- Email: [email protected]
Developments in Image-Guided Deep Circumflex Iliac Artery Flap Harvest: A Step-by-Step Guide and Literature Review
SURGICAL ONCOLOGY AND RECONSTRUCTION
Developments in Image-Guided Deep Circumflex Iliac Artery Flap Harvest: A Step-by-Step Guide and Literature Review Jeannette W. C. Ting, MBBS, MS,* Warren M. Rozen, MBBS, BMedSci, PGDipSurgAnat, MD, PhD,y Vachara Niumsawatt, MBBS,z Charles Baillieu, MBBS, FRACS,x Michael Leung, MBBS, FRACS,jj and James C. Leong, MBBS, MS, FRACS{ Purpose:
The deep circumflex iliac artery (DCIA) flap has evolved significantly over time in the intricacies of flap design and breadth of surgical application. This has been facilitated by advances in preoperative imaging and planning, in particular, computed tomographic angiography. Studies have highlighted that advanced imaging modalities and other technologies such as image-guided stereolithographic biomodeling can substantially improve flap planning, flap harvest, and operative outcomes.
Patients and Methods:
The present report comprises a combined literature review and clinical cohort study of 20 consecutive patients to assess the modern technologies applied to DCIA flap planning and harvest. We have also described a step-by-step guide for the implementation of these techniques into clinical practice.
Results:
The protocol for a single, standardized technique of computed tomographic angiography scanning is presented and was applied to a range of techniques in the preoperative planning of DCIA flaps. These include 1) bony and vascular imaging analysis of both donor and recipient sites, 2) stereolithographic ‘ biomodeling’’ of both donor and recipient bony and vascular anatomy, and 3) the use of preoperative ‘ virtual surgery’’ with image-guided stereotactic navigation. The application and role of each technique was explored. Conclusions: Modern imaging and stereolithographic techniques are innovations that can substantially improve surgical outcomes in DCIA flap surgery, such as has been highlighted in our clinical experience and in published studies. Notably, few outcome studies have been reported, and the need for larger case series and comparative studies is apparent. Crown Copyright Ó 2014 Published by Elsevier Inc on behalf of the American Association of Oral and Maxillofacial Surgeons. All rights reserved J Oral Maxillofac Surg 72:186-197, 2014
The deep circumflex iliac artery (DCIA) flap, a composite osteomusculocutaneous flap of the iliac crest, abdominal wall musculature, and overlying skin, has
evolved significantly during the previous 30 years since its inception in the late 1970s. With an increasingly reported role for a range of facial, lower limb, and upper {Plastic and Reconstructive Surgical Consultant, Department of
Received from the Department of Surgery, Monash University Faculty of Medicine, Clayton, Victoria, Australia.
Plastic and Reconstructive Surgery, Dandenong Hospital, Southern
*Plastic and Reconstructive Surgical Registrar, Department of
Health, Dandenong, Victoria, Australia.
Plastic and Reconstructive Surgery, Dandenong Hospital, Southern
Address correspondence and reprint requests to Dr Ting: Depart-
Health, Dandenong, Victoria, Australia.
ment of Plastic and Reconstructive Surgery, Dandenong Hospital,
yPlastic and Reconstructive Surgical Registrar, Department of
Southern Health, David St, Dandenong, Victoria 3175, Australia;
Plastic and Reconstructive Surgery, Dandenong Hospital, Southern
e-mail: [email protected]
Health, Dandenong, Victoria, Australia. zPlastic and Reconstructive Surgical Registrar.
Received April 13 2013 Accepted June 27 2013
xPlastic and Reconstructive Surgical Consultant, Department of
Crown Copyright Ó 2014 Published by Elsevier Inc on behalf of the American
Plastic and Reconstructive Surgery, Dandenong Hospital, Southern
Association of Oral and Maxillofacial Surgeons. All rights reserved
Health, Dandenong, Victoria, Australia.
0278-2391/13/00826-4$36.00/0
jjPlastic and Reconstructive Surgical Consultant, Department of
http://dx.doi.org/10.1016/j.joms.2013.06.219
Plastic and Reconstructive Surgery, Dandenong Hospital, Southern Health, Dandenong, Victoria, Australia.
186
187
TING ET AL
limb reconstructions,1-5 its most widespread utility has been for hemimandibular defect reconstruction.6-10 Although the iliac crest has long been used for these various bony reconstructions, its versatility as a composite flap has largely been limited by an understanding of the finer vascular anatomy of the region. Initial attempts to harvest the iliac crest flap using the superficial circumflex iliac artery as its vascular supply in 1978 met with less than ideal results.11 Although greater success was achieved with the DCIA pedicle flap after the landmark report by Taylor and Townsend12 in 1979, detailing the DCIA as the main blood supply to the iliac crest, a lack of familiarity with the DCIA perforators in these early studies limited the use of the DCIA flap as a composite flap. Recently, anatomic studies have definitively demonstrated the anatomy of the perforators to the skin overlying the iliac crest,13 and subsequent clinical case series have highlighted that preoperative computed tomographic angiography (CTA) can be an essential tool to guide surgeons in harvesting the DCIA perforator flap.14 Similarly, computer software technology has advanced exponentially, allowing a greater ability to manipulate the data obtained from CTA that can be of practical use to surgeons. Recent studies have described this and other technological advances to aid DCIA flap surgery, including the use of image-guided stereolithographic biomodeling for both bony and vascular modeling pre- and intraoperatively.15,16 A stereolithographic biomodel can be used as a preplanned and prefabricated instrument, thus reducing the operative time with improved precision. Although techniques such as color duplex ultrasound and magnetic resonance imaging have been described for vascular imaging, we have preferred to use CTA as our modality of choice for vascular mapping because of its availability, objectivity, accuracy, and efficacy.17 The present report describes a step-by-step guide for implementation of these techniques into clinical practice, described from both our clinical experience and literature review.
Patients and Methods Our clinical approach as described was based on implementation of these techniques in a cohort of 20 consecutive patients (12 men and 8 women) who underwent DCIA flap reconstruction for an oncologic defect. The patients had an age range of 39 to 75 years and a range of body habitus (Table 1). All patients provided written informed consent for inclusion, and their cases were deemed clinically appropriate for DCIA flap surgery, with no exclusions. The technique described was based on a standardized approach to preoperative imaging and interpretation, with the imaging findings reported in all cases by 1 of us. The
Table 1. DEMOGRAPHIC DATA
Demographic Data
Value
Patients (n) Age (yr) Mean Range Gender Male Female Diagnosis Ameloblastoma Squamous cell carcinoma Neuroblastoma Trauma Deep circumflex iliac artery flap Flap constituents Osseous flap Osseomuscular Osseomusculocutaneous Osseocutaneous Preoperative imaging: 64-slice multidetector row CTA Operative guidance based on imaging Pedicle Perforators Concordance between imaging and operative findings Pedicle Perforators Flap-related complications Donor site complications
20 60.2 39-75 12 8 2 15 1 2 20
3 2 5 10 20
20 15
19 15 0 1 (hematoma)
Abbreviation: CTA, computed tomographic angiography. Ting et al. Image-Guided DCIA Flap Harvest. J Oral Maxillofac Surg 2014.
imaging and technique of application of modern adjuncts have been described qualitatively and compared in the discussion to the findings from the literature review. The Melbourne Health HREC provided institutional ethical approval for evaluating the role of CTA (no. 2006.231). This and other aspects of the study complied with the Declaration of Helsinki, 1995. All patients gave informed consent for all imaging, surgery, and inclusion in the report, and patient anonymity was preserved. The results are reported qualitatively, with the methodologic approach to image-guided DCIA harvesting and insetting presented. TECHNIQUE
Each patient underwent CTA of both their defect site and donor site. The CTA hardware used included a Siemens Somatom Sensation 64 multidetector row CT scanner (Siemens Medical Solutions, Erlangen,
188 Germany). Intravenous contrast was used in all cases, with no oral contrast used, and included nonionic iodinated contrast media (Ultravist 370, Schering, Berlin, Germany; or Omnipaque 350, Amersham Health, Princeton, NJ). Intravenous access was accessed through a cubital fossa vein, with an 18-gauge cannula, and injection performed using a biphasic power injection pump at a flow rate of 4 to 6 mL/s. A bolus tracking technique was used to identify filling of the appropriate vessels with contrast as a method to initiate scanning. The scan was timed from the external iliac artery for scan triggering (to initiate arterial phase filling). At that point, the minimum delay in scanning was undertaken (this was defined by the scanner software and was typically 4 seconds). These techniques aim to provide pure ‘‘arterial phase’’ scans, with no or minimal venous opacification. A similar technique was used for the recipient site. Image reformatting software was achieved with 2 software programs: Siemens Syngo InSpace (InSpace 2004A_PRE_19, Siemens Australia, Victoria, Australia) and Osirix (Osirix Medical Imaging Software, GPL Licensing Open Source Initiative, http://www.osirixviewer.com). Multiplanar 3-dimensional reconstruc-
IMAGE-GUIDED DCIA FLAP HARVEST
tions were achieved with maximum intensity projection and volume-rendered technique reconstructions. All specific measurements of the diameter or distances were performed using the raw fine axial images. Stereotactic navigation was achieved from the same raw data using the technique described below. Step 1: Preoperative Imaging of Recipient Site Anatomy A single head and neck CT scan can assess tumor oncology in terms of the soft tissue and bony extent and involvement and relation to the surrounding anatomy and to plan the degree of soft tissue and bony resection (Fig 1).15,16 The same scan can be used to map the bony anatomy of the mandible to create a template for bony reconstruction. If CTA has been performed, the same scan can be used to map the locoregional vascular anatomy for the free flap recipient vasculature. Step 2: Stereolithographic Modeling of Recipient Anatomy Using the same volumetric scan data from step 1, the image data were transferred as raw data for the production of the image-guided stereolithographic models
FIGURE 1. Preoperative 3-dimensional, multiplanar computed tomography image of the mandible for resection planning. Reproduced in part with permission from: Rozen WM, Ting JWC, Leung M, et al: Advancing image-guided surgery in microvascular mandibular reconstruction: Combining bony and vascular imaging with CT guided stereolithographic bone modelling. Plast Reconstr Surg 130:227e, 2012. Ting et al. Image-Guided DCIA Flap Harvest. J Oral Maxillofac Surg 2014.
189
TING ET AL
FIGURE 2. Image-guided, stereolithographic biomodel (BioModels, Anatomics) of the mandible for resection planning (note comparison to equivalent computed tomography scan in Fig 1). Reproduced in part with permission from: Rozen WM, Ting JWC, Leung M, et al: Advancing image-guided surgery in microvascular mandibular reconstruction: Combining bony and vascular imaging with CT guided stereolithographic bone modelling. Plast Reconstr Surg 130:227e, 2012. Ting et al. Image-Guided DCIA Flap Harvest. J Oral Maxillofac Surg 2014.
(BioModels, Anatomics, Victoria, Australia). These defect-site models can visually demonstrate the site, size, and shape of the defect, with the tumor able to be ‘‘pre-resected’’ to form the template models (Fig 2).15,16 These models were brought to the multidisciplinary meeting before surgery to allow the resection and reconstructive surgeons to use these models to familiarize themselves with the defect. Resection templates were then made according to
the surgical plans, and the fixation plates can be premolded and modeled to the defect preoperatively. These models were all sterilized and made available during the operation. Step 3: Preoperative Imaging of Donor Site Anatomy A donor site scan can be used to map all aspects of the donor anatomy and to create a ‘ virtual guide’’
FIGURE 3. Preoperative computed tomographic angiogram with coronal maximum intensity projection reformatting, highlighting the deep circumflex iliac artery and its course, with a branch to the iliac crest (red arrow) and ascending branch (yellow arrow) shown. Ting et al. Image-Guided DCIA Flap Harvest. J Oral Maxillofac Surg 2014.
190 through each step of flap harvest. The DCIA pedicle itself can be demonstrated, showing the origin of the vessel, the location relative to other soft tissue and bony landmarks, the diameter throughout its course, the branching pattern, and its course toward the iliac crest (Fig 3). Furthermore, specific branches to the iliac crest can be followed, enabling the selection of the optimally vascularized part of the iliac crest for harvest.4 The accessory branch of the DCIA, which supplies the internal oblique and transverses the abdominis muscles, can be assessed if a muscle cuff is required. Cutaneous perforators can also be assessed for a vascular supply to a skin paddle, if required (Fig 4).13,14 The perforator number, location, size, and course can be clearly evaluated using CTA. The individual branches to each lamina of a composite flap can be traced, with branches to the skin, muscle, and bone often separately identifiable (Fig 5).15,16 Step 4: Stereolithographic Modeling of Flap (Donor Site) Anatomy The same volumetric scan data from step 3 can also be used to plan surgery, with the image data transferred as raw data for the production of image-guided stereoli-
IMAGE-GUIDED DCIA FLAP HARVEST
thographic models of the donor anatomy (Figs 5 and 6), similar to the method used for the recipient site (BioModels, Anatomics, Victoria, Australia). These can then directly facilitate the operative approach (Figs 7-9). Step 5: Preoperative ‘‘Virtual Surgery’’ With ImageGuided Stereotactic Navigation Image-guided stereotactic navigation is a technique that uses the same preoperative scan (CTA or magnetic resonance angiography [MRA]) performed for routine preoperative imaging, with no additional imaging required, and transfers the data through infrared and computer analysis for additional surgical analysis.18 The technique of image-guided stereotaxy comprises several key components: ‘ registration’’ of the patient anatomy, the transfer of data to the computer systems, and navigation through the scan anatomy in ‘ real time.’’ Patient ‘ registration’’ was achieved using fiducial marker registration, with radiopaque markers applied to the skin surface. After CTA scanning, the images generated were exported as Digital Imaging and Communications in Medicine images and transferred to the stereotactic computer system (BrainLAB, Vector Vision, Cranial, version 7.0, Sydney, Australia). Real-time navigation through the scan data was achieved by placing the
FIGURE 4. Preoperative computed tomographic angiogram with 3-dimensional, volume-rendered reconstruction, highlighting the deep circumflex iliac artery perforators (arrows), emerging from the deep fascia posterosuperiorly to the iliac crest, for planning a skin paddle in an osteocutaneous deep circumflex iliac artery flap. Ting et al. Image-Guided DCIA Flap Harvest. J Oral Maxillofac Surg 2014.
191
TING ET AL
multiple computer platforms. ImageJ is free to download and use, with numerous plug-in functions that allow image analysis.19 The process involves the use of volumetric scan data from steps 1 and 3, with the image data able to be processed with different functional plug-ins. The anatomy of the DCIA vessels and iliac crest can be assessed and quantified accurately using the software. It allows for an evaluation of the vessels, not only of their quantity and diameter, but also the density and measurements of distances to bony landmarks. The evaluated images can then be stored in several image formats such as JPEG, TIFF, BMP, or RAW (Fig 10). The stored images can be reassessed by different individuals, improving their sensitivity and specificity and reproducibility. MIMICS BIOMEDICAL SOFTWARE
FIGURE 5. Preoperative 3-dimensional, multiplanar computed tomographic angiogram of the donor iliac crest and deep circumflex iliac artery to plan donor site harvest, highlighting pedicle branches to the iliac crest. Reproduced with permission from Rozen WM, Ting JW, Baillieu C, Leong J: Stereolithographic modeling of the deep circumflex iliac artery and its vascular branching: A further advance in computed tomography-guided flap planning. Plast Reconstr Surg 130:380e, 2012 Ting et al. Image-Guided DCIA Flap Harvest. J Oral Maxillofac Surg 2014.
patients in a supine position, mimicking both the position during CT scanning and the operative position, with a ‘ stereotactic pointer’’ or ‘ wand’’ used to navigate through the 3-dimensional anatomy. Intraoperatively, surgeons were able to use the navigation system to perform accurate osteotomies of both the donor and recipient sites in accordance with the preoperative plans. It was also used to identify the pedicle and perforator location throughout dissection of the vasculature to reduce the risk of inadvertently damaging the vessels. Step 6: Preoperative Vascular Assessment With ImageJ ImageJ is a Java-based image analysis application for high output processing and image manipulation on
The process of preoperative planning, anatomic analysis, visual 3-dimensional modeling, and anatomic models has been widely accepted and used, with many software companies adopting these concepts and producing various products that are commercially available. One such software product that has been developed is called ‘‘MIMICS innovation suite,’’ which incorporates 4 components to assist reconstructive surgeons in preoperative planning. The suite has 3 components: virtual image reconstruction; computer-aided design tools; and anatomic models. The process begins with reconstruction of 2-dimensional images that can be reconstructed into 3-dimensional virtual images similar to those of ‘ Osirix.’ These images can be manipulated to assess and analyze the internal anatomy. The reconstructed images can then be transferred to computer software, where a prosthesis or virtual surgery can be performed using a computer console. The surgeon can thus practice, design, and implement patient-specific osteotomies and ostectomies and placement of fixation devices before any surgery. The end result can be viewed in 3 dimensions and any necessary adjustment can be made at the same time (either preoperatively or intraoperatively). Ultimately, the designed images can then be reproduced into an anatomical 3-dimensional model. The product is tangible, accurate, and realistic, allowing surgeons to explore and evaluate the patient anatomy to gain better insight into specific pathologic features. It also allows for prefabrication of plates and the osteotomy design, which has been shown to reduce the operative time and increase the accuracy in flap positioning and planning.20,21
Results CTA and stereolithographic models were performed and made for 20 consecutive patients undergoing DCIA reconstruction (Table 1). All patients had clear oncologic clearance. Intraoperative adjustments of
192 the resections were all within 1 cm of the preoperatively planned models. All patients underwent successful dissection and elevation of their DCIA flap, and no complete flap failures occurred. The concordance between the imaging and operative findings of the pedicle was 95%, with 100% concordance for the perforators. This was determined by the number and location of the pedicles and perforators seen on the model compared with the intraoperative findings using the anterior superior iliac spine as the bony landmark. Minor adjustments of the preoperative molded plates were made during surgery for a small number of patients, with no significant increase in the operation time. One donor site complication occurred, a donor site hematoma. The patients were followed for 18 months to 4 years, with 1 recurrence found during the follow-up period.
Discussion The iliac crest has long been recognized to be an excellent source of bone graft material and was widely used as a nonvascularized graft as early as World War I, with poor results.22,23 It was not until the 1970s, with the developments in microsurgery and vascularized transfers, that surgeons were able to markedly improve the iliac bone graft survival rate. The vascularized DCIA flap was popularized in association with an inconspicuous donor scar and low donor site morbidity.24-26 In addition, it has a dependable vascular pedicle length and diameter and a wide arc of rotation for the flap.12 Although the free fibula flap is commonly used to reconstruct osteocutaneous defects in the head and neck region, the DCIA flap is particularly useful for hemimandibular defects. Osteotomies of the bone are not needed, and the anterior inferior iliac spine can be used to reconstruct the mandibular condyle and hence the temporomandibular joint. Its versatility can be increased for defects that require variable amounts of bone, soft tissue, and skin that were once thought too small and the flap too bulky to reconstruct. This includes the iliac cortex splitting technique first described by Taylor9 in 1982 and again described by Shenaq and Klebuc8 in 1994. This was introduced to harvest smaller, thinner bone from the iliac crest and to reduce the rate of hernia after removal of part of the iliac crest. The muscle cuffsparing variation, aiming to reduce the bulk of the flap, was attempted by Safak et al25 in 1997 and Kimata27 in 2003, with limited success. Although no anterior mandibular defects involved the mandibular symphysis in our study, such a defect would require reconstruction of the anterior prominence of the mandible. Our preference for reconstruction of such a defect would be to use the straight segment of the superolateral iliac crest with or without the need for osteotomy. Using preoperative
IMAGE-GUIDED DCIA FLAP HARVEST
imaging and 3-dimensional reconstruction of both the donor (iliac crest) and recipient (mandible) sites, such reconstruction can be performed to assess and implement the contour geometry. However, should an adjustment to the curvature for the defect require an osteotomy of the iliac crest, this can be preplanned and, indeed, preplated in a similar manner as for mandibular ramus or body reconstruction. Furthermore, separate bony branches of the DCIA can also be identified on imaging and incorporated into the flap to increase the reliability of the segmental osteotomies. As perforator flaps became more popular in the 1980s, the DCIA perforator flap was used cautiously, because it had developed a reputation of having unreliable perforators. Without perforator skeletonization, it was a bulky flap that depended on the underlying abdominal musculature to ensure skin survival.28-30 However, the view that DCIA perforators were unreliable was found questionable in a large study of these perforators in 2009. It found that reliable perforators greater than 0.8 mm were present in more than 60% of hemiabdomens.13 A subsequent case series demonstrated that with preoperative CTA, 6 of 7 patients who underwent planning for a DCIA perforator flap had it raised successfully without any flap loss or significant related morbidity.14 Preoperative imaging was able to highlight specific branches of the DCIA to the iliac crest, facilitating more precise bone harvest and selection of optimally vascularized bone. This was highlighted in a case report in which a small amount of vascularized iliac crest was used to reconstruct a radial bony defect.4 The DCIA flap should be considered a valuable tool in the reconstructive surgeons’ repertoire of composite free flaps. It provides strong cortical bone with a reliable vascular source and is especially effective for the mandible, where solid bone is required to withstand the normal forces of mastication and for osseointegration. The cortex splitting technique, in which only the inner cortex is used, ensures a large amount of bone will also be available for thinner bony reconstruction such as the orbital floor and maxilla. The associated internal oblique muscle can be used to close maxillectomy fistulas and becomes epithelialized with minimal bulk.31 The skin paddle is often independent of the overlying bone (unlike the free fibula flap) and muscle (supplied by the separate accessory branch) and can be used to reconstruct a skin defect that is a small distance from the bony defect. In addition, the skin paddle, unlike the free fibula flap, is not hairy and is well tolerated by patients when used to line the oral mucosa. Although CTA has been the investigation modality of choice for imaging of the perforators, concerns have often been raised about radiation exposure and the accuracy of the data compared with other imaging techniques. With scanning protocol modifications, we
TING ET AL
(and other groups) have been able to ensure that patients are only exposed to 6 mSv of radiation, equivalent to 2 plain radiographic abdominal series or less than one quarter of the radiation of a full trauma CT scan.32 CTA has been shown to be superior in imaging abdominal wall perforators compared with ultrasound or MRA.33 Useful perforators of 1 mm in diameter might not be highlighted with MRA,34 and the validation of MRA in the identification of perforators has not been as widely described. For investigation of the venous anatomy, only CTA has been shown to be an effective option.35 In addition, although studies have suggested that MRA might be preferential to CTA because of the lack of ionizing radiation, the true clinical efficacy of MRA has not been validated.34,35 Studies have also conceded that MRA is more operator dependent than CTA, is more difficult to process, and is less available than CTA,36 making it a less accessible imaging modality. A recent study by Schaverien et al34 has also demonstrated an almost 10% incidence of respiratory artifact with MRA that has not been reported with CTA of the abdomen. With MRA technology constantly evolving, the use of MRA to produce stereolithographic models might be explored in the future. Technological advances in imaging and microsurgery, in particular, with modern modalities and stereolithographic biomodels, have been described in published studies for a number of years and with reproductive accuracy.37 However, their application to mandibular reconstruction38-43 and iliac crest harvesting15,44-46 has only recently been reported. Stereolithographic biomodels add an extra element of tangibility that allows the surgeon to ‘ see’’ and ‘ feel’’ the surgery beforehand and, combined with imaging, can enable ‘ virtual surgery’’ preoperatively. Surgeons can also use the models as guides during surgery, and modifications to the plan can be easily communicated between the surgical teams15,16 (Fig 6). The use of these biomodels can minimize bone harvest by selection of the most appropriately vascularized and best-fitting bone, spare donor site morbidity through more precise harvesting, and improve bone-to-bone contact at the recipient site (Figs 7–9). This can facilitate increased predictability of the anticipated surgery, especially in complex craniomaxillofacial reconstructions. In addition, surgical teams can anticipate the potential difficulties and have contingency plans in the event of unexpected complications. Imaging and biomodels can also be used in the consent process to explain the surgery and reconstruction to the patient and can be a valuable teaching tool for the training surgeon. Premolding of plates and measurement of screws can reduce the operation time and allow the resection and reconstruction teams to work simultaneously. The premolded plate can also serve as
193
FIGURE 6. Image-guided stereolithographic bone and vascular biomodel (BioModels, Anatomics) of the donor iliac crest and deep circumflex iliac artery to plan donor site harvest (note, comparison to equivalent computed tomography angiogram scan in Fig 5). Reproduced with permission from Rozen WM, Ting JW, Baillieu C, Leong J: Stereolithographic modeling of the deep circumflex iliac artery and its vascular branching: A further advance in computed tomography-guided flap planning. Plast Reconstr Surg 130:380e, 2012. Ting et al. Image-Guided DCIA Flap Harvest. J Oral Maxillofac Surg 2014.
a trial, allowing reconstructive surgeons to perform as many preoperative adjustments as needed. Before the operation, the premolded plate can be sterilized and used in the operation. However, the current status of virtual 3 dimensions is still in its early state and expensive and requires expertise to operate the system. This planning can enable a degree of independent movement of the individual flap components. The pedicle length and, therefore, the arc of movement can also be measured preoperatively to determine whether the length is adequate. Other related techniques have been offered for computermodeled DCIA flap harvest,45,46 with reported benefits. Image-guided stereotaxy is a recent advance in imaging technology, with computer guidance facilitating improved surgical planning and accuracy. We have
194
IMAGE-GUIDED DCIA FLAP HARVEST
FIGURE 7. Operative hemimandibular resection specimen. The resection was performed for ameloblastoma. Ting et al. Image-Guided DCIA Flap Harvest. J Oral Maxillofac Surg 2014.
found that image-guided stereotactic navigation can accurately identify the location and course of the DCIA perforators, potentially to a greater degree of accuracy than CTA alone. This might be because the
localization is achieved in 3 dimensions (unlike the 2-dimensional analysis with CTA), and the data collection occurs in a ‘‘live’’ patient, potentially taking into account the mobility of the integument during data
FIGURE 8. Deep circumflex iliac artery flap after harvest (note, match to resected specimen in Fig 7). Ting et al. Image-Guided DCIA Flap Harvest. J Oral Maxillofac Surg 2014.
195
TING ET AL
FIGURE 9. Postoperative computed tomographic angiogram of the mandible after flap inset. Ting et al. Image-Guided DCIA Flap Harvest. J Oral Maxillofac Surg 2014.
FIGURE 10. Screenshot of image analysis of the reconstructed computed tomographic angiographic images using ImageJ. Ting et al. Image-Guided DCIA Flap Harvest. J Oral Maxillofac Surg 2014.
196 collection. By combining the findings of routine CTA and CTA-guided stereotaxy, the efficacy of scanning can be improved. The adoption of these new technologies has currently been largely limited by the financial costs of computer hardware and software and the costs in the production of biomodels. Although it might not be financially feasible to use such models in all reconstructive cases, for complex reconstruction cases, the financial cost of the model and the extra time taken to plan the operation will be offset by the reduced intraoperative time, the more precise harvesting of the donor bone (and therefore reduced morbidity),43 and increased patient satisfaction with the final operative results.38,45 A recent study examining the use of CT and biomodels to aid in free bone flap harvest found that this modeling reduced the operative time by 7 to 22 minutes, with no major adjustments required for iliac crest harvesting and shaping intraoperatively.45 With improved vascular modeling, as highlighted, these outcomes could be improved further.15,16 In our experience, imaging was thought to substantially reduce the operative time; however, without a comparative group, the actual calculation of the operative time was outside the scope of the present study, although it will form the basis of future investigation. Gillie’s principle of carefully planning before operating is doctrine to most facial surgeons. In the past 40 years, with the familiarity of microsurgery, the question is no longer about survival of the free flap or perforator flap. Innovations should aim to improve and refine free flap techniques to reduce donor site morbidity, reduce the surgical times, and reduce the incidence and severity of overall complications. In conclusion, modern imaging and stereolithographic techniques are innovations that can substantially improve the surgical outcomes of DCIA flap surgery. Future advances are likely to include finer resolution imaging techniques, advances in in vivo realtime anatomic navigation, and improved computer software and analysis systems. The published data support the use of existing technologies, and, with few outcome studies reported, the need for larger case series and comparative studies is apparent.
References 1. Peek A, Giessle GA: Functional total and subtotal heel reconstruction with free composite osteofasciocutaneous groin flaps of the deep circumflex iliac vessels. Ann Plast Surg 56:628, 2006 2. Sanders R, Mayou BJ: A new vascularized bone graft transferred by microvascular anastomosis as a free flap. Br J Surg 66:787, 1979 3. Leung PC, Chan KT: Giant cell tumour of the distal end of the radius treated by the resection and free vascularized iliac crest graft. Clin Orthop Relat Res 202:232, 1986
IMAGE-GUIDED DCIA FLAP HARVEST 4. Ting JWC, Rozen WM, Leong J, Crock J: Free deep circumflex iliac artery vascularised bone flap for reconstruction of the distal radius: Planning with CT angiography. Microsurgery 30:163, 2010 5. Gabl M, Reinhart C, Lutz M, et al: Vascularized bone graft from the iliac crest for the treatment of nonunion of the proximal part of the scaphoid with an avascular fragment. J Bone Joint Surg 81:1414, 1999 6. Urken L, Vickery C, Weinberg H, Buchbinder DM: The internal oblique iliac crest osseomyocutaneous free flap in oromandibular reconstruction: Report of 20 cases. Arch Otolaryngol Head Neck Surg 115:339, 1989 7. Brown JS: Reconstruction of the total maxillectomy defect using a pedicled coronoid flap and deep circumflex iliac artery free flap. Br J Oral Maxillofac Surg 26:423, 2008 8. Shenaq SM, Klebuc M: The iliac crest microsurgical free flap in mandibular reconstruction. Clin Plast Surg 21:37, 1994 9. Taylor GI: Reconstruction of the mandible with free composite iliac bone graft. Ann Plast Surg 9:361, 1982 10. Taylor GI, Townsend P, Corlett R: Superiority of the deep circumflex iliac vessels as the supply for free groin flaps: Clinical work. Plast Reconstr Surg 64:745, 1979 11. Daniel RK, Kerrigan CL: Mandibular reconstruction with free tissue transfers. Ann Plast Surg 1:346, 1978 12. Taylor GI, Townsend P, Corlett R: Superiority of the deep circumflex iliac vessels as the supply for free groin flaps: Experimental work. Plast Reconstr Surg 64:595, 1979 13. Ting JWC, Rozen WM, Grinsell D, et al: The in vivo anatomy of the deep circumflex iliac artery perforators: Defining the role for the DCIA perforator flap. Microsurgery 29:326, 2009 14. Ting JWC, Rozen W, Chubb D, et al: Improving the utility and reliability of the deep circumflex iliac artery perforator flap: The use of preoperative planning with CT angiography. Microsurgery 31:603, 2011 15. Rozen WM, Ting JW, Baillieu C, Leong J: Stereolithographic modeling of the deep circumflex iliac artery and its vascular branching: A further advance in computed tomography-guided flap planning. Plast Reconstr Surg 130:380e, 2012 16. Rozen WM, Ting JWC, Leung M, et al: Advancing image-guided surgery in microvascular mandibular reconstruction: Combining bony and vascular imaging with CT guided stereolithographic bone modelling. Plast Reconstr Surg 130:227e, 2012 17. Rozen WM, Ribuffo D, Atzeni M, et al: Current state of the art in perforator flap imaging with computed tomographic angiography. Surg Radiol Anat 31:631, 2009 18. Rozen WM, Buckland A, Ashton MW, et al: Image-guided stereotactic perforator flap surgery: A prospective comparison of current techniques and review of the literature. Surg Radiol Anat 31:401, 2009 19. Abramoff MD, Magelhaes PJ, Ram SJ: Image processing with image. J Biophotonics Intl 11:36, 2004 20. Markiewicz MR, Bell RB: The use of 3D imaging tools in facial plastic surgery. Facial Plast Surg Clin North Am 19:655, 2011 21. Anuja 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 128:1080, 2011 22. Disa JJ, Cordeiro PG: Mandible reconstruction with microvascular surgery. Semin Surg Oncol 19:226, 2000 23. Phillips CM: Primary and secondary reconstruction of the mandible after ablative surgery. Am J Surg 114:601, 1967 24. Kudo K, Fujioka Y: Review of bone grafting for reconstruction of discontinuity defects of the mandible. J Oral Surg 36:791, 1978 25. Safak T, Klebuc MJ, Mavili E, Shenaq SM: A new design of the iliac crest microsurgical free flap without including the ‘ obligatory’’ muscle cuff. Plast Reconstr Surg 100:1703, 1997 26. Bitter K, Schlesinger S, Westerman U: The iliac bone or osteocutaneous transplant pedicled to the deep circumflex iliac artery. J Maxillofac Surg 11:241, 1983 27. Kimata Y: Deep circumflex iliac perforator flap. Clin Plast Surg 30:433, 2003 28. Akyurek M, Conejero A, Dunn R: Deep circumflex iliac artery perforator flap without iliac crest. Plast Reconstr Surg 122: 1790, 2008
TING ET AL 29. Goldie SJ, Thompson R, Soutar DS, Shaw-Dunn J: Arterial supply of the deep circumflex iliac artery (DCIA) flap. Eur J Plast Surg 31:311, 2008 30. Akyurek M, Morse A, Safak T, et al: Deep circumflex iliac artery perforator skin flap: Anatomic study and clinical application. J Reconstr Microsurg 22:A031, 2006 31. Brown JS: Deep circumflex iliac artery free flap with internal oblique muscle as a new method of immediate reconstruction of maxillectomy defect. Head Neck 18:412, 1996 32. Rozen WM, Whitaker IS, Stella DL, et al: The radiation exposure of computed tomographic angiography (CTA) in DIEP flap planning: Low dose but high impact. J Plast Reconstr Aesthet Surg 62:e654, 2009 33. Rozen WM, Stella DL, Bowden J, et al: Advances in the preoperative planning of deep inferior epigastric artery perforator flaps: Magnetic resonance angiography. Microsurgery 29:119, 2009 34. Schaverien MV, Ludman CN, Neil-Dwyer J, et al: Contrast-enhanced magnetic resonance angiography for perioperative imaging in DIEP flap breast reconstruction. Plast Reconstr Surg 128: 56, 2011 35. Phillips TJ, Stella DL, Rozen WM: Multidetector CT evaluation of abdominal wall for breast reconstruction: Take a look at the veins (Response to Letters to the Editor). Radiology 251:947, 2009 36. Schaverien MV, Ludman CN, Neil-Dwyer J, McCulley SJ: Contrast-enhanced magnetic resonance angiography for preoperative imaging of deep inferior epigastric artery perforator flaps: Advantages and disadvantages compared with computed tomography angiography. A United Kingdom perspective. Ann Plast Surg 67:671, 2011
197 37. Bouyssie JF, Bouyssie S, Sharrock P: Stereolithographic models derived from x-ray computed tomography: Reproduction accuracy. Surg Radiol Anat 19:193, 1997 38. Ro E, Ridge JA, Topham NS: Using stereolithographic models to plan mandibular reconstruction for advanced oral cavity cancer. Laryngoscope 117:759, 2007 39. Mavili ME, Canter HI, Salglam-Aydinatay B, Kamaci S, et al: Use of three-dimensional medical modeling methods for precise planning of orthognathic surgery. J Craniofac Surg 18:740, 2007 40. Hallermann W, Olsen S, Bardyn T, et al: A new method for computer-aided planning for extensive mandibular reconstruction. Plast Reconstr Surg 117:2431, 2006 41. Liu XJ, Gui L, Mao C, et al: Applying computer techniques in maxillofacial reconstruction using a fibula flap: A messenger and an evaluation method. J Craniofac Surg 20:372, 2009 42. Vakharia K, Natoli N, Johnson T: Stereolithography-aided reconstruction of the mandible. Plast Reconstr Surg 192:194e, 2012 43. Antony A, Wei FC, Kolokythas A, et al: Use of virtual surgery and stereolithography-guided osteotomy for mandibular reconstruction with free fibula. Plast Reconstr Surg 128:1080, 2011 44. Juergen P, Klug C, Krol Z, et al: Navigation guided harvesting of autologous iliac crest graft for mandibular reconstruction. J Oral Maxillofac Surg 69:2915, 2011 45. Modabber A, Gerressen M, Stiller MB, et al: Computer-assisted mandibular reconstruction with vascularized iliac crest bone graft. Aesthetic Plast Surg 36:653, 2012 46. Shen Y, Sun J, Li J, et al: Using computer simulation and stereomodel for accurate mandibular reconstruction with vascularized iliac crest flap. Oral Surg Oral Med Oral Pathol Oral Surg Oral Med Oral Pathol Oral Radiol 114:175, 2012
Developments in Image-Guided Deep Circumflex Iliac Artery Flap Harvest: A Step-by-Step Guide and Literature Review Jeannette W. C. Ting, MBBS, MS,* Warren M. Rozen, MBBS, BMedSci, PGDipSurgAnat, MD, PhD,y Vachara Niumsawatt, MBBS,z Charles Baillieu, MBBS, FRACS,x Michael Leung, MBBS, FRACS,jj and James C. Leong, MBBS, MS, FRACS{ Purpose:
The deep circumflex iliac artery (DCIA) flap has evolved significantly over time in the intricacies of flap design and breadth of surgical application. This has been facilitated by advances in preoperative imaging and planning, in particular, computed tomographic angiography. Studies have highlighted that advanced imaging modalities and other technologies such as image-guided stereolithographic biomodeling can substantially improve flap planning, flap harvest, and operative outcomes.
Patients and Methods:
The present report comprises a combined literature review and clinical cohort study of 20 consecutive patients to assess the modern technologies applied to DCIA flap planning and harvest. We have also described a step-by-step guide for the implementation of these techniques into clinical practice.
Results:
The protocol for a single, standardized technique of computed tomographic angiography scanning is presented and was applied to a range of techniques in the preoperative planning of DCIA flaps. These include 1) bony and vascular imaging analysis of both donor and recipient sites, 2) stereolithographic ‘ biomodeling’’ of both donor and recipient bony and vascular anatomy, and 3) the use of preoperative ‘ virtual surgery’’ with image-guided stereotactic navigation. The application and role of each technique was explored. Conclusions: Modern imaging and stereolithographic techniques are innovations that can substantially improve surgical outcomes in DCIA flap surgery, such as has been highlighted in our clinical experience and in published studies. Notably, few outcome studies have been reported, and the need for larger case series and comparative studies is apparent. Crown Copyright Ó 2014 Published by Elsevier Inc on behalf of the American Association of Oral and Maxillofacial Surgeons. All rights reserved J Oral Maxillofac Surg 72:186-197, 2014
The deep circumflex iliac artery (DCIA) flap, a composite osteomusculocutaneous flap of the iliac crest, abdominal wall musculature, and overlying skin, has
evolved significantly during the previous 30 years since its inception in the late 1970s. With an increasingly reported role for a range of facial, lower limb, and upper {Plastic and Reconstructive Surgical Consultant, Department of
Received from the Department of Surgery, Monash University Faculty of Medicine, Clayton, Victoria, Australia.
Plastic and Reconstructive Surgery, Dandenong Hospital, Southern
*Plastic and Reconstructive Surgical Registrar, Department of
Health, Dandenong, Victoria, Australia.
Plastic and Reconstructive Surgery, Dandenong Hospital, Southern
Address correspondence and reprint requests to Dr Ting: Depart-
Health, Dandenong, Victoria, Australia.
ment of Plastic and Reconstructive Surgery, Dandenong Hospital,
yPlastic and Reconstructive Surgical Registrar, Department of
Southern Health, David St, Dandenong, Victoria 3175, Australia;
Plastic and Reconstructive Surgery, Dandenong Hospital, Southern
e-mail: [email protected]
Health, Dandenong, Victoria, Australia. zPlastic and Reconstructive Surgical Registrar.
Received April 13 2013 Accepted June 27 2013
xPlastic and Reconstructive Surgical Consultant, Department of
Crown Copyright Ó 2014 Published by Elsevier Inc on behalf of the American
Plastic and Reconstructive Surgery, Dandenong Hospital, Southern
Association of Oral and Maxillofacial Surgeons. All rights reserved
Health, Dandenong, Victoria, Australia.
0278-2391/13/00826-4$36.00/0
jjPlastic and Reconstructive Surgical Consultant, Department of
http://dx.doi.org/10.1016/j.joms.2013.06.219
Plastic and Reconstructive Surgery, Dandenong Hospital, Southern Health, Dandenong, Victoria, Australia.
186
187
TING ET AL
limb reconstructions,1-5 its most widespread utility has been for hemimandibular defect reconstruction.6-10 Although the iliac crest has long been used for these various bony reconstructions, its versatility as a composite flap has largely been limited by an understanding of the finer vascular anatomy of the region. Initial attempts to harvest the iliac crest flap using the superficial circumflex iliac artery as its vascular supply in 1978 met with less than ideal results.11 Although greater success was achieved with the DCIA pedicle flap after the landmark report by Taylor and Townsend12 in 1979, detailing the DCIA as the main blood supply to the iliac crest, a lack of familiarity with the DCIA perforators in these early studies limited the use of the DCIA flap as a composite flap. Recently, anatomic studies have definitively demonstrated the anatomy of the perforators to the skin overlying the iliac crest,13 and subsequent clinical case series have highlighted that preoperative computed tomographic angiography (CTA) can be an essential tool to guide surgeons in harvesting the DCIA perforator flap.14 Similarly, computer software technology has advanced exponentially, allowing a greater ability to manipulate the data obtained from CTA that can be of practical use to surgeons. Recent studies have described this and other technological advances to aid DCIA flap surgery, including the use of image-guided stereolithographic biomodeling for both bony and vascular modeling pre- and intraoperatively.15,16 A stereolithographic biomodel can be used as a preplanned and prefabricated instrument, thus reducing the operative time with improved precision. Although techniques such as color duplex ultrasound and magnetic resonance imaging have been described for vascular imaging, we have preferred to use CTA as our modality of choice for vascular mapping because of its availability, objectivity, accuracy, and efficacy.17 The present report describes a step-by-step guide for implementation of these techniques into clinical practice, described from both our clinical experience and literature review.
Patients and Methods Our clinical approach as described was based on implementation of these techniques in a cohort of 20 consecutive patients (12 men and 8 women) who underwent DCIA flap reconstruction for an oncologic defect. The patients had an age range of 39 to 75 years and a range of body habitus (Table 1). All patients provided written informed consent for inclusion, and their cases were deemed clinically appropriate for DCIA flap surgery, with no exclusions. The technique described was based on a standardized approach to preoperative imaging and interpretation, with the imaging findings reported in all cases by 1 of us. The
Table 1. DEMOGRAPHIC DATA
Demographic Data
Value
Patients (n) Age (yr) Mean Range Gender Male Female Diagnosis Ameloblastoma Squamous cell carcinoma Neuroblastoma Trauma Deep circumflex iliac artery flap Flap constituents Osseous flap Osseomuscular Osseomusculocutaneous Osseocutaneous Preoperative imaging: 64-slice multidetector row CTA Operative guidance based on imaging Pedicle Perforators Concordance between imaging and operative findings Pedicle Perforators Flap-related complications Donor site complications
20 60.2 39-75 12 8 2 15 1 2 20
3 2 5 10 20
20 15
19 15 0 1 (hematoma)
Abbreviation: CTA, computed tomographic angiography. Ting et al. Image-Guided DCIA Flap Harvest. J Oral Maxillofac Surg 2014.
imaging and technique of application of modern adjuncts have been described qualitatively and compared in the discussion to the findings from the literature review. The Melbourne Health HREC provided institutional ethical approval for evaluating the role of CTA (no. 2006.231). This and other aspects of the study complied with the Declaration of Helsinki, 1995. All patients gave informed consent for all imaging, surgery, and inclusion in the report, and patient anonymity was preserved. The results are reported qualitatively, with the methodologic approach to image-guided DCIA harvesting and insetting presented. TECHNIQUE
Each patient underwent CTA of both their defect site and donor site. The CTA hardware used included a Siemens Somatom Sensation 64 multidetector row CT scanner (Siemens Medical Solutions, Erlangen,
188 Germany). Intravenous contrast was used in all cases, with no oral contrast used, and included nonionic iodinated contrast media (Ultravist 370, Schering, Berlin, Germany; or Omnipaque 350, Amersham Health, Princeton, NJ). Intravenous access was accessed through a cubital fossa vein, with an 18-gauge cannula, and injection performed using a biphasic power injection pump at a flow rate of 4 to 6 mL/s. A bolus tracking technique was used to identify filling of the appropriate vessels with contrast as a method to initiate scanning. The scan was timed from the external iliac artery for scan triggering (to initiate arterial phase filling). At that point, the minimum delay in scanning was undertaken (this was defined by the scanner software and was typically 4 seconds). These techniques aim to provide pure ‘‘arterial phase’’ scans, with no or minimal venous opacification. A similar technique was used for the recipient site. Image reformatting software was achieved with 2 software programs: Siemens Syngo InSpace (InSpace 2004A_PRE_19, Siemens Australia, Victoria, Australia) and Osirix (Osirix Medical Imaging Software, GPL Licensing Open Source Initiative, http://www.osirixviewer.com). Multiplanar 3-dimensional reconstruc-
IMAGE-GUIDED DCIA FLAP HARVEST
tions were achieved with maximum intensity projection and volume-rendered technique reconstructions. All specific measurements of the diameter or distances were performed using the raw fine axial images. Stereotactic navigation was achieved from the same raw data using the technique described below. Step 1: Preoperative Imaging of Recipient Site Anatomy A single head and neck CT scan can assess tumor oncology in terms of the soft tissue and bony extent and involvement and relation to the surrounding anatomy and to plan the degree of soft tissue and bony resection (Fig 1).15,16 The same scan can be used to map the bony anatomy of the mandible to create a template for bony reconstruction. If CTA has been performed, the same scan can be used to map the locoregional vascular anatomy for the free flap recipient vasculature. Step 2: Stereolithographic Modeling of Recipient Anatomy Using the same volumetric scan data from step 1, the image data were transferred as raw data for the production of the image-guided stereolithographic models
FIGURE 1. Preoperative 3-dimensional, multiplanar computed tomography image of the mandible for resection planning. Reproduced in part with permission from: Rozen WM, Ting JWC, Leung M, et al: Advancing image-guided surgery in microvascular mandibular reconstruction: Combining bony and vascular imaging with CT guided stereolithographic bone modelling. Plast Reconstr Surg 130:227e, 2012. Ting et al. Image-Guided DCIA Flap Harvest. J Oral Maxillofac Surg 2014.
189
TING ET AL
FIGURE 2. Image-guided, stereolithographic biomodel (BioModels, Anatomics) of the mandible for resection planning (note comparison to equivalent computed tomography scan in Fig 1). Reproduced in part with permission from: Rozen WM, Ting JWC, Leung M, et al: Advancing image-guided surgery in microvascular mandibular reconstruction: Combining bony and vascular imaging with CT guided stereolithographic bone modelling. Plast Reconstr Surg 130:227e, 2012. Ting et al. Image-Guided DCIA Flap Harvest. J Oral Maxillofac Surg 2014.
(BioModels, Anatomics, Victoria, Australia). These defect-site models can visually demonstrate the site, size, and shape of the defect, with the tumor able to be ‘‘pre-resected’’ to form the template models (Fig 2).15,16 These models were brought to the multidisciplinary meeting before surgery to allow the resection and reconstructive surgeons to use these models to familiarize themselves with the defect. Resection templates were then made according to
the surgical plans, and the fixation plates can be premolded and modeled to the defect preoperatively. These models were all sterilized and made available during the operation. Step 3: Preoperative Imaging of Donor Site Anatomy A donor site scan can be used to map all aspects of the donor anatomy and to create a ‘ virtual guide’’
FIGURE 3. Preoperative computed tomographic angiogram with coronal maximum intensity projection reformatting, highlighting the deep circumflex iliac artery and its course, with a branch to the iliac crest (red arrow) and ascending branch (yellow arrow) shown. Ting et al. Image-Guided DCIA Flap Harvest. J Oral Maxillofac Surg 2014.
190 through each step of flap harvest. The DCIA pedicle itself can be demonstrated, showing the origin of the vessel, the location relative to other soft tissue and bony landmarks, the diameter throughout its course, the branching pattern, and its course toward the iliac crest (Fig 3). Furthermore, specific branches to the iliac crest can be followed, enabling the selection of the optimally vascularized part of the iliac crest for harvest.4 The accessory branch of the DCIA, which supplies the internal oblique and transverses the abdominis muscles, can be assessed if a muscle cuff is required. Cutaneous perforators can also be assessed for a vascular supply to a skin paddle, if required (Fig 4).13,14 The perforator number, location, size, and course can be clearly evaluated using CTA. The individual branches to each lamina of a composite flap can be traced, with branches to the skin, muscle, and bone often separately identifiable (Fig 5).15,16 Step 4: Stereolithographic Modeling of Flap (Donor Site) Anatomy The same volumetric scan data from step 3 can also be used to plan surgery, with the image data transferred as raw data for the production of image-guided stereoli-
IMAGE-GUIDED DCIA FLAP HARVEST
thographic models of the donor anatomy (Figs 5 and 6), similar to the method used for the recipient site (BioModels, Anatomics, Victoria, Australia). These can then directly facilitate the operative approach (Figs 7-9). Step 5: Preoperative ‘‘Virtual Surgery’’ With ImageGuided Stereotactic Navigation Image-guided stereotactic navigation is a technique that uses the same preoperative scan (CTA or magnetic resonance angiography [MRA]) performed for routine preoperative imaging, with no additional imaging required, and transfers the data through infrared and computer analysis for additional surgical analysis.18 The technique of image-guided stereotaxy comprises several key components: ‘ registration’’ of the patient anatomy, the transfer of data to the computer systems, and navigation through the scan anatomy in ‘ real time.’’ Patient ‘ registration’’ was achieved using fiducial marker registration, with radiopaque markers applied to the skin surface. After CTA scanning, the images generated were exported as Digital Imaging and Communications in Medicine images and transferred to the stereotactic computer system (BrainLAB, Vector Vision, Cranial, version 7.0, Sydney, Australia). Real-time navigation through the scan data was achieved by placing the
FIGURE 4. Preoperative computed tomographic angiogram with 3-dimensional, volume-rendered reconstruction, highlighting the deep circumflex iliac artery perforators (arrows), emerging from the deep fascia posterosuperiorly to the iliac crest, for planning a skin paddle in an osteocutaneous deep circumflex iliac artery flap. Ting et al. Image-Guided DCIA Flap Harvest. J Oral Maxillofac Surg 2014.
191
TING ET AL
multiple computer platforms. ImageJ is free to download and use, with numerous plug-in functions that allow image analysis.19 The process involves the use of volumetric scan data from steps 1 and 3, with the image data able to be processed with different functional plug-ins. The anatomy of the DCIA vessels and iliac crest can be assessed and quantified accurately using the software. It allows for an evaluation of the vessels, not only of their quantity and diameter, but also the density and measurements of distances to bony landmarks. The evaluated images can then be stored in several image formats such as JPEG, TIFF, BMP, or RAW (Fig 10). The stored images can be reassessed by different individuals, improving their sensitivity and specificity and reproducibility. MIMICS BIOMEDICAL SOFTWARE
FIGURE 5. Preoperative 3-dimensional, multiplanar computed tomographic angiogram of the donor iliac crest and deep circumflex iliac artery to plan donor site harvest, highlighting pedicle branches to the iliac crest. Reproduced with permission from Rozen WM, Ting JW, Baillieu C, Leong J: Stereolithographic modeling of the deep circumflex iliac artery and its vascular branching: A further advance in computed tomography-guided flap planning. Plast Reconstr Surg 130:380e, 2012 Ting et al. Image-Guided DCIA Flap Harvest. J Oral Maxillofac Surg 2014.
patients in a supine position, mimicking both the position during CT scanning and the operative position, with a ‘ stereotactic pointer’’ or ‘ wand’’ used to navigate through the 3-dimensional anatomy. Intraoperatively, surgeons were able to use the navigation system to perform accurate osteotomies of both the donor and recipient sites in accordance with the preoperative plans. It was also used to identify the pedicle and perforator location throughout dissection of the vasculature to reduce the risk of inadvertently damaging the vessels. Step 6: Preoperative Vascular Assessment With ImageJ ImageJ is a Java-based image analysis application for high output processing and image manipulation on
The process of preoperative planning, anatomic analysis, visual 3-dimensional modeling, and anatomic models has been widely accepted and used, with many software companies adopting these concepts and producing various products that are commercially available. One such software product that has been developed is called ‘‘MIMICS innovation suite,’’ which incorporates 4 components to assist reconstructive surgeons in preoperative planning. The suite has 3 components: virtual image reconstruction; computer-aided design tools; and anatomic models. The process begins with reconstruction of 2-dimensional images that can be reconstructed into 3-dimensional virtual images similar to those of ‘ Osirix.’ These images can be manipulated to assess and analyze the internal anatomy. The reconstructed images can then be transferred to computer software, where a prosthesis or virtual surgery can be performed using a computer console. The surgeon can thus practice, design, and implement patient-specific osteotomies and ostectomies and placement of fixation devices before any surgery. The end result can be viewed in 3 dimensions and any necessary adjustment can be made at the same time (either preoperatively or intraoperatively). Ultimately, the designed images can then be reproduced into an anatomical 3-dimensional model. The product is tangible, accurate, and realistic, allowing surgeons to explore and evaluate the patient anatomy to gain better insight into specific pathologic features. It also allows for prefabrication of plates and the osteotomy design, which has been shown to reduce the operative time and increase the accuracy in flap positioning and planning.20,21
Results CTA and stereolithographic models were performed and made for 20 consecutive patients undergoing DCIA reconstruction (Table 1). All patients had clear oncologic clearance. Intraoperative adjustments of
192 the resections were all within 1 cm of the preoperatively planned models. All patients underwent successful dissection and elevation of their DCIA flap, and no complete flap failures occurred. The concordance between the imaging and operative findings of the pedicle was 95%, with 100% concordance for the perforators. This was determined by the number and location of the pedicles and perforators seen on the model compared with the intraoperative findings using the anterior superior iliac spine as the bony landmark. Minor adjustments of the preoperative molded plates were made during surgery for a small number of patients, with no significant increase in the operation time. One donor site complication occurred, a donor site hematoma. The patients were followed for 18 months to 4 years, with 1 recurrence found during the follow-up period.
Discussion The iliac crest has long been recognized to be an excellent source of bone graft material and was widely used as a nonvascularized graft as early as World War I, with poor results.22,23 It was not until the 1970s, with the developments in microsurgery and vascularized transfers, that surgeons were able to markedly improve the iliac bone graft survival rate. The vascularized DCIA flap was popularized in association with an inconspicuous donor scar and low donor site morbidity.24-26 In addition, it has a dependable vascular pedicle length and diameter and a wide arc of rotation for the flap.12 Although the free fibula flap is commonly used to reconstruct osteocutaneous defects in the head and neck region, the DCIA flap is particularly useful for hemimandibular defects. Osteotomies of the bone are not needed, and the anterior inferior iliac spine can be used to reconstruct the mandibular condyle and hence the temporomandibular joint. Its versatility can be increased for defects that require variable amounts of bone, soft tissue, and skin that were once thought too small and the flap too bulky to reconstruct. This includes the iliac cortex splitting technique first described by Taylor9 in 1982 and again described by Shenaq and Klebuc8 in 1994. This was introduced to harvest smaller, thinner bone from the iliac crest and to reduce the rate of hernia after removal of part of the iliac crest. The muscle cuffsparing variation, aiming to reduce the bulk of the flap, was attempted by Safak et al25 in 1997 and Kimata27 in 2003, with limited success. Although no anterior mandibular defects involved the mandibular symphysis in our study, such a defect would require reconstruction of the anterior prominence of the mandible. Our preference for reconstruction of such a defect would be to use the straight segment of the superolateral iliac crest with or without the need for osteotomy. Using preoperative
IMAGE-GUIDED DCIA FLAP HARVEST
imaging and 3-dimensional reconstruction of both the donor (iliac crest) and recipient (mandible) sites, such reconstruction can be performed to assess and implement the contour geometry. However, should an adjustment to the curvature for the defect require an osteotomy of the iliac crest, this can be preplanned and, indeed, preplated in a similar manner as for mandibular ramus or body reconstruction. Furthermore, separate bony branches of the DCIA can also be identified on imaging and incorporated into the flap to increase the reliability of the segmental osteotomies. As perforator flaps became more popular in the 1980s, the DCIA perforator flap was used cautiously, because it had developed a reputation of having unreliable perforators. Without perforator skeletonization, it was a bulky flap that depended on the underlying abdominal musculature to ensure skin survival.28-30 However, the view that DCIA perforators were unreliable was found questionable in a large study of these perforators in 2009. It found that reliable perforators greater than 0.8 mm were present in more than 60% of hemiabdomens.13 A subsequent case series demonstrated that with preoperative CTA, 6 of 7 patients who underwent planning for a DCIA perforator flap had it raised successfully without any flap loss or significant related morbidity.14 Preoperative imaging was able to highlight specific branches of the DCIA to the iliac crest, facilitating more precise bone harvest and selection of optimally vascularized bone. This was highlighted in a case report in which a small amount of vascularized iliac crest was used to reconstruct a radial bony defect.4 The DCIA flap should be considered a valuable tool in the reconstructive surgeons’ repertoire of composite free flaps. It provides strong cortical bone with a reliable vascular source and is especially effective for the mandible, where solid bone is required to withstand the normal forces of mastication and for osseointegration. The cortex splitting technique, in which only the inner cortex is used, ensures a large amount of bone will also be available for thinner bony reconstruction such as the orbital floor and maxilla. The associated internal oblique muscle can be used to close maxillectomy fistulas and becomes epithelialized with minimal bulk.31 The skin paddle is often independent of the overlying bone (unlike the free fibula flap) and muscle (supplied by the separate accessory branch) and can be used to reconstruct a skin defect that is a small distance from the bony defect. In addition, the skin paddle, unlike the free fibula flap, is not hairy and is well tolerated by patients when used to line the oral mucosa. Although CTA has been the investigation modality of choice for imaging of the perforators, concerns have often been raised about radiation exposure and the accuracy of the data compared with other imaging techniques. With scanning protocol modifications, we
TING ET AL
(and other groups) have been able to ensure that patients are only exposed to 6 mSv of radiation, equivalent to 2 plain radiographic abdominal series or less than one quarter of the radiation of a full trauma CT scan.32 CTA has been shown to be superior in imaging abdominal wall perforators compared with ultrasound or MRA.33 Useful perforators of 1 mm in diameter might not be highlighted with MRA,34 and the validation of MRA in the identification of perforators has not been as widely described. For investigation of the venous anatomy, only CTA has been shown to be an effective option.35 In addition, although studies have suggested that MRA might be preferential to CTA because of the lack of ionizing radiation, the true clinical efficacy of MRA has not been validated.34,35 Studies have also conceded that MRA is more operator dependent than CTA, is more difficult to process, and is less available than CTA,36 making it a less accessible imaging modality. A recent study by Schaverien et al34 has also demonstrated an almost 10% incidence of respiratory artifact with MRA that has not been reported with CTA of the abdomen. With MRA technology constantly evolving, the use of MRA to produce stereolithographic models might be explored in the future. Technological advances in imaging and microsurgery, in particular, with modern modalities and stereolithographic biomodels, have been described in published studies for a number of years and with reproductive accuracy.37 However, their application to mandibular reconstruction38-43 and iliac crest harvesting15,44-46 has only recently been reported. Stereolithographic biomodels add an extra element of tangibility that allows the surgeon to ‘ see’’ and ‘ feel’’ the surgery beforehand and, combined with imaging, can enable ‘ virtual surgery’’ preoperatively. Surgeons can also use the models as guides during surgery, and modifications to the plan can be easily communicated between the surgical teams15,16 (Fig 6). The use of these biomodels can minimize bone harvest by selection of the most appropriately vascularized and best-fitting bone, spare donor site morbidity through more precise harvesting, and improve bone-to-bone contact at the recipient site (Figs 7–9). This can facilitate increased predictability of the anticipated surgery, especially in complex craniomaxillofacial reconstructions. In addition, surgical teams can anticipate the potential difficulties and have contingency plans in the event of unexpected complications. Imaging and biomodels can also be used in the consent process to explain the surgery and reconstruction to the patient and can be a valuable teaching tool for the training surgeon. Premolding of plates and measurement of screws can reduce the operation time and allow the resection and reconstruction teams to work simultaneously. The premolded plate can also serve as
193
FIGURE 6. Image-guided stereolithographic bone and vascular biomodel (BioModels, Anatomics) of the donor iliac crest and deep circumflex iliac artery to plan donor site harvest (note, comparison to equivalent computed tomography angiogram scan in Fig 5). Reproduced with permission from Rozen WM, Ting JW, Baillieu C, Leong J: Stereolithographic modeling of the deep circumflex iliac artery and its vascular branching: A further advance in computed tomography-guided flap planning. Plast Reconstr Surg 130:380e, 2012. Ting et al. Image-Guided DCIA Flap Harvest. J Oral Maxillofac Surg 2014.
a trial, allowing reconstructive surgeons to perform as many preoperative adjustments as needed. Before the operation, the premolded plate can be sterilized and used in the operation. However, the current status of virtual 3 dimensions is still in its early state and expensive and requires expertise to operate the system. This planning can enable a degree of independent movement of the individual flap components. The pedicle length and, therefore, the arc of movement can also be measured preoperatively to determine whether the length is adequate. Other related techniques have been offered for computermodeled DCIA flap harvest,45,46 with reported benefits. Image-guided stereotaxy is a recent advance in imaging technology, with computer guidance facilitating improved surgical planning and accuracy. We have
194
IMAGE-GUIDED DCIA FLAP HARVEST
FIGURE 7. Operative hemimandibular resection specimen. The resection was performed for ameloblastoma. Ting et al. Image-Guided DCIA Flap Harvest. J Oral Maxillofac Surg 2014.
found that image-guided stereotactic navigation can accurately identify the location and course of the DCIA perforators, potentially to a greater degree of accuracy than CTA alone. This might be because the
localization is achieved in 3 dimensions (unlike the 2-dimensional analysis with CTA), and the data collection occurs in a ‘‘live’’ patient, potentially taking into account the mobility of the integument during data
FIGURE 8. Deep circumflex iliac artery flap after harvest (note, match to resected specimen in Fig 7). Ting et al. Image-Guided DCIA Flap Harvest. J Oral Maxillofac Surg 2014.
195
TING ET AL
FIGURE 9. Postoperative computed tomographic angiogram of the mandible after flap inset. Ting et al. Image-Guided DCIA Flap Harvest. J Oral Maxillofac Surg 2014.
FIGURE 10. Screenshot of image analysis of the reconstructed computed tomographic angiographic images using ImageJ. Ting et al. Image-Guided DCIA Flap Harvest. J Oral Maxillofac Surg 2014.
196 collection. By combining the findings of routine CTA and CTA-guided stereotaxy, the efficacy of scanning can be improved. The adoption of these new technologies has currently been largely limited by the financial costs of computer hardware and software and the costs in the production of biomodels. Although it might not be financially feasible to use such models in all reconstructive cases, for complex reconstruction cases, the financial cost of the model and the extra time taken to plan the operation will be offset by the reduced intraoperative time, the more precise harvesting of the donor bone (and therefore reduced morbidity),43 and increased patient satisfaction with the final operative results.38,45 A recent study examining the use of CT and biomodels to aid in free bone flap harvest found that this modeling reduced the operative time by 7 to 22 minutes, with no major adjustments required for iliac crest harvesting and shaping intraoperatively.45 With improved vascular modeling, as highlighted, these outcomes could be improved further.15,16 In our experience, imaging was thought to substantially reduce the operative time; however, without a comparative group, the actual calculation of the operative time was outside the scope of the present study, although it will form the basis of future investigation. Gillie’s principle of carefully planning before operating is doctrine to most facial surgeons. In the past 40 years, with the familiarity of microsurgery, the question is no longer about survival of the free flap or perforator flap. Innovations should aim to improve and refine free flap techniques to reduce donor site morbidity, reduce the surgical times, and reduce the incidence and severity of overall complications. In conclusion, modern imaging and stereolithographic techniques are innovations that can substantially improve the surgical outcomes of DCIA flap surgery. Future advances are likely to include finer resolution imaging techniques, advances in in vivo realtime anatomic navigation, and improved computer software and analysis systems. The published data support the use of existing technologies, and, with few outcome studies reported, the need for larger case series and comparative studies is apparent.
References 1. Peek A, Giessle GA: Functional total and subtotal heel reconstruction with free composite osteofasciocutaneous groin flaps of the deep circumflex iliac vessels. Ann Plast Surg 56:628, 2006 2. Sanders R, Mayou BJ: A new vascularized bone graft transferred by microvascular anastomosis as a free flap. Br J Surg 66:787, 1979 3. Leung PC, Chan KT: Giant cell tumour of the distal end of the radius treated by the resection and free vascularized iliac crest graft. Clin Orthop Relat Res 202:232, 1986
IMAGE-GUIDED DCIA FLAP HARVEST 4. Ting JWC, Rozen WM, Leong J, Crock J: Free deep circumflex iliac artery vascularised bone flap for reconstruction of the distal radius: Planning with CT angiography. Microsurgery 30:163, 2010 5. Gabl M, Reinhart C, Lutz M, et al: Vascularized bone graft from the iliac crest for the treatment of nonunion of the proximal part of the scaphoid with an avascular fragment. J Bone Joint Surg 81:1414, 1999 6. Urken L, Vickery C, Weinberg H, Buchbinder DM: The internal oblique iliac crest osseomyocutaneous free flap in oromandibular reconstruction: Report of 20 cases. Arch Otolaryngol Head Neck Surg 115:339, 1989 7. Brown JS: Reconstruction of the total maxillectomy defect using a pedicled coronoid flap and deep circumflex iliac artery free flap. Br J Oral Maxillofac Surg 26:423, 2008 8. Shenaq SM, Klebuc M: The iliac crest microsurgical free flap in mandibular reconstruction. Clin Plast Surg 21:37, 1994 9. Taylor GI: Reconstruction of the mandible with free composite iliac bone graft. Ann Plast Surg 9:361, 1982 10. Taylor GI, Townsend P, Corlett R: Superiority of the deep circumflex iliac vessels as the supply for free groin flaps: Clinical work. Plast Reconstr Surg 64:745, 1979 11. Daniel RK, Kerrigan CL: Mandibular reconstruction with free tissue transfers. Ann Plast Surg 1:346, 1978 12. Taylor GI, Townsend P, Corlett R: Superiority of the deep circumflex iliac vessels as the supply for free groin flaps: Experimental work. Plast Reconstr Surg 64:595, 1979 13. Ting JWC, Rozen WM, Grinsell D, et al: The in vivo anatomy of the deep circumflex iliac artery perforators: Defining the role for the DCIA perforator flap. Microsurgery 29:326, 2009 14. Ting JWC, Rozen W, Chubb D, et al: Improving the utility and reliability of the deep circumflex iliac artery perforator flap: The use of preoperative planning with CT angiography. Microsurgery 31:603, 2011 15. Rozen WM, Ting JW, Baillieu C, Leong J: Stereolithographic modeling of the deep circumflex iliac artery and its vascular branching: A further advance in computed tomography-guided flap planning. Plast Reconstr Surg 130:380e, 2012 16. Rozen WM, Ting JWC, Leung M, et al: Advancing image-guided surgery in microvascular mandibular reconstruction: Combining bony and vascular imaging with CT guided stereolithographic bone modelling. Plast Reconstr Surg 130:227e, 2012 17. Rozen WM, Ribuffo D, Atzeni M, et al: Current state of the art in perforator flap imaging with computed tomographic angiography. Surg Radiol Anat 31:631, 2009 18. Rozen WM, Buckland A, Ashton MW, et al: Image-guided stereotactic perforator flap surgery: A prospective comparison of current techniques and review of the literature. Surg Radiol Anat 31:401, 2009 19. Abramoff MD, Magelhaes PJ, Ram SJ: Image processing with image. J Biophotonics Intl 11:36, 2004 20. Markiewicz MR, Bell RB: The use of 3D imaging tools in facial plastic surgery. Facial Plast Surg Clin North Am 19:655, 2011 21. Anuja 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 128:1080, 2011 22. Disa JJ, Cordeiro PG: Mandible reconstruction with microvascular surgery. Semin Surg Oncol 19:226, 2000 23. Phillips CM: Primary and secondary reconstruction of the mandible after ablative surgery. Am J Surg 114:601, 1967 24. Kudo K, Fujioka Y: Review of bone grafting for reconstruction of discontinuity defects of the mandible. J Oral Surg 36:791, 1978 25. Safak T, Klebuc MJ, Mavili E, Shenaq SM: A new design of the iliac crest microsurgical free flap without including the ‘ obligatory’’ muscle cuff. Plast Reconstr Surg 100:1703, 1997 26. Bitter K, Schlesinger S, Westerman U: The iliac bone or osteocutaneous transplant pedicled to the deep circumflex iliac artery. J Maxillofac Surg 11:241, 1983 27. Kimata Y: Deep circumflex iliac perforator flap. Clin Plast Surg 30:433, 2003 28. Akyurek M, Conejero A, Dunn R: Deep circumflex iliac artery perforator flap without iliac crest. Plast Reconstr Surg 122: 1790, 2008
TING ET AL 29. Goldie SJ, Thompson R, Soutar DS, Shaw-Dunn J: Arterial supply of the deep circumflex iliac artery (DCIA) flap. Eur J Plast Surg 31:311, 2008 30. Akyurek M, Morse A, Safak T, et al: Deep circumflex iliac artery perforator skin flap: Anatomic study and clinical application. J Reconstr Microsurg 22:A031, 2006 31. Brown JS: Deep circumflex iliac artery free flap with internal oblique muscle as a new method of immediate reconstruction of maxillectomy defect. Head Neck 18:412, 1996 32. Rozen WM, Whitaker IS, Stella DL, et al: The radiation exposure of computed tomographic angiography (CTA) in DIEP flap planning: Low dose but high impact. J Plast Reconstr Aesthet Surg 62:e654, 2009 33. Rozen WM, Stella DL, Bowden J, et al: Advances in the preoperative planning of deep inferior epigastric artery perforator flaps: Magnetic resonance angiography. Microsurgery 29:119, 2009 34. Schaverien MV, Ludman CN, Neil-Dwyer J, et al: Contrast-enhanced magnetic resonance angiography for perioperative imaging in DIEP flap breast reconstruction. Plast Reconstr Surg 128: 56, 2011 35. Phillips TJ, Stella DL, Rozen WM: Multidetector CT evaluation of abdominal wall for breast reconstruction: Take a look at the veins (Response to Letters to the Editor). Radiology 251:947, 2009 36. Schaverien MV, Ludman CN, Neil-Dwyer J, McCulley SJ: Contrast-enhanced magnetic resonance angiography for preoperative imaging of deep inferior epigastric artery perforator flaps: Advantages and disadvantages compared with computed tomography angiography. A United Kingdom perspective. Ann Plast Surg 67:671, 2011
197 37. Bouyssie JF, Bouyssie S, Sharrock P: Stereolithographic models derived from x-ray computed tomography: Reproduction accuracy. Surg Radiol Anat 19:193, 1997 38. Ro E, Ridge JA, Topham NS: Using stereolithographic models to plan mandibular reconstruction for advanced oral cavity cancer. Laryngoscope 117:759, 2007 39. Mavili ME, Canter HI, Salglam-Aydinatay B, Kamaci S, et al: Use of three-dimensional medical modeling methods for precise planning of orthognathic surgery. J Craniofac Surg 18:740, 2007 40. Hallermann W, Olsen S, Bardyn T, et al: A new method for computer-aided planning for extensive mandibular reconstruction. Plast Reconstr Surg 117:2431, 2006 41. Liu XJ, Gui L, Mao C, et al: Applying computer techniques in maxillofacial reconstruction using a fibula flap: A messenger and an evaluation method. J Craniofac Surg 20:372, 2009 42. Vakharia K, Natoli N, Johnson T: Stereolithography-aided reconstruction of the mandible. Plast Reconstr Surg 192:194e, 2012 43. Antony A, Wei FC, Kolokythas A, et al: Use of virtual surgery and stereolithography-guided osteotomy for mandibular reconstruction with free fibula. Plast Reconstr Surg 128:1080, 2011 44. Juergen P, Klug C, Krol Z, et al: Navigation guided harvesting of autologous iliac crest graft for mandibular reconstruction. J Oral Maxillofac Surg 69:2915, 2011 45. Modabber A, Gerressen M, Stiller MB, et al: Computer-assisted mandibular reconstruction with vascularized iliac crest bone graft. Aesthetic Plast Surg 36:653, 2012 46. Shen Y, Sun J, Li J, et al: Using computer simulation and stereomodel for accurate mandibular reconstruction with vascularized iliac crest flap. Oral Surg Oral Med Oral Pathol Oral Surg Oral Med Oral Pathol Oral Radiol 114:175, 2012