ComputerizedTomographic andMagnetic ResonanceAngiography forPerforator-BasedFree Flaps: TechnicalConsiderations

Justin S. Lee, MDa, Ketan M. Patel, MDb, Zhitong Zou, MDc,Martin R. Prince, MD, PhDc, Emil I. Cohen, MDa,*

KEYWORDS

� Computed tomographic angiography� Magnetic resonance angiography� Preoperative planning � Surgical flaps

The advent of multidetector computed tomog-raphy (MDCT), with an ever-increasing number ofdetectors and faster gantry rotations, has revolu-tionized diagnostic radiology, allowing for rapidimaging at an increased resolution.1 CouplingMDCT technology with improvements to intrave-nous contrast has made it possible to performprecise imaging during the arterial phase ofcontrast infusion increasing resolution of smallervessels.2 Computed tomographic angiography(CTA) is now routinely used in vascular, abdominal,and transplant surgery for its ability to provideaccurate vascular anatomic detail. In mostsettings it has supplanted conventional invasivecatheter angiography as the diagnostic imagingmodality of choice for imaging blood vessels.CTA is a noninvasive method for preoperativeplanning, such as determining tumor resectability,arterial anatomy before organ donation, and

a Division of Vascular and Interventional Radiology,Hospital, 3800 Reservoir Road NW, Washington, DC 2000b Department of Plastic Surgery, Johns Hopkins UniversitUSAc Department of Radiology, Weill Cornell Imaging atColumbia University, 416 East 55th Street, New York, NY* Corresponding author.E-mail address: Emil.I.Cohen@gunet.georgetown.edu

Clin Plastic Surg 38 (2011) 219–228doi:10.1016/j.cps.2011.03.0020094-1298/11/$ – see front matter � 2011 Elsevier Inc. All

extent of peripheral vascular disease.3–5 Asadvanced CTA has become mainstream, it is notsurprising that new and novel applications havebeen developed in other specialties, includingreconstructive surgery.

Over the last several decades, the use ofperforator-based free flaps has gained appealbecause of the reduction in donor-site morbiditycommon with conventional musculocutaneousflaps.6,7 Successful perforator-based free flapsrely on selection of the appropriate dominantvessel supplying the vascular territory of the flap.Generally, anatomic variability increases in distalbranches beyond the parent vessel. In addition,anatomic variability tends to increase as vesselsize decreases. Improvements to surgical tech-nique, allowing for the harvest of smaller, distalvascular segments has made knowledge of thenative vascular anatomy critical during surgical

Department of Radiology, Georgetown University7, USAy, Georgetown University Hospital, Washington, DC,

New York Presbyterian Hospital, Weil Cornell and10022, USA

rights reserved. plasticsurgery.th

eclinics.com

Lee et al220

dissection. Anatomic course and perforator diam-eter are important determining factors for flapperfusion.8 Furthermore, understanding the perfo-rator distribution and blood supply of a flap canimpact the operative time and perfusion of theflap, ultimately affecting the outcome.9

Doppler ultrasound is used extensively for free-flap operative planning. Duplex sonography addsthe ability to assess the rate of flow in additionto the acoustic characteristics of a vessel.10 Unfor-tunately, ultrasound of any form is limited by itssubjective nature, time required to performa quality evaluation, and ultimately its reproduc-ibility. Digital subtraction angiography was at onetime considered the gold standard for the identifi-cation and mapping of vessels. However, it is aninvasive procedure, which always has the risk ofcomplications, such as vessel dissection andaccess-site hemorrhage. Evaluation of small,perforator-sized vessels requires selective angiog-raphy, which increases the risk of vessel injury andstill fails to accurately depict the course of thevessel of interest in the surrounding soft tissue.11

Magnetic resonance angiography (MRA) andCTA have the advantages of being noninvasivemethods of imaging vascular anatomy with highspatial resolution and soft-tissue detail that iseasily reproducible. MRA has the advantage ofnot using ionizing radiation, which is a consider-ation, particularly when dealing with youngerpatient populations. These two imaging modalitiesare discussed further as tools for preoperativeevaluation in perforator-based flap reconstruction.Alonso-Burgos and colleagues12 published one

of the first reports using CTA for reconstructivesurgery in which 6 patients were evaluated usingCT for deep inferior epigastric perforator (DIEP)tissue flap planning. A 4-detector row MDCTscanner was used with 150 mL of iodixanolcontrast medium. The investigators obtaineda slice thickness of 1.25 mm and reformattedimages into multiplanar reformats, maximumintensity projections (MIP), and 3-dimensional(3-D) volume-rendered images. Arterial perforatorswere identified and evaluated for vessel diameter,fascial penetration pattern, intramuscular course,origin from the deep inferior epigastric artery,and other anatomic variations. In all 6 patients,accurate main perforators were identified on CTAwith no additional vessels found at the time ofthe surgery. In addition, CTA provided importantadjunctive preoperative information, such asmuscular diastasis, abdominal wall hernia, andfatty infiltration of potential flaps. This early experi-ence with an early 4-detector row scanner showedpromising results for CTA as a noninvasive meansfor presurgical planning.

The same year Masia and colleagues13 pub-lished a retrospective review of 66 DIEP flap recon-structions in which CTA was used for preoperativeplanning. The investigators found an average timesavedof 1hour and40minutes in caseswithpreop-erative CTA. There were 2 cases with partialnecrosis and1 total failure in the groupwithout priorCTA and only 1 partial necrosis case in the CTAgroup. The investigators thought that CTA offereda high sensitivity, specificity, and 100% positivepredictive value. Furthermore, they were able tohighlight the value of CTA as a tool to reduce oper-ative time by identifying the most suitable perfo-rator allowing safe ligation of other smaller vessels.In 2008, Rozen and colleagues14 published

a series of articles using CTA for DIEP and superfi-cial inferior epigastric artery (SIEA) flaps (Table 1).The first report included 75 patients using a 64-detector row scanner the images reconstructed as1-mm slices. Of the 75 patients, they specificallydescribe seven cases in which CTA actuallychanged the operative plan due to the anatomy,particularly patients with prior abdominal wallsurgical history. Later, a second study performedby the same group evaluated 104 reconstructionstodetermine if therewereoutcomedifferencesafterpreoperative CTA.15 The investigators concludedthat preoperative CTAwas associatedwith a statis-tically significant decrease in flap complications,donor-site morbidity and operative stress for thesurgeon. In a similar study, Smit and colleagues16

also demonstrated a trend toward the reduction insurgical time. Their study compared 70 patientswho were evaluated preoperatively with CTA and68 by preoperative Doppler ultrasound. There wasa statistically significant decrease in surgical timeand no flap complications in the CTA group.Despite CTA strengths in arterial imaging, limita-

tions do exist. A comparison study of Doppler toCTA for DIEP flap planning in 2010 evaluated45 patients preoperatively examined with bothDoppler and CTA. In this series, the dominantperforator used for the flapwas found in 44 patientswith Doppler and 41 patients with CTA.17 Addition-ally, among CTA patients, there was a disagree-ment in perforator size described on the CTAcompared with what was found during surgery.Overestimation on the CTA was attributed toa summation of the perforating artery and adjacentvein during measurement, likely secondary tovolume averaging. The investigators did find thatCTA provided a better analysis of the intramuscularcourse of the vessels as well as assessment ofsuperficial venous communication, and that overallCTA provided a global picture to the surgeon.MRA has continued to improve on its ability to

visualize vessels distinctly within adjacent soft

Table 1Comparison of recently available publications evaluating CTA for DIEP flap harvesting

Authors YearNumber ofPatients/Flaps Free Flap

ReportedSuccess Rate (%)

Alonso-Burgos et al12 2006 6 DIEP 100

Masia et al13 2006 66 DIEP 100

Rozen et al14,15,a 2008 75 DIEP/SIEA 100

Rozen et al14,15 2008 88 DIEP 100

Schaverien et al27 2008 12 DIEP —

Rozen et al14,15 2008 10 DIEP —

Phillips et al21 2008 65 DIEP/SIEA —

Clavero et al28 2009 126 DIEP 100

Masia et al29 2009 162 DIEP 99

Rozen et al30 2009 26 DIEP —

Smit et al16 2009 70 DIEP 100

Pacifico et al31 2009 60 DIEP —

Rozen et alb 2009 6 DIEP 100

Whitaker et al32 2009 325 DIEP —

Rozen et al33 2009 10 DIEP 100

Kim et al34 2010 58 ALT —

Chen et al35 2010 32 ALT 83

Gattaura et al36 2010 100 DIEP —

Masia et al26 2010 357 DIEP 100

Zhang et al37 2010 4 ALT —

Visscher et al38 2010 10 DIEP —

Katz et al39,c 2010 86 DIEP —

Ting et al40,a 2010 1 DCIA —

Rad et al41,c 2010 12 SGAP/LSGAP —

Gacto-Sanchez et al42,b 2010 70 DIEP 10

Cina et al17,b 2010 45 DIEP 91

Ribuffo et al43 2010 41 OFFF 100

a Case reports.b Connotes a comparison of Doppler ultrasound and CTA.c Studies using CTA but not evaluating CTA outcome.

Abbreviations: ALT, anterolateral thigh flap; OFFF, osteocutaneous fibula free-flap; SGAP/LSGAP, superior gluteal arteryperforator/lateral supragenicular artery.

CTA and MRA: Technical Considerations 221

tissue and lack of ionizing radiation. This ability isin large part because of faster gradients andimproved sequences. A recent study confirmedthe potential of MRA for preoperative planning.18

Continued work in this area especially with newertechnological advancements, including the intro-duction of blood-pool contrast agents, should ulti-mately give MRA a distinct advantage in this field.

CTA TECHNIQUE

Proper acquisition of the raw data is critical forcorrect vascular assessment of the donor area.Although CTA acquisition involves the manipula-tion of a small set of variables, familiarity with

them insures correct image acquisition and there-fore reduces the risk for repeated contrast bolusesand exposures to ionizing radiation.

PATIENT PREPARATION

Enteric contrast is avoided before any CTA acquisi-tion. Patients are asked to disrobe and only weara hospital gown without tying it. Because 3-Dshaded surface rendering will be used to depictanatomic location of the perforators in relation totheoverlyingskin, anymaterial distorting thenormalcontour of the abdomen is avoided. Furthermore,patients are positioned with arms at their sides toagain replicate the natural neutral position of the

Lee et al222

abdominal soft tissue during the future surgery. Forsimilar reasons no retention devices are used.A large-gauge (18 gauge or larger) peripheral

intravenous line is placed in the antecubital regionof the arm. The large size is required for the rapidinfusion of contrast, which is essential for adequateopacification of small vessels. Such vessels, as inthe case of abdominal perforators, which range insize between 1 to 4 mm, require an adequatecontrast bolus to be properly imaged. Extensiveliterature regarding contrast bolus is available inthe coronary CTA literature regarding the idealrate of contrast infusion for vessels of similar sizein that region of the body. As a general rule, approx-imately 1 to 2 g of iodine is injected per second toallow the best visualization of vessels of thiscaliber.2,19,20 Although iodine concentrations permilliliter of contrast vary for CTA, the highest avail-able concentration is used. This concentration isgenerally 350 to 370 mg/mL in the United Statesand 400 mg/mL in Europe. The principle can besimply stated: a lower iodine concentration willrequire a higher injection rate to obtain equivalentimage quality. Furthermore, most contrast agentshave a high viscosity and therefore generate highpressures at the suggested high flow rates. It istherefore prudent to place the largest intravenousaccess feasible and adjust the flow rate accord-ingly. In most patients flow rates between 4 to6 mL/s should be attainable keeping in mind thatpatientswith a larger body habituswill require high-er flow rates. Finally, when performing examinationwith high injection rates it is recommended to usea test bolus of at least 10 mL of normal saline totest the intravenous access before infusion ofcontrast bolus. This practice will minimize the riskof infiltration of the iodinated contrast, which inturn may cause significant morbidity ranging fromarm discomfort to a compartment syndrome.

CT EXAMINATION

Although a routine abdomen and pelvis CTA canprovide adequate coverage, this would exposepatients to unnecessary additional ionizingradiation. The area of interest is from the origin ofthe inferior epigastric artery to a level approxi-mately 4 cm above the umbilicus. In the interestof lowering exposure to ionizing radiation, it ispossible to obtain all the necessary informationfor preoperative planning by limiting the scanarea to this region when acquiring the images.The study should be performed on a CT machinecapable of rapid image acquisition at a high reso-lution. Generally, most machines on the markettoday, with the capability of 16 or greater

simultaneous slice acquisition, have the necessaryspecifications to perform a quality examination.The examination may be performed in the tradi-

tional cranio-caudad mode after an aorticthreshold of 100 Hounsfield units is reached orfurther optimized by placing the region of intereston the common femoral artery and scanning ina caudo-craniad fashion as has been describedpreviously.21 This later modification has beenadvocated for better timing of the examinationfor the arterial phase of the abdominal perforatorvessels by taking advantage of the direction thesevessels enhance.The amount of contrast will vary depending on

the speed of image acquisition. Thus, in a fastermachine less contrast can be used. Empirically,80 mL of contrast will suffice for most examina-tions because of the narrow area of interest andthe possibility to complete most examinations inless than 10 seconds. However, calculation ofthe exact dose required may be obtained by thefollowing formula:

Contrast volume (mL) 5 scan duration (delay toscan start from threshold trigger 1 scan time) Xinjection rate (typically 4–5 mL/s).

Note on Dose Modulation

Most scanners today modulate the x-ray beamenergy (defined as milliampere-second) based onthe topogram obtained at the beginning of the ex-amination or a similar algorithm. Subjective report-ing of blurring of the vessels of interest caused bythis technique has led investigators to disable thisoption during image acquisition of DIEP flap CTA.22

MRA TECHNIQUE

Despite recent excitement about the potential of3.0-T machines when compared with the standard1.5-T MR imaging for various parts of the body,23

for perforator flap MRA, characteristics other thansignal-to-noise ratio takeprecedence.For instance,it ismore important to have goodsuppressionof thesubcutaneous fat over awide field of view to includeall the abdominal subcutaneous fat and tominimizeartifacts from adjacent air containing viscera.Accordingly, thescanner that has thebestmagneticfield homogeneity, shimming, and largest field ofview is generally better; in most cases, this isa scanner with a 1.5-T field scanner.Because imaging small perforators on MRA over

a large fieldof view iscapitalizingonMRtechnology,a newer state-of-the-art scanner with top gradientperformance and a high-quality body or cardiac,8- to 32-channel phased array coil are essential.24

Fig. 1. Coronal MIP image from a CTA showing thepatency of the inferior epigastric system on a patient be-ing evaluated for a DIEP flap reconstruction. This imageplane can aid in differentiating the branching patterns.

CTA and MRA: Technical Considerations 223

PATIENT PREPARATION

Unlike CTA, fasting is not necessary before MRAbecause the gadolinium contrast rarely causesnausea/vomiting. However, high-protein foods cansimulate contrast and therefore interfere with thepostcontrast imaging essential for a quality exami-nation. Accordingly, some have advocated a mildbowel preparation or avoidance of high-proteinmeals the day beforehand or even 0.5 mg glucagonintravenously to decrease peristalsis. A phase en-coding gradient from right to left can also minimizethese artifacts. The ideal gadolinium preparation,gadofosveset trisodium, sometimes causes transitpelvic tingling. Informing patients about this beforethe examination helps to avoid surprises that cancause motion during the arterial phase of contrastinjection.

The umbilicus is used as a reference for locatingthe abdominal perforators. Its marking witha vitamin E capsule may be undertaken to aid inbetter localization before patient positioning inthe MR gantry. Additional important landmarksmay also be marked, such as the gluteal crease,pubic symphysis, and sternal notch.25

PATIENT POSITIONING IN SCANNER

Normally MRA scanning is performed in the supineposition. However, positioning patients proneminimizes motion artifact in the anterior subcuta-neous fat and abdominis rectus musclesespecially if patients are not able to suspendbreathing for the long MRA scan time with bloodpool contrast agents, such as gadofosveset triso-dium. In fact, eliminating the need for breathholding allows acquisition of high-resolution512� 512 matrix 3-D images over a 3- to 4-minutescan duration during the blood pool phase.

The umbilicus is a critical reference landmark forlocalizing each perforator in the anterior abdominalwall, it is therefore important to ensure its midlinepositioning before scanning patients. If need be,breast coils may be used in addition to the abdom-inal coils to image the breast tissue for possiblevolume analysis and graft size planning. Aftercompleting prone imaging, patients can beswitched to supine position for a final 3-D acquisi-tion of medial thigh perforators, if necessary.

PULSE SEQUENCES

As with all MRA, single-shot T2 sequences may beobtained initially to aid in better localization of thearea of interest as well as further characterizationof any abnormalities. The primary sequence forvisualizing the perforators is a 3-D spoiled gradientecho sequence with separation of fat and water

signals. Ideally this is amulti-echo sequence recon-structed with 2-point or 3-point Dixon method-ology. If the latter is not available, traditional fatsuppression or subtraction imaging can be used.But no matter what technique is used ultimately,a noncontrast series must be obtained to ensurethat the fat suppression is working properly in theareas of interest. Any drift of the fat suppressiononto the water peak in the area of interest shouldbe corrected because this will cause obscurationof the vessels of interest. Techniques used to mini-mize poor fat saturation include a reduced field ofview, shimming, or adjusting the center frequency.

Typically, slices are 3 mm thick with 2-fold zeropadding for a 1.5-mm spacing and 1.5-mmoverlapof reconstructed images. Timing of the gadofosve-set trisodium bolus to the arterial phase can bedone with fluoroscopic triggering. Gadofosvesettrisodium can be hand injected at about 1 mL/s.The arteries and veins generally run in pairs andare better visualized on a higher resolution, 512 �512 matrix equilibrium blood pool phase axialimage that typically is acquired over 3 to 4 minuteswithout using the parallel acquisition technique.The latter is often used to shorten acquisition timesat the expense of signal intensity. Near the end ofthe examination, delayed imaging in sagittal andcoronal planes, at a lower resolution,with activationof 2-fold parallel imaging and breath holding is ob-tained to provide an overview of the anatomy.26

IMAGE RECONSTRUCTION

As with any CTA/MRA evaluation, most of the rele-vant information can be ascertained from the

Lee et al224

source images where a general impression of thevascular anatomy of the donor site is obtained.Subsequently, axial and coronal MIP images(Fig. 1) are compiled to confirm the anatomy andto generate images for the operating surgeon.This later set of images is often supplemented byvolume-rendered 3-D images demonstrating theanatomic relationship of the vessels of interestwith body landmarks, such as skin and the umbi-licus (Fig. 2). Many workstations also have thecapability of overlapping a grid on the obtainedimages to allow the surgeon to accurately calcu-late the course of the donor blood vessel relativeto landmarks, such as the umbilicus.

Fig. 2. Composite CTA image represents 3-D shaded surfacrators in relation to the umbilicus. Similarly, the coronal imwithout the adjacent soft-tissue structures. Composite imagrator of interest.

Various color schemes can be used to denotethe vessel of interest, muscle, and facial planesto facilitate surgical planning. The points thatmust be addressed are the course and branchingof the arteries, locations of perforators, andobvious venous anomalies. Of the 3, the courseof the artery is most important and special notemust be made of any long intramuscular course.The diameter of the inferior epigastric artery andperforators must also be noted, specifically if it isfelt that the vessels are of small caliber (<1 mm).When reviewing MIP images, a slab that is 8 to

20 mm in thickness is used in multiple obliqueplanes. Most workstations allow placement of

e image illustrating the location of the patient’s perfo-age shows the course and size of the vessels of interestes can allow more accurate localization of each perfo-

Fig. 3. A right abdominal perforator R1 (vessel diameter 2.2 mm) coming through the right rectus muscle on(A) axial and (B) sagittal maximum-intensity projection images from an MRA examination.

CTA and MRA: Technical Considerations 225

arrows or other such annotations in a selectedplane and transfer them to other orthogonal planestomark the course of a selected vessel (see Fig. 2).

In MRA evaluation, high-resolution 3-D bloodpool phase images give the best delineation ofperforating vessels. For each perforator, intramus-cular course can be depicted on a thin slab MIP,which combines several slices to show the perfo-rator over a longer path. The diameter anddistance between the marker and perforatorsalso need to be recorded (Figs. 3–6).

COMPARISON OF TECHNIQUES

When comparing duplex ultrasound, CT angiog-raphy, and MR angiography there are benefitsand drawbacks of each technique. Thesestrengths and weaknesses are related to thespatial and contrast resolution of each respectivemodality. Ultrasound is ideally suited to assess

Fig. 4. A right gluteal perforator R1 (vessel diameter 1.8MRA examination. On Figure 4A, the distance from thewas measured as 176.2 mm.

superficial structures in great detail because ofits high spatial resolution of superficial structuresand the ability to map vessels by using duplexexamination. The limitations of ultrasound arecaused by its low contrast resolution, which mani-fests as operator dependence and interobservervariability of the technique. Furthermore, patientattributes can affect the examination. Forexample, body habitus of an obese patient canlimit evaluation as the spatial resolution of ultra-sound is significantly diminished by object depth.This in turn may result in under description of thenumber of vessels present in the area of interest.A further limitation is the inability of the techniqueto generate information relating to the 3-D courseof the vessels is an additional limitation. Ultra-sound does have the advantage of the lowest totalcost of the group.

CTA acquisitions are rapid and thorough withlittle dependence on the operator. It has the highest

mm) on (A) axial and (B) coronal MIP images from angluteal crease to R1 perforating the gluteus medius

Fig. 5. A left gracilis muscle perforator L1 on MRAwasshown on Fig. 5. The fat thickness from the innersurface of adductor longus muscle, L1 perforationand adductor magnus muscle to the inner skin borderwas measured 34.7, 19.8 and 42.4 mm separately.

Lee et al226

spatial resolution, although it ranks behindMRA forcontrast resolution. Given the inherent densitydifferences between vessels, muscle and fat alongwith the small caliber of the vessels of interest, CTAcan easily visualize the vessels of interest. Themain drawbacks of CTA are ionizing radiation andthe need for intravenous contrast, limiting its usein patients with marginal renal function. Currentdose reduction algorithms in use significantlyreduce but do not eliminate this weakness.

Fig. 6. A right latissimus dosi muscle perforator R1(vessel diameter 1.0 mm) from an MRA examination.

MRA has the highest contrast resolution but thelowest spatial resolution and highest overall cost.The availability of high-quality MR machines withadequate technical specifications to allow forhigh-resolution MRA imaging is not uniform. Delin-eating which machines are adequate to perform anexamination is an arduous task requiring familiaritywith details about the individual machines. Almostall CT scanners possessing 16 or greater detectorrows are adequate for noncardiac CTA, the samecannot be said of all 1.5-T scanners prevalent inthe clinical practice. Factors vital to obtaininghigh-resolution MRA images, such as the gradienttype, are not readily available. Furthermore, MRexaminations are more arduous for patientsbecause of their scan durations.MRA does have the advantage of using

magnetic fields rather than ionizing radiation,which has little known deleterious effects andholds great promise for vascular imaging givenits unparalleled contrast resolution. Research into image quality, vessel sharpness, and numberof perforators visualized has shown great promisefor gadofosveset trisodium compared with tradi-tional extracellular contrast agents, such asgadobenate dimeglumine. Future MRA imagingwill focus on improving image quality; using tech-niques, such as homogeneous fat suppression;diminishing bright bowel signal on the anteriorabdominal wall; and using blood pool contrastagents to acquire a sharper edge of the perforatorflap. As with CTA, the cost of the examinationsmay be justified by the time saved during theactual harvesting procedure by having a clearvascular map before the procedure.Studies comparing vessel diameter measured

on CTA, MRA, and ultrasound have repeatedlyshown that size estimations based on cross-sectional imaging (CTA/MRA) is less accuratethan on ultrasound.17 This finding is not surprisingbecause the spatial resolution of an ultrasoundexamination is much higher when dealing withsuperficial structures, such as the inferior epigas-tric arteries and their perforators. Nevertheless,in clinical practice it is often adequate to simplydifferentiate vessels, which are of less thanadequate diameter for the pending procedure.

SUMMARY

Perforator-based free flaps, such as the DIEP flap,have become the standard approach for recon-structive surgery. As the volume of reconstructionincreases, the need for a quick, reliable, and repro-ducible method for preoperative evaluation isnecessary. Adequate evaluation of the vascularsupply of free flaps is critical to the success of

CTA and MRA: Technical Considerations 227

the operation. MDCT technology currentlyprovides the ability to perform accurate imagingof vascular anatomy, improving preoperative infor-mation. The result is diminished operative time andimproved outcomes. The limitations of CTAregarding intravenous contrast and spatial resolu-tion compared with Doppler ultrasound suggestthat the technologies need to be applied basedon patient and surgeon needs. MRA technologymay eliminate some of the deleterious effects ofCTA while maintaining superior imaging qualityneeded for reconstruction planning. Radiationexposure is controversial, implying at best thatthe overall benefit of the study needs to beweighted against the individual need of each case.

REFERENCES

1. Rubin GD, Walker PJ, Dake MD, et al. Three-

dimensional spiral computed tomographic angiog-

raphy: an alternative imaging modality for the

abdominal aorta and its branches. J Vasc Surg

1993;18:656–64 [discussion: 665].

2. Kim DJ, Kim TH, Kim SJ, et al. Saline flush effect for

enhancement of aorta and coronary arteries at multi-

detector CT coronary angiography. Radiology 2008;

246:110–5.

3. Foley WD, Stonely T. CT angiography of the

lower extremities. Radiol Clin North Am 2010;

48:367–96, ix.

4. Vrtiska TJ, Fletcher JG, McCollough CH. State-of-

the-art imaging with 64-channel multidetector CT

angiography. Perspect Vasc Surg Endovasc Ther

2005;17:3–8.

5. Winston CB, Lee NA, Jarnagin WR, et al. CTangiog-

raphy for delineation of celiac and superior mesen-

teric artery variants in patients undergoing

hepatobiliary and pancreatic surgery. AJR Am J

Roentgenol 2007;189:W13–9.

6. Nahabedian MY, Tsangaris T, Momen B. Breast

reconstruction with the DIEP flap or the muscle-

sparing (MS-2) free TRAM flap: is there a difference?

Plast Reconstr Surg 2005;115:436–44 [discussion:

445–6].

7. Man LX, Selber JC, Serletti JM. Abdominal wall

following free TRAM or DIEP flap reconstruction:

a meta-analysis and critical review. Plast Reconstr

Surg 2009;124:752–64.

8. Patel SA, Keller A. A theoretical model describing

arterial flow in the DIEP flap related to number and

size of perforator vessels. J Plast Reconstr Aesthet

Surg 2008;61:1316–20 [discussion: 1320].

9. Rozen WM, Ashton MW. Improving outcomes in

autologous breast reconstruction. Aesthetic Plast

Surg 2009;33:327–35.

10. Giunta RE, Geisweid A, Feller AM. The value of

preoperative Doppler sonography for planning free

perforator flaps. Plast Reconstr Surg 2000;105:

2381–6.

11. Lell M, Tomandl BF, Anders K, et al. Computed

tomography angiography versus digital subtraction

angiography in vascular mapping for planning of

microsurgical reconstruction of the mandible. Eur

Radiol 2005;15:1514–20.

12. Alonso-Burgos A, Garcıa-Tutor E, Bastarrika G, et al.

Preoperative planning of deep inferior epigastric

artery perforator flap reconstruction with multislice-

CT angiography: imaging findings and initial experi-

ence. J Plast Reconstr Aesthet Surg 2006;59:585–93.

13. Masia J, Clavero JA, Larranaga JR, et al. Multidetec-

tor-row computed tomography in the planning of

abdominal perforator flaps. J Plast Reconstr Aesthet

Surg 2006;59:594–9.

14. Rozen WM, Ashton MW, Grinsell D, et al. Establish-

ing the case for CT angiography in the preoperative

imaging of abdominal wall perforators. Microsurgery

2008;28:306–13.

15. Rozen WM, Anavekar NS, Ashton MW, et al. Does

the preoperative imaging of perforators with CT

angiography improve operative outcomes in breast

reconstruction? Microsurgery 2008;28:516–23.

16. Smit JM, Dimopoulou A, Liss AG, et al. Preoperative

CT angiography reduces surgery time in perforator

flap reconstruction. J Plast Reconstr Aesthet Surg

2009;62:1112–7.

17. Cina A, Salgarello M, Barone-Adesi L, et al. Planning

breast reconstruction with deep inferior epigastric

artery perforating vessels: multidetector CT angiog-

raphy versus color Doppler US. Radiology 2010;

255:979–87.

18. Neil-Dwyer JG, Ludman CN, Schaverien M, et al.

Magnetic resonance angiography in preoperative

planning of deep inferior epigastric artery perforator

flaps. J Plast Reconstr Aesthet Surg 2009;62:1661–5.

19. Cademartiri F, Mollet NR, van der Lugt A, et al. Intra-

venous contrast material administration at helical

16-detector row CT coronary angiography: effect

of iodine concentration on vascular attenuation.

Radiology 2005;236:661–5.

20. Becker CR, Hong C, Knez A, et al. Optimal contrast

application for cardiac 4-detector-row computed

tomography. Invest Radiol 2003;38:690–4.

21. Phillips TJ, Stella DL, Rozen WM, et al. Abdominal

wall CT angiography: a detailed account of a newly

established preoperative imaging technique. Radi-

ology 2008;249:32–44.

22. Cademartiri F, Luccichenti G, van Der Lugt A, et al.

Sixteen-row multislice computed tomography: basic

concepts, protocols, and enhanced clinical applica-

tions. Semin Ultrasound CT MR 2004;25:2–16.

23. Alonso-Burgos A, Garcıa-Tutor E, Bastarrika G, et al.

Preoperative planning of DIEP and SGAP flaps:

preliminary experiencewithmagnetic resonance angi-

ography using 3-tesla equipment and blood-pool

Lee et al228

contrast medium. J Plast Reconstr Aesthet Surg 2010;

63:298–304.

24. Greenspun D, Vasile J, Levine JL, et al. Anatomic

imaging of abdominal perforator flaps without

ionizing radiation: seeing is believing with magnetic

resonance imaging angiography. J Reconstr Micro-

surg 2010;26:37–44.

25. Vasile JV, Newman T, Rusch DG, et al. Anatomic

imaging of gluteal perforator flaps without ionizing

radiation: seeing isbelievingwithmagnetic resonance

angiography. J Reconstr Microsurg 2010;26:45–57.

26. Masia J, Kosutic D, Cervelli D, et al. In search of the

ideal method in perforator mapping: noncontrast

magnetic resonance imaging. J Reconstr Microsurg

2010;26:29–35.

27. Schaverien M, Saint-Cyr M, Arbique G, et al. Arterial

and venous anatomies of the deep inferior epigastric

perforator and superficial inferior epigastric artery

flaps. Plast Reconstr Surg 2008;121:1909–19.

28. Clavero JA, Masia J, Larranaga J, et al. MDCT in the

preoperative planning of abdominal perforator

surgery for postmastectomy breast reconstruction.

AJR Am J Roentgenol 2008;191:670–6.

29. Masia J, Larranaga J, Clavero JA, et al. The value of

the multidetector row computed tomography for the

preoperative planning of deep inferior epigastric

artery perforator flap: our experience in 162 cases.

Ann Plast Surg 2008;60:29–36.

30. Rozen WM, Murray AC, Ashton MW, et al. The cuta-

neous course of deep inferior epigastric perforators:

implications for flap thinning. J Plast Reconstr Aes-

thet Surg 2009;62:986–90.

31. Pacifico MD, See MS, Cavale N, et al. Preoperative

planning for DIEP breast reconstruction: early expe-

rience of the use of computerised tomography angi-

ography with VoNavix 3D software for perforator

navigation. J Plast Reconstr Aesthet Surg 2009;62:

1464–9.

32. Whitaker IS, Rozen WM, Smit JM, et al. Peritoneo-

cutaneous perforators in deep inferior epigastric

perforator flaps: a cadaveric dissection and

computed tomographic angiography study. Micro-

surgery 2009;29:124–7.

33. Rozen WM, Stella DL, Bowden J, et al. Advances in

the pre-operative planning of deep inferior epigastric

artery perforator flaps: magnetic resonance angiog-

raphy. Microsurgery 2009;29:119–23.

34. Kim EK, Kang BS, Hong JP. The distribution of the

perforators in the anterolateral thigh and the utility

of multidetector row computed tomography angiog-

raphy in preoperative planning. Ann Plast Surg

2010;65:155–60.

35. Chen SY, Lin WC, Deng SC, et al. Assessment of

the perforators of anterolateral thigh flaps using

64-section multidetector computed tomographic

angiography in head and neck cancer reconstruc-

tion. Eur J Surg Oncol 2010.

36. Ghattaura A, Henton J, Jallali N, et al. One hundred

cases of abdominal-based free flaps in breast recon-

struction. The impact of preoperative computed

tomographic angiography. J Plast Reconstr Aesthet

Surg 2010;63:1597–601.

37. Zhang Q, Qiao Q, Yang X, et al. Clinical application

of the anterolateral thigh flap for soft tissue recon-

struction. J Reconstr Microsurg 2010;26:87–94.

38. Visscher K, Boyd K, Ross DC, et al. Refining perfo-

rator selection for DIEP breast reconstruction using

transit time flow volume measurements. J Reconstr

Microsurg 2010;26:285–90.

39. Katz RD, Manahan MA, Rad AN, et al. Classification

schema for anatomic variations of the inferior

epigastric vasculature evaluated by abdominal CT

angiograms for breast reconstruction. Microsurgery

2010;30(8):593–602.

40. Ting JW, Rozen WM, Leong J, et al. Free deep

circumflex iliac artery vascularised bone flap for

reconstruction of the distal radius: planning with

CT angiography. Microsurgery 2010;30:163–7.

41. Rad AN, Flores JI, Prucz RB, et al. Clinical experi-

ence with the lateral septocutaneous superior

gluteal artery perforator flap for autologous breast

reconstruction. Microsurgery 2010;30:339–47.

42. Gacto-Sanchez P, Sicilia-Castro D, Gomez-Cıa T,

et al. Use of a three-dimensional virtual reality model

for preoperative imaging in DIEP flap breast recon-

struction. J Surg Res 2010;162:140–7.

43. Ribuffo D, Atzeni M, Saba L, et al. Clinical study of

peroneal artery perforators with computed tomo-

graphic angiography: implications for fibular flap

harvest. Surg Radiol Anat 2010;32:329–34.