Incorporating Cone-beam CT into the Treatment Planning for...

Incorporating Cone-beam CT into the TreatmentPlanning for Yttrium-90 RadioembolizationJohn D. Louie, MD, Nishita Kothary, MD, William T. Kuo, MD, Gloria L. Hwang, MD,

Lawrence V. Hofmann, MD, Michael L. Goris, MD, Andrei H. Iagaru, MD, and Daniel Y. Sze, MD, PhD

PURPOSE: To prepare for yttrium-90 (90Y) microsphere radioembolization therapy, digital subtraction angiography(DSA) and technetium- 99m–labeled macroaggregated albumin (99mTc MAA) scintigraphy are used for treatmentplanning and detection of potential nontarget embolization. The present study was performed to determine ifcone-beam computed tomography (CBCT) affects treatment planning as an adjunct to these conventional imagingmodalities.

MATERIALS AND METHODS: From March 2007 to August 2008, 42 consecutive patients (21 men, 21 women; meanage, 59 years; range, 21–75 y) who underwent radioembolization were evaluated by CBCT in addition to DSA and99mTc MAA scintigraphy during treatment planning, and their records were retrospectively reviewed. The contrast-enhanced territories shown by CBCT with selective intraarterial contrast agent administration were used to predictintrahepatic and possible extrahepatic distribution of microspheres.

RESULTS: In 22 of 42 cases (52%), extrahepatic enhancement or incomplete tumor perfusion seen on CBCT affectedthe treatment plan. In 14 patients (33%), the findings were evident exclusively on CBCT and not detected by DSA.When comparing CBCT versus 99mTc MAA scintigraphy, CBCT showed eight cases of extrahepatic enhancement (19%)that were not evident on 99mTc MAA imaging. CBCT findings directed the additional embolization of vessels orrepositioning of the catheter for better contrast agent and microsphere distribution. One case of gastric ulcer fromnontarget embolization caused by reader error was observed.

CONCLUSIONS: CBCT can provide additional information about tumor and tissue perfusion not currently detectableby DSA or 99mTc MAA imaging, which should optimize 90Y microsphere delivery and reduce nontarget embolization.

J Vasc Interv Radiol 2009; 20:606–613

99m
Abbreviations: CBCT � cone-beam CT, DSA � digital subtraction angiography, GDA � gastroduodenal artery, Tc MAA � technetium-99m-macroaggre-gated albumin, RGA � right gastric artery
CONE-beam computed tomography(CBCT), which produces CT-like im-ages in an angiography suite, is un-

From the Division of Vascular and InterventionalRadiology, Department of Radiology, Stanford Uni-versity Medical Center, 300 Pasteur Drive, H-3630,Stanford, CA 94305-5642. Received September 8,2008; final revision received January 7, 2009; ac-cepted January 9, 2009. Address correspondence toJ.D.L.; E-mail: [email protected]

N.K. and L.V.H. have received research grants fromSiemens (Erlangen, Germany); D.Y.S. serves as apaid consultant to MediGene (San Diego, Califor-nia), Pain Therapeutics (San Francisco, California),and Jennerex Biotherapeutics (San Francisco, Cali-fornia). None of the other authors have identified aconflict of interest.

© SIR, 2009

DOI: 10.1016/j.jvir.2009.01.021

606

dergoing rapidly increasing use inthe field of interventional radiology.Early case reports indicated the use-fulness of CBCT in difficult neuro-vascular interventions, transjugularintrahepatic portosystemic shuntcreation, endoleak repair, splenicembolization, and hepatic malig-nancy treatment (1–3). An interestingoffshoot of modern flat-panel fluo-roscopy technology, CBCT has beenshown to provide anatomic informa-tion in addition to that detected bydigital subtraction angiography(DSA) during hepatic arterial inter-ventions, and this information canimpact patient management in 19%of hepatic malignancy treatments (4).
For transarterial chemoembolization,
CBCT has improved catheter place-ment planning and diagnostic confi-dence in the selected catheter posi-tion for more selective chemotherapyinfusion (5,6). These benefits of ap-plying CBCT technology to chemo-embolization may have even greaterimpact on patients undergoing ra-dioembolization.

Before radioembolization, meticu-lous planning with DSA and techne-tium- 99m–labeled macroaggregatedalbumin (99mTc MAA) scintigraphyneeds to be performed to avoid non-target embolization and to ensureadequate delivery of microspheres toall the tumors. This attention to thevascular anatomy is important, be-
cause the microspheres themselves
mailto:[email protected]

Louie et al • 607Volume 20 Number 5

are radiolucent and their eventualdistribution is only inferred from thepreprocedural imaging. Hepaticoen-teric anastomoses are preemptivelycoil-embolized to prevent nontargetembolization, which could otherwiseproduce severe gastrointestinal com-plications (7). In a recent review ofthe literature, the median rate of ul-ceration was 8%, with 6% requiringsurgical intervention (8). The spatialresolution and three-dimensionaldataset produced by CBCT couldtheoretically supplement the tradi-tional tools of DSA and 99mTc MAAscintigraphy to prevent extrahepatic90Y microsphere deposition and topredict which tissues will be af-fected. This study examines whetherCBCT changes the treatment plan-ning for radioembolization, andwhether its routine use is warranted.

MATERIALS AND METHODS

Between March 2007 and August2008, 42 consecutive patients (21 men,21 women; mean age, 59 years; agerange, 21–75 y) who underwent prepro-cedural CBCT evaluation for radioem-bolization of hepatic-dominant malig-nancies were included in thisretrospective study. The underlyingprimary tumors, in descending orderof frequency, included colon, neuroen-docrine, melanoma, cholangiocarci-noma, lung, hepatocellular, ovarian,pancreatic, granulosa cell, sarcoma,bladder, ampullary, hemangiopericy-toma, endometrial, breast, leiomyosar-coma, and adenocarcinoma of un-known primary lesion. Selectioncriteria were based on multidisci-plinary consensus panel report fromthe radioembolization brachytherapyoncology consortium (9). Data werehandled in accordance with Health In-surance Portability and AccountabilityAct regulations, and the study wasgranted an exemption from obtainingpatient consent by our institutional re-view board.

The radioembolization process wasdivided into two outpatient sessions,performed 1–2 weeks apart. During thefirst session, the patients underwent he-patic DSA imaging and coil emboliza-tions to close natural hepaticoentericanastomoses and accessory arteries,such as the gastroduodenal artery(GDA), right gastric artery (RGA), su-
praduodenal artery, falciform artery,
and accessory phrenic artery. Duringthe initial vascular mapping, DSA im-ages were obtained with a 5-F catheterplaced at the common or proper hepaticartery, injecting contrast medium (Visi-paque 320; GE Healthcare, Princeton,New Jersey) at 3–5 mL/sec for 2–4 sec-onds. Superselective DSA images wereobtained with microcatheters and handinjections in vessels that potentiallyneeded embolization. CBCT was usedas an adjunct to DSA as a problem-solving tool when tissue perfusion orvascular anatomy was unclear. In addi-tion, at the conclusion of each angio-graphic session, CBCT was performedto delineate the vascular territory servedby the planned catheter placement, andtherefore to predict distribution of themicrospheres. The microcatheter wasplaced in the anticipated microsphereimplantation vessel, which was thecommon or proper hepatic artery forwhole-liver treatment. More selectivelobar treatment required the microcath-eter to be placed in the right or left he-patic artery. Anatomic variants, such asa replaced right hepatic artery, wouldalso be evaluated with CBCT if it wasthe planned delivery vessel or if therewas uncertainty of tumor perfusion. Ifthere was a particular artery of uncer-tain tissue perfusion, CBCT imageswere obtained with superselective injec-tion in the questionable artery. To en-sure no contrast agent reflux, whichwould yield erroneous findings, theCBCT injection rates were approxi-mately half the DSA injection rates. Inaddition, all the raw images from theCBCT acquisition were inspected for re-flux. After CBCT, the patients under-went 99mTc MAA arterial administra-tion and planar and single-photonemission CT scintigraphy to measurethe proportion shunted into the lungsand to evaluate for any extrahepaticdeposition.

During the second session of the ra-dioembolization process, the patientsunderwent confirmatory DSA andCBCT studies to evaluate for any in-terim changes in perfusion before ad-ministration of the 90Y resin micro-spheres (SIR-Spheres; Sirtex Medical,Sydney, Australia). If any extrahepaticperfusion was detected on the 99mTcMAA scintigraphy or on the currentDSA or CBCT, additional coil emboliza-tion or catheter reposition was per-formed until CBCT showed no evidence
of persistent extrahepatic enhancement.
Yttrium-90 resin microspheres werethen administered as described in detailelsewhere, including a multidisciplinaryconsensus report (7,9). The prescribedactivity calculation was based on theconsensus recommendation of the bodysurface area method, which has a favor-able toxicity profile and similar re-sponse and survival outcomes as theempiric method. In the dose calculationsfor 90Y resin microspheres, exact volumemeasurements are not required, unlikefor glass microspheres. However, thepercentage of perfused liver in relationto the whole liver as seen on CBCT wasused in the calculation of the deliverydose, especially in cases of partial liverradioembolization. No dose adjustmentwas made for catheter repositions, al-though there was a risk of reflux, espe-cially for resin microspheres. If near-stasis of forward blood flow wasobserved during implantation, the en-tire dose was not delivered to preventreflux. After delivery of the micro-spheres, patients were discharged thesame day with a 10-day oral methyl-prednisolone taper regimen, 30-daycourse of proton pump inhibitor, andanalgesic and antiemetic agents asneeded. Patients were seen in clinic 1month after radioembolization and fol-lowed for at least 3 months to monitorfor possible complications.

DSA and CBCT images were ob-tained with use of a ceiling mountedangiographic C-arm system with 30-cm � 40-cm flat-panel detector (AxiomArtis dTA; Siemens, Erlangen, Ger-many). Patients were positioned witharms abducted, and extraneous metallicobjects such as oxygen saturation andcardiac monitoring leads and cableswere kept out of the field. The CBCTimages (ie, DynaCT) were obtained byrotating the C-arm 200° around the pa-tient in 8 seconds. The detector acquiredimages every 0.5° for a total of 419 im-ages. Contrast medium (Visipaque 320;GE Healthcare) was diluted 1:1 withnormal saline solution and injected atapproximately half the rate of a DSArun, from 0.1 to 4 mL/sec depending onthe size of the artery, rate of flow, andtype of microcatheter. For all CBCTstudies, a 4-second imaging delay wasprescribed to allow injected contrast me-dium to enhance parenchymal tissue be-fore data acquisition, and contrast me-dium was injected throughout the dataacquisition period to keep the arterial
tree enhanced. The 4-second imaging

608 • Cone-beam CT in Treatment Planning for 90Y Radioembolization May 2009 JVIR

delay, combined with the 8-second ac-quisition period, made the total injectiontime 12 seconds. The volumetric data setwas transferred to a separate postpro-cessing workstation (Syngo-X; Siemens)for three-dimensional reconstruction.Image correction algorithms were ap-plied for scatter, beam hardening, ringartifact, and truncation. The CT-typedataset was viewed as three-dimen-sional and multiplanar reconstructedimages.

The DSA, 99mTc MAA scintigraphy,and CBCT studies were retrospectivelyevaluated to determine whether CBCTchanged the treatment plan for radio-embolization. Specifically, additionalembolization of hepaticoenteric anasto-moses, embolization of accessory he-patic arteries, embolization of parasit-ized nonhepatic arteries, and changes inplanned positioning of the microcath-eter at the time of 90Y resin microsphereadministration were tallied.

RESULTS

A total of 132 CBCT acquisitionswere obtained in 42 patients, for an av-erage of 3.1 studies per patient (range,1–7 studies), during 91 outpatient ses-sions. In this group, 30 patients (71%)received single-session whole-livertreatment, 10 (24%) received partial livertreatment, and two (5%) did not receivetreatment as a result of worsening liverfunction. In 22 of 42 patients (52%), find-ings of extrahepatic enhancement (n �18; 43%), incomplete perfusion of thetumor (n � 8; 19%), or both (n � 4; 10%)affected the treatment plan. Extrahe-patic enhancement most often involvedstomach, duodenum, or pancreas(Figs 1,2). Incomplete perfusion of thetumors was most often a result of pre-viously unrecognized accessory he-patic arteries or parasitized extrahe-patic arteries (Fig 3). Embolization ofthese unrecognized vessels redirects theblood flow away from the accessory andextrahepatic vessels to the artery inwhich the microspheres are injected, asevidenced by altered contrast enhance-ment, leading to presumed depositionof microspheres into areas of tumor. Ofthe 22 cases in which CBCT affected thetreatment plan, 14 patients (33%) hadunexpected CBCT findings that actuallyresulted in real change of the treatmentplan. Additional embolization was per-formed in eight patients, catheter posi-
tion was moved in four patients, and
both were performed in two patients(Table 1). These unexpected findings ofpotential nontarget embolization or in-complete inclusion of the treatment areawere not demonstrated by DSA or 99mTcMAA imaging, even in retrospect. In theother eight cases (19%), the CBCT find-ings were also detected on DSA, butidentification of the end organ perfusedwas equivocal (Table 2). CBCT allowedthe operator to confirm hepatic or non-hepatic tissue enhancement, and there-fore embolize vessels or reposition cath-eters with greater confidence (Fig 4).

Figure 1. Images from a 59-year-old manously treated with hepatic resection and muradioembolization. (a) Proper hepatic angilized GDA. Embolization coils were placebranch of the left gastric artery, and the athrough collateral circulation from segmentwere identified by DSA. (b) Arterial contlocation showed enhancement in the posdetected by DSA. Extensive beam-hardenincoils and catheters, partially obscuring the gcould not be identified by DSA despite angthe left gastric artery was selected and co(arrow) via the left-to-right gastric anastomshowed no residual enhancement in the gaslinear beam-hardening streak artifacts exgastric walls.

When comparing CBCT versus 99mTc

MAA single-photon emission CT stud-ies during the same treatment period,CBCT showed nine cases of extrahepaticenhancement, of which only one wasdetected by 99mTc MAA scintigraphy.Scintigraphy failed to identify the othereight as a result of limitations in spatialresolution.

During the 3-month follow-up ofthe patients, one case of nontarget ra-dioembolization–induced gastric ulcerwas seen with endoscopic biopsy con-firmation. On retrospective review ofthe CBCT images, enhancement of the

th bilobar hepatocellular carcinoma, previ-le chemoembolizations, who presented foram shows stasis of flow to the coil-embo-n the accessory left hepatic artery and aries to segments 2 and 3 were now filledNo remaining hepaticoenteric anastomosest-enhanced CBCT from the same catheterior gastric antrum (arrow) that was notrtifacts were also seen from the radiopaquetric parenchyma. (c) The origin of the RGAraphy in multiple obliquities. As a result,were deposited at the origin of the RGAis. (d) Follow-up contrast-enhanced CBCTwall (arrow) after the coil placement. Note

ding through the anterior and posterior

wiltipogrd irte4.

rasterg aasiog

ilsos

tricten

gastric wall was identified, probably


supplied by a collateral vessel feedingthe coil-embolized RGA. This en-hancement, partly obscured by beam-hardening artifact, was missed by theauthors at the time of the CBCT study.However, in retrospect, CBCT accu-rately portrayed the nontarget perfu-sion of the stomach in this single caseof a complication.

DISCUSSION

CBCT provides three-dimensionaldata and higher sensitivity to x-ray at-tenuation compared with DSA, and hassuperior spatial resolution to 99mTcMAA scintigraphy. DSA shows intralu-minal enhancement of arteries with su-perior spatial and temporal resolution,but territorial parenchymal perfusion islimited by the projectional nature of thetechnology. In contrast, CBCT is able todelineate territorial tissue perfusion inthree dimensions, with an attenuationresolution of approximately 10 HU. Theclinical impact of these capabilities isonly just being explored.

We found that these theoretical ad-vantages translate into real changes inpatient care in a high proportion of pa-tients undergoing radioembolization. In

Figure 2. Images from a 57-year-old womfor radioembolization. (a) Common hepaextrahepatic enhancement. (b) Contrast-ennot seen on DSA. This was treated by remicrospheres into the right and left hepatienhancement before radioembolization.

33% of the cases, CBCT showed the po-

tential for hazardous nontarget emboli-zation and/or incomplete tumor perfu-sion that conventional DSA was unableto portray. Delivery of the 90Y micro-spheres without recognizing these vas-cular patterns could have led to ulcersor pancreatitis, or to incomplete tumortreatment (10). The CBCT informationprompted additional embolizations andcatheter repositions that would not haveoccurred with DSA alone. CBCT helpsoptimize catheter placement to maxi-mize the number of treated tumorswhile minimizing the risk for nontargetembolization. Although routine selec-tive radioembolization would reducenontarget embolization, most patientshave multifocal disease in most seg-ments of the liver, which requires amore central injection to treat all tumors.Whole-liver treatment, which was per-formed in 71% of the patients in thepresent series, is more likely to benefitfrom CBCT because of the more proxi-mal point of delivery.

The utility of increased sensitivity totissue perfusion was first documentedwith the use of CBCT for transarterialchemoembolization. Small hepatocellu-lar carcinomas were not well visualizedwith DSA, but could be identified with

with chemotherapy-refractory lung metasarteriogram shows no flow into the GDced CBCT shows enhancement in the duoitioning the microcatheter more distally irteries. CBCT in each hepatic vessel confir

CBCT (11). CBCT’s ability to recognize

subtle tissue perfusion offers potentialincremental improvement in the field ofradioembolization. Compared with che-moembolization, radioembolization hasthe common potential of ischemic injuryto bowel, but poses the added risk ofpersistent radiation damage from afixed radiation source. As a result, thereis an inherent higher risk of morbidityfrom nontarget embolization with 90Yresin microspheres, which can be miti-gated with CBCT. In addition, contrast-enhanced CBCT with contrast mediuminjected at the site of planned micro-sphere delivery can also reveal territo-ries where radioembolization will be in-effective because unenhanced regionsare unlikely to receive microsphere dep-osition, a finding in 19% of our patients.These unenhanced regions with tumorwere often supplied by small accessoryhepatic arteries, which were coil-embo-lized before microsphere administra-tion.

The use of two-dimensional projec-tional DSA to map arterial anatomymay be confusing because of superim-positions. However, as is true of othercross-sectional modalities, CBCT allowsfor multiplanar reconstructions that en-able manipulation of the images to dif-

s to both lobes of the liver who presentednd RGA after coil embolization and noum and pancreatic head (arrow) that wasthe liver and sequentially delivering the

d absence of the duodenal and pancreatic

an tasetic A ahan denpos ntoc a me

ferentiate adjacent or overlapping or-


gans and arteries. For example,parasitized blood supply from thephrenic or intercostal arteries could beclearly shown to supply intrahepatic tu-mors on CBCT. In addition, parenchy-mal blushes of gastric, esophageal,pancreatic, or duodenal DSA imagesin patients with cachexia who haveempty gastrointestinal tracts weresometimes difficult to differentiatefrom enhancement of adjacent liver,particularly if the hepatic enhance-ment pattern was distorted by tumorinvolvement. In 19% of our patients,DSA interpretation was equivocalabout the identity of the end organsperfused, and CBCT clarified the per-fusion question and increased confi-dence in the treatment plan.

Figure 3. Images from a 44-year-old wopresented for radioembolization. (a) CommGDA, RGA, and falciform artery after coappeared to be the entire liver. (b) Contrasdevoid of enhancement (arrow). (c) A smalmesenteric artery supplied collateral flowwhich arose from the previously coil-embnoticed on the original flush aortogram,coil-embolized. (d) Contrast-enhanced CBcommon hepatic artery after coil embolizaticontrast agent to segment 6 (arrow).

Differentiation between adjacent or-

gans is an even greater challenge for99mTc MAA scintigraphy, even withthree-dimensional single-photon emis-sion CT. Although scintigraphy is excel-lent for lung shunt evaluation, identifi-cation of extrahepatic enhancement isextremely limited by poor spatial reso-lution. In this series, there were ninecases of extrahepatic enhancement, allof which were detected by CBCT. Incontrast, the 99mTc MAA scintigraphystudies missed eight of the nine casesthat were detected by CBCT.

Despite the increased sensitivity toextrahepatic perfusion afforded byCBCT, one patient (2.4%) did have agastric ulcer from nontarget radioembo-lization, which was confirmed by endo-scopic biopsy. The patient recovered af-

n with bilobar colorectal metastases whohepatic arteriogram shows no flow into theembolization, and enhancement of whathanced CBCT shows region in segment 6

ancreaticoduodenal branch of the superioran accessory right hepatic artery (arrow),zed GDA. This collateral branch was nott with the aid of CBCT the artery waswith injection of contrast medium in theof the collateral vessel showed rerouting of

ter conservative management with a

high-dose pantoprazole and sucralfate.Retrospective review of the DSA andCBCT images showed subtle perfusionof the gastric antrum via a newly ap-peared collateral vessel from the left he-patic artery, feeding the already coil-embolized RGA. This accessory RGAwas not recognized prospectively, rep-resenting a human error rather than adeficiency of the technology. Becauseour sample size is small, a larger num-ber of patients will need to be studied toconfirm that a significant reduction inthe complication rate can be achieved byroutine use of CBCT.

Although CBCT has the advantage ofmultiplanar reconstruction and in-creased sensitivity to attenuation differ-ences, it is subject to serious artifacts.For instance, it is more prone to beam-hardening streak artifacts than is con-ventional CT. These artifacts can becaused by the routine placement of coilsadjacent to organs of interest, by thepresence of gas within hollow viscera,and by the presence of radiopaque cath-eters and surgical clips. These artifactscan mimic enhancement of visceralwalls. One can usually distinguish be-tween artifact and true enhancement bylooking for nonlinear enhancement thatdoes not extend beyond anatomicboundaries.

Other potential disadvantages of in-corporating routine use of CBCT intothe radioembolization protocol includeadditional contrast medium adminis-tered, additional radiation exposure tothe patient, additional time expended,and discomfort in patients with limitedshoulder range of motion. However, be-cause of the higher sensitivity of CBCTto attenuation differences, the amountof contrast medium used for one CBCTacquisition is actually identical to thatused in one DSA acquisition. One CBCTacquisition also results in patient radia-tion exposure equivalent to only one ortwo DSA acquisitions. Our institution’sequipment is adjusted to acquire DSAimages at 3.60 �Gy per image andCBCT images at 0.36 �Gy per image.For an 8-second rotational scan in which419 images are acquired, the dosage isequivalent to that of 42 DSA images, ora 10.5-second DSA run at four framesper second. The radiation is distributedover a larger surface area and volumebecause of the 200° rotation, whereasDSA delivers the radiation more focally.Although the abdomen dosage has not

maonil

t-enl ptoolibu

CTon

been studied, calculations for CBCT of

Note.—SMA � superior mesenteric artery.


Table 2Equivocal End-organ Perfusion Clarified by CBCT that Led to Corresponding Adjustment of Treatment Plan*

Pt. No. Vessel Injected Extrahepatic EnhancementIncomplete Hepatic

Perfusion Additional Embolizations

2 Pancreatic artery Pancreas — Dorsal pancreatic artery8 Left hepatic artery Stomach — Accessory left gastric artery

11 Cystic artery Gallbladder and hilar lymph nodes — Cystic artery16 Distal left hepatic

arteryLeft hemidiaphragm — Accessory left phrenic artery

19 RGA Stomach — RGA25 Common hepatic

arteryStomach — RGA

31 Pancreatic artery Pancreas — Dorsal pancreatic artery38 Segmental splenic

artery— Missing perfusion of

portions ofsegments 2 and 3

Parasitized superior branches ofsplenic artery supplying leftlobe of liver

Table 1Unexpected CBCT Findings Not Seen in DSA or 99mTc MAA Scintigraphy and Corresponding Changes in TreatmentPlanning before Radioembolization

Pt. No. Vessel InjectedExtrahepaticEnhancement

Incomplete HepaticPerfusion Additional Embolizations

Reposition ofCatheter

9 Proper hepaticartery

Stomach andduodenum

— — Distally insegment 4a/4b

14 Commonhepatic artery

Pancreas andduodenum

— — Distally in properhepatic artery


Omentum, anteriorabdominal wall,stomach

Missing enhancement ofsegment 6

Falciform artery and accessoryright hepatic artery branchof SMA

—

17 Replacedproperhepatic artery

Pancreas andduodenum

— — More distally

18 Replaced righthepatic artery

— Missing enhancement ofmedial aspect oftumor when replacedright hepatic arteryinjected

Middle hepatic artery arisingfrom the celiac axis

—


— Missing enhancement ofsegments 2 and 3

Accessory left hepatic artery —


Stomach — RGA —


Duodenum — — More distally


Hilar lymph nodes andgreater omentum

Missing enhancement ofpart of segment 6

Dorsal pancreatic, right T9,T10, T12 intercostal arteries

—


Stomach Missing peripheralenhancement inmultiple segments

Right inferior phrenic artery More distally toavoid stomach


Extrahepatic drainingvein

Missing peripheralenhancement inmultiple segments

Right inferior phrenic artery Repositioned forbetterevaluation


— Missing enhancement ofportions of segments5 and 6

Accessory right hepatic arterybranch of SMA

—


Duodenum — Supraduodenal andpancreaticoduodenal arteries

—


Stomach andesophagus

— Accessory leftgastric/esophageal artery

—

* No repositioning of the catheter was necessary in any of the cases listed.


the head and neck indicate a reductionin total radiation dose of as much as90% compared with DSA. The reductionreflects primarily the lower milliamper-age and thus the dose per frame (12).The amount of time required to acquireand process CBCT images is also equiv-alent to that of approximately two tothree DSA acquisitions, but the CBCTdata acquired may actually supplant thedata conventionally obtained from aneven greater number of DSA acquisi-tions, resulting in a net decrease in timeand radiation dose.

Our study is limited by the noncon-trolled retrospective design and itssmall sample size. We assumed that em-bolization of parasitized and accessory

Figure 4. (a) Right inferior phrenic arterigiocarcinoma shows enhancement of thehepatic enhancement. (b) Contrast-enhancement of a hepatic tumor (arrow) fed by tresolve on DSA. (c) Selective distal left hepbilobar colorectal metastases shows a tortuouwere already placed in the GDA in this patstrated anterior abdominal wall (arrow) and osuperior epigastric artery branches that werevessel represented a falciform artery.

arteries would consolidate blood flow to

all tumors via the main hepatic artery.CBCT showed successful consolidationby complete parenchymal and tumorenhancement during one central con-trast agent injection. However, the mean30-�m diameter and specific gravity of1.6 mg/mL of the 90Y resin micro-spheres may make them behave differ-ently from the smaller contrast agentmolecules as they flow through intrahe-patic collateral vessels of unknown size.Therefore, distribution of microspheresmay or may not be the same as distri-bution of contrast medium. Contrast en-hancement predicted the pattern of mi-crosphere deposition, but this was notconfirmed by explant pathologic exam-ination. Further studies need to be per-

am in a 74-year-old woman with cholan-phragm, but is equivocal in identifyingBCT in the same patient shows enhance-

right phrenic artery that was difficult toarteriogram in a 44-year-old woman with

essel of uncertain destination (arrow). Coilst, and contrast-enhanced CBCT (d) demon-ntal enhancement with communication witht seen on DSA, confirming that this tortuous

formed to evaluate the assumption that

microspheres would travel through col-lateral vessels to regions previously sup-plied by the embolized accessory andparasitized arteries. These smaller ves-sels that supply tumor would otherwisebe untreated becasue of the increasedrisk of reflux.

In conclusion, we found CBCT to bea sensitive and useful problem-solvingtool when questions of perfusion wereraised in patients undergoing radioem-bolization. Treatment plans were al-tered in one third of our patients basedon findings suggesting potential nontar-get embolization or incomplete tumortreatment. Change in catheter place-ment or additional embolizations wereperformed in response to the findings,which were generally beneath thethreshold of detectability by DSA andby nuclear scintigraphy.

References1. Heran NS, Song JK, Namba K, Smith

W, Niimi Y, Berenstein A. The utilityof DynaCT in neuroendovascular pro-cedures. AJNR Am J Neuroradiol 2006;27:330–332.

2. Sze DY, Strobel N, Fahrig R, Moore T,Busque S, Frisoli JK. Transjugular in-trahepatic portosystemic shunt creationin a polycystic liver facilitated by hybridcrosssectional/angiographic imaging. JVasc Interv Radiol 2006; 17:711–715.

3. Hirota S, Nakao N, Yamamoto S, et al.Cone-beam CT with flat-detector digi-tal angiography system: early experi-ence in abdominal interventional pro-cedures. Cardiovasc Intervent Radiol2006; 29:1034–1038.

4. Wallace MJ, Murthy R, Kamat PP,et al. Impact of C-arm CT on hepaticarterial interventions for hepatic ma-lignancies. J Vasc Interv Radiol 2007;18:1500 –1507.

5. Virmani S, Ryu RK, Sato KT, et al.Effect of C-arm angiographic CT ontranscatheter arterial chemoemboliza-tion of liver tumors. J Vasc Interv Ra-diol 2007; 18:1305–1309.

6. Liapi E, Hong K, Georgiades CS, Ge-schwind JFH. Three-dimensional rota-tional angiography: introduction of anadjunctive tool for successful transarte-rial chemoembolization. J Vasc IntervRadiol 2005; 16:1241–1245.

7. Murthy R, Nunez R, Szklaruk F, et al.Yttrium-90 microsphere therapy forhepatic malignancy: devices, indica-tions, technical considerations, and po-tential complications. Radiographics2005; 25(suppl):S41–S55.

8. Murthy R, Brown DB, Salem R, et al.Gastrointestinal complications associ-

ogrdiad Cheatics v

ienmeno

ated with hepatic arterial yttrium-90


microsphere therapy. J Vasc IntervRadiol 2007; 18:553–561.

9. Kennedy A, Nag S, Salem R, et al.Recommendations for radioemboliza-tion of hepatic malignancies using yt-trium-90 microsphere brachytherapy: aconsensus panel report from the radio-
embolization brachytherapy oncology
consortium. Int J Radiat Oncol BiolPhys 2007; 68:13–23.

10. Liu D, Salem R, Bui JT, et al. Angio-graphic considerations in patients un-dergoing liver directed therapy. J VascInterv Radiol 2005; 16:911–935.

11. Kakeda S, Korogi Y, Ohnari N, et al.
Usefulness of cone-beam volume CT
with flat panel detectors in conjunctionwith catheter angiography for trans-catheter arterial embolization. J VascInterv Radiol 2007; 18:1508–1516.

12. Bridcut RR, Murphy E, Workman A,Flynn P, Winder RJ. Patient dosefrom 3D rotational neurovascular stud-
ies. Br J Radiol 2007; 80:362–366.

Incorporating Cone-beam CT into the Treatment Planning for...

Documents

Transcript of Incorporating Cone-beam CT into the Treatment Planning for...