Modern Surgery-First Approach Concept in Cleft-Orthognathic Surgery: A Comparative Cohort Study with 3D Quantitative Analysis of Surgical-Occlusion Setup

Despite the evident benefits of the modern surgery-first orthognathic surgery approach (reduced treatment time, efficient tooth decompensation, and early improvement in facial esthetics), the challenge of the surgical-occlusion setup acts as a hindering factor for the widespread and global adoption of this therapeutic modality, especially for the management of cleft-skeletofacial deformity. This is the first study to assess three-dimensional (3D) quantitative data of the surgical-occlusion setup in surgery-first cleft-orthognathic surgery. This comparative retrospective study was performed on 3D image datasets from consecutive patients with skeletal Class III deformity who had a unilateral cleft lip/palate (cleft cohort, n = 44) or a noncleft dentofacial deformity (noncleft cohort, n = 22) and underwent 3D computer-assisted single-splint two-jaw surgery by a single multidisciplinary team between 2014 and 2018. They received conventional orthodontics-first or surgery-first approaches. 3D quantitative characterization (linear, angular, and positional measurements) of the final surgical-occlusion setup was performed and adopted for comparative analyses. In the cleft cohort, the occlusion setup in the surgery-first approach had a significantly (all p < 0.05) smaller number of anterior teeth contacts and larger incisor overjet compared to the conventional approach. Considering the surgery-first approach, the cleft cohort presented significantly (all p < 0.05) larger (canine lateral overjet parameter) and smaller (incisor overjet, maxillary intercanine distance, maxillary intermolar distance, ratio of intercanine distance, and ratio of intermolar distance parameters) values than the noncleft cohort. This study contributes to the literature by providing 3D quantitative data of the surgical-occlusion setup in surgery-first cleft-orthognathic surgery, and delivers information that may assist multidisciplinary teams to adopt the surgery-first concept to optimize cleft care.


Introduction
Cleft lip, with or without a cleft palate, is the second most common global birth defect, affecting 1.7 in every 1000 births [1]. A substantial percentage of skeletally mature patients

Material and Methods
The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines were used for reporting the results of this institutional review board-approved (Chang Gung Medical Foundation, protocol 201900008B0) comparative retrospective study. Consecutive skeletally mature patients (finished their growth spurt with no more increase in body height) with Class III occlusion were recruited who had unilateral complete cleft lip/palate (cleft cohort) or noncleft deformity (noncleft cohort) undergoing orthodontic-surgical treatment by the Chang Gung multidisciplinary team between 2014 and 2018. All included patients were treated by the same orthodontist (B.C.-J.P.) and surgeon (L.-J.L.). Demographics (age and gender), orthodontic-surgical (dental characteristics, type of orthognathic approach, time of presurgical therapy, and need for revision bone and/or soft tissue surgery to improve occlusal, maxillary, mandibular, and/or chin morphology within the follow-up), and cone beam computed tomography (CBCT) scan data were retrieved from the Craniofacial Research Center database.
Exclusion criteria were patients with Class I or II skeletal patterns; bilateral cleft deformity; any syndromic diagnosis; previous orthognathic surgery; and/or an incomplete medical/3D image record. Patients who underwent segmental osteotomies (surgically assisted maxillary expansion, anterior subapical osteotomy, and/or segmented maxillary osteotomies) were also excluded.

Orthognathic-Surgery Treatment
Full descriptions of the pre-and postorthognathic-treatment principles used in this center were previously published [22,23,28,[32][33][34][35][36][37][38][39][40][41][42][43]. After a short period of orthodontic preparation for secondary alveolar bone grafting (9-year-old patients), no further surgical intervention or orthodontic treatment have been performed in patients with unilateral cleft lip/palate who present signs of skeletofacial deformity such as Class III skeletal pattern. The orthognathic surgery treatment process has been started when the patients reach the skeletal maturity (>15 and 18 years for female and male patients, respectively).
All included patients were surgically treated by using 3D computer-assisted single-splint two-jaw orthognathic surgery [38][39][40][41]. Selected patients were orthodontically managed by using conventional orthodontics-first or surgery-first approaches [22,23,28,[32][33][34][35][36][37] based on dentition status at presentation: (a) In the conventional approach, surgery was performed after a period of at least seven months of complete orthodontic therapy, including the leveling and alignment of dental arches to eliminate any occlusal interference at surgery, and the removal of all dental compensations to maximize optimal surgical repositioning of the jaws. (b) Patients with different compositions of the dental conditions (Table 1) received the surgery-first orthognathic treatment based on a patient-specific therapeutic planning ( Figure 1). Patients with more and less favorable dental conditions received the surgery-first treatment as it was based on the orthodontist's judgment of achievement of a practicable surgical-occlusion setup as well as anticipation of a feasible postoperative orthodontic treatment. Our team stratified the surgery-first approach into two models. In the standard surgery-first model, surgery was performed with no need for presurgical orthodontic therapy. In the modified surgery-first model, a short period (≤6 months) of orthodontic therapy was performed preoperatively. This presurgical dental adjustment was exclusively implemented for the reduction of potential dental collisions and the minimal decompression of mandibular teeth.
In this study, the 3D quantitative data of standard and modified surgery-first models were initially compared. Both models were then compiled as a unique dataset (surgery-first approach) and adopted for comparative analysis between surgery-first and conventional approaches. (middle) anterior-teeth alignment; and (right) incisor inclination, adopted to distingue (top) more favorable (minimal anterior dental crowding, flat-to-mild curve of Spee, and normal range of angle between basal bone and upper and lower incisors) and (bottom) less favorable dentition status for management by the surgery-first orthognathic approach. As there is broad clinical presentation with variable degrees of association between these dental elements, patient-specific diagnosis and tailored therapeutic planning should be established in a case-by-case basis. Table 1. Spectrum of dental characteristics in cleft cohort before surgery-first model treatment.

3D-Image Acquisition
One month before surgery, a standard craniofacial CBCT scan was performed for each patient using an i-CAT 3D Dental Imaging System (Imaging Sciences International, Hatfield, PA, USA) with the following parameters: 120 kVp, 0.4 mm × 0.4 mm × 0.4 mm voxel size, 40 s scan time, and a 22 cm × 16 cm field of view. The patients' head was positioned with the Frankfort horizontal plane parallel to the ground. Throughout the scan, patients were instructed not to swallow, to keep their mouth closed, and to maintain a centric occlusion bite [32,33,38,40]. Images were stored in Digital Imaging and Communications in the Medicine format and rendered into 3D volumetric images using the Dolphin 3D software package (Dolphin Imaging and Management Solutions, Chatsworth, CA, USA).

Final Surgical-Occlusion Setup
Two weeks before surgery, the final surgical-occlusion setup was manually performed using the dental-cast model method, considering dental midline coincidence, canting, and the relative position of dentitions between maxillary and mandibular arches. The target was to avoid severe postoperative occlusal instability, incomplete or excessive skeletal correction, or skeletofacial deformities such as asymmetry while defining the surgical-occlusion setup.
To transfer the dental-cast method into a digital image, the maxillary and mandibular dental casts and the defined surgical occlusion were scanned ( Figure 2) by using a surface scanner (3-Shape, Copenhagen, Denmark). Using Dolphin software, the dentition in CBCT was superimposed and replaced by a digitalized dental image [42,43]. The 3D skull models were oriented according to the Frankfort horizontal and midsagittal planes. Osteotomy lines were created by segmenting the maxilla (Le Fort I) and mandible (bilateral sagittal split osteotomy). The digitalized dental image was then manipulated by moving the osteotomized distal mandible segment to the fixed maxilla for achievement of the desired final surgical occlusion (Figures 3-5).
variable degrees of association between these dental elements, patient-specific diagnosis and tailored therapeutic planning should be established in a case-by-case basis.
In this study, the 3D quantitative data of standard and modified surgery-first models were initially compared. Both models were then compiled as a unique dataset (surgery-first approach) and adopted for comparative analysis between surgery-first and conventional approaches.

3D-Image Acquisition
One month before surgery, a standard craniofacial CBCT scan was performed for each patient using an i-CAT 3D Dental Imaging System (Imaging Sciences International, Hatfield, PA, USA) with the following parameters: 120 kVp, 0.4 mm × 0.4 mm × 0.4 mm voxel size, 40 s scan time, and a 22 cm × 16 cm field of view. The patients' head was positioned with the Frankfort horizontal plane parallel to the ground. Throughout the scan, patients were instructed not to swallow, to keep their mouth closed, and to maintain a centric occlusion bite [32,33,38,40]. Images were stored in Digital Imaging and Communications in the Medicine format and rendered into 3D volumetric images using the Dolphin 3D software package (Dolphin Imaging and Management Solutions, Chatsworth, CA, USA).

Final Surgical-Occlusion Setup
Two weeks before surgery, the final surgical-occlusion setup was manually performed using the dental-cast model method, considering dental midline coincidence, canting, and the relative position of dentitions between maxillary and mandibular arches. The target was to avoid severe postoperative occlusal instability, incomplete or excessive skeletal correction, or skeletofacial deformities such as asymmetry while defining the surgical-occlusion setup.
To transfer the dental-cast method into a digital image, the maxillary and mandibular dental casts and the defined surgical occlusion were scanned ( Figure 2) by using a surface scanner (3-Shape, Copenhagen, Denmark). Using Dolphin software, the dentition in CBCT was superimposed and replaced by a digitalized dental image [42,43]. The 3D skull models were oriented according to the Frankfort horizontal and midsagittal planes. Osteotomy lines were created by segmenting the maxilla (Le Fort I) and mandible (bilateral sagittal split osteotomy). The digitalized dental image was then manipulated by moving the osteotomized distal mandible segment to the fixed maxilla for achievement of the desired final surgical occlusion (Figures 3-5).        Occlusogram with color map tool displaying 3D (top) maxillary and (bottom) mandibular dental-arch models (left) before and (right) after mandible mobilization for occlusion setup, with surgical occlusal contact on three segments and six teeth. Note the reduction of tooth contact in posterior region (red and green color) due to the creation of anterior-tooth contact (green color), which is a characteristic step adopted for surgical-occlusion setup in surgery-first approach. Red indicates degree of (left) dental collision, which can be (right) thoroughly adjusted before finishing surgicalocclusion setup. For the surgery-first approach, the orthodontic brackets were bonded preoperatively but with no orthodontic tooth movement; wires were placed one day before surgery; tooth #14 was extracted during surgery; orthodontic treatment started during the healing stage by addressing the curve of Spee (a large overjet was designed for the surgery) and the constrict upper posterior teeth with trans-palatal arch appliance. Dental images displayed in Figures 2 and 3.

Virtual Planning
One week before surgery, 3D computer-assisted surgical planning was performed by the surgeon and orthodontist. To achieve a normal jaw relationship with skeletofacial harmony and symmetry, the maxillomandibular complex was mobilized while preserving the surgical-occlusion setup ( Figure 6) [38,40]. If any modification in the surgical-occlusion setup was judged as necessary (e.g., surgical unfeasibility due to inappropriate initial occlusion setup), the overall process was redesigned (including a new final surgical-occlusion setup) until consensus was achieved between surgeon and orthodontist. The fabrication of computer-generated 3D surgical-occlusion splints ( Figure 7) was accomplished by only adopting the final surgical-occlusion setup as template for adjustment of thickness (OrthoAnalyzer software package; 3Shape, Copenhagen, Denmark) and printing (Objet30 OrthoDesk 3D Printer, Stratasys Ltd., Rehovot, Israel). These 3D-printed final surgical-occlusion splints were adopted in the surgical procedures. No intermediate surgicalocclusion setup or intermediate occlusal splint was adopted in the authors' approach. Occlusogram with color map tool displaying 3D (top) maxillary and (bottom) mandibular dental-arch models (left) before and (right) after mandible mobilization for occlusion setup, with surgical occlusal contact on three segments and six teeth. Note the reduction of tooth contact in posterior region (red and green color) due to the creation of anterior-tooth contact (green color), which is a characteristic step adopted for surgical-occlusion setup in surgery-first approach. Red indicates degree of (left) dental collision, which can be (right) thoroughly adjusted before finishing surgical-occlusion setup. For the surgery-first approach, the orthodontic brackets were bonded preoperatively but with no orthodontic tooth movement; wires were placed one day before surgery; tooth #14 was extracted during surgery; orthodontic treatment started during the healing stage by addressing the curve of Spee (a large overjet was designed for the surgery) and the constrict upper posterior teeth with trans-palatal arch appliance. Dental images displayed in Figures 2 and 3.

Virtual Planning
One week before surgery, 3D computer-assisted surgical planning was performed by the surgeon and orthodontist. To achieve a normal jaw relationship with skeletofacial harmony and symmetry, the maxillomandibular complex was mobilized while preserving the surgical-occlusion setup ( Figure 6) [38,40]. If any modification in the surgical-occlusion setup was judged as necessary (e.g., surgical unfeasibility due to inappropriate initial occlusion setup), the overall process was redesigned (including a new final surgical-occlusion setup) until consensus was achieved between surgeon and orthodontist. The fabrication of computer-generated 3D surgical-occlusion splints ( Figure 7) was accomplished by only adopting the final surgical-occlusion setup as template for adjustment of thickness (OrthoAnalyzer software package; 3Shape, Copenhagen, Denmark) and printing (Objet30 OrthoDesk 3D Printer, Stratasys Ltd., Rehovot, Israel). These 3D-printed final surgical-occlusion splints were adopted in the surgical procedures. No intermediate surgical-occlusion setup or intermediate occlusal splint was adopted in the authors' approach.

Surgical Approach
All the included patients received 3D computer-assisted single-splint two-jaw orthognathic surgery (final occlusal splint, 1-piece Le Fort 1 maxillary osteotomy, and bilateral sagittal split osteotomy) according to the previously described by our team [38][39][40]. Using the 3D simulated image as a guiding template (Figure 8), the maxillomandibular complex with 3D-printed final surgicalocclusion splint was moved to the desired position ( Figure 9). The Le Fort I was initially fixed by using 2-mm titanium miniplates placed on the nasomaxillary and zygomatico-maxillary pillars bilaterally, with no rigid fixation in the anterior maxillary walls. Three-hole miniplates and 6 mm screws were routinely bent to match the maxillary contour at the Le Fort I osteotomy line, ensuring the desired position of the maxillomandibular complex. Longer miniplates, i.e., four or five holes, were alternatively employed to overcome potential drawbacks related to the presence of weak maxillary bone or osteotomy-induced fracture in the medial or lateral maxillary pillar region [39]. After Le Fort I fixation, the proximal ramus segment was placed in a relaxed position and gently pushed up to ensure the position of the condylar head in the glenoid fossa. Percutaneous insertion of three bicortical screws 14-16 mm long was performed in the ramus. No interpositional bone graft was used. Intermaxillary fixation was released and the occlusion was evaluated. Genioplasty was finally executed as planned, along with intraoperative judgement. The patients with no intermaxillary fixation were admitted in regular ward for two days following the surgery and then

Surgical Approach
All the included patients received 3D computer-assisted single-splint two-jaw orthognathic surgery (final occlusal splint, 1-piece Le Fort 1 maxillary osteotomy, and bilateral sagittal split osteotomy) according to the previously described by our team [38][39][40]. Using the 3D simulated image as a guiding template (Figure 8), the maxillomandibular complex with 3D-printed final surgical-occlusion splint was moved to the desired position ( Figure 9). The Le Fort I was initially fixed by using 2-mm titanium miniplates placed on the nasomaxillary and zygomatico-maxillary pillars bilaterally, with no rigid fixation in the anterior maxillary walls. Three-hole miniplates and 6 mm screws were routinely bent to match the maxillary contour at the Le Fort I osteotomy line, ensuring the desired position of the maxillomandibular complex. Longer miniplates, i.e., four or five holes, were alternatively employed to overcome potential drawbacks related to the presence of weak maxillary bone or osteotomy-induced fracture in the medial or lateral maxillary pillar region [39]. After Le Fort I fixation, the proximal ramus segment was placed in a relaxed position and gently pushed up to ensure the position of the condylar head in the glenoid fossa. Percutaneous insertion of three bicortical screws 14-16 mm long was performed in the ramus. No interpositional bone graft was used. Intermaxillary fixation was released and the occlusion was evaluated. Genioplasty was finally executed as planned, along with intraoperative judgement. The patients with no intermaxillary fixation were admitted in regular ward for two days following the surgery and then clinically examined based on regular surgical and orthodontic appointments. A liquid diet was recommended in the first week, followed by a soft diet in the second week. clinically examined based on regular surgical and orthodontic appointments. A liquid diet was recommended in the first week, followed by a soft diet in the second week.   . Single-splint two-jaw orthognathic surgery principle. Both maxilla (Le Fort I segment) and mandible (two proximal ramus segments and one distal segment) were completely osteotomized, fixed in the final occlusion using the 3D-printed final surgical-occlusion splint, and moved as an integrated maxillomandibular complex to the 3D-simulated position. To transfer the 3D planning to actual surgery, measurements in maxillary pillars bilaterally (Figure 8), face bow-based midline checking (nasal dorsum and tip, lips, maxilla, dental arches, and chin areas), and middle and lower facial third proportions judgments were used as reference. For this, the maxillomandibular complex was moved in six potential directions, including pitch, roll, and yaw rotations (blue arrows) and enbloc linear horizontal (left or right shifts and advancements or setbacks in the antero-posterior direction) and vertical (extrusion or intrusion) movements (green arrows).

3D Quantitative Analysis
The 3D image datasets displaying the final surgical-occlusion setup adopted for surgery were included for analysis as they represented the occlusion setup in the context of surgical feasibility. 3D quantitative analyses of occlusion characteristics were performed based on dentoskeletofacial parameters defined in a previous investigation [32]: dental-occlusion contacts (number and location), overjet/overbite, angle molar relation (Class I, II, or III), posterior open bite, transverse arch coordination, dental inclination, midline and transverse discrepancies, and jaw relationship (ANB angle and A-point-nasion-B-point angle). All 3D image datasets were analyzed by an investigator with no information about the type of orthodontic approach by using Dolphin software tools (line, angle, and occlusogram with a color map). Twenty randomly selected patients' CBCT scans were measured in duplicate, with one-month interval between each measurement.
Accuracy of surgical occlusion was determined by assessing the number of occlusions requiring two setups [32]. The 3D CBCT-derived cephalometric normative data for the Taiwanese Chinese Figure 9. Single-splint two-jaw orthognathic surgery principle. Both maxilla (Le Fort I segment) and mandible (two proximal ramus segments and one distal segment) were completely osteotomized, fixed in the final occlusion using the 3D-printed final surgical-occlusion splint, and moved as an integrated maxillomandibular complex to the 3D-simulated position. To transfer the 3D planning to actual surgery, measurements in maxillary pillars bilaterally (Figure 8), face bow-based midline checking (nasal dorsum and tip, lips, maxilla, dental arches, and chin areas), and middle and lower facial third proportions judgments were used as reference. For this, the maxillomandibular complex was moved in six potential directions, including pitch, roll, and yaw rotations (blue arrows) and en-bloc linear horizontal (left or right shifts and advancements or setbacks in the antero-posterior direction) and vertical (extrusion or intrusion) movements (green arrows).

3D Quantitative Analysis
The 3D image datasets displaying the final surgical-occlusion setup adopted for surgery were included for analysis as they represented the occlusion setup in the context of surgical feasibility. 3D quantitative analyses of occlusion characteristics were performed based on dentoskeletofacial parameters defined in a previous investigation [32]: dental-occlusion contacts (number and location), overjet/overbite, angle molar relation (Class I, II, or III), posterior open bite, transverse arch coordination, dental inclination, midline and transverse discrepancies, and jaw relationship (ANB angle and A-point-nasion-B-point angle). All 3D image datasets were analyzed by an investigator with no information about the type of orthodontic approach by using Dolphin software tools (line, angle, and occlusogram with a color map). Twenty randomly selected patients' CBCT scans were measured in duplicate, with one-month interval between each measurement.
Accuracy of surgical occlusion was determined by assessing the number of occlusions requiring two setups [32]. The 3D CBCT-derived cephalometric normative data for the Taiwanese Chinese population were adopted as the reference value of the jaw relationship (ANB angle = 3.3 ± 1.6 (0.5-6.2) degrees [44]).

Statistical Analysis
In descriptive analysis, data were presented as means ± standard deviations. It was verified that the data were normally distributed by using the Kolmogorov-Smirnov test. The Student t-test and chi-square test were used for the comparative analyses. Two-sided values of p < 0.05 were considered statistically significant. All analyses were performed using SPSS Version 17.0 (IBM Corp., Armonk, NY, USA).

Results
No new surgical-occlusion setup was required during virtual planning. All patients were treated by two-jaw orthognathic surgery, with no intraoperative problems with the 3D-printed final-occlusion splints. On average, a normal jaw relationship was noticed after virtual planning in the cleft and noncleft cohorts (ANB angle = 3.4 and 3.2 degrees, respectively; p > 0.05). Three patients with cleft and one patient with noncleft presented with lip or chin numbness at 1-6 months postoperatively, with full recovery at long-term evaluations. No wound infection, postoperative hemorrhage/hematoma, or requirement or request for revisionary surgery during follow-up was observed in the cleft and noncleft cohorts.

Time of Presurgical Orthodontic Therapy
The time for presurgical orthodontic therapy was similar between cohorts, with 11.7 ± 3.8 and 10.2 ± 7.4 (p = 0.447) months in the conventional orthodontics-first approach, and 4.9 ± 1.6 and 4.0 ± 1.0 (p = 0.171) months in the modified surgery-first model for the cleft and noncleft cohorts, respectively. Patients who underwent the standard surgery-first model had no presurgical orthodontic therapy.

Primary Endpoint
No significant difference was observed in comparison between standard and modified surgery-first models for all parameters ( Table 2).  There were significant (all p < 0.05) differences in the comparison between surgery-first and conventional orthodontics-first approaches in the number of anterior-tooth contacts and incisor overjet parameters, with no significant difference for the remainder of the tested parameters (Table 3).   (Table 4).

Secondary Endpoint
Considering the surgery-first approach, the cleft cohort presented significantly (all p < 0.05) larger (canine lateral overjet parameter) and smaller (incisor overjet, maxillary intercanine distance, maxillary intermolar distance, ratio of intercanine distance, and ratio of intermolar distance parameters) values than the noncleft cohort (Table 5).

Reliability
Intra-investigator reliability was considered excellent (intraclass correlation coefficients = 0.898 to 0.975) for all quantitative parameters.

Discussion
In this comparative study of occlusion setup, the primary endpoint-related data releveled a smaller number of anterior teeth contacts and larger incisor overjet in the surgery-first cleft-orthognathic surgery approach than the conventional cleft-orthognathic surgery approach, which clinically represents an incisor decompensation postponed after surgery. Patients with cleft also had smaller overjet and higher anterior contacts in cleft side with conventional orthognathic surgery than surgery first approach, which characterizes the typical status of dentition in surgery first-treated patients who had the upper incisors positioned in a more upright position due to surgical procedure-derived scar contraction. Moreover, the secondary endpoint-related data demonstrated a larger canine lateral overjet and smaller incisor overjet and maxillary transversal-related distances in the surgery-first cleft-orthognathic surgery approach than surgery-first noncleft-orthognathic surgery approach, which clinically represents the cleft-associated dental anomalies and transverse maxillary collapse.
It is important to emphasize that the surgical-occlusion setup is certainly more technically demanding in cleft than non-cleft deformities due to the complex cleft-related dental abnormalities, such as irregular arch form and shape as well as teeth anomalies [11,55,56]. In the conventional orthodontics-first approach, as presurgical therapy brings maxillary and mandibular teeth into ideal relationships to the underlying skeletal bases, the surgical-occlusion setup is very close to the final occlusion, i.e., the ideal occlusion [11][12][13]. When embracing the surgery-first approach, dental alignment, arch leveling, and coordination, and incisor decompensation are deferred for postsurgical management; the surgical-occlusion setup is consequently different from the final (ideal) occlusion at the end of treatment [22,23,28]. Not only can anteroposterior dental movements be orthodontically adjusted postoperatively, but also transverse and vertical dental movements can be successfully achieved due to the increased metabolic turnover of the regional acceleratory phenomenon [36]. Surgical occlusion in surgery-first treatment was thus set as a treatable malocclusion [22,23,28]. A major concern for this setup is the accurate estimation of the required space for postsurgical dental movement with many combinations of potential directions [22,23,28]. These challenges are probably the major hampering factors for the widespread use of this technique in cleft centers globally.
In this center, the indication of surgery-first orthognathic surgery treatment has been variable [22,23,[57][58][59]. In our orthognathic surgery workflow, the combination of accurate clinical examinations and high-quality 3D imaging has permitted a precise preoperative diagnosis that encompasses the many deviations of involvement of the dental, skeletal, and facial soft tissue elements, with less favorable patterns of dental characteristics being not considered contraindication for the surgery-first approach. The rate of indication of this surgery-first protocol has mainly been determined by level of orthodontic experiences to evaluate the accomplishment of a workable surgical-occlusion setup and to anticipate an achievable postoperative orthodontic treatment planning [22,23,[57][58][59], with senior experienced professionals (not included as co-authors of this current study) reaching a rate of 100% for surgery-first-based treatments [59]. Therefore, due to the accumulated experience of our team with a high-volume of surgery-first noncleft orthognathic-surgery procedures [22,23,28,[32][33][34][35][36][37], the surgery-first approach has progressively been adopted for cleft-skeletofacial deformity. The regular use of virtual simulation has also assisted the change of our cleft practice, as CBCT-derived images allow three-dimensionally appraising the precision of the surgical-occlusion setup in terms of residual or induced skeletal deformity with a designation of surgical feasibility before the actual procedure [40]. For the surgery-first approach, we have indicated the surgical procedure in patients with no need for presurgical therapy or requiring a short period of therapy (standard and modified models, respectively). Other proponents of the surgery-first approach have also adapted models for a short preparatory phase (e.g., "minimal" and "early"), with presurgical therapy ranging 1-6 months, and the preservation of key advantages of the surgery-first pathway (i.e., immediate postoperative improvement of facial appearance with substantial reduction of total treatment time) [45][46][47][48][49][50][51]. In the current study, no differences were found between the cleft and noncleft cohorts for presurgical-therapy time in the modified surgery-first model. Moreover, no need for revisionary surgery was observed during follow-up. These aspects reveal that it is clinically feasible to apply the principles of selection of patient dentition in the cleft scenario with achievement of the desired surgical-occlusion setup, but with no compromise of time and surgical achievability parameters.
As expected, the standard and modified surgery-first models had no significant difference for all tested parameters, reinforcing that the same principles were adopted during the surgical-occlusion setup of the surgery-first approach regardless of a short period of presurgical orthodontic therapy. Logically, patients managed with the modified model had slight differences in dentition status at presentation in our center compared to patients managed by the standard model. Importantly, the main target of the modified surgery-first treatment was not to transform a patient's dentition with indication for conventional orthodontics-first approach into a favorable dentition to receive a surgery-first approach. Presurgical orthodontics was actually only applied to reduce potential premature contact between maxillary and mandible teeth with the removal of severe occlusal interference enhancing stable surgical occlusion. Minimal decompression of mandibular teeth was also performed when necessary, but with no attempt for decompression of the maxillary teeth.
Different strategies have been employed during surgical-occlusion setup to compensate for space for dental alignment, and arch leveling and coordination after surgery, but with no consensus among advocates of the surgery-first approach and no quantitative data for cleft-related treatment [22,23,[27][28][29][30][31][32][33][34][35][36][37][38]. Appraisal of comparative analyses of this study reveals 3D quantitative-based practical fundamentals for surgical-occlusion setup in surgery-first cleft-orthognathic surgery. In the conventional orthodontics-first approach, surgical occlusion was ideally set as a normal overjet (2 mm) and overbite (2 mm) and Class I molar relationship [60]. Because compensation of horizontal mandibular relapse was planned for with a 2 mm overcorrection in surgery-first treatment [28,32], the Class II molar relationship was frequently set in the occlusion of both cleft and noncleft sides. As incisor decompensation was deferred after surgery [28,32], analyses of the cleft cohort exhibited a significantly larger overjet in the surgery-first than the conventional approach. However, the noncleft cohort (mean of 4.37 mm) had a significantly larger overjet than the cleft cohort (mean of 3.31 mm). This is not surprising because the upper incisors of patients with clefts are positioned in a more upright position due to scar contracture from previous surgical interventions [11,55,56].
Definitions of stable occlusion were previously described [15,16,27,30]. To achieve proper tooth contacts with at least three-point teeth contact (preferably one and two at the anterior and bilateral posterior regions, respectively), increasing the posterior open bite by pitch counterclockwise rotation of the distal mandibular segment was formerly recommended [28]. This compromise of superoinferior dental position in the posterior region to attain a better setup in the anterior region was adopted in our cohorts in a case-by-case basis, with the posterior open bite respecting the limit of postoperative orthodontic tooth movement (<10 mm) [22,23,28,32]. Based on the current quantitative data, stable occlusion can be achieved by occlusal contact on one (anterior region), two, or three (most frequent pattern in both cleft and noncleft cohorts) segments or occlusal contact on five to seven teeth, with all of the included patients presenting with at least one point of contact in the anterior maxillary segment. This quantitative pattern of surgical-occlusion setup is similar to a previous report showing quantitative data in noncleft Class III skeletal deformity [28], suggesting that it is possible to achieve stable occlusion even in patients with clefts and associated dental anomalies.
It was advised to not include posterior crossbite at setup [27,31], but our strategies in the transverse dimension emphasized the coordination of jaw midlines instead of the dental arch to avoid positional jaw asymmetry [28]. Current data show that the cleft cohort had significantly smaller values for maxillary transversal-related distances than the noncleft cohort, but with no difference for the mandible region. Transversal deficiencies secondary to scar-tissue contraction are one of the major concerns for professionals treating patients with clefts [61][62][63]. In our center, the rate of maxillary segmental surgery to correct maxillary transversal-related issues has decreased over the years, as selection of patients for each type of procedure, and the technical details have accordingly evolved. Segmental surgery has only been indicated in patients with severe skeletal crossbite. Arch coordination is deferred after surgery with the surgery-first approach, with posterior dental crossbite and mild skeletal crossbite being orthodontically corrected, for example, by bending orthodontic archwire, or inter-or intra-arch elastics.
The potential limitations of this study should be addressed. Due to the adopted study design, we do not provide inter-investigator reliability for quantitative analysis and intra-and inter-professional reliability for occlusion setup, deserving future investigation by using a distinct methodological approach [64,65]. Our results are restricted to a relatively limited number of patients. An a priori sample size and power calculation could not be defined due to the methodological heterogeneity between the current study and prior literature. We also do not provide post-hoc power analysis due to the inadequacy of this particular method [66,67]. In addition, our findings are based on a specific subgroup of young adult patients with unilateral complete cleft lip/palate who were managed by the same orthodontic and surgeon professionals by using two types of orthodontic approaches and a particular surgical technique (3D-assisted single-splint two-jaw procedure). Moreover, patients with variable degree of the dental presentation (curve of Spee, anterior-teeth alignment, incisor inclination, and present dentition) were included in this study. This represents the orthognathic surgery practice in this center and further details of the patient-specific approach have previously been described [22,23,28,[32][33][34][35][36][37]39,[57][58][59]. This study presented 3D quantitative data derived from the final surgical-occlusion setups of patients who actually received orthognathic treatment and presented no need for revision procedure during follow-up, which denoted satisfactory results. As the patient cohorts were not selected based on surgical results (satisfactory or unsatisfactory), the bias related to analyses performed only on the satisfactory results was considerably reduced, which infers therapeutic feasibility in the cohorts reported here. Moreover, as the applied surgical-orthodontic treatment would considerably vary depending on the dentition status of each patient, orthodontists and surgeons should be aware of the principles and limits of the surgery-first approach during the definition of patient-specific diagnosis and the therapeutic plan (including the prediction of postsurgical change). However, this study does not provide postoperative stability-related statistics, long-term follow up data on results, or information that may guide postoperative arch coordination and dental decompensation, deserving further investigation.
This study did not answer all issues about the addressed topic; however, the current 3D quantitative data can be adopted and adapted by other multidisciplinary teams as initial benchmark values for surgical-occlusion setup in surgery-first cleft-orthognathic surgery. The indication barriers for the surgery-first approach are continuously changing, and we expect that a higher number of patients with clefts would benefit from this modern approach in the future. This may result in changes of the current delivery of cleft-orthognathic surgery care.

Conclusions
This comparative study of occlusion setup showed: (1) similar 3D quantitative characteristics in standard and modified surgery-first models for the cleft cohort; (2) a smaller number of anterior teeth contacts and larger incisor overjet in the surgery-first cleft-orthognathic surgery approach than the conventional cleft-orthognathic surgery approach; and (3) a larger canine lateral overjet and smaller incisor overjet and maxillary transversal-related distances in the surgery-first cleft-orthognathic surgery approach than surgery-first noncleft-orthognathic surgery approach.