Effect of Skin Electrode Location on Radiofrequency Ablation Lesions: An In Vivo and a Three-Dimensional Finite Element Study

of 12
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Information Report
Category:

Services

Published:

Views: 35 | Pages: 12

Extension: PDF | Download: 0

Share
Description
Skin Electrode Location Affects RF Ablation Lesions. Introduction: Objectives: To assess the effect of skin electrode location on radiofrequency (RF) ablation lesion dimensions and energy requirements. Background: Little is known about the effects of
Tags
Transcript
  1325 Effect  of  Skin Electrode Location  on  RadiofrequencyAblation Lesions:  An In  Vivo  and a  Three-DimensionalFinite Element Study MUDIT  K.  JAIN,  B.E.,  GERY TOMASSONI,  M.D.. RICHARD  E.  RILEY, M.S.,*  and  PATRICK  D.  WOLF,  PH.D. From  the  NSF/ERC  and the  Department  of  Biomedical Engineering. Duke University.Durham, North Carolina:  and  *Cardiac Pathways Corp.. Sunnyvale, California Skin Electrode Location Affects  RF  Ablation Lesions,  introduction:  Objectives:  To assess  the  efTect  of  skin electrode location  on  radiofrequency  (RF)  ablation lesion dimensionsand energy requirements. Background: Little  is  known about  tbe  effects  of  skin electrode loca-tion  on  RF ablation lesion dimensions and efficiency. Methods  and  Results Temperature-controlled ablation  at 60 C for 60  seconds  was per- formed  in six  sbeep. Paired lesions were created  in tbe  lateral, anterior, posterior,  and  septalwalls  of  both  the  ventricles.  For  group 1 lesions,  tbe  skin electrode  was  positioned directly  op- po.site  tbe  catbeter  tip  (optimal).  For  group  2  lesions,  we  used eitber  the  standard posterior  lo- cation  or an  anterior location  if the  posterior skin electrode location  was  used  for  group  1. Group  1  lesions were  5.8 ± 0.8 mm  deep  and 9.3 ± 1.9 mm  wide, compared with  4.6 ± 1.0 mm deep  and 7.7 ± 1.9 mm  wide group  2  lesions  (P   0.001). Group  1  lesion dimensions also  bad less variability.  A  finite element model  was  ased  to  simulate temperature-controlled ablationand  to  study  the  effect  of  skin electrode locations  on  lesion dimensions, ablation efficiency,  and blood beating.  The  optimal location  was 1.6  times more efficient,  and tbe  volume  of  bloodbeated  to >  90 C was 0.005 mm' for optimal versus  2.2  mm'  for  the nonoptimal location. Conelusion Optimal skin electrode placement:  (1)  creates deeper  and  larger lesions;  (2) re- duces lesion size variability;  and (3)  decreases blood beating.  U  Cardiovmc Electrophysiol.  Vol. 9,  pp. 1325-1335, December  199H) radiofrequency, catheter ahlation. tachycardia, computer  model,  finite element,  .skin  electrode,ground electrode Introduction Radiofrequency (RF) catheter ablation is theprimary interventional therapy for the treatment of many  cardiac tachyanhythmias. In particular, RFablation has been highly successful in the treat- This work  was  supported  in  part  by  NSF/ERC Grant  CDR- 8622201  and a  grant  of  computer lime from  the  North CarolinaSupercompLling Center.  The  siudy  was  approved  hy the  Institu-tional Animal  C^e and Use  Committee  at  Duke University,  and conforms  to the  po.siiions  of ihe  American Hean AsscK'iittion  on research animal  use. Address  tor  correspondence: Patrick  D.  Wolf,  Box  W28I. Depan-ment  of  Biomedical Kngineering, Duke University. Durham,  NC 27708-0281.  Fax:  919-660-541)5: E-mail: pdw'c^eel-mail.mc.duke.eduManuscript received  .'^0  June  99S:  Accepted  for  publication  2 September 1998. ment of supraventricular tachycardias.'-However,the success of this technique has been limited inthe treatment of ventricular tachycardia (VT) forthe following reasons: (I) the presence of multi-ple types of VT; (2) the inability to induce clini-cally relevant VT; (3) the difficulty in mapping  a VT that causes hypotension; and (4) the inabilityto ablate deeply located and/or diffuse ventriculararrhythmogenic sites.''-'' Addressing this last issue,RF ablation might be more effective in the treat-ment of ventricular tachyarrhythmias if larger anddeeper lesions could be created from the endo-cardial surface.Increasing the applied power to create a largerlesion is not feasible because of coagulum for-mation and the resulting sudden impedance rise.'' '"  Chilled RF ablation., improved electrode tip  1326 Journal of Cardiovascular Eiectrophysioiogy  Vol.  9.  No. 12.  December 1998 design, and use of niiiltipolar catheters" '-' aresome  ot"  the new lechniques used to limit the in-cidence of sudden impedance rise while increas-ing the lesion size. In addition to the above men-tioned techniques, we propose that placing theskin electrode directly opposite the catheter tipcan significantly improve the efficiency of the ab-lation procedure by affecting the current densitydistribution, hence creating a deeper and largerlesion.Santoro et al.'-* investigated the effect of skinelectrode location on impedance in humans, butthe effect of skin electrode location on lesion di-mensions and ablation efficiency has not beenmeasured.The -Specific objectives of this study of tem-perature-controlled RF ablation in the ventriclewere: (I) to directly compare the size of lesionscreated by the standard (posterior) and the opti-mal skin electrode location; (2) to study the effectof skin electrode placement on lesion size vari-ability; (3) to simulate the effect of the skin elec-trode location on temperature-controlled RF ab-lation using three-dimensional finite element analy- sis;  (4) to compare the in vivo and computermodeling results; and (5) to evaluate, using a com-puter model, the effects of skin electrode place-ment on the heating of intracavitary blood and en-ergy efficiency.Methods Animal Preparation and RF Current Delivery Setup Six adult female sheep (weight 122.75 ± 5.14Ib) were studied at tbe Experimental Electro-physiology Lab at Duke Utiiversity. Tbe sheepwere sheered to level the skin surface (for properelectrical contact) and sedated with ketamine hy-drochloride (15 to 22 mg/kg IM). After an IVwas established, isotlurane gas (1% to  59c)  wasused to anesthetize. The animal was intubatedand placed on a mechanical respirator (modelSAV. North American Drager Inc.. Telford. PA,USA) with the administration of supplementaloxygen. Intravenous Lactated Ringers was in-fused. An intravenous lidocaine bolus dose of100 mg was administered followed by a 2mg/min drip. A 10.000 unit bolus dose of hep-arin was infused with a subsequent dose of 2.000units/hour. Continuous blood pressure and arte-rial blood sampling via the right femoral arterywere performed. Electrolyte and respiratorchanges were made depending on serial elec-trolyte and arterial blood gas measurements.Blood pressure, lead II electrogram. and bodytemperature were continuously monitored.An RF generator (Cardiac Pathways Corp..Sunnyvale, CA. USA) was used to deliver un-modulated 50()-kHz sinusoidal current betweenthe ablalitm electrode and a 139.5-cm- skin elec-trode (model Polyhesive, Valleylab Inc., Boul-der. CO. USA) that was placed at the skin sur- face.  The long axis of the skin electrode wasoriented perpendicular to the median sternum(cranial to caudal is defined as parallel to themedian sternum). A 7-French 4-nim long distalelectrode tip steerable catheter (RADII-T. Car-diac Pathways Corp.) was used for RF ablation.Tbis catheter has a thermocouple at the elec-trode tip to provide feedback to the RF genera-tor for temperature-controlled ablation. The RFgenerator was set to maintain the electrode tiptemperature at 60°C. and to shutdown if the im-pedance exceeded 250  il  or if the maximum de-livered power exceeded 70 W. Ablation Protocol A 7-French quadripolar steerable ablation cath-eter was advanced into the right ventricle viathe right femoral vein, and the left ventriclevia the right femoral artery. Digital recordingsof two fluoroscopic views for each burn wereobtained to localize the catheter tip and to fa-cilitate the identification of lesions when theheart was excised. Endocardial electrode con-tact was guided using tluoroscopy. pacing thresh-olds (< 1 V at 0.5-msec pulse width), and in-tracardiac electrocardiogram amplitudes (R wavesensitivity > 10 mV).Right and left ventricular lesion pairs werecreated on the anterior, posterior, septal, and lat-eral walls. A maximum of two lesion pairs werecreated on a wall. For all lesion pairs on a givenwall, two skin electrode locations were tested(group I: skin electrode directly opposite tbecatheter tip; group 2: (a) the standard posteriorskin electrode, or (b) if the posterior skin elec-trode location was used for group I. then an an-terior skin electrode. For eacb lesion, properelectrical contact between the skin surface andthe electrode was ensured by applying conduc-tive gel to the electrode surface and by holdingthe electrode firmly in the desired location us-ing medical adhesive and elastic tape wrappedaround the animal's torso.  Jaiii.  et  al.  Skin Electrode Location Affects  RF  Ablation Lesions 1327With the catheter properly positioned at thechosen site for ablation, temperature-controlledRF ablation at 60°C was performed for 60 sec- onds.  Recordings of power, maximum imped;ince,and tip temperature were obtained during the RFcurrent application. Eacb group 2 lesion was pro-duced before tbe corresponding group I lesion.The group I and group 2 lesions were producedin close proximity in order to minimize variabil-ity in lesion size due to anatomic location. To aidin identifying the lesion for measurement, thegroup 1 lesion was always placed superiorly tothe group 2 lesion. If the maximum power limitwas exceeded, the catheter was withdrawn,checked for coagulation, and cleaned. The cath-eter was then repositioned and a new pair of le-sions created. The srcinal lesion pair was ex-cluded from the size analysis but enumerated forthe safety statistics.After completion of each study, with the ani-mal under general anesthesia, euthanasia was per-formed with the use of potassium chloride, as rec-ommended by the Panel on Euthanasia of theAmerican Veterinary Medical Association. Theheart was excised and grossly dissected, takingcare not to damage the lesions. Each lesion wasidentified, resected, tagged, and fixed in 109f for-malin. After fixation, measurements were madeof the width, length, and depth of the lesions bya blinded investigator. Lesion Measurements To measure the lesion depth, 1-mm thick slicesof the tissue were cut parallel to the endocardiumusing a microtome (model Wolfe 628164. Car-olina Science and Math, Burlington, NC, USA).If the lesion was not apparent on both surfacesof a 1-mm slice, then tbat slice was further dis-sected into two 0.5-mm slices to measure thedepth of the lesion in that slice of tissue. Thetotal lesion depth was determined to a resolu-tion of 0.5 mm by summing the contributionsfrom each slice.Tbe slices for eacb lesion were photographedwith length scales for calibration. The eolor pho-tographs were digitally scanned using a NikonCoolscan scanner and the digital image down-loaded to a Macintosh computer {Power Macin-tosh 8500/150. Apple Computer Corp.. Cuper- tino,  CA, USA). The tissue in the region of RFablation lesions bad a well-demarcated area oftissue necrosis surrounded by a zone of hemor-rhage and intlammatory cells. The outer bound-ary of the thick dark brown region (hemorrhagicregion), which encloses the central white region,on each lesion slice was outlined, and the en-scribed area evaluated using the public domainNIH Image program (developed at the U.S. Na-tional Institutes of Health and available on theInternet at http://rsb.info.nih.gov/nih-image/down-load.html). The volume of an individual lesionslice was obtained by multiplying the lesion areawith the slice thickness. Individual slice volumeswere added to evaluate the total lesion volume.A maximum difference of 35% was found in thevolume measurements made using this technique,and the volume evaluated using the ellipsoidalapproximation'-'^ of lesion shape  {Volume =2B-nahc.  where  a  and  h  are the lesion half-lengthand tbe half-width, respectively, and  c  is the le-sion depth). The lesion width, depth, volume,and efficiency index were compared for group Iand group 2. Efficiency index (mmVJ) is definedas the volume of lesion produced per unit of de-livered energy. Statistical Analysis Continuous values are expressed as the mean± SD unless otherwise stated. Lesion depth, max-imum width, volume, maximum power, and to-tal energy were compared using a paired Stu-dent's r-test. P values < 0.05 were consideredsignificant. Lesion size for the two groups wasstudied using a repeated measures analysis ofvariance (ANOVA) that included an effect forskin electrode location, lesion location, and theinteraction term. ANOVA was used to testchanges in variance; however, the coefficient ofvariation was used to demonstrate these differ-ences. Chi-square statistics were used to test ifthere were any differences in the frequency ofimpedance rises and exceeding the maximumpower limit between group 1 and group 2. Finite Element Model Description Heat dissipation in the cardiac tissue caus-ing dehydration and protein denaturation pro-duces a myocardial lesion. The shape, size,and depth of this lesion depend on the temper-ature distribution in the myocardium. The tem-perature distribution in turn depends on the cur-rent density distribution in the tissue. Thus, thefinite element model was based on a time-do-main analysis of a coupled electric-thermal fieldproblem.  1328 Journal  of  Cardiovascular Eiectrophysioiogy  Vol. 9. Nti. 12.  December I99H G H B Catheter BodjPosteriorCardiac AblatingTissueElectrodeSkin P;itch 40nim 25mm 40mni 20mm80mm o T3 60.050.060.0 I 40.0 U 30.0 20.0 10 20 30 40 50 Time (seconds) Figure 1.  A  view ihrough  the  mid-section of the finite element model left panel).  The  blood volume represents  the  ventricularcavity.  In  this figure, skin electrode  is  .vhown  at  .surface  A  base), which  was  moved  to  surface  B and  siufaces  C. 1). and E to .simulate  the top. and the  .side  positions, respectively. Plots right panel) .show  the  voltafie levels applied  at ihe  electrode  in the finite element model) during  the ^  seconds ofRF current delivery for three ca.tes  and  typical electrode  tip  temperature.Voltage levels were reduced wilh  the  pas.safje  of time  to  maintain  the  electrode  tip at a  constant temperature  of56°C. A diagram of the geometry used for computersimulations is shown in Figure 1. A block of tho-rax was considered for the finite element model.An earlier two-dimensional model proposed byLabonte'"'•''' was extended to three dimensionsfor these computer simulations. Two blocks oftissue (80 mm X 40 mm X 40 mm) were con-sidered to model the posterior and anterior car-diac tissue. The ventricular cavity was modeledby the blood volume (80 mm x 25 mm x 40mm) shown between the anterior and tbe poste-rior endocardial tissue (Fig. I). A hemispheri-cal 2.5-mm diameter ablating electrode wasplaced on the endocardium to a depth of 1.25mm. Blood circulated around the ablating elec-trode, which was perpendicularly in contact withthe endocardia surface.An 80 mm X 40 mm skin electrode was sim-ulated at the following surfaces (Fig. I) to studythe effects of skin electrode location: (I) surfaceA (base); (2) surface B (top): and (3) surfacesC. D. and F (side). Base position is equivalentto placing the skin electrode opposite the abla-tion electrode lip. as in group 1 in vivo le-sions. Top and side positions simulate the group2 in vivo lesions.The temperature distribution in the tissue dur-ing ablation was evaluated by solving Pennes"bio-heat equation.'^ Metabolic heat generationand myocardial perfusion were neglected as theyhave been shown to be negligible for ablation.''-'^The voltage distribution in the model was eval-uated using Laplace's equation. The details ofthese equations and the underlying a.ssumptionshave been discussed in detail by Labonte."' Table1 gives the material properties used in thisThe temperattire variation of the electricalconductivity of the cardiac tissue (-t-2%/°C)-^was used for both the healthy and the thermallydamaged tissue, as no data exist for the ther-mally damaged tissue. In the 0.5- to 1-MHzrange, the electrical impedatice of the tissue ismostly resistive,'' hence a quasi static analysiswas used. Because the heart and the lungssimilar electrical conductivities  ^a  = 0.61 l">;  Volume TissueBkKHlElcctRxicCalheter BodyMaterial Properties  (at Mutcriiil Tissue BloodPt.-lr. Polytireihane1  MHz/•[k^1.06 1 .05221.4570 TABLK Frequency) •/m'l X  10' X  10' X  10' IUsed  in ihe  Finiie  HI k|W/mKl 0.70.5 0.26  X 10 ' iLMiienl Mtxlcl c[.l/kK.Kl 3.68  X  10' 3.8 X  to' 130.53 1.045 X 10' -r  |S/m| 0.610.95 9.4 X 10* 10"  Jain,  et  at.  Skin Electrode Location Affects  RF  Ablation Lesions  1329 S/m  at I  Mhz"^), use  of a  homogeneous modelwas justified.ANSYS  5.2A  finite element software pack-age (ANSYS, Pittsburgh. PA. USA), which  al- lows coupled thermal electric transient analysis,was used  for  the finite element solution. The  fi- nite element model  had  10,200 nodes,  and used both brick  and  hexahedral elements. Con-vergence tests were performed  to  ensure  ade- quate .spatial  and  temporal discretization.  The smallest grid size  was  0.1  mm at tbe  ablatingelectrode surface, and  6  mm  at  the model bound- aries.  Previous studies-^ have shown that  the maximum rate  of  rise  in  temperature occurs dur-ing the first  few  seconds  of  the ablation proce-dure. Thus,  a  time step size  of  0.5 seconds wasused  for  the first  20  seconds,  and  subsequentlya time step  of  1  second was used  for a  total timeof 60 seconds.used classically  in  thermal modeling^^ and  specif- ically  in  ablation modeling"^'''"  to  simulate  the cooling provided  by  the circulating blood. Whilethe  use of a  convective boundary condition  is an approximation, we feel the bltwd temperatures ob-tained with this approximation convey the relativedifferences in blood temperatures for the three skinelectrode locations.In  the  model,  the  lesion deptb was measuredfrom the tissue-blood interface surface along  the axis  of  the electrode, and the diameter  of  tbe  le- sion  was  measured  at its  greatest point.  Any element with  a  temperature  >:  50"C  at the  endof  60  seconds  was  considered part  of the le- sion.-''"' Applied power was evaluated from  the voltage and current values (P  =  VI). These powercalculations were then used  to  evaluate  the to- tal delivered energy  by  integrating  the  powerover time. Boundary Conditions Constant voltage (Dirichlet) boundary condi-tions were applied  at the  electrode surfaces.  To simulate temperature-controlled  RF  ablation,  a few trial simulations were used  to  determine thetip temperature required  to  maintain  the  maxi-mum tissue temperature constantly  at  —95°C.From these simulations,  an  electrode  tip tem- perature  of  —56°C was observed  to be  optimal. Thus,  the voltage levels applied  at  the electrodetip were modulated during  the  simulations  to maintain the electrode tip temperature  at  =56°C.Similar (6O''C) electrode tip temperature settingshave been used  in  previously reported clinicalstudies.-^  The  applied voltage  and  typical elec-trode  tip  temperature versus time curves  are shown  in  Figure  .  Similar curves have been  re- ported  in  vivo  by  McRury  and  Haines.-" Zeroelectric current (Neumann) boundary conditionswere applied  at  all the outer surfaces  of  the modelexcept  on the  skin electrode.  The  voltage  on the surfaces where the skin electrode was locatedwas  set to 0 V. The temperature  at all  the outer surfaces  of  themodel was fixed  at  37°C. as these surfaces are dis-tant from the ablating electrode. Convective bound-ary conditions were applied  at  the tissue-blocxl andelectrode-blood interfaces  to  account  for  coolingby  the  circulating blood  in the  cardiac chamber.The evaluation  of  the average heat transfer  coef- ficient has been discussed by Labonte  (/i = 1.8 X U)-'  W/nim-rC).'" The use of a convective bound-ary condition  is an  approximation that  has  beenResults In Vivo Lesions A  total  of  35 pairs  of  ventricular lesions wereobtained  in 6  adult sheep. However, only  27 pairs  of  lesions  (17  left ventricular  and 10 right ventricular lesion pairs) were used for com-parison  in  this work.  In 7 of  the unused lesion E  I (I E o.n *.  , •   'TI   •   t •• 1 300.0  - 5 100,0 t r  f \ 1   .,11 • • •  i i   1 1   - Antenor .SL pI;ilPiisltTior Lateral Figure  2.  Lesion volume and efficiency  inde.x  for  group  I and group  2  plotted according  to the  location  of  the ven-tricular lesion. Each lesion pair is  shown  connected  with  a line. Of the total 27 lesion pairs, in 26 pairs, the lesion vol-ume  for  group  I  was larger than that  for  group  2.  Onlyfour group  2  lesions were produced more efficiently thangroup  I  lesions.
Recommended
View more...
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks
SAVE OUR EARTH

We need your sign to support Project to invent "SMART AND CONTROLLABLE REFLECTIVE BALLOONS" to cover the Sun and Save Our Earth.

More details...

Sign Now!

We are very appreciated for your Prompt Action!

x