1055

Transvenous Cold Mapping and Cryoablation of the AVNode in Dogs: Observations of Chronic Lesions and

Comparison to Those Obtained UsingRadiofrequency Ablation

LUZ-MARIA RODRIGUEZ. M.D., JET LEUNISSEN, R.N.,AREND HOEKSTRA. PH.D.,t BART-JAN KORTELING, B . S c t

JOEP L.R.M. SMEETS. M.D.. GARL TIMMERMANS, M.D., MARC VOS, PH.D.,

MAT DAEMEN. M.D.,* atid HEIN J.J. WELLENS, M.D.

From the Departments of Cardiology and ^Pathology. Academic Hospilal, Maastricht,and tCordis Eumpa NV. Roden. The Netherlands

Cryoablation of the Proximal AV Node. Introduction: Radiofrequency (RF) is the mostcommonly used energy source for the treatment of cardiac arrhythmias. Surgical experiencehas shown that cryoablation also is effective for ablating arrhythmias. The aims of this studywere to (I) investigate the feasibility of inducing permanent complete AV block (CA VB). (2) in-vestigate the value of cold mapping to select the cryoablation site to produce permanentCAVB, (3) study the macro- and microscopic lesion characteristics 6 weeks later, and (4) com-pare them to those produced with RF energy.

Methods and Results: A new steerable 8.5-French bipolar electrode catheter having a ther-mocouple with a 3-mni tip using N,O as the refrigerant controlled by a cryoconsole was used.Six mongrel dogs were anesthetized, and the catheter was positioned via the femoral veinacross the tricuspid valve to record a large low right atrial and a small His-hundle potential.After cold mapping (-15" to -2O''C tip temperature) resulted in fX'G modiflcations, cryother-mia (-70'O was given twice, lasting 5 minutes each, to create permanent CAVB (Cryo group).Additionally, RF catheter ablation of the AV node was performed in two anesthetii^ed mongreldogs (RF group). In tfie Cryo group, a permanent proximal CAVB was created in four dogs(block occurred within 10 to 20 sec of cryothermia). Permanent right bundle branch block wasobtained in one dog and transient CAVB in the remaining dog. In f)oth dogs of the RF group,permanent CAVB was obtained. The cryolesions consisted of well-circumscribed, fiomoge-neous areas of fibrofic tissue without viable cardiomyocytes. Lesions produced with RF wereless circumscribed and inhomogeneous, with clear evidence of viable cardiomyocytes and carti-lage formation (patchy lesions).

Conclusions: (1) Permanent CAVB can he created by using a steerable cryoablation cathe-ter. (2) Histologically, cryoablated sites were homogeneous and showed fibrotic tissue withoutsigns of chronic inflammation and no evidence of viable myocytes. (3) Lesions created with RFwere less homogenous and still contained viable myocytes within the lesion and cartilage for-mation. (4) The arrhythmogenic significance of these differences requires further study. (5)The technology of using reversible cold mapping has the potential to identify the successful ab-lation site and warrants further clinical study. (J Cardiovasc Eiectrophysiot. Vol. 9. pp. 1055-J06I, October 1998)

cryoablation, radiofrequency ablation, ablation of the proxitnal His, transvenous cryoablation

Address Tor correspondence: Luz-Maria Rodriguez. M.D.. Depart- lm.rodrigiiez@cardio.azni.nlmeni of Cardiology. Academic Hospital Maasirichl. CARIM (Car-diovascular Research, tnstitutc Maastricht), P.O. Box 58W), Maas- Manuscript received 18 April 1998; Accepted for puhlication 2tricht. The Netherlands. Fax: 3 N43-3875104; E-mail: July 1998.

1056 Journal of Cardiovascular Eiectrophysiology Vol. 9, No. 10, October 1998

Introduction

Radiofrequency (RF) is the most commonlyused energy source for the treatment of ciirdiac ar-rhythmias.'- Surgical experience has shown thatcryoablation also is effective for ablating arrhyth-mias.̂ ^̂ Some years ago, the introduction of per-cutaneous cryoablation for the management of car-diac arrhythmias was attempted, but without goodresults.'' Failure was due to catheter thickness, lackof steerability, and the inability to record electro-grams from the cryocatheter. Previous studies inanimals'"'" and humans" have investigated thetissue characteristics of lesions after surgical andtransveneous cryoablation. Although different tech-niques were used in those studies., all agreed thatthe cryolesions are small., homogeneous, and welldelineated. The aims of this study were to (1) studythe feasibility of producing permanent completeAV bkK;k (CAVB) using a new steerable 8.5-Frenchbipolar electrode catheter with a 3-mm tip, (2)investigate cold mapping to select the cryoablationsite to produce permanent CAVB, (3) study thehistologic characteristics of cryoablation-inducedchronic lesions, and (4) compare such lesions tothose produced using RF energy.

Methods

Experimental Preparation

The protocol was approved by the Committeefor Experiments on Animals of the University ofLimburg, Maastricht. The Netherlands. Eight mon-grel dogs of either sex weighing between 21 to 34kg were used in this study. After overnight fast-ing, anesthesia was induced by intramuscularpremedication {I mL per 5 kg: lO-mg oxycodon,1-mg acepromazine, and 0.5-mg atropine), andsodium pentobiirbital (20 mg/kg body weight IV).Under aseptic conditions., the dogs were artificiallyventilated through a cutted endotracheal tube us-ing a mixture of oxygen, nitrous oxide (1:2), andhalothane (vapor concentration 0.5% to 1 %) by arespirator. Ventilation was controlled by keepingthe carbon dioxide concentration in the expired airbetween 3.8% and 4.2%. A thcnnal mattress wasused to maintain adequate body temperature. He-mostatic sheaths were introduced into the rightfemoral vein. Six surface ECG leads and endo-cardial recordings were recorded simultaneouslyand stored on optical disk. After creation of CAVB,a pennanent pacing electrode (Stimline ISP 13B,8-French [Vitatron, Dieren, The Netherlands]) wasinserted at the apex of the right ventricle via the

extemal jugular vein. Six dogs underwent cryoab-lation and two dogs RF catheter ablation. Duringthe experiment, the dog received 0.5 to 1.0 L of0.9% NaCI through the cephalic vein to preventvolume depletion.

Cryoablation Procedure

A new steerable 8.5-French bipolar electrodecatheter including a thermocouple (Cordis Eu-ropa NV, Roden, The Netherlands) with a 3-mmtip (Fig. 1) using N^O as refrigerant controlledby a cryoconsole (Spembly, Hampshire, UK) wasused to create CAVB. The steerable catheter waspositioned across the tricuspid valve to record alarge low right atrial and small His bundle po-tential, guided by fluoroscopic views from theright and left anterior oblique angles. Cryother-mia -10" lo -20°C tip temperature was applieduntil ECG changes occurred (prolonged PR,Wenckebach conduction. 2:1 AV block [Fig.2]) and/or noise on the His-bundle electrogramdue to ice formation were observed. After coldmapping, cryothermia -70°C was given twice,lasting 5 minutes each, allowing recovery toapproximate body temperature between each ap-plication. If CAVB did not occur in the first 45seconds, the application was stopped, the cathe-ter repositioned, and a second attempt was made.After CAVB was created, atropine 0.5 mg IVwas given to check for residual AV conduction.

RF Ablation

A 7-French quadripolar steerable catheter witha 4-mm tip (CordisAVebster, Baldwin Park, CA,USA) was positioned across the tricuspid valve torecord a large low right atrial and a small-His bun-dle potential, guided by tluoroscopic views fromthe right and left anterior oblique angles. RF cur-rent was delivered from a 500-Hz generator (HAT200 [Dr. Osypka, Grenzach-Wyhlen, Germany]) at70°C for 60 seconds between the large-tip ther-mocouple electrode of the ablation catheter and anadhesive electrosurgical dispersion pad applied onthe back of the dog.

Follow-Up

To confirm persistence of chronic CAVB, ECGswere recorded 6, 24, and 48 hours and once everyweek for 6 weeks after the procedure.

The animals were sacrificed at 6 weeks afterthe prtKedure by 2(X) mg/kg euthesate (Apharmo,

Rodriguez, et al. Cryoablation of the Proximal AV Node 1057

Figure 1. The 8.5-Erench bipolar steerable catheter.

The Netherlands). Tbe hearts were removed, andthe lesions were analyzed macroscopically and mi-croscopically.

Histologic Examination

Tbe beart was flushed with saline and fixed in10% phospbate-buleered fomiaiin (pH 7.4). Aftergross inspection, including quantification of lesionsize by lengtb, width, and depth (in mm) of tis-sue necrosis and photography of the AV junction,a rectangular block was taken with relevant land-marks such as the septal leaflet of the tricuspidvalve and tbe septa! leaflet of the mitral valve. Se-rial slices were cut perpendicularly through the AVconduction system at 10-mm thickness. Tbe tissueslices were processed and embedded in paraffin(4-fj.m thickness); every 20th section was mountedfor Azan staining. The tissue sections were ex-amined for accuracy of the location of tbe lesionin the AV junctional area, homogeneity of the scar,and damage to any surrounding structures.

Results

Cryoablation

In 4 of 6 dogs, cold mapping resulted in AVconduction disturbances. In these four dogs {no. 1,2, 4, 5) permanent proximal CAVB was achievedafter two cryo applications of 5 tninutes each (Fig.2). CAVB in these dogs occurred within 10 to 20seconds, witb an escape rbythm of 50 to 55beats/min. The width of the QRS of tbe escapebeats was narrow (60 to 80 msec), and a His-bun-dle potential preceded the QRS complex, indicat-

ing complete block in the AV node. One dog (no.3) had a transient CAVB six times. In this dog, thetime of block was late (after 40 sec), and therewere no AV conduction changes observed duringcold mapping. The final re.sult was pennanent rigbtbundle branch block (RBBB). In tbe remainingdog (no. 6), only transient AV block was obtained.In this dog, the time of block also was late (range40 to 50 sec), and no AV conduction cbanges wereobserved during cold mapping (Table 1). Therewere no complications observed during or fol-lowing the procedure.

RFAhlaHon

In both dogs, a permanent CAVB was obtained,after one (dog I) or four (dog 2) applications at70°C for 60 seconds. An escape rhythm of 50 to55 beats/min with namjw QRS complexes (80 msec)was observed. A His-bundle potential precededthe QRS complex, indicating an AV ncxlal origin ofthe block. The temperature ranged from 60"C to65°C and the power ranged from 15W to 25 W.Tbere were no complications noted during RF ab-lation and during the follow-up period.

Tissue Characteristics

Cryoablation

Macroscopic findings. After 6 weeks ofcryoablation, all animals showed a discrete, oval-shaped lesion above or along the septal leaflet ofthe tricuspid valve. The lesion areas ranged in sizefrom 2 cm X 1.8 cm to 0.3 cm X 0.3 cm (mean1.22 ± 0.66 cm X 1.10 ± 0.68 cm). Adjacentregions of the septal leaflet of tbe tricuspid valve

1058 Journal of Cardiovascular Electrophysiulogy Vol. 9, No. 10, October 1998

Ill

AVR

AVL

AVF

HBE

TIME (sec)DEGREES (°c)

AH = 70 ms

PRE0

AH = 150 ms

COLO MAPPING5

-15

COLD MAPPING

-15

CRYOTHEJIMIA300-70

Figure 2. Pte 0 = endocardial recordings at precryoablation site. Note that the atnplitude of the atrium is bigger than that ofthe His recording. AH = 70 tns. Cold trapping 5-15 = application of 5 .seconds of cryothermia at -I5''C. Note the prolonga-tion of the AH ititerval frotn 70 ms {ba.tal) to ISO ms. Cold mapping 15-15 = effect of 15 seconds of cryothermia at -I5°C.Note the presences of 2:1 AV block. At this moment of application, note the disappearance of the His-bundle electrogramrecordings due to ice formation on the tip of the catheter. Ctyothermia = end of 5-minute application of cryothermia at-70"C. Note the appearance of cotnplete A V block with an e.tcape rhythm of 55 beats/min. The width of the QRS duting theescape rhythm is 60 ms.

were somewbat thickened in two animals. TTierewas no evidence of damage of the tricuspid valve,aortic valve, t>r mitral valve. In dogs 1, 2, 4, and 5(with CAVB), tbe lesions were observed in theregion of tbe compact node. In dog 3 (with RBBB),the lesion was observed in the apex of tbe triangleof Kcx:h. In the remaining dog (with transient AVblock), the lesion was observed in the base of thetriangle of Koch (more on tbe atrial aspect).

system (atriutn) to destroy tbe compact node. Theseptal leaflet of the tricuspid valve was free in allanimals. Tbe mitral and aortic valves were notaffected by cryoablation.

RF ablation

Macroscopic findings. The lesion area rangedfrom 2.7 cm X 2.5 cm to 2.0 cm X 2.5 cm. Tbe

Histologic findings. Tbe histology findings aresummarized in Table 1. In general, the ablativelesions were well delineated, discrete, and showedbomogeneous den.se fibrous tissue without viablemyocardium interspersed (Fig. 3A). These lesionsinvolved tbe compact AV node area and weretransmural (dogs I, 2, 4, 5). In dog 3, in wbiehonly pennanent RBBB was obtained, tbe lesionwas localized in the penetrating bundles. In dog 6,in which a transient AV block was obtained, thelesion was created too high in the conduction

No.

I23456

TABLE 1Microscopic Dimensions of Cryoablaiion

Result

CAVBCAVBRBBBCAVBCAVBTrCAVB

Length(cm)

0.70.80.50.6O.«0.8

VVidlli(cm)

0.60.20.70.30.40.4

Lesions

Depth

TMTMTMTMTMTM

CAVB = complete .\\ block; RBBB = right biiiiLlie branch bloclc;TM = ImnMnLiral; Tr = transient.

Rodriguez, et al. Cryoablaiion of (he Proximal A V Node 1059

Figure 3. L^: Chronic lesion 6 weeks after cryoablation in dog I fcryo sgroup). stained with Azan. Note the homogeneity offt-hrotic tissue at the ablation site extending from the right atrial wall across the central filmms body to encroach upon the muscularintcrventricutar septum. Inset: Higher tnagntftcation phototnicrogntphs of principal lusti>togic findings (Azan stain): homogeneousfibrotic ti.s.stic without .stratids of viable myocardium. Note the well-delineated border between fibrotic and tionnal myocardium.Blue color indicates fibrotic ti.ssue atid red color indicates normal ti.ssue. Right: Chronic lesion 6 weeks after radiofrequency (RF)ahlation in dog / f RF group), stained with Azan. Note the presence of inhomogeneous fibrous tissue with scattered .strands of vi-able myocardium within the .scar tis.sue and cartilage formation. Inset: Higher magnificatum phouunirrographs of principal histo-logic finditigs (Azan stain). Note that the borders of the le.sion are not well delineated. The le.sitm contains inhomogeneous fibrotictissue with strands of viable myocytes and cartilage fotmation. Blue color indicates fibrotic tissue and ted color indicates nonnalti.s.sue. Ao = aorta; C = cartilage: IVS = inteirentricular .septutn: N = compact A V node; RA = right atrium: TV = tricuspid valve.

lesion was localized to the region of the AV nodeand did not affect the tricuspid or aortic valves.

Histologic findings. The lengths of the lesionarea created using RF were larger in comparisonwith tbose created using cryolesions. The lesionareas ranged from 1.8 cm X 0.5 cm to 1.5 cm X

0.5 cm (Table 2). All lesions were transmural andhad a triangular shape (base on the right side andapex on the left side). The lesions showedinhomogeneous dense fibrous tissue withscattered strands of viable myocardium within thescar tissue and cartilage formation (Fig. 3B).

TABLE 2Microscopic Dimensions ot Radiofrequency

Nil. Result

I CAVB2 CAVB

Length(cm)

\-51.8

Width(cm)

0.50.5

Lesions

Depth

TMTM

CAVB = complulc AV block: TM = transmural.

Discussion

Main Findings

Cryoablation

Tbis study shows that cryoablation of the AVnode with the use of a new bipolar steerablecatheter is effective in creating chronic CAVB. Pre-

1060 Journal of Cardiovascular Electrophysiologj Vol. 9, No. 10, October 1998

dictors for proximal permanent CAVB were theoccurrence of AV conduction abnormalities dur-ing cold mapping and tbe occurrence of completeblock within 20 seconds during cryoablation at-70'^C. Adjacent regions of tbe septal leaflet of thetricuspid valve or the tricuspid valve itself werenot damaged nor was the aortic or the mitral valve.The bistopathologic lesions were discrete andshowed homogeneous, dense fibrous tissue with-out viable myocardium. Tbese lesions weretransmural in all dogs, showing more or less thesame distance fix)ni the right and left septum. Wbena CAVB was obtained, the lesions were localizedin the compact AV node.

Previous Studies

Tbe first study on tbe feasibility of cooling toprcxluce reversible block of the specialized cardiactissues in dogs was reported by Lister and Hoff-man.'' In that study, a silver "U" bent tubing 4 mmin length was sutured in the atrial septum in tberegion of the His bundle. Passage of the coolingmixture (alcohol-carbon dioxide mixture), initiallyat -10° to -ZO^C, inhibited impulse forinafion oftbe sinoatrial node. To suppress AV conduction,the cooling niixure was cooled to temperatures aslow as -45°C. By progressive cooling, ECGchanges such as prolongation of tbe PR intervalwere observed. When tbe cooling was maintainedat -45°C, a stable 5:1 block was observed. Almostimmediately after stopping cooling, AV conduc-tion resumed to the state prior to cooling.'^

Transvenous cryoablation of the His bundle wasattempted 7 years ago by Gillette et al.'̂ Using an1 i -French nonsteerable catheter not containing elec-trodes, CAVB was obtained in 4 of 5 miniatureswines. However, CAVB lasted for only 1 hour.Tbe feasibility of cold mapping was not studied byGillette et al.'' The tecbnique of these autbors con-sisted of one application at -60°C for 3 minutes.The macroscopic acute lesions consisted of a cir-cular, red-purple area 0.8 to I cm in diameter lo-calized at tbe apex of the triangle of Koch. No grossdamage of tbe tricuspid valve was visible. Histcvlogically, these acute lesions were cbaracterized byextensive extravasation of blood cells and coagu-lation necrosis of myofibrils, sharply demarcatedfrom adjacent nonaffected cardiomyocytes. Thesame group'" reported on tbe histology of chroniceryolesions. In that study, cryothermia (-60°C) wasgiven three times for 3 minutes each. Six weekspost ablation, the hearts were analyzed. Macro-scopically. the lesions were similar to the lesions

reported in our study. Histologically, lesions wereobserved that wens well circumscribed, having densefibrous tissue with a stnall number ol" chronic cel-lular infiltrates, involving the compact node and thepenetrating bundle to a depth of 2 to 3 mm in thoseanimals with successful pennanent CAVB. The le-sions involved only the atrial tissue above the con-duction system in those hearts that resutned AVconduction. In our study, lesions were transtnural.The discrepancy between those and our findingscan be explained by the differences in tecbniqttcsand catheters used. In humans, chronic lesions pro-duced by surgical cryoablation of the His bundlealso showed fibrotic atid homogeneous lesions withsharp borders of demarcation between nonnal andinjured tissue." Cryoablation in the normothermicbeating heart in patients with acces.sory path-ways"*̂ ' bas sbown good results. The cryolesions inthose studies were described to be sharply iind welldemarcated, homogeneous, and not arrhythmo-genic.^^ Interestingly, chronic cryolesions in pa-tients with accessoi-y pathways did not produce rup-ture or fonnation of an aneurysm,'''' even when cry-olesions were applied in the coronajy sinus or tothe thin atrial wall.-'̂ '̂ These aspects suggest a pos-sible advantage of cryoablation over other abla-five techniques.

Size of the Cryolesions

Animal studies have shown that the size of thecryoablation lesion depends on four factors: (I) thetemperature of the cryoprobe, (2) the diameter ofthe cryoprobe head, (3) tbe exposure time, and (4)myocardial temperature.'^ '̂ Small lesions are ob-tained using probes with a 4-mm head, cryother-mia (-10°C) for 3 minutes, and a nonnal myocar-dial temperature.'^^^ The reduction in lesion di-ameter results from thawing of a marginal zone ofmyocardium in which the temperature has achievedthe freezing point but not -21°C (tbe intracellulartemperature required to produce permanent de-struction)."' The size of this rim dcjjends on the sur-rounding myocardial temperature rather than on thecryoprobe temperature.'^ The lesions may be en-larged by using lower tetnperature (-196°C), pre-cooling the beart to 6° to 12"C before applying thecryoprobe, cryothemiic exposure of 4 to 6 minutes,and using larger probes with circular heads."'''

Comparison Between Cryogenic and RF lesions

Histopathologic effects of RF ablation on ani-mal my(.K;ardium include surface pitting, granulation.

Rodriguez, et al. Cryoablation of the Proximal A V Node 1061

cartilage formation, and chronic inflammation as seenat 2 months after ablationj^'** in contrast to the uni-form fibro.sis observed with cryolesions.''"-'^''^ Al-though RF ablation was performed in only two dogs,Ihe lesions of both dogs showed inhomogeneousdense fibrous tissue with scattered strands of viablemyocardium within the .scar tissue. Also., as previ-ously reported, the lesions contained cartilage for-mation.'''" The shapes of these lesions were trian-gular, with the base in the right septum and the apexin the left septum. Theoretically, this type of lesioncould be a potential arrhythmogenic substrate, un-like the homogeneous and sharp borders of demar-cation between nonnal and injured tissue producedby cryogenic lesions. Furtiiermore, in view of thelack of an inflammatory component of the cryole-sions, this technique could be used to create lesionsaround the pulmonary veins for ablation of focalatrial tachycardia and atrial fibrillation.

Cryoablation has some limitations, including alower procedural efficacy, one of the reasons be-ing the small .size of tbe cryolesion. Even thoughtwo applications of 5 minutes each were given, thecooling effects were attenuated by circulating warmbkxid. However, these limitations Ciin be overcomein the future by designing new catheters. The newcatheter design should (1) include a new configu-ration and shape of the tip. (2) be able to have bet-ter energy transport, (3) create a lower tissuetemperature in a larger area, and (4) improve tbecooling and thawing rate. All these parameters willaffect the size of the cryolesion.

Conclusions

(1) Permanent CAVB ean be created by using abipolar steerable cryoablation catheter (2) Histolog-icalJy, cryoablated sites are discrete and show fibrotictissue (transmural) without signs of chronic inflam-mation and no evidence of viable cells. (3) Lesionscreated with RF were less homogenous and still con-tained viable myocytes within tbe lesion. (4) Thistechnology by using reversible cold mapping has thepotential to identify the successful ablation site iuidwiirrants furtiier clinical .study. (5) The iirrhythmo-genie significance of differences between cryolesionsand RF lesions and the possible iuJvantages of cryoab-lation require further study in the buman heart.

References

1. Jackman W. Wang X, Friday KJ. et al: Catheter abla-iion of accessory pathways (Wolff-Parkinson-WhiteSyndrome) by radiofrequency current. N Engi J Med1991 ;324:1605-1611.

2. Kuck HK. Schlutter M, Geiger M. el al: Radiofre-quency current catheter ablation of accessory path-ways. Lancet i 991 ;337:1557-1561.

3. Gallagher JJ, Sealy WC. Anderson W. et al: Cryosurgi-cai ablation of accessory atriovenlricukir connections.A method for correction of the preexcitaiion syndrome.Circulation 1977:55:471-478.

4. Camm J, Ward DE, Spurrell RAJ, et al: Cryttthermalmapping and cryoablation in treatment of refractorycardiac arrhythmias. Circulation l9S{);62:67-74.

5. Guiraudon GM, Klein GJ, Gulamhusein S, et al: Surgicalrepair of the Wolff-Parkinson-White syndrome. A newclosed-heart technique. Ann Thorac Surg I984;37:67-7I.

6. Guiraudon GM, Klein GJ, Sharma AD. et al: Closed-heart technique for the Woift-Parkinst)n-White syn-drome. Further experience and potential limitations.Ann Thorac Surg 1986;42:65l-657.

7. Klein GJ. Sealy WC. Pritchett ELC. et al: Cryosurgicalablation of the atrioventricular node-His bundle: Long-term follow-up and properties of the junctional pace-maker. Circulation I98O;6I:8-I2.

8. Cox JL, Holnian WL, Cain ME: Cryosurgical treat-ment of atrioventricular node reentrant tachycardia.Circulation 1987:76:1329-1336.

9. Gillette PC. Swindle M. Thompson RP. et al: Transve-nous cryoablation of the bundle of His. PACE 1991;14:504-510.

10. Fujino H, Thompson RP. Germroth PG. et ai: Histo-logic study of chronic catheter cryoablation oi' alrio-ventricular conduction in swine. Am Heart J1993,125:1632-1637.

11. Ohkawa S, Hackel DB. Mikat EM, et al: Anatomic ef-fects of cryoablation of the atrioventricular conductionsystem. Circulation 1982;65:l 155-1 i62.

12. Lister JW, Hoffman BF: Reversible cold block nf thespecialized cardiac tissues of tbe unanesthetized dog.Science 1964:145:723-725.

13. Holman W], Ikeshita M. Douglas JM, et al: Cardiaccryosurgery: Effects of myocardiai temperature on cry-oiesion size. Ann Thorac Surg 1983:35:386-393,

14. Hunt GB. Chard RB. Johnson DC. et al: Comparisonof eariy and late dimensions and arrhythmogenicity ofcryoiesions in the normothermic canine heart. J ThoracCardiovasc Surg i 989;97:313-318.

15. Peiffert B, Feldman L. Villemot JP, et al: Cry-ochirurgie des tachycardies ventHcuiaires. ChirurgieI992;i 18:137-143.

16. Le Pivert PJ: Basic considerations of the cryolesion. InAhlin RJ, ed: Handbook of Cryosurgery: Science andPractice of Surgery I. Marcel Dekker. New York,1979. p. 15.

17. Huang SK. Bharati S, Lev M. et al: Eiectrophysiologicand histologic observations of chronic atrioventricularblock induced by closed-chest catheter desiccationwith radiofrequency energy. FACE 1987:10:805-816.

18. Huang SK. Bahari S. Graham AR. et al: Closed chestcatheter desiccation of the atrioventricular junction us-ing radiofrequency energy. A new method of catheterablation. J Am Coll Cardiol 1987:9:349-358.