Introduction

Disorders of the temporomandibular joint (TMJ) are frequent and generally coexist with other abnormalities of the maxillofacial skeleton [1,2,3]. When surgical correction is required, single-stage procedures are feasible and beneficial for patients [1, 4,5,6,7]. Surgical corrective goals may be achieved with adequate planning and surgical execution, preoperative virtual planning (PVP), 3D model printing and intraoperative virtual navigation (IVN) which are useful tools that facilitate treatment of complex disorders. Such tools, generically called “computer-assisted surgery” (CAS); primary endpoint is achieving more predictable and precise postoperative results [8, 9]. They facilitate simple procedures, like removal of a foreign body, as well as complex ones such as those performed for facial trauma, correction of craniomaxillary dysmorphisms, orthognathic surgery, oncological resections or TMJ replacements [10,11,12].

PVP allows to virtually design osteotomies and prostheses, while INV is useful for implementing decisions made during PVP and guiding the surgeon during surgery by providing continuous imaging assessment of the anatomy. Additionally, 3D printing is used to make splints, prostheses and 3D models [11].

The goal of this study is to present the protocol used in our institution for orthognathic surgery associated to unilateral and bilateral TMJ replacement with customized prostheses, performed in a single stage procedure with computer assistance; we present two cases to illustrate this protocol and its results. As secondary aim, we wish to assess the accuracy of the procedure and postoperative clinical course.

Protocol Presentation

In both patients, the general treatment aims were to achieve harmony in the mid- and lower thirds of the face, adequate vertical dimension and facial projection, and general functional rehabilitation, including TMJ (Table 1).

Table 1 Surgical protocol summary
Full size table

Orthognathic surgery and total TMJ replacement, unilateral and bilateral, were performed in a single surgical stage. Preoperative planning was performed with a 64 slice-high resolution computed tomography (CT) of the craniofacial skeleton and neck, without contrast enhancement (Multislice 64, Aquilion, Toshiba Medical Solutions, Erlangen, Germany); the CT scan protocol includes 1-mm sections, obtained with the patient in occlusion and a wax up model in centric relation to determine the condylar location.

Despite the retrospective nature of this study, it was approved by the hospital’s institutional review board (IRB). We have read and followed the guidelines stated in the Helsinki Declaration when treating these patients and preparing this article. The IRB stated no need of informed consent for the developing of this paper.

Case 1

The patient is a 49-year-old lady with a history of recurred condylar hyperplasia, referred due to progressive occlusion dysfunction and lateral deviation of the mandible. On physical examination, her skeletal class was type I, with a slight occlusion overjet and lateral deviation. Metabolic condylar activity was assessed in three bone SPECT exams at 8-month intervals, which showed a 16% difference between both condyles and a CT of the facial skeleton with a 16.3 × 13.6 mm bone lesion, associated to arthrosis of the TMJ and no condyle excursion with open mouth technique. Orthodontic treatment was implemented for presurgical alignment and leveling of dental arches and subsequently perform a total TMJ replacement associated to a sagittal osteotomy of both ramuses. For orthodontic planning, VTO (visual treatment objective) and 3D images were used (PVP).

The VTO and PVP showed a maxillary canted occlusal plane that limited transverse rotation of the mandible (Fig. 1a); therefore, it was decided to combine the bilateral mandibular osteotomies with a type I Le Fort osteotomy. The surgical objectives were:

  • A 3 mm advancement in the maxilla with a 2 mm impaction in the right side and no impaction in the left side.

  • Bilateral sagittal osteotomy of the mandible for transverse rotation

  • Centralization of the lower midline.

  • Total right-sided TMJ replacement.

Fig. 1
figure 1

A 3D reconstruction of the patient’s facial skeleton. a The lines show the maxillary canted plane. b Virtual planning of the mandibular osteotomies. The referral lines are also shown. c The VTO movements were applied to the 3D model; notice that the inferior dental nerve is colored in yellow

Full size image

The PVP was performed with assistance of the MediCAS Software Virtual Platform.

  1. 1.

    Frontal and lateral photographs of the patient were obtained by the orthodontist in the office, with a natural head position and a marker located in the right side of the image.

  2. 2.

    A 3D reconstruction of the patient’s facial skeleton was performed (skin and bone) and a bite plaster model based on the CT.

  3. 3.

    With the use of specific tools, the position of the head was reoriented in the three spatial axes, matching the frontal and lateral photographs, correcting pitch, yaw and roll. Subsequently, reference lines were traced: a true vertical line, an intercanthal line and a true horizontal line (Fig. 1b).

  4. 4.

    The patient’s bite model was fused with his/her CT (match).

  5. 5.

    A 3D reconstruction of the inferior dental nerve was obtained.

  6. 6.

    A Le Fort I osteotomy and bilateral sagittal ramus osteotomies were virtually performed (Fig. 1b, c). The lower view allowed assessing the left mandibular basal bone.

  7. 7.

    The VTO movements were applied to the 3D model (Fig. 1c).

  8. 8.

    The intermediate and final splints were designed.

  9. 9.

    A patient-fitted TMJ prosthesis was designed and manufactured (Fig. 2a); the glenoid component was made of ultra-high molecular weight polyethylene (UHMWPE), using at least 6 2.0-mm screws for fixation (Fig. 2b, c). The mandibular component was built in titanium, and at least 8, 2.4-mm screws were used for fixation (Fig. 2d).

Fig. 2
figure 2

a Virtual design and virtual final position of the prosthesis. Location of mandible component screws. The osteotomy is shown underneath the mandible component. b, c Design of the glenoid component and its screws. d Design of the mandible component fixed with screws

Full size image

3D Printing

The splint was printed in photopolymerizable resin (Stratasys Object 30 printer), the biomodels of the upper and lower maxillaries in powder (3D system Pro Jet 360 printer), the prosthetic models (Stratasys F 270 printer) in PLA (polylactic acid), and the final prosthesis in titanium with a high molecular weight polyethylene glenoid component.

Intraoperative Navigation

For the IVN, the navigator used was a Stryker Navigator, Chart II, Freiburg, Germany.

  1. 1.

    Two 3-mm pins were placed in the chin, to position the active navigation pointer using an Ortholock™ fixation system.

  2. 2.

    A correspondence system between the 3D image and the actual patient’s anatomy was employed, with 4 reference points, and the surface was refined with approximately 40 additional points.

  3. 3.

    The PVP was implemented using the navigation pointer to delineate the osteotomy lines, drill the base of the skull and to position the screws in the glenoid and condylar prostheses.

Surgical Procedure

Through lateral preauricular and retromandibular approaches, the TMJ and mandibular ramus were exposed, and the integrity of the VII cranial nerve and its branches was confirmed with neurostimulation.

A Le Fort I osteotomy was performed in the maxilla, with fixation using an intermediate splint, plus bilateral sagittal osteotomy of the mandibular rami and their fixation with three bicortical screws, using the definitive splint.

The patient was placed in intermaxillary fixation (IMF) using the intermediate splint designed with PVP. Primary and surface registrations of the mandible were performed; the pins were located in the chin. The medial pterygoid, masseter and temporal muscles were detached, and coronoidectomy plus condylectomy were performed according to the planning. The TMJ was exposed, and the capsule and articular meniscus were resected. Primary and surface registrations of the glenoid fossa were performed. Drilling of the base of the skull was performed with IVN assistance, avoiding the middle cranial fossa. The prosthesis was positioned and fixation to the zygomatic arch was performed with at least 6, 2.0-mm screws. A third registration stage followed and at least 8, 2.4-mm screws were used to fix the mandibular component. The IMF was removed and the condyle’s position was checked with opening and closure movements of the mandible. Before concluding the procedure, IVN was used to compare intraoperative execution to PVP.

Case 2

A 29-year-old man without a significant medical history was referred because of abnormal occlusion and sleep apnea. On physical examination his skeletal class was type II with an open bite; the centric relationship of condyles cannot be achieved manually with occlusion. A panoramic X-ray and CT of the facial skeleton showed bilateral condylar resorption. A bone SPECT confirmed the diagnosis. A MRI evidenced absence of corticalization in both condyles. His diagnosis was rheumatoid arthritis.

The goals of VTO were:

  • A 9 mm advancement of the upper maxillary

  • An 8 mm superior impaction for correction of gummy smile

  • A 1 mm correction of the superior midline from right to left

  • A 7 mm mandibular autorotation

  • A 10 mm advancement for genioplasty

  • Total bilateral TMJ replacement

In this patient, preoperative planning and intraoperative sequences were similar to those followed in the previous case. The main difference was that this patient required bilateral TMJ replacement, and therefore, a bilateral approach was used. In terms of registration, a skull post replaced the pins.

No immediate complications occurred, and both patients started rehabilitation the first day post surgery; they were discharged on postoperative days 4 and 7 respectively.

Measurement of Accuracy

Measurements were performed overlapping pre- and postoperative CT scans; for all screws implanted, the head center of the position planned was compared to the actual position.

Results

The difference in mm was 1.72 mm for the glenoid component and 2.16 mm for the condylar prosthesis or mandibular component in the first case (Fig. 3). In the second case, differences in the right side were 2.59 for the glenoid component and 2.06 for the ramus, and in the left side due to the anatomy the difference was greater (3.03 mm for the glenoid component and 6.30 mm for the ramus), although not clinically significant (Fig. 4).

Fig. 3
figure 3

Assessment of outcomes for patient 1. A 3D tomographic reconstruction and axial images were overlapped and compared with the pre- and postoperative CT scans used for surgical planning

Full size image
Fig. 4
figure 4

Measurement of results for patient 2. Tomographic image overlapping shows a mild difference between planning and final outcome, but without clinical impact

Full size image

No joint abnormalities were reported in the long-term follow-up (42 months for the first patient and 16 months for the second patient). Both patients objectively recovered mouth opening and occlusion (32 and 30 mm, respectively) and mild lateralization movements improved subjectively (Figs. 5, 6).

Fig. 5
figure 5

Postoperative occlusion and mouth opening of patient 1

Full size image
Fig. 6
figure 6

Postoperative occlusion and mouth opening of patient 2

Full size image

Discussion

The prevalence of TMJ dysfunction in the general population may vary, and reports range between 5 and 30% [13]. Several of these disorders have an indication of total TMJ replacement, among them condylar hyperplasia, benign or malignant articular neoplasias, keratocysts, trauma, idiopathic condylar resorption, ankylosis, connective tissue diseases, arthritis, etc. Clinically, they may present or be associated to myofascial pain, otalgia, malocclusion, headaches, sleep apnea, etc. [1, 2, 6, 7].

Joint replacement entails restoring function and consequently, an improvement in patient´s quality of life, allowing them to return to their social activities, with less pain, better mouth opening and easier feeding [14, 15]. Regarding quality of life, Mercuri et al. [16] described an 85% improvement compared to their condition prior to TMJ replacement. Additionally, replacement with alloplastic prostheses prevents morbidity of the donor site and maintains a stable occlusion [15, 16].

Disorders of the TMJ are not isolated, and often coexist with other abnormalities of the facial skeleton; therefore, simultaneous correction of both malformations in a single-stage surgical procedure may prove beneficial for the patient [4,5,6,7]. Nadershah et al. [2] considered that the main indications for concomitant surgery are for patients who failed conservative treatment and patients with advanced TMJ condylar resorption or active TMJ pathologic conditions like condylar hyperplasia, osteochondroma, who require surgical TMJ treatment.

Such complex approach requires an adequate PVP and precise surgical execution. The development of PVP software and assistance with IVN has improved the results in selected craniomaxillofacial surgery, like the ones described above. Many authors have assessed the benefits of using these technologies and agreed that the combination of PVP and IVN delivers more accurate and predictable postoperative results when translating virtual surgery to the operating room [4, 5, 9, 17,18,19]. Besides, the design and development of customized prostheses with the use of these tools improves fitting and osseointegration, thus resulting in less friction and micromovements, better stability and possibly longer mean durability [20]. For the first case presented, the prosthesis was designed and manufactured according to planning, by Biomet ® (Jacksonville, Florida, USA), and for the second case, it was designed at our institution and manufactured by KeyMed Devices ® (Buenos Aires, Argentina). Long-term results have shown adequate fitting with subjective improvement in the quality of life in both patients.

TMJ replacement and orthognathic surgery in a single stage has been well-described [1, 4, 5, 7, 21,22,23,24], but none have reported the use of IVN (Table 2). However, all agree that virtual planning is a valuable contribution to surgical accuracy and predictability. Movahed et al. [4] described their staged protocol and concluded that the use of CAS in this type of procedures shortens the laboratory work previous to surgery (fabrication of prostheses, and for setting the stereolithic models) and surgical times and increases accuracy. Wolford [1], from the same study group, presented posteriorly their experience with similar conclusions; they underscore that single stage surgery may be performed but requires expert surgeons as well as precise and careful diagnosis and planning. Also, they addressed that TMJ should always be treated first. Mohaved et al. [4] also noticed that there are yet some areas to develop, such as the elimination of dental models by using laser scanning, in order to recontour the rami and fossae and to perform accurate maxillary surgical movements using CAS technology. In our patients, PVP was performed with mediCAS planner ® (www.medicas3d.com), estimating the necessary advancements and impactions for the upper maxillary in each patient and defining the location of the inferior dental nerve and its relation with the sagittal ramus osteotomies and the overlaps and gaps between them in the inferior view. The PVP also determined the location and fixation means of prosthetic components. Contrary to Wolford’s approach [1], in our patients the orthognathic surgery was performed prior to TMJ replacement, and we associate IVN for further reliability.

Table 2 Review of literature: computed assisted surgery in concurrent TMJ replacement with orthognathic surgery
Full size table

A successful IVN requires an accurate registration; in our experience, similar to other authors, fixed fiducial points are the more stable markers for navigation [12, 17, 19, 25, 26]. We have used bone landmarks for primary and secondary registrations, avoiding soft tissue areas. In the first case, registration was performed three times, relocating the pins each time; in the second case we decided to change the type of navigation pointer to a skull post and the patient was set in IMF; as a result, the registration was made only twice, navigating the maxilla and mandible at the same time. It is worth noticing that registration accuracy is defined by the difference, in millimeters, between the virtual and real coordinate [27], so that the accuracy of optical navigation should be between 0.4 and 1 mm to decrease the margin of error and ensure precision; registration errors can occur but the range of acceptable error is less than 1 mm [25, 27]. Under this statement, we performed the registration on both patients.

The use of IVN allows transferring the surgical plan to the patient [10, 12, 19, 25]. Intraoperatively, it was useful to avoid injuring the base of the skull while drilling the temporal fossa and to perform the osteotomies of the coronoid process and condyle. It was also helpful to drill the mandibular basal bone excess in unilateral condylar hyperplasia and also to correctly positioning and fixating the prosthesis. Additionally, as described in the literature, it allowed confirming, during surgery, the agreement between planning and surgical execution [8, 12, 25].

A postoperative CT scan was used to evaluate the surgical outcomes by comparing and matching the PVP with the post-operative results. According to Levine et al. [8], excellent accuracy is achieved within 1–5 mm. Azarmehr et al. [26], considered a maximum difference of 2 mm between planning and actual results. This difference is accepted in order to call an orthognathic surgery successful (success criterion). In our case, the postoperative outcome had a minimal difference with no functional impairments.

Among the disadvantages of PVP and IVN are the long learning curve (about 30–40 h are required) and their cost [25]. By contrast, prospective studies have shown that, particularly in orthognathic surgery, 3D planning versus the conventional technique decreases duration of the procedure; thus, it is time saving [4, 28]. Therefore, in spite of the learning time required; it will eventually decrease the total procedure time.

In regard to performing both procedures simultaneously, it is worth noting that it prolongs overall intraoperative time, which could eventually be related to infectious complications [29]. Despite this, in our cases, no infections were reported. However, we are aware of this important issue and our aim is to shorten surgical time in the near future.

Conclusion

Complex surgeries of the midface and mandible simultaneously performed with total TMJ replacement are feasible and beneficial for the patient, but their success will depend mainly on an appropriate planning and intraoperative execution. These are complex procedures, usually in the context of a marked distortion of the usual anatomy; hence, the use of PVP, 3D printing and IVN may facilitate their execution.