. Oral Max-
necessary? J. Oral Maxillofac. Surg. 52, 536e552. https://doi.org/10.1016/0278- illofac. Surg. 37, 1089e1093. https://doi.org/10.1016/j.ijom.2008.06.011.
2391(94)90085-X. Solow, B., Tallgren, A., 1971. Natural head position in standing subjects. Acta
Epker, B.N., 1977. Modifications in the sagittal osteotomy of the mandible. J. Oral Odontol. Scand. 29, 591e607. https://doi.org/10.3109/00016357109026337.
Surg. 35, 157e159. Stokbro, K., Aagaard, E., Torkov, P., Bell, R.B., Thygesen, T., 2016. Surgical accuracy of
Falter, B., Schepers, S., Vrielinck, L., Lambrichts, I., Politis, C., 2013. Predicted versus three-dimensional virtual planning: a pilot study of bimaxillary orthognathic
executed surgical orthognathic treatment. J. Cranio-Maxillo-Fac. Surg. 41, procedures including maxillary segmentation. Int. J. Oral Maxillofac. Surg. 45,
547e551. https://doi.org/10.1016/j.jcms.2012.11.026. 8e18. https://doi.org/10.1016/j.ijom.2015.07.010.
Fedorov, A., Beichel, R., Kalpathy-Cramer, J., Finet, J., Fillion-Robin, J.-C., Pujol, S., Sun, Y., Luebbers, H.-T., Agbaje, J.O., Schepers, S., Vrielinck, L., Lambrichts, I.,
Bauer, C., Jennings, D., Fennessy, F., Sonka, M., Buatti, J., Aylward, S., Miller, J.V., Politis, C., 2013. Accuracy of upper jaw positioning with intermediate splint
Pieper, S., Kikinis, R., 2012. 3D slicer as an image computing platform for the fabrication after virtual planning in bimaxillary orthognathic surgery.
quantitative imaging network. Magn. Reson. Imaging 30, 1323e1341. https:// J. Craniofac. Surg. 24, 1871e1876. https://doi.org/10.1097/
doi.org/10.1016/j.mri.2012.05.001. SCS.0b013e31829a80d9.
Gaber, R.M., Shaheen, E., Falter, B., Araya, S., Politis, C., Swennen, G.R.J., Jacobs, R., Titiz, I., Laubinger, M., Keller, T., Hertrich, K., Hirschfelder, U., 2012. Repeatability and
2017. A systematic review to uncover a universal protocol for accuracy assess- reproducibility of landmarksea three-dimensional computed tomography
ment of 3-dimensional virtually planned orthognathic surgery. J. Oral Max- study. Eur. J. Orthod. 34, 276e286. https://doi.org/10.1093/ejo/cjq190.
illofac. Surg. 75, 2430e2440. https://doi.org/10.1016/j.joms.2017.05.025. Tng, T.T., Chan, T.C., Ha €gg, U., Cooke, M.S., 1994. Validity of cephalometric land-
Ghoneima, A., Cho, H., Farouk, K., Kula, K., 2017. Accuracy and reliability of marks. An experimental study on human skulls. Eur. J. Orthod. 16, 110e120.
landmark-based, surface-based and voxel-based 3D cone-beam computed to- https://doi.org/10.1093/ejo/16.2.110.
mography superimposition methods. Orthod. Craniofac. Res. 20, 227e236. Tucker, S., Cevidanes, L.H.S., Styner, M., Kim, H., Reyes, M., Proffit, W., Turvey, T.,
https://doi.org/10.1111/ocr.12205. 2010. Comparison of actual surgical outcomes and 3-dimensional surgical
Haas Junior, O.L., Guijarro-Martínez, R., Sousa Gil, A.P. de, Me
ndez-Manjo
n, I., Valls- simulations. J. Oral Maxillofac. Surg. 68, 2412e2421. https://doi.org/10.1016/
Otan ~o

n, A., de Oliveira, R.B., Herna
ndez-Alfaro, F., 2019. Cranial base superim- j.joms.2009.09.058.
position of cone-beam computed tomography images: a voxel-based protocol Van Sickels, J.E., Richardson, D.A., 1996. Stability of orthognathic surgery: a review
validation. J. Craniofac. Surg. 30, 1809e1814. https://doi.org/10.1097/ of rigid fixation. Br. J. Oral Maxillofac. Surg. 34, 279e285. https://doi.org/
SCS.0000000000005503. 10.1016/S0266-4356(96)90002-9.
Hsu, S.S.-P., Gateno, J., Bell, R.B., Hirsch, D.L., Markiewicz, M.R., Teichgraeber, J.F., Zhang, N., Liu, S., Hu, Z., Hu, J., Zhu, S., Li, Y., 2016. Accuracy of virtual surgical
Zhou, X., Xia, J.J., 2013. Accuracy of a computer-aided surgical simulation pro- planning in two-jaw orthognathic surgery: comparison of planned and actual
tocol for orthognathic surgery: a prospective multicenter study. J. Oral Max- results. Oral Surg Oral Med Oral Pathol Oral Radiol 122, 143e151. https://
illofac. Surg. 71, 128e142. https://doi.org/10.1016/j.joms.2012.03.027. doi.org/10.1016/j.oooo.2016.03.004.
Kim, Y.-I., Cho, B.-H., Jung, Y.-H., Son, W.-S., Park, S.-B., 2011. Cone-beam comput-
erized tomography evaluation of condylar changes and stability following two-

287