Advancing Image-Guided Surgery in Microvascular Mandibular Reconstruction

Sir:FigureMandibular reconstruction is challenging, with optimal outcomes difficult to achieve. As such, meticulous preoperative planning is often sought. The fibula and deep circumflex iliac artery flaps are frequent options for such defects, and imaging is used to aid in designing complex bony and soft-tissue modifications. Bony planning for multiple osteotomies and contouring to match mandibular architecture, and vascular mapping of the pedicle and its branches to bone and soft tissues, have led to preoperative computed tomographic scanning offering substantially improved outcomes.1 We describe our technique for combining existing computed tomographic scanning with the creation of image-guided stereolithographic BioModels (Anatomics, Melbourne, Victoria, Australia) to optimize outcomes, used preoperatively and intraoperatively to enhance bony modeling. The only imaging performed is high-resolution, multidetector row computed tomographic angiography, performed according to previously described techniques.1 Using the same volumetric scan data from established presurgical scans, raw image data are transferred for production of image-guided stereolithographic models (BioModels). The BioModels are used both preoperatively and intraoperatively to guide operative technique (Fig. 1).Fig. 1: Three-dimensional, multiplanar computed tomographic imaging of the mandible for resection and the donor iliac crest (above) used for production of image-guided stereolithographic BioModels (below).A single head and neck scan is used to assess tumor oncology, map mandibular bony anatomy, plan the bony resection, and formulate a bone reconstruction template. A donor-site computed tomographic scan maps the donor bone flap, the donor vascular pedicle and arterial branches from the pedicle to the planned bone flap, and the skin and muscle for composite flaps. The models thus created can demonstrate the site, size, and shape of the defect, with the tumor able to be “preresected” to form template models. Donor-site models can minimize bone harvest, spare donor-site morbidity, and improve bone-to-bone contact.2 Reconstruction plates can be premolded for improved fit and screws measured preoperatively, reducing the trial-and-error process commonly experienced in complex reconstructions and reducing operative times by up to 1.5 hours.3 By correlating vascular imaging to the models, selection of the most appropriately vascularized bone can be achieved. Virtual reconstruction using preoperative scans has been described for mandibular reconstruction,2,4,5 and although a series of complex geometrical and mathematical algorithms together with real-time navigation can be used instead of these models, these software programs are complex, and biomedical engineers are often required to view and manipulate the scans.2 In fact, stereolithographic models are often constructed to supplement the software. Of note, several enhancements to previously described stereolithographic modeling can improve outcomes. First, the literature predominantly describes its use in defining the mandibular defect alone3; however, the combination with donor-site models contributes more to accurate flap harvest. Second, recreation of the defect in one of two ways is advantageous: either a “resection template” created preoperatively, or an intraoperative template excised from the donor model. Lastly, combined use of the computed tomographic angiogram with the stereolithographic models can optimize the harvest of optimally vascularized bone. Currently performed preoperative scans can be further used to optimize outcomes in mandibular reconstruction, with stereolithographic models able to improve the quality of reconstruction when used preoperatively and intraoperatively. Warren M. Rozen, M.B.B.S., P.G.Dip.Surg.Anat., Ph.D. Jeannette W. C. Ting, M.B.B.S. Michael Leung, F.R.A.C.S. Terry Wu, F.R.A.C.S. David Ying, F.R.A.C.S. James Leong, F.R.A.C.S. Department of Plastic and Reconstructive Surgery, Dandenong Hospital, Southern Health, Dandenong, Victoria, Australia, Department of Surgery, Faculty of Medicine, Monash University, Clayton, Victoria, Australia DISCLOSURE The authors have no financial interest to declare in relation to the content of this article. No outside funding was received.