Modern surgical pathology of the head and neck has roots from the late 19th century, but the most significant leaps in diagnosis and treatment of these disease entities have rapidly developed over the last 35 years. Significant advances in clinical research were transferred to the operating room environment, and clinicians were able to treat benign and malignant maxillofacial lesions with increased predictability. The advent of microvascular free tissue transfer transformed severely debilitating and disfiguring operations into single-stage ablative–reconstructive procedures. The addition of this type of reconstruction has allowed ablative surgeons to take on the most difficult cases and obtain extremely functional and aesthetic results. At New York University Langone Medical Center, the authors are carrying these principles forward and are using digital technology to improve surgical treatment of head and neck pathology. This article will focus on how computer-aided design and computer aided manufacturing (CAD-CAM) has greatly impacted surgical pathology and reconstruction of the head and neck. The methods are described, and several cases are illustrated.
At New York University Langone Medical Center, the authors have been utilizing 3-dimensional facial analysis and virtual surgical simulation in all of their craniomaxillofacial reconstructive and ablative cases. Over the past 3 years, over 90 cases have been planned, modeled, and executed in this manner, leading to more reliable and predictable outcomes. The authors have been continuously refining these techniques, which have truly revolutionized the way they diagnose, treat, and reconstruct head and neck diseases and defects. In the modern computer era, digital planning has been the standard in architectural design, engineering, and biomedical fabrication. In all aspects of surgery, proper planning facilitates more predictable operative results, but prior to the use of virtual planning, much of this relied on 2-dimensional imaging and surgical trial and error. These techniques are teachable, have a shallow learning curve, allow for precise and anatomic bony resection and reconstruction, and ultimately decrease surgical time.
The goal of this article is to illustrate the ease in which virtual surgery and computer-aided design can be utilized by the surgeon to decrease operative time and create accurate postoperative results when compared with traditional craniomaxillofacial surgical treatment planning. The authors will review their methodology in approaching some of these problems and illustrate application of these techniques. This will only be a small representation of the types of cases to which the authors have applied these techniques, and the authors feel the applications will broaden as surgeons find the ease in which their operations can be planned and executed. These techniques have become the authors’ preferred method for complex craniomaxilliofacial surgery and reconstruction.
Evolution of technique
The evolution of the authors’ current technique initially involved the use of stereolithographic models as templates on which to preoperatively bend plates around the presumed bony resection, to develop plates to work around exophytic lesions, and to utilize for intraoperative reference to help us to preoperatively plan operations. The authors first utilized this technique and these models for mandibular tumor ablation and reconstruction. This was an area where the authors felt there was great variability in outcome based off of how well all the bony segments were reconstructed. Although helpful, this technique was still quite labor intensive, and there was still often guesswork and room for error, both in performing osteotomies of the mandible and fibula correctly, aligning the released jaw segments, and in setting of the fibula into the resection site.
The authors’ current technique involves virtual planning on CAD-CAM software and using custom fabricated guides, templates, and cutting jigs based on their virtual plan. These techniques allow for surgically efficient and highly predictable outcomes as far as bone and soft tissue positioning. The authors continue to refine these techniques in regards to cutting guide design and surgical planning, including placement of permanent implants, dentures, and ideal bone positioning. The basic process of planning and using CAM-CAM technology for pathology and reconstruction will be described.
Virtual surgical planning
Computer-aided mandibular ablation and reconstruction involves 4 distinct phases: planning, modeling, surgical, and evaluative. Planning begins with a high-resolution computed tomography (CT) scan of the patient’s craniofacial skeleton according to standard scanning protocols. CT scan of donor sites for flaps and grafts (fibula, iliac crest, scapula, and others) are also obtained. These images are then forwarded to the desired modeling company (the authors have primarily worked with and evolved their techniques with Medical Modeling Incorporated, Golden, Colorado). In cases where dental occlusion is being manipulated, stone dental models are shipped to the modeling company as well. The scans are converted into 3-dimensional images of the craniomaxillofacial skeleton and the donor site if required. Dental models are laser scanned and integrated into the reconstructed images to create a high-resolution replica of the teeth. A Web meeting is then held with biomedical engineers from the modeling company and the surgical team. The key parameters of the planning phase for mandibular reconstruction are the margins of resection and location of free flap placement in relation to the remaining mandible and craniofacial skeleton. These inputs are determined by the surgeon and marked during the Web meeting by the modeling engineer on the 3-dimensional reconstructed image. Control of the mouse pointer can be transitioned between both teams, and real-time cephalometric, volumetric, and linear analysis can be extrapolated as bony segments are being virtually manipulated. The engineers have the tools and skills required to manipulate any bony segment in the desired fashion.
The authors were able to overcome many challenges they faced in traditional fibula free flap reconstruction with implementation of CAD/CAM technology while educating the engineers in what they were trying to accomplish surgically. Communication between both teams is necessary in translating the surgical resection and reconstruction into a virtual model that can then be used to create cutting jigs and templates that are individually adapted to the patient’s bony anatomy and ultimately allow one to create seamless fibula–mandible continuity. By incorporating the engineers into the surgical planning, they understand the importance and reality of tissue positioning including bone segments, soft tissue, and vascular pedicles.
The virtual resection of the mandible is completed first. The desired osteotomy designs are marked, and the segment is virtually removed. The 3-dimensional reconstructed fibula image is then superimposed on the mandibular defect in its desired vascular and soft tissue orientation. Virtual fibular osteotomies are designed to fit the idealized reconstruction. The first osteotomy is designed to precisely fit the proximal angle of resection on the native mandible. Additional osteotomies are created, as needed, to recreate the shape of the resected portion of the native mandible. The engineers can use the geometry of the virtually resected mandible or mirror the contralateral disease free mandible in order to create ideal orthognathic relationships. The shape, make, and size of the reconstruction plate and the number and lengths of fibula segments can be modified to optimize the shape of the neomandible, maintain well-vascularized segments of fibula, provide appropriate bone-to-plate relationships for positioning of implants, provide seamless bony approximation, and maintain a perfect occlusal arrangement. The virtual osteotomies in both the mandible and fibula are planned to optimize bone apposition for subsequent bony union and to ease positioning and placement intraoperatively. Bone positioning is seamless between the osteotomies created on the mandible for the resection and the osteotomies created on the fibula for the reconstruction. Currently, for all benign cases, the authors will also plan precise dental endosseous implant placement by choosing the desired position to obtain postoperative prosthetic occlusion. The authors have also successfully completed 2 cases with immediate dental prosthesis application at time of surgery.
The modeling phase involves stereolithographic manufacturing of the planned components. This includes a model of the native craniofacial skeleton for intraoperative reference and to augment the education of residents, surgeons, and the patient. Next, cutting guides are produced that fit flush onto the native mandible and fibula and allow the osteotomies to precisely match those created during the planning phase. A reconstruction plate template and a model of the neomandible with screw hole reference pegs are designed that facilitates bending of the desired reconstruction plate preoperatively. The cutting guides are anatomically contoured and fit with tactile feedback. They are secured with bone screws that may also serve as plate reference holes. Once fastened, osteotomies are created through the cutting slot, creating seamless matchup of the fibula to the mandibular defect. The learning curve that is required to perform the appropriate osteotomies is removed, and the authors believe the results obtained from using these techniques are consistently better than any other.
The surgical phase of the reconstruction proceeds in typical fashion and is described in the authors’ previous publications on this topic. All cutting guides and implants are sterilized. Access to the mandible is based on location and severity of tumor or pathology. After access to the mandible is obtained, the cutting jigs are secured, and osteotomies are completed with a sagittal saw through the cutting slots to duplicate the virtually planned cuts. The jigs are removed, and the remaining mandible is placed into maxillomandibular fixation if possible. The screw holes from the cutting jigs are designed to fit specific holes on the prebent reconstruction plate in order to maintain accurate maxillomandibular relationships of nontooth-bearing segments and provide additional anatomic reference points for hardware fixation. The fibula guide is secured to the harvested fibula while maintaining the vascular pedicle. If planned, dental implant osteotomies are created through the drill guides on the jig. The fibula shaping is done with the pedicle intact and perfusion uninterrupted. The fibula cutting guide is used to replicate the cuts for both the end and closing wedge osteotomies that were planned previously. A sagittal saw is used to create the proper osteotomies with protection of the pedicle provided by a narrow malleable retractor placed between the fibula and the vascular pedicle.
The evaluation phase has been performed in nearly all of the authors’ virtually planned cases. This includes a CT scan along with standard postoperative follow-up. The CT scan is used as the ultimate comparison with the operative plan that was virtually generated. This has allowed the authors to critically assess their accuracy and refine the techniques for simplification. In general, accuracy is excellent and is within 1 to 5 mm. Any sources of error are probably due to the hand bending of the reconstruction plates when translating them from the virtual model. Because no premilled plates currently exist, attention to detail is important when prebending the plates from the models. This not only helps the accuracy of the outcome but shortens the length on the case.
For all benign pathology cases and malignant cases not requiring postoperative radiotherapy, the authors virtually plan and then place dental implants intraoperatively. Most of patients advance to complete dental rehabilitation within the first year postoperatively. The authors believe, even if not placed intraoperatively, placement of dental implants and prosthetic dental reconstruction is facilitated by the precise alignment and positioning of the fibula obtained from virtual planning.
Four case studies demonstrating the application of these digital technologies are described.