Chemical Tumor Ablation with Use of a NovelMultiple-tine Infusion System in a CanineSarcoma ModelAndrew Hines-Peralta, MD, Zheng-jun Liu, MD, Clare Horkan, MD, Stephanie Solazzo, BA, and

S. Nahum Goldberg, MD

PURPOSE: To determine whether larger confluent zones of ablation can be achieved in chemical ablation with use ofa multiple-tine infusion device compared with standard needle infusion in a solid tumor model.

MATERIALS AND METHODS: Multiple canine venereal sarcomas (N � 42) were implanted in nine mildly immu-nosuppressed dogs (treated with 10 mg/kg cyclosporin A twice daily). Tumors incubated for 8–12 weeks grew to adiameter of 5.4 cm � 1.0. With ultrasound guidance, 8–56 mL of 100% ethanol or 15% acetic acid (diluted in saturatedsaline solution) were injected in aliquots (2–8 mL) at multiple distances (radius of 0–2 cm) from the needle axis withuse of a multiple-tine infusion device. Presence of fluid reflux at the needle puncture site and resultant coagulationdiameters were measured within 1 hour and compared with the results of infusion with a standard 18-gauge needle.

RESULTS: Multiple-tine infusion enabled greater fluid infusion (15 mL � 3 to 53 mL � 3 depending on protocol) thanstandard needle injection (8 mL � 1) before reflux was observed at the puncture site (P < .01). Additionally,progressive gains in contiguous tumor coagulation were achieved because acetic acid was infused as far as 2 cm fromthe needle axis with the multiple-tine device (P < .01; R2 � 0.59; y � 0.5x � 2.9). Optimal coagulation was achieved withthe infusion of 4-mL aliquots at 0.5 cm and 1.0 cm from the needle, followed by three 4-mL or 8-mL aliquots (40° rotationbetween infusions) at 1.5 cm and 2.0 cm from the needle (32 mL � 0 and 53 mL � 3 total, respectively). This yieldedconfluent short-axis coagulation diameters of 4.9 cm � 1.0 and 5.4 cm � 1.0, respectively, which were significantly greaterthan the measurement of 3.1 cm � 0.4 achieved with standard needle infusion (P < .01). Smaller and noncontiguous fociof coagulation foci (1.7 cm � 0.5) were seen with the use of ethanol for standard needle and multiple-tine infusions.

CONCLUSIONS: Chemical ablation with 15% acetic acid with use of a multiple-tine infusion device resulted in largerdiameters of contiguous tumor coagulation and enabled greater volumes of infusion than standard needle infusion orethanol ablation. This suggests that chemical ablation with acetic acid infused with use of a multiple-tine device mayovercome some of the difficulties seen with the use of conventional needle chemical ablation injection alone, such asirregular ablation and fluid reflux up the needle tract.

J Vasc Interv Radiol 2006; 17:351–358

Abbreviation: RF � radiofrequency

CHEMICAL ablation with percutane-ous ethanol instillation for small (�3cm) primary hepatocellular carcino-mas has been shown to be as effectiveas surgical resection (1–5). However,

many focal hepatomas are too large atdiagnosis to be effectively treated withchemical ablation alone (2,6,7). Disap-pointing results have also been re-ported for metastatic disease to the

liver (8–11). This limitation has fueledthe development of thermal tumorablation methods, including radiofre-quency (RF) (11–13), microwave (14),and cryoablation (15) devices.

The poor clinical performance ofchemical ablation for larger tumorshas been primarily attributed to un-predictable and irregular zones of co-agulation achieved with large-volumechemical instillation, which in turn iscaused by nonspherical, nonsymmet-ric diffusion of the agent via paths ofleast resistance. This leads to higher

From the Department of Radiology, Beth Israel Dea-coness Medical Center, Harvard Medical School, 330Brookline Avenue, Boston, Massachusetts 02215. Re-ceived June 16, 2005; accepted November 2. Addresscorrespondence to S.N.G.; E-mail: sgoldber@car-egroup.harvard.edu

Research supported in part by RexMedical, LP, Con-

shohocken, PA. None of the authors have identifieda conflict of interest.

© SIR, 2006

DOI: 10.1097/01.RVI.0000196355.31419.1E

351

rates of local tumor progression com-pared with RF ablation despite an in-crease in the number of treatment ses-sions (16,17). In addition, for densesolid tumors, the path of least resis-tance for chemical reflux may occuralong the needle shaft (ie, the path ofleast resistance) into the peritoneum,leading to pain and other side effects(18). Irregular and unpredictable pen-etration with ethanol has also stimu-lated research into other chemicalagents such as acetic acid, which hasshown superior tissue penetration inlaboratory studies (19,20) and clinicalstudies (21).

One solution that has been pro-posed for RF ablation to overcome in-adequate tissue coagulation is thedevelopment of a multiple-tine RF ap-plicator (22). Several such devices arenow commercially available. In thisstudy, we apply this concept to chem-ical ablation as a possible means toimprove chemical ablation diffusionby performing chemical ablation withuse of a multiple-tine instillation de-vice. In these studies, we first opti-mized a multiple-tine chemical abla-tion device for use in a solid caninesarcoma tumor model for acetic acidand ethanol. We then compared ourresults to those of standard needle in-fusion by assessing tumor coagula-tion, total infusible volume, and chem-ical reflux.

MATERIALS AND METHODS

Overall Experimental Design

Percutaneous chemical ablationwas performed in large (�5 cm) ca-nine venereal sarcoma tumors (N � 42in nine dogs) with a standard 18-gauge needle (n � 6) or a multiple-tinedevice (n � 36). Ablation was per-formed with 100% ethanol (n � 9) or15% acetic acid diluted in a saturatedsaline solution (n � 33), a mixture thathas been shown to have antitumor ef-fects in this animal model (19). Withthe multiple-tine device, chemical in-stillation was performed by injecting2–8-mL aliquots at multiple distances(0–2 cm) perpendicular to the needleaxis. Resultant coagulation diameters,quality of ablation (ie, confluence andshape), total infusible volume, andfluid reflux were compared versus theresults of standard 18-gauge needle in-fusion.

Device Design

Chemical infusion was performedwith use of a multiple-tine infusiondevice consisting of three deployabletines embedded within an 18-gaugeshaft (QuadraFuse; RexMedical, Con-shohocken, PA; Fig 1). The tines areseparated by 120° and deploy in anumbrella fashion. Infusion of a chem-ical agent occurs simultaneouslythrough the end of each tine at a radialdistance 0.5–2 cm perpendicular to theneedle shaft depending on how far thetines are deployed. The deploymentsettings described here refer to the di-ameter distance between tines (de-ployed 1–4 cm). Given the design, thetines also travel approximately 1 cmdistally along the needle axis direction(ie, z axis) as they are deployed fromthe 1-cm setting to the 4-cm setting.The ablative agent is delivered manu-ally via a syringe attached to the de-vice.

Tumor Preparation and Animal Care

Use of animals was conductedthrough the approval of the institu-tional animal care and use committee.Multiple canine venereal sarcomaswere implanted subcutaneously onthe back and torso in a canine model.Canine venereal sarcoma cell line haswell-characterized growth and hasbeen used for chemical ablation andRF ablation studies (20,23). Femalemongrel dogs (N � 9; Harlan Farms,

North Rose, NY) were mildly immu-nosuppressed (10–20 mg/kg cyclo-sporin A twice daily; Neoral; Novartis,East Hanover, NJ) 5 days before tumortransplantation until the end of the ex-periment. Under sterile technique, tu-mor from a live carrier was harvestedand homogenized with a tissuegrinder (Model 23; Kontes Glass, Vine-land, NJ) and suspended in DulbeccoModified Eagle Medium (INC Bio-medicals, Aurora, IL) to a concentra-tion of approximately 1 � 108 cells/mL. Animals were anesthetized withTelazol (Lederle, Carolina, PR), andtumor injection sites were shaved anddisinfected with Betadine (PurdueFrederick, Norwalk, CT) and 100%ethanol (Fisher Scientific, Fairlawn,NJ). The tumor suspension was in-jected under direct visualization via an18-gauge needle into 12 sites on theback and torso. The tumors were al-lowed to grow to a mean diameter of5.4 cm � 1.0, approximately 10–12weeks after inoculation. Animals weremonitored daily and tumor sizes weremeasured weekly. Adequate tumoryield per animal (tumor diameter �4cm) was approximately 33%–67%(four to eight tumors per 12 injected),given variation in tumor growth andshape.

Tumor Size

The tumors reached a mean diam-eter of 5.4 cm � 1.0 (short-axis diam-eter range, 3.9–6.2 cm; long-axis diam-eter range, 4.8–7.5 cm). All tumorswere solid with a thin surroundingmembrane. Histologically, the tumorswere composed of a dense cellularstroma. These tumors are hypovascu-lar tumors with more vascularity out-side the surrounding membrane. Tu-mor appearance was similar to thatseen in previous studies (20).

Chemical Ablation

Two chemical agents, acetic acidand ethanol, were used to performchemical ablation. One hundred per-cent ethanol and 50% acetic acid wereacquired from the manufacturer(Fisher Scientific). Ethanol was used assupplied, but the 15% acetic acid solu-tion was created through the dilutionof 50% stock solution in saturated 36%saline solution until a 15% acetic acidconcentration was achieved. This dilu-

Figure 1. Multiple-tine chemical ablationinfusion device. The three tines extendingfrom the trocar shaft (arrows) are deployedto a diameter of 2 cm around the needleaxis. This lateral extension enables deliveryof sclerosant deeper into the tissues.

352 • RF Ablation with a New Multiple-tine System in a Canine Model February 2006 JVIR

tion mixture was chosen based on op-timal results achieved previously inthis tumor model (20).

Percutaneous multiple-tine infu-sion was performed under ultrasound(US) guidance. Tine deployment andproper tine spacing was confirmed byUS. Infusion of the entire volume wasperformed according to a stepwise in-jection strategy comprising variouscombination deployments includingdiameters of 1 cm, 2 cm, 3 cm, and 4cm perpendicular to the needle axis.For 1-cm and 2-cm tine deployment, asingle aliquot injection was used.However, for infusion at distances of 3cm and 4 cm, three aliquots were in-fused, allowing for 40° tine rotationamong aliquots to enable superior dis-tribution (based on unpublished pre-liminary data). The total volume ofchemical agent infused per tumorranged from 8 mL to 56 mL.

Percutaneous standard needle infu-sion was performed with use of a stan-dard 18-gauge, 3-cm needle (FisherScientific) that was introduced underUS guidance into the center of the tu-mor. For all needle types, each bolus of100% ethanol or 15% acetic acid wasslowly infused over a period of 30 sec-onds with use of a 10-mL syringe. In-fusion was discontinued when refluxof fluid was observed at the surface ofthe tumor at the needle puncture site.

Pathologic Studies of Necrosis

Each subject animal was eutha-nized immediately after ablation and

tumors were removed and sectionedwithin 1 hour after the animal’s death.Tumors were sectioned perpendicularto the device axis in 1-cm-thick slabswith use of a sharpened tissue-section-ing knife. Immediately after section-ing, mitochondrial enzyme activitywas used as an indicator of tumor vi-ability, and was assessed by incubat-ing tissue sections for 30 minutes in2% 2,3,5-triphenyl tetrazolium chlo-ride (Fisher Scientific) at 20°C–25°C.The absence of mitochondrial enzymeactivity has been shown to accuratelyreveal irreversible cellular injury in-duced by many forms of percutaneoustumor ablation (24). After the 30-minute incubation, the extent of visi-ble gross coagulation in the short axisdiameter (ie, perpendicular to the nee-dle axis) was then measured with cal-ipers in cross-sectioned x and y axes.Tissue that did not visibly stain formitochondrial enzyme was defined ascoagulated as per previous studies(24).

Statistical Analysis

For all experiments, each parameterwas fulfilled three separate times inseparate tumors, and all findings arereported as means � SD. Primary out-come measures included total volumeinfusible before visualization of fluidreflux at the needle puncture site. An-other primary outcome measure usedfor statistical analyses included coag-ulation diameter perpendicular to theapplicator shaft (ie, short axis) (25). In

tumors with asymmetric shapes, the xaxis was defined as the longest cross-sectional diameter from the analyzedslabs. The y axis diameter was definedas the perpendicular coagulation di-ameter of the same slab. Student t tests(significance threshold of P � .05; two-tailed test; Excel 2002; Microsoft, Red-mond, WA) or analysis of variance(version 6.1; Origin, Northampton,MA) was used to assess significant dif-ference for comparison groups as ap-propriate. Multivariate regressionswith use of linear and higher-orderregression models were performedwith use of Origin 6.1 software withthe strength of the best-fit regressioncurves reported as R2 computations.

RESULTS

Fluid Reflux

Multiple-tine infusion at 1 cm de-ployment enabled significantly moreacetic acid to be infused (15 mL � 3)than standard needle infusion (8 mL �1; P � .01) before reflux of fluid wasobserved at the tumor surface (Ta-ble 1). However, single infusions atgreater deployment distances did notenable significantly more fluid to beinfused without reflux (Table 1). Thegreatest amount of infusible fluid wasobtained with use of the stepped ali-quot infusion protocol (as describedlater) in which greater total infusionvolumes could be infused as furtherdeployments were performed. The to-tal infusible volume with a stepped

Table 1Total Infusible Volume before Fluid Reflux at Puncture Site

Distance

Device Chemical0 cm(mL)

1 cm(mL)

2 cm(mL)

3 cm(mL)

4 cm(mL)

Total(mL)

Mean CoagulationDiameter (cm)

18-g needle 15% acetic acid 8 �1 0 0 0 0 8 � 1 3.1 � 0.4Multiple-tine 15% acetic acid NA 15 � 3 0 0 0 15 � 3 3.5 � 0.8Multiple-tine 15% acetic acid NA 0 19 � 6 0 0 19 � 6 2.4 � 1.1*Multiple-tine 15% acetic acid NA 0 0 19 � 4 0 19 � 4 1.6 � 0.4*Multiple-tine 15% acetic acid NA 4 4 8 � 0 8 � 0 53 � 3 5.4 � 0.7

8 � 0 7 � 18 � 0 6 � 2

18-g needle 100% ethanol 8 � 4 0 0 0 0 8 � 4 2.0 � 0.3Multiple-tine 100% ethanol NA 4 4 4 � 0 4 � 0 31 � 0 1.7 � 0.5*

4 � 0 4 � 04 � 0 3 � 1

* Noncontiguous coagulation.

Hines-Peralta et al • 353Volume 17 Number 2

aliquot approach to a maximum de-ployment of 4 cm was 53 mL � 3.

For ethanol, standard needle infu-sion of 100% ethanol enabled infusionof 8 mL � 4 before reflux was noted atthe puncture site. More total infusiblevolume was achievable with use of themultiple-tine device (P � .05) in astepped aliquot approach in which 31mL � 1 represented the limit of infus-ible volume without reflux. The great-est infusion possible with 100% etha-nol (31 mL � 1) was notably less than

the volume achievable with 15% aceticacid (53 mL � 3; P � .01).

Optimization of Multiple-tineInfusion

To create contiguous zones of abla-tion in this tumor model, multiple in-jections of 2–8 mL acetic acid wererequired at each deployment setting(Table 2). Bolus infusion (19 mL � 4)at 3-cm multiple-tine deploymentyielded noncontiguous coagulation

(Fig 2). To achieve contiguous ablationfor 2-cm, 3-cm, and 4-cm tine deploy-ment, (ie, tines extending 1.0–2.0 cmfrom the central cannula), additionalinfusions were necessary at 1 cm and 2cm from the needle tip; otherwise, cen-tral residual viable tumor was seen atthe center of the tumor (Fig 2). At the3-cm and 4-cm settings, three aliquots(with tine rotation of 40° among ali-quots) were needed to achieve roundconfluent coagulation, as single ali-quot infusions at the 2-cm and 3-cmsettings yielded irregular borders. Op-timal infusions (ie, continuous coagu-lation without reflux) in this hard tu-mor model for spherical confluentchemical ablation were 4 mL for 1-cmand 2-cm deployments and three 4-mL40° rotated aliquots at 3 cm and 4 cm(Table 2).

Comparison of Extent ofCoagulation

Significant gains in coagulationwere achieved as more fluid was in-fused with use of our optimized mul-tiple-aliquot algorithm at further dis-tances from the needle axis (P � .05;linear R2 � 0.59; Figs 3–4). With anoptimal multiple-tine infusion proto-col of 32 mL and 15% acetic acid infu-sion (1 cm, 4 mL; 2 cm, 4 mL; 3 cm,three 4-mL aliquots at 40° rotation; 4cm, three 4-mL aliquots at 40° rota-tion), the mean coagulation diametermeasured 4.9 cm � 1.1, which wassignificantly larger than the coagula-

Figure 2. Achievement of contiguous coagulation with 15% acetic acid. (a) Singleinfusion at 3 cm of 19 � 4 mL results in nonconfluent coagulation with viable patches oftissue (red tissue; arrows) observed throughout the ablation. Conversely, stepped infu-sions at 1 cm (4 mL), 2 cm (4 mL), 3 cm (12 mL), 4 cm (12 mL; 32 mL total) produce largezones of confluent coagulation in this hard tumor model (b).

Table 2Optimization of Multiple-tine Infusion for 15% Acetic Acid

Distance

TumorsTreated

1 cm(mL)

2 cm(mL)

3 cm(mL)

4 cm(mL)

Total(mL)

Mean Coagulation(cm)

3 4 � 0 4 � 0 0 0 8 � 0 3.2 � 0.93 0 12 � 0 4 � 0 0 16 � 0 3.2 � 1.0*3 4 � 0 4 � 0 4 � 0 4 � 0 16 � 0 3.3 � 1.0*3 4 � 0 4 � 0 8 � 0 8 � 0 24 � 0 3.8 � 1.2*3 4 � 0 4 � 0 2 � 0 2 � 0 20 � 0 4.0 � 0.3*

2 � 0 2 � 02 � 0 2 � 0

3 4 � 0 4 � 0 4 � 0 4 � 0 32 � 0 4.9 � 1.14 � 0 4 � 04 � 0 4 � 0

3 4 � 0 4 � 0 8 � 0 8 � 0 53 � 3 5.4 � 0.78 � 0 7 � 18 � 0 6 � 2

* Noncontiguous ablation.Note.—Triplicate numbers at 3 cm and 4 cm deployment denote that the device wasrotated 40° among three smaller aliquot infusions at that diameter.

Figure 3. With stepped aliquot infusionof 15% acetic acid, the needle and resultantcoagulation diameter with use of the mul-tiple-tine device increases linearly with theincrease in chemical volume. Use of greaterdeployment distances enabled greater totalinfusion volume, resulting in larger zonesof coagulation.

354 • RF Ablation with a New Multiple-tine System in a Canine Model February 2006 JVIR

tion diameter with maximum stan-dard needle infusion of 8 mL � 1,which yielded 3.1 cm � 0.4 (P � .01;Table 3).

Fifteen percent acetic acid achievedsignificantly larger and more conflu-ent zones of ablation than 100% etha-nol for standard needle infusion andoptimal 32-mL multiple-tine infusion.

For standard needle infusion, aceticacid achieved ablation zones of 3.1 cm� 0.4 compared with 2.0 cm � 0.2 for100% ethanol (P � .05; Table 3). Ad-ditionally, multiple-tine infusion of100% ethanol at the optimal parame-ters that yielded 4.9 cm of coagulationfor acetic acid did not yield a confluentablation zone but instead produced

three small (1.7 cm � 0.5) separatezones of coagulation (Fig 5) in a clo-verleaf pattern. These areas of chemi-cal ablation corresponded to the pathsof the three tines.

Multiple-tine infusion of 15% aceticacid also resulted in more sphericalzones of ablation than standard needleinjection alone. Injection via a stan-dard single needle of 8 mL � 1 (ie,maximum infusible volume withoutreflux) resulted in elliptical coagula-tion measuring 3.9 cm � 1.0 (along theneedle shaft) by 2.2 cm � 0.3 in diam-eter. However, infusion of 8 mL withuse of the multiple-tine device (1 cm, 4mL; 2 cm, 4 mL) created a relativelyspherical lesion measuring 3.2 cm �0.9 by 3.0 cm � 0.9.

Adverse Effects

Some local adverse effects were ob-served after infusion of 53 mL � 3acetic acid (ie, 8-mL aliquots at 3 cmand 4 cm). Specifically, sclerosis ofnormal blood vessels at the base oftumor was observed in 67% of cases(two of three ablations), suggestingthat systemic chemical reflux was oc-curring with use of the larger 8-mLaliquots (Fig 6). No adverse effects as-sociated with chemical uptake wereobserved under all other infusion pro-tocols.

DISCUSSION

Multiple-tine Infusion Results inGreater Total Infused Volume

The data suggest that larger zonesof ablation can be achieved with largervolumes of injected sclerosing com-pounds (26). However, for hard tumormodels, the total amount of infusiblevolume is limited by chemical refluxoutside the tumor, which usually oc-curs at the needle puncture site. In theclinical setting, this results in un-wanted spillage of ablative agents intothe peritoneal cavity or surroundingtissue, causing pain or other negativeside effects (18). The data here suggestthat multiple-tine infusion with use ofa stepped aliquot protocol reliably en-ables significantly more infusible vol-ume while minimizing the risk ofchemical reflux. The result is signifi-cantly greater coagulation with morepredictable and spherical contours.Standard needle infusion is likely lim-

Figure 4. (a) Multiple-tine infusion with 15 mL � 3 of 15% acetic acid at the 1-cm setting.(b) Stepped multiple-tine infusion with the following protocol: 1 cm, 4 mL; 2 cm, 4 mL;3 cm, 12 mL; 4 cm, 12 mL (32 mL total). Infusion further away from the needle enablesmore total infusion and yields a significantly larger and confluent lesion.

Figure 5. Ethanol ablation: in this hard tumor model, treatment with 100% ethanol withan optimal multiple-tine infusion protocol (32 mL, stepped delivery) results in a cloverleafpattern of foci of ablation (arrows).

Hines-Peralta et al • 355Volume 17 Number 2

ited because fluid easily tracks backalong the needle axis when high resis-tance is encountered in tumor paren-chyma. This problem is potentiallycompounded because this fluid in-duces chemical sclerosis in tissues ad-jacent to the needle tract. However, inmultiple-tine infusion, multiple diffu-sion paths are simultaneously createdaway from the needle axis, enablinggreater penetration into tumor paren-chyma. Stepwise extension of the tinesfurther brings the chemical agent to

new areas of the tumor that have notundergone sclerosis.

Acetic Acid versus EthanolChemical Ablation

Acetic acid diluted to 15% in satu-rated saline solution was far superiorto 100% ethanol for standard needleinfusion and multiple-tine infusion.This supports previous data indicatingthat even dilute acetic acid achievessuperior coagulation to 100% ethanol

(19–21). The larger confluent zones ofcoagulation from 15% acetic acidlikely result from superior penetrationand diffusion into the solid tumor pa-renchyma. This contrasts with multi-ple-tine infusion of 100% ethanol,which yielded small zones of noncon-tiguous coagulation surrounding tinedeployment sites. This suggests that,in some hard tumor types, ethanol isnot sufficient to achieve adequatetreatment even when aided by simul-taneous multiple-tine infusion.

Although the greatest coagulationoccurred with greatest infused volume(53 mL � 3), this was also associatedwith adverse events. As expected, thissuggests that there is an upper bound-ary of safety even for this technique.The maximum safe infusible volumefor chemical infusion may be higher insoft hepatocellular carcinoma, othertumor types, or larger tumors, butnevertheless, it may be prudent to gaininitial clinical experience with the useof protocols like the one suggestedherein. Similar recommendations arereported by other investigators forethanol infusion (1–3). Regardless, it ismost important to determine the min-imal amount of any chemical neces-sary to achieve satisfactory results tominimize the possibility of these ad-verse risks.

Limitations of the Study

Given a limited number of avail-able tumors to study, we assessed onlytwo concentrations of ablation fluid,100% ethanol and 15% acetic acid di-luted in saturated saline solution, anda finite number of injection algo-rithms. Therefore, it is possible that

Table 3Coagulation with Use of Optimal Parameters versus Standard Needle Infusion

Algorithm

Distance 15% Acetic Acid 100% Ethanol P Value(100% ethanol

vs 15%acetic acid)

1 cm(mL)

2 cm(mL)

3 cm(mL)

4 cm(mL)

Total(mL)*

X Axis(cm)

Y Axis(cm)

Mean(cm)

X Axis(cm)

Y Axis(cm)

Mean(cm)

18-g needle 0 0 0 0 8 � 1 (8 � 4) 3.9 � 1.0 2.2 � 0.3 3.1 � 0.4 2.3 � 0.3 1.7 � 0.3 2.0 � 0.3 �.05Multiple-tine

infusion4 � 0 4 � 0 0 0 8 � 0 (8 � 0) 3.3 � 1.0 3.0 � 0.9 3.2 � 0.9 1.4 � 0.3 0.9 � 0.2 1.1 � 0.4† �.05

Multiple-tineinfusion

4 � 0 4 � 0 4 � 04 � 04 � 0

4 � 04 � 04 � 0

32 � 0 (31 � 1) 4.9 � 1.0 4.8 � 1.0 4.9 � 1.1 2.1 � 0.2 1.4 � 0.4 1.7 � 0.5† �.05

* Values in parentheses represent 100% ethanol.† Noncontiguous coagulation.

Figure 6. Skin specimens that overlay a tumor treated with multiple-tine infusion of 53mL � 3 of 15% acetic acid. Sclerosis of the vessels (arrows) in the skin surrounding thetumor occurs from uptake of acid in the skin, a negative side effect of treatment. Thisfinding suggests that one must be judicious not to overinject chemical agents.

356 • RF Ablation with a New Multiple-tine System in a Canine Model February 2006 JVIR

further modification of the ethanol oracetic acid regimen may yield superiorresults. In addition to different infu-sion fluid, the tumor type itself islikely to influence the size and qualityof ablation for multiple-tine infusion,given that standard needle infusionyields different results for hepatocel-lular carcinoma and colorectal carci-noma (2,3,10). Results will need to berefined for each tumor and clinical sce-nario.

Size limitations of our model per-mitted study of tine extension to nogreater than 4 cm because our tumorsizes were almost always limited to ashort-axis diameter of 4–5 cm givenanimal care committee stipulations.Fortunately, these sizes are currentlyat the upper limits of tumors in whichchemical ablation treatment canachieve favorable clinical results. Inaddition to tumor type, tumor size islikely to affect the total infusible vol-ume and coagulation size. Therefore,further clinical study will be needed todetermine the effect of tumor size andtype as well as the optimal numberand volume of injection aliquots for5-cm deployment. In these experi-ments, calculation of the z axis of co-agulation (ie, along the needle axis)would also be advantageous to enablecalculation of the volume of ablationin addition to the coagulation diame-ters.

In addition, we did not assess coag-ulation, allowing for repeat infusionafter repositioning of a standard nee-dle, which is commonly performed inclinical practice. In such an experi-ment, allowing for repositioning intro-duces multiple variables that may bedifficult to assess given the limitedavailability for large animal studies. Afuture area of interest may be the com-parison of single or multiple multiple-tine infusions versus multiple reposi-tioned standard needle infusions in aclinical study. Under such circum-stances, we hypothesize that repetitiveneedle repositioning may lead to moreirregular results and necessitate treat-ment with more total infusion volumeto achieve similar results. In additionto this, clinical outcome research andcost–benefit studies will likely be war-ranted, as the very low costs of stan-dard needle infusion must be weighedagainst the higher cost of precisionchemical ablation with various nichechemical ablation methods such as the

multiple-tine infusion investigatedhere. Ultimately, comparison of out-comes, cost effectiveness, and physi-cian and patient preferences will needto be compared for these chemical ab-lation devices against other ablationtechniques such as RF and microwaveablation. This will likely reveal variousstrengths of individual modalities forthe treatment of different tumor types.

CONCLUSION

Multiple-tine chemical infusion im-proved the size and quality of chemi-cal ablation compared with standardneedle infusion. In a large solid animaltumor, this suggests that chemical ab-lation with acetic acid infused with amultiple-tine device may yield a largerextent of coagulation than conven-tional needle injection alone in clinicalpractice, even for some hard solid tu-mors. Given the significant superiorityof 15% acetic acid compared with100% ethanol, it is probably worth-while to continue to pursue infusionstrategies with the use of acetic acid,most likely with the optimized injec-tion algorithm reported herein as astarting point. More investigation isnecessary to determine if the preciseinfusion achieved with multiple-tinedevices enables more effective and ef-ficient volume use than repositioninga standard single needle. However,use of the single needle stick offeredby this multiple-tine device is likely tobe more straightforward and poten-tially will reduce the procedure time,and theoretically can reduce the risksof some complications such as bleeding.As such, further research in the form ofpreliminary clinical trials with use of astepped algorithm as suggested hereinis likely warranted to determine the util-ity of multiple-tine chemical ablation inthe clinical setting.

References1. Livraghi T, Giorgio A, Marin G, et al.

Hepatocellular carcinoma and cirrhosisin 746 patients: long-term results ofpercutaneous ethanol injection. Radiol-ogy 1995; 197:101–108.

2. Shiina S, TagawaK, Niwa Y, et al.Percutaneous ethanol injection therapyfor hepatocellular carcinoma: results in146 patients. AJR Am J Roentgenol1993; 160:1023–1028.

3. Lencioni R, Pinto F, Armillotta N, et al.Long-term results of percutaneous eth-anol injection therapy for hepatocellu-

lar carcinoma in cirrhosis: a Europeanexperience. Eur Radiol 1997; 7:514–519.

4. Livraghi T. Radiofrequency ablation,PEIT, and TACE for hepatocellular car-cinoma. J Hepatobiliary Pancreat Surg2003; 10:67–76.

5. Lencioni R, Cioni D, Crocetti L, et al.Percutaneous ablation of hepatocellu-lar carcinoma: state-of-the-art. LiverTranspl. 2004; 10(2 suppl 1):S91–S97.

6. Livraghi T. Percutaneous ethanol in-jection in the treatment of hepatocellu-lar carcinoma in cirrhosis. Hepatogas-troenterology 2001; 48:20–24.

7. Lencioni R, Pinto F, Armillotta N, et al.Long-term results of percutaneous eth-anol injection therapy for hepatocellu-lar carcinoma in cirrhosis: a Europeanexperience. Eur Radiol 1997; 7:514–519.

8. McGahan JP, Dodd GD III. Radiofre-quency ablation of liver tumors: cur-rent status. AJR Am J Roentgenol 2001;176:3–16.

9. Gillams AR, Lees WR. Survival afterpercutaneous, image-guided, thermalablation of hepatic metastases fromcolorectal cancer. Dis Colon Rectum2000; 43:656–661.

10. Giovannini M, Seitz JF. Ultrasound-guided percutaneous alcohol injectionof small liver metastases: results in 40patients. Cancer 1994; 73:294–7.

11. Liu Z, Lobo SM, Humphries S, et al.Radiofrequency tumor ablation: in-sight into improved efficacy usingcomputer modeling. AJR Am J Roent-genol 2005; 184:1347–1352.

12. Pereira PL, Trubenbach J, Schenk M, etal. Radiofrequency ablation: in vivocomparison of four commerciallyavailable devices in pig livers. Radiol-ogy. 2004; 232(2):482–490.

13. Rhim H, Goldberg SN, Dodd GD III, etal. Essential techniques for successfulradio-frequency thermal ablation ofmalignant hepatic tumors. Radio-graphics 2001; 21(suppl):S17–S35.

14. Shock SA, Meredith K, Warner TF, etal. Microwave ablation with loop an-tenna: in vivo porcine liver model. Ra-diology. 2004; 231:143–149.

15. Lau WY, Leung TW, Yu SC, et al. En-hanced renal parenchymal cryoabla-tion with novel 17-gauge cryoprobes.Urology 2004; 64:173–175.

16. Arata S, Tanaka K, Okazaki H, et al.Risk factors for recurrence of largeHCC in patients treated by combinedTAE and PEI. Hepatogastroenterology2001; 48:480–485.

17. Ikeda M, Okada S, Ueno H, et al. Ra-diofrequency ablation and percutane-ous ethanol injection in patients withsmall hepatocellular carcinoma: a com-parative study. Jpn J Clin Oncol. 2001;31:322–326.

Hines-Peralta et al • 357Volume 17 Number 2

18. Vehmas T. Reflux of ethanol duringexperimental liver ethanol injections.Invest Radiol. 1992; 27:918–921.

19. Shah SS, Jacobs DL, Krasinkas AM, etal. Percutaneous ablation of VX2 car-cinoma-induced liver tumors with useof ethanol versus acetic acid: pilotstudy in a rabbit model. J Vasc IntervRadiol 2004; 15:63–67.

20. Ahmed M, Weinstein J, Liu Z, et al.Image-guided percutaneous chemicaland radiofrequency tumor ablation inan animal model. J Vasc Interv Radiol2003; 14:1045–1052.

21. Huo TI, Huang YH, Wu JC, et al.Comparison of percutaneous acetic

acid injection and percutaneous etha-nol injection for hepatocellular carci-noma in cirrhotic patients: a prospec-tive study. Scand J Gastroenterol 2003;38:770–778.

22. Rossi S, Buscarini E, Garbagnati F.Percutaneous treatment of small he-patic tumors by an expandable RF nee-dle electrode. AJR Am J Roentgenol1998; 170:1015–1022.

23. Ahmed M, Liu Z, Afzal KS, et al.Radiofrequency ablation: effect ofsurrounding tissue composition oncoagulation necrosis in a canine tu-mor model. Radiology 2004; 230:761–767.

24. Liszczak TM, Hedley-Whyte ET, Ad-ams JF, et al. Limitations of tetrazo-lium salts in delineating infractedbrain. Acta Neuropathol (Berl) 1984;65:150–157.

25. Goldberg SN, Charboneau JW, DoddGD III, et al. International WorkingGroup on Image-Guided Tumor Abla-tion. Image-guided tumor ablation:proposal for standardization of termsand reporting criteria. Radiology 2003;228:335–345.

26. Meloni F, Lazzaroni S, Livraghi T.Percutaneous ethanol injection: singlesession treatment. Eur J Ultrasound2001; 13:107–115.

358 • RF Ablation with a New Multiple-tine System in a Canine Model February 2006 JVIR