The Combined Use of Virtual Surgical Planning, Clinical Transfer Tools, and Free Tissue Transfer Improves Surgical Treatment of Complex Craniofacial Diagnoses
- Patrick K Kelley, MD1-4
- Steven L Henry, MD1-2,4
- Patrick D Combs, MD1-2,4
- Raymond J Harshbarger, MD1-4
- Institute of Reconstructive Plastic Surgery of Central Texas, Austin, TX
- Pediatric Specialty Services, Inc., Austin, TX
- University of Texas Southwestern-Austin, Austin, TX
- No financial disclosures
Concepts presented at International Society of Craniofacial Surgeons, Jackson Hole, WY, 2013
Statement on Human Investigation
This manuscript and the clinical work associated with its creation and publication is consistent with the ethical and moral principles set forth in the Declaration of Helsinki as amended and updated in 2013. All patients whose images are published in the manuscript were properly informed and consented for the use of their facial images in this publication.
Abstract
Virtual surgical planning (VSP) provides surgeons a new and powerful tool for the accurate and predictable reshaping of the craniofacial form, and when combined with free tissue transfer, represents a quantum leap in the treatment of the most severe craniofacial diagnoses. This paper details the fundamental concepts, strategies, and techniques of VSP for craniofacial reconstruction, based upon our experience in over 130 cases. Three representative cases are presented.
Level of Evidence: Therapeutic, Level V
Introduction
Common among many of the complex craniofacial diagnoses are both malposition of key anatomic elements and deficiency of tissue. To correct such deformities the surgeon must therefore both reposition structures and add volume. Simply “borrowing” tissue from regional sources (i.e., other regions of the face) is often not feasible, especially in severe phenotypes, thus necessitating free tissue transfer from elsewhere in the body. As suitable donor sites, such as the fibula, are typically available for such transfer, the challenge then becomes establishing proper form and position. To this end, virtual surgical planning (VSP) software has recently emerged as an extremely useful tool in planning surgical procedures and conceptualizing outcomes, and, due to its inherently complex form, has proven particularly helpful in the craniofacial region.1-3
The initial step in VSP is to establish the ideal virtual form for the particular patient. The ability to reconstruct this ideal form may be limited by concomitant uncorrectable deformities (e.g., skull base deformity in congenital diagnoses), and by the desires of the patient, which may limit the scope of procedures that the patient is willing to undergo and consequently limit the degree of correction the surgeon can produce. For these reasons the ideal virtual form does not always approach what would be considered normal or average craniofacial form (Fig. 1). Once the ideal virtual form is established, the next step is to digitally morph the tissue that to be transferred into a form as close to the ideal virtual form as possible, resulting in the virtual surgical plan. Surgical execution of the virtual plan, known as clinical transfer, is accomplished through the use of a variety of implements (cutting jigs, positioning guides, pre-bent plates, etc.) generated through the VSP process.
The combination of VSP and free tissue transfer represents a quantum leap in the ability to treat the most severe craniofacial diagnoses, including congenital, traumatic, and oncologic etiologies.4-6 Deformities heretofore considered uncorrectable can now be addressed with accurate and reproducible outcomes. Our center has been engaged in VSP since its inception, completing more than 130 cases to date. Our skills and strategies in VSP and clinical transfer, along with other centers utilizing these techniques, have evolved with experience into its present form.7-15 Representative cases among the various diagnoses will be presented to demonstrate current concepts and processes.
Planning Process
Initiating the Process
The VSP process begins with the obtainment of a high-resolution (≤1 mm cuts) CT scan of the craniofacial skeleton. As will be discussed later, the peculiarities of a given case may necessitate a CT scan of the donor site, although generic data exist and are often adequate for the common donor sites (e.g., fibula, scapula, iliac crest). The CT data are sent to a biomedical engineering company that operates the VSP software and produces the clinical transfer implements. (Medical Modeling Corp., Golden, CO). Planning is performed jointly between the surgical team and the biomedical engineer during a series of webinars.
Generating the Ideal Virtual Form
The ideal virtual form is the form in which all relevant anatomic elements (primarily bony structures) are positioned properly. This is achieved by digitally removing all malformed, damaged, or diseased structures and provisionally replacing them with structures of normal shape and position, without strict regard to how this can be achieved surgically. In tumor cases, for example, this step involves the bony tumor’s virtual resection with proper margins and its virtual replacement with normal bone. It may be necessary to collaborate with the oncologic surgeon to plan the resection margins, and, in cases of aggressive disease, it is very important to utilize a recent CT scan to avoid underestimating the extent of the tumor. The virtual replacement of the resected bone can be straightforward if the defect is small and simple, but if large or complex, it may be helpful to superimpose the contralateral side, if that side is normal, in a process called mirror imaging (Fig. 2). Alternatively, in cases where the contralateral side is not normal, generic data from a size-matched cohort can used in a process termed cohort substitution. Cohort substitution is often useful in cases of massive trauma or severe craniofacial malformation (Fig. 3). A critical consideration in predicting the actual defect is to anticipate potential bony interferences and other obstacles that will need to be addressed at the time of surgery. An example would be a mandibular ramus that is rotated secondary to contracture of the temporalis; this situation would require a coronoidectomy to permit proper positioning of the ramus. Such decisions about what is to be resected and what is to be preserved and/or modified are the most crucial elements of this step and are the primary determinants of the learning curve of the entire process.
Generating the Virtual Surgical Plan
The ideal virtual form is merely a template; the next step in the process is to modify the transferred tissue to fit this template, thus creating a representation of what is actually going to be done in the operating room. For traditional craniofacial osteotomies (e.g., Lefort I advancement), the process involves simply the creation of virtual osteotomies and the repositioning of bony segments in a way that best approximates the ideal form (Fig. 4), or, if distraction osteogenesis is to be used, the positioning of the distraction device in a way that establishes the proper vector for the transport segment (Fig. 5). For cases in which tissue must be imported, the donor bone, most commonly the fibula, is similarly osteotomized and positioned.
Patient-specific data for the donor tissue (i.e., data obtained from CT scan of the patient’s fibula) are often not necessary. However, this option does increase the accuracy of the virtual plan and is advisable in cases when the patient’s anatomy is unlikely to match cohort data or when the utmost precision is required for the reconstruction, as in a very large mandibular defect that would require the entire available length of the fibula. Another example would be a case in which a skin paddle is to be included with a specific segment of the fibula; in these situations a CT angiogram can enable the segments to be created in relation to the location of the perforating cutaneous vessels.
Modifying the straight fibula to replicate the curved bones of the face is complicated, particularly in light of the need to protect the vascular pedicle. Experience has taught us to eliminate no more than 1 cm of bone between segments along the margin to be collapsed in a closing wedge osteotomy, and to eliminate 2 cm of bone between segments that will be turned 180 degrees (e.g., in a double-barrel mandible construct).
One of the primary benefits of VSP is that the segmental interfaces can be made to be perfectly precise and flush (Figs. 21 and 22, mandible). Optimal bone contact enables faster and stronger union, which affords a more structurally sound and durable reconstruction and a shorter latency for the initiation of radiation therapy. However, optimal bone contact can still be difficult to achieve at the margin where the transferred bone interfaces with the recipient bone. Planning should include an attempt to optimize the recipient bone for that interface.
Clinical Transfer
The printed image of the final plan is itself a powerful clinical transfer tool for documenting and diagramming the various maneuvers that must occur to achieve the goals of the reconstruction. In addition, a CAD/CAM rapid prototyping process can be used to translate the VSP image into sterilizable acrylic models, cutting/drilling guides, and splints. Resection guides are used to establish appropriate resection margins and, by extension, the precise interfaces between the resection margins and the reconstruction segment (Fig. 6a). This is an extremely important step, since the imported tissue has been planned to morph into this very precise defect. The biggest challenge here is the accurate registration of the resection guide to the native anatomy for proper positioning of the osteotomy; the incorporation of as much unique anatomical detail in the region of the resection guide is therefore recommended. While guides that are too bulky are not useful because of the limited access incisions used in the craniofacial region, guides should be made large enough to incorporate unique surface anatomy (e.g., prominent genial tubercles), dental anatomy, or curves/angles in order to ensure accurate in situ positioning of the clinical tool. The merging of dental model data with the craniofacial CT data, vis-a-vis virtual orthognathic surgery, can greatly improve accuracy by permitting the incorporation of bite surface architecture into cutting guides and occlusal splints for control of occlusion (Fig. 6b).
Osteotomies of the donor bone are guided by a separate reconstruction-cutting guide (Fig. 7). These osteotomies are made at the donor site after the flap has been fully dissected but is still being perfused. The flap is then taken to the recipient site and the proximal segment of the flap is inset, which establishes the proper length and lay of the pedicle before the anastomoses are performed. This sequence (dissection of flap, osteotomy, division of pedicle, inset of first segment, anastomoses) minimizes ischemia time and optimizes anastomotic positioning. A common belief is that the segments of the fibula should be made no smaller than 2 cm in width. Our experience has shown that segments as small as 1 cm do very well (evidenced by efficient healing) as long as the periosteum remains well attached. Additionally, the strategic placement of the “extra” periosteum that is left at the regions of the flap where the bone was removed can facilitate bone healing and neo-osteogenesis.
Pre-bent plates are also invaluable tools for clinical transfer. These plates can be bent by the surgeon using a sterelithographic model of the virtual plan, or, if time permits, most plating companies can bend the plate remotely and send it to the hospital in time for the case. Another approach is to manufacture a stereolithographic model of the plate itself; this can be used much like the plating template that is provided with most mandible reconstruction sets. The pre-bending of plates not only reduces time in the operating room but also helps provide additional cues to proper orientation of transferred bone segments, again improving accuracy and decreasing time spent in the inset step. A further improvement on this process is the employment of a strategy referred to as predictive hole placement. This strategy positions virtual plates—generated from digitized data of actual plates, including realistic bending specifications—on the virtually planned reconstruction, avoiding tooth roots and nerves. Holes used to secure the resection guides are made to match holes in the plate, and the pre-bent plate is laser-etched to mark the holes that match the resection guide holes (Fig. 8). Ideally this step eliminates the need to put the patient in occlusion prior to plating. That being said, we often still put patients into occlusion using a CAD/CAM occlusal splint to further ensure the accuracy of our reconstruction. Importantly, we use smaller pilot holes and screws to secure the resection guide (using 2 mm screws) so that the holes will not be over-expanded when the reconstruction plate is applied (using 2.3-2.5 mm screws).
Additional clinical transfer tools include bone graft templates that help harvest a bone graft with the precise dimensions needed to fill a preplanned defect. Positioning guides—much like those used to position a maxilla without the use of an occlusal splint—can be used to direct a segment into a precisely planned position (Fig. 9). Distractor positioning guides direct the position and orientation of the osteotomy and the location of the distractor footplates, thereby establishing the vector of distraction (Fig. 5). Lastly, occlusal splints like those used in orthognathic surgery assist in the positioning of the basal bone of the mandible to ensure a proper occlusal outcome.
Together, these clinical transfer tools dramatically improve the accuracy and efficiency of the reconstruction procedure. Because all components—cutting guides, bone segments, and plates—are designed to fit together perfectly, no time is wasted whittling bone or bending plates, and the ultimate coaptation of bone segments and hardware far exceeds what could be accomplished by “freehand” techniques. Moreover, because the components are pre-cut/pre-bent, inset of the bone and hardware can be accomplished with relatively minimal exposure, providing an additional cosmetic benefit. These advantages realized in the operating room, in our opinion, more than justify the additional time, effort, and expense incurred during the preoperative planning process.
Case Illustrations
Case 1
A 12-year-old female presented with a large ameloblastoma of the mandible (Fig. 10). Single-staged composite resection and reconstruction was virtually planned. Cohort mandibular data was used to establish an ideal virtual form after virtual resection of the tumor (Fig. 11). The virtual surgical plan involved the transfer of an osseous fibula flap, morphing it into two rows to re-establish proper alveolar height (Fig. 12). Clinical transfer tools included a fibula cutting guide, upper- and lower-row pre-bent plates, and an occlusal splint. Surgery was accomplished successfully, and dental implants were incorporated nine months post-operatively to complete the reconstruction with an implant-retained prosthesis. Oral opening of 40 mm and restoration of her pre-disease form was accomplished (Fig. 13).
Case 2
A 16-year-old female presented with a severe phenotype of hemifacial microsomia, causing severe sleep apnea, masticatory embarrassment, and accelerated focal dental wear (Fig. 14). A two-staged correction was employed. The first stage was planned virtually as a double-jaw procedure involving LeFort I correction of the maxilla, left-sided sagittal split osteotomy, and resection of the vestigial right ramus to allow for proper advancement and rotation of the mandible (Fig. 15). The correction was stabilized with an RED device (KLS Martin, Jackson, FL) and a temporary tracheostomy was employed for airway protection for a period of two months (Fig. 16). The second stage was planned virtually using a fibula flap in multiple segments to reconstruct the right ramus, with the goals of preventing relapse and establishing proper form to the right posterior face. The plan was clinically transferred using a fibula cutting guide, occlusal splints, predictive hole placement, and a pre-bent reconstruction plate (Fig. 17). The combined surgical plan resulted in a functional bite, much improve dental show, an oral opening of 35 mm, a structurally sound and stable union, a significant improvement in facial form, and resolution of sleep apnea (Fig. 18).
Case 3
A 26-year-old male presented with a facial gunshot wound with submental entrance and midfacial exit (Fig. 19). After serial debridements, the upper midface and orbits were reconstructed and the mandible was stabilized with an external fixator. One week later, the mandible was stabilized internally with a reconstruction plate spanning a large defect left of midline. Oral domain was re-established with an anterolateral free flap to the floor of mouth. Three months later a virtual plan was established to reconstruct the mandible using a fibula flap (Fig. 20). Maxillary and palatal fistula reconstruction occurred as an additional stage several months later using the contralateral fibula (Fig. 21). Oral opening of 43 mm, unrestricted mastication, unobstructed nasal and oral breathing, and normal speech resonance were achieved 1.5 years post-injury.
Conclusions
The standard surgical treatment of many complex craniofacial diagnoses is suboptimal in terms of form, function, stability, healing efficiency, and number of surgical procedures. Much of this is related to the tissue volume deficiencies associated with these diagnoses. Tissue transfer helps solve this problem, but at the same time is limited by the accuracy with which these tissues can be morphed into the correct form, which is paramount for not only appearance but also proper function in the craniofacial region.
Virtual planning has greatly improved the accuracy of the planning phase of craniofacial reconstruction by facilitating the creation of an ideal virtual image that serves as a template for the eventual creation of a virtual surgical plan. The accurate execution of the virtual plan is further facilitated by the use of a number of clinical transfer tools. Our center has experienced notably improved outcomes with regard to accuracy of form, efficiency of healing, and stability of results, with commensurate improvements in functional outcomes with regard to breathing, eating, speaking, and vision.
References
1. Marmulla R, Hoppe H, Mühling J, Hassfeld S. New augmented reality concepts for craniofacial surgical procedures. Plast Reconstr Surg. 2005; 115(4):1124-1128.
2. Roser SM, Ramachandra S, Blair H, et al. The accuracy of virtual surgical planning in free fibula mandibular reconstruction: comparison of planned and final results. J Oral Maxillofacial Surg. 2010; 68:2824–2832.
3. Antony AK, Chen WF, Kolokythas A, et al. Use of virtual surgery and stereolithography-guided osteotomy for mandibular reconstruction with the free fibula. Plast Reconstr Surg. 2011; 128:1080–1084.
4. Hanasono MM Jacob R, Bidaut L, Skoracki R. Complex mid-facial reconstruction using virtual planning, rapid prototype modeling, and stereotactic navigation. Plast Reconstr Surg. 2010; 126(6): 2002-2006.
5. Saad A, Winters R, Wise M, et al. Virtual surgical planning in complex composite maxillofacial reconstruction. Plast Reconstr Surg. 2013; 132(3):626-633.
6. Nikkhah D, Ponniah A, Ruff, C, et al. Planning surgical reconstruction in Treacher-Collins Syndrome using virtual simulation. Plast Reconstr Surg. 2013; 132(5):790e-805e.
7. Foley BD, Thayer WP, Honeybrook A, et al. Mandibular reconstruction using computer-aided design and computer-aided manufacturing: an analysis of surgical results. J Oral Maxillofac Surg. 2013 Feb; 71(2):e111-e119.
8. Mazzoni S, Marchetti C, Sgarzani R, et al. Prostehtically guided maxillofacial surgery: evaluation of the accuracy of a surgical guide and custom-made bone plate in oncology patients after mandible reconstruction. Plast Reconstr Surg. 2013 Jun; 131(6):1376-1385.
9. Levine JP, Bae JS, Soares M, et al. Jaw in a day: total maxillofacial reconstruction using digital technology. Plast Reconstr Surg. 2013 Jun; 131(6):1386-1391.
10. Hou JS, Chen M, Pan CB, et al. Application of CAD/CAM-assisted technique with surgical treatment in reconstruction of the mandible. J Craniomaxillofac Surg. 2012 Dec; 40(8):e432-e437.
11. Ciocca L, Mazzoni S, Fantini M, et al. CAD/CAM guided secondary mandible reconstruction of a discontinuity defect after ablative cancer surgery. J Craniomaxillofac Surg. 2012 Dec; 40(8):e511-e515.
12. Adophs N, Haberl EJ, Liu W, et al. Virtual planning for craniomaxillofacial surgery—7 years of experience. J Craniomaxillofac Surg. 2013 Nov 5.
13. Matros E, Albornoz CR, Rensberger M, et al. Computer-assisted design and computer-assisted modeling technique optimization and advantages over traditional methods of osseous flap reconstruction. J Reconstr Microsurg. 2013 Dec 9.
14. Mardini S, Alsubaie S, Cayci C, et al. Three-dimensional preoperative virtual planning and template use for surgical correction of craniosynostosis. J Plast Reconstr Aesthet Surg. 2013 Nov 21.
15. Modabber A, Gerressen M, Ayoub N, et al. Computer-assisted zygoma reconstruction with vascularized iliac crest bone graft. Int J Med Robot. 2013 Dec; 9(45):497-502.
Figure Legend
Figure 1. Ideal virtual form established for a severe Treacher-Collin patient to reconstruct zygomas to treat ocular exposure.
Figure 2. Virtual resection of deformed or diseased tissue to establish the defect and mirror imaging of contralateral unaffected side to establish an ideal virtual form. a. Mandible tumor, b. Orbital and skull base neurofibromatosis.
Figure 3. Use of cohort data to generate an ideal virtual form (seen in Fig 1.) in a process called cohort substitution.
Figure 4. Several digital methods of generating a virtual plan. a. segmentation and transfer of a fibula flap into a mandible defect to reconstitute the ideal virtual form. b. traditional osteotomies and virtual repositioning into an ideal form established from orthognathic standards
Figure 5. Virtual plan established for distraction osteogenesis that focuses on proper positioning and angulation of the distractors to move the transport segment to the ideal virtual form.
Figure 6. a. Designing resection guides to prepare the defect for the proper interface with the imported tissue. b. Incorporation of dental surface anatomy into guides helps precisely orient guides at the cost of increasing bulk and adding intraoral incisions.
Figure 7. Clinical transfer of the virtual plan by the manufacture of an acrylic reconstruction cutting guide for use in the operating room to duplicate the precision cuts established in the virtual plan.
Figure 8. Predictive hole placement registers the holes used to secure the resection cutting guides with specific holes on a pre-bent plate which can be laser etched to facilitate accurate recognition of the coincident holes at the time of surgery.
Figure 9. Clinical transfer via CAD/CAM designed acrylic positioning guides. a. Lefort I positioning guide to eliminate use of intermediate splint b. Positioning guides used to mobilize a monobloc segment into the ideal virtual position.
Figure 10. Case 1: Preoperative photo and imaging of patient with large ameloblastoma mandible.
Figure 11. Virtual ideal form established for patient in Fig. 10 via virtual resection of tumor and merging with cohort imaging data in a process termed cohort substitution.
Figure 12. Image demonstrating the use of the ideal virtual form as a template for developing the virtual surgical plan which included the transfer of an osseous fibula flap in multiple rows and segments.
Figure 13. Post-operative photos and images of the patient in Fig. 10 demonstrating the successful use of virtual surgical planning process to reconstruct the patient with a complex osseous flap, resulting in normal facial form, proper occlusion, normal oral opening, normal mastication, based on a stable and complete union of the flap.
Figure 14. Case 2: Preoperative photo and imaging of patient with severe hemifacial microsomia
Figure 15. First stage virtual plan demonstrating a double jaw correction of maxillary and mandibular position of patient in case two. Plan was clinically transferred with intermediate and final occlusal splints and resection cutting guides.
Figure 16. Stabilization of the movements in step one was achieved with the use of a RED as an external fixator (no distraction employed).
Figure 17. Virtual plan for stage two in case two reconstructing the congenitally deficient ramus using a osseous fibula flap to prevent relapse from the corrected position of the maxilla and mandible in stage one.
Figure 18. Photo and CT of patient in case two after two staged correcting, prior to ear reconstruction and soft tissue revisions.
Figure 19. Photograph and CT image on presentation of case three.
Figure 20. The ideal virtual form was created by mirror imaging establishing a template for the fibula flap in the virtual plan. Clinical transfer was accomplished with an occlusal splint, resection cutting guides, reconstruction cutting guides, predictive hole placement, pre-bent mandibular reconstruction plates.
Figure 21. Virtual plan for maxillary and palatal fistula reconstruction using an osteocutaneous fibula flap. Clinical transfer achieved with a reconstruction cutting guide.
Figure 22. Final post-operative photograph and CT images, 1.5 years post-injury.
Social Media