The value of catheter mapping during sinus rhythm to localize site of origin

ELECTROPHYSIOLOGY

The value of catheter mapping during

sinus rhythm

to

localize site of origin

of

ventricular

tachycardia

DENNIS M. CASSIDY, M.D., JOSEPH A. VASSALLO, M.D., ALFRED E. BUXTON, M.D.,

JOHN U. DOHERTY, M.D., FRANCIS E. MARCHLINSKI, M.D., ANDMARK E. JOSEPHSON, M.D.

ABSTRACT We assessed the value of endocardial catheter mapping in 52 patients in sinus rhythm and with 102morphologically distinct ventricular tachycardias. The local bipolar electrograms from various regions of the left ventricle were assessed and quantitatively classified with respect to the

characteristicsof amplitude and duration. With theuseof thisassessmentwefound that electrograms

from the site of originwere ofsignificantly lower amplitude and longer duration; however, because

suchanoverlapoccurred with electrograms thatwerenotfrom sitesof origin, this doesnotserve as a

useful clinical marker. Various types of electrograms, including normal, abnormal, fractionated, abnormal late,fractionated late, and longest,wereevaluated withrespect tosensitivity, specificity,and

positivepredictive value. None of thesetypespossessed the abilitytoreliably localize the site of origin of ventricular tachycardia. We therefore conclude that

endocardial

catheter mapping during sinus

rhythm isnotusefulas aguide in localized surgical therapy of ventricular tachycardia. Surgery guided

only by the results of mapping during sinus rhythm would result ina moreextensiveexcision than that directed by maps obtained during ventricular tachycardia and in some cases would result in the

exclusion ofthe area considered to be the site oforigin of the tachycardia. Circulation 69, No. 6, 1103-1110, 1984.

THE MANAGEMENT of ventricular tachycardia has been improved with the application of new surgical techniques.1Paramount to the success of surgery is the

abilitytolocalize the siteoforigin of ventricular tachy-cardia by endocardial mapping procedures.2, 3

Map-ping

during ventricular tachycardia has been accom-plished both before surgery in the catheterization laboratoryandduring surgery. It is not, however, pos-sibleto map allmorphologically distinct tachycardias inall patients. This is related either to an inability to

FromtheClinicalElectrophysiologyLaboratory,Hospitalof the Uni-versity of Pennsylvania,CardiovascularSection, Department of

Medi-cine, UniversityofPennsylvaniaSchool ofMedicine, Philadelphia.

Supported in partbygrantsfrom The American.HeartAssociation,

SoutheasternPennsylvaniaChapter,Philadelphia;grant Nos.HL00361 andHL24278from the NationalHeart,LungandBloodInstitute, Be-thesda;andbyTheFannie E. Ripple Foundation,Morristown, NJ.

Address forcorrespondence: MarkE.Josephson,M.D.,Chief, Car-diovascularSection, Room 658 RavdinBuilding, Hospitalof the

Uni-versity ofPennsylvania, 3400Spruce St.,Philadelphia, PA19104. Received Nov. 4, 1983; revisionacceptedMarch8, 1984.

Dr.Cassidyissupported byaCanadian Heart Foundation Research

Fellowship, Ottawa.

Drs. Buxton and Marchlinski are supported by the University of PennsylvaniaMeasey Foundation.

Dr.Dohertyissupported byagrant-in-aidfromthe AmericanHeart

Association, Southeastern Pennsylvania Chapter.

Dr. Josephson is the Robinette Foundation Associate Professor of Medicine(CardiovascularDiseases).

induceventriculartachycardia or aninducible ventric-ulartachycardia that is nottolerated hemodynamical-ly. This limitation could be overcome if endocardial maps obtainedduringsinusrhythmcould supply infor-mation that would be ofpredictivevalue in localizing the site of origin of ventriculartachycardia. Although such an approach has been

suggested,4'

no data has been available to test the usefulness of such tech-niques. Ourstudywastherefore undertaken to critical-ly evaluate the role of endocardial mapping during sinusrhythmtolocalize the site of

origin

of ventricular tachycardia. The specific aims were (1) to quantita-tively assess the

amplitude

and duration of local left ventricular endocardialelectrogramsinapopulationof patients with coronary artery disease and ventricular tachycardias, and(2)todetermine if theseorany other qualitative electrographic property

possessed

charac-teristics that could localize the site of

origin

of ventric-ular tachycardia.

Methods

Patientpopulation.Fifty-twopatientsreferredtothe

hospi-tal of the University ofPennsylvania for electrophysiologic

evaluation ofventricular tachycardiaunderwent sinusrhythm

endocardialcathetermapping duringthecourseof their studies.

There were45 men and seven women with amean age of 58 years (range 34 to 75).

Allpatients had coronary artery disease andprevious histo-ries ofmyocardial infarction. Theelectrocardiographic pattern

ofinfarction wasanterior in38, inferior ineight, and mixed in

six. Inallthe patientscoronary arterydisease was assessedby

coronary angiography and left ventriculography. All patients

had left ventricular aneurysms and reduced left ventricular

func-tionasdeterminedbythefindingofdepressed ejectionfraction

(mean29%, range 12%to45%). All patients had histories and documentation ofrecurrent sustained ventricular tachycardia.

Electrophysiologic study. Studies were performed in pa-tientsinnonsedated,postabsorptivestateafterinformed written consent had been obtained. Thirty-three patients were not on anyantiarrhythmic therapy,while theremaining 19wereonone

antiarrhythmic drug each. The patients who were receiving

drugsunderwentmappingbothwhile in sinusrhythmand

dur-ing ventricular tachycardia while on the same antiarrhythmic

drug. Wheneverpossiblethe twomapswereobtained with the catheteratthesamesite;thiswasaccomplished by stoppingand

starting the ventricular tachycardia. In otherpatients the map

during sinus rhythm was obtained first, followed by the map

duringventriculartachycardia.Ifmorethanone

morphological-ly distinct ventricular tachycardia was induced it was also

mapped.

Onequadripolarcatheter(No. 6FUSCI)wasinserted

percu-taneouslyintothefemoral artery and advancedtotheleft

ventri-cle of each patient under fluoroscopic guidance. One to two

quadripolar catheters were inserted percutaneously and were

advanced to the right ventricular apex and right ventricular outflow tract. Catheters had a5 mm interelectrode distance.

Heparin,5000unitsas abolus,followedby 1000units/hr,was

administeredafterinsertionofthe arterialcatheter in each ofthe

patients.

The left ventricular mapping scheme used is illustratedin

figure 1.The 12 sitesrepresentsegmentalareasof the heart of

approximately

5 to 10 cm2. Eightto 32electrograms were

re-corded in eachpatient. The catheter sites were independently verified by multiple-planefluoroscopy in the presence oftwo

experienced

physiciansfamiliar with thismappingscheme. The

catheterwas repositionedtoanothersitetoconfirm

reproduc-ibility

ofrecordingsandonlyminimalmorphologicdifferences between localelectrogramswere noted. Stability wasensured

by

recording from each site for a minimum of 5 to 30 sec.

Electrogramswererecordedwiththeuseofa 10mm

interelec-trodedistance. Theintracardiacelectrogramswererecordedat

variablegainto achieve the bestelectrographicdefinition and were accompanied by a 1 mV calibration signal. A 10 mm

bipolar fixed-gainsignalwasalso recordedata1cm/mV ampli-ficationateachsite.Electrogramswerefilteredat30to500 Hz.

Intracardiac recordings were simultaneously displayed on a

multichanneloscilloscope(Electronicsfor Medicine VR 16) and

were stored on analog magnetic tape (Honeywell 5600) and

recordedona16-channelMingograf(SiemensElema)at apaper

speedof 200mm/sec.

Definitions. Electrographic amplitude (inmV) was defined

asthepeak-to-peakdeflectionmeasured in the 10 mm

variable-gain

bipolar electrogram.

Electrographic duration (in msec) was defined as the time

fromthe earliest electrical activitythatdeviatedfrom a stable baseline totheonset of theamplification signaldecay artifact

(anartifactcausedbyelectronicdecayofanamplifiedfiltered

signal)measured inthe 10mmfixed-gain bipolarelectrogram

(figure 2).Theamplificationanddurationmeasurementswere

combined to give an amplitude/duration ratio to allow equal

emphasistobeplacedoneach ofthese values.

Local activation timeat asitewasdefinedasthetime fromthe

FIGURE 1. Schema of mapping sites in left ventricle (see textfor

discussion).

onsetof the surface QRS to the time at whichthelargestrapid

deflection of the local electrogramcrossed the baseline. This

measurementwasobtained from the 10 mmvariable-gain

bipo-larelectrogram.

Electrogramsweredefined as normal basedonthe

measure-mentsobtained from 112electrograms inpatients with normal

left ventricles. Thenormal values representthose from95%of

allelectrogramsobtainedin thesepatients.6Normalamplitude

wasdefined as one of 3 mV or more and normal duration was definedasoneof 70 msec orless. The normal

amplitude/dura-tion ratio was defined as 0.046 or greater.In casesinwhichonly

oneof theamplitudeordurationmeasurements wereabnormal,

theamplitude/duration ratio was used to definenormality. Electrogramsfailing tosatisfytheabovecriteriafornormality

weredefined asabnormal. Basic statistical methods werethen

usedtodefine the mean and SDvaluesfor amplitude,duration,

andamplitude/durationratio for theelectrograms of this

popula-tion.

Theelectrogramsthat werethe most abnormal were defined

asfractionated and werecharacterized by low amplitude,high frequency, multicomponent signalsof prolonged duration.This

groupwasquantitatively defined asconsisting of those abnor-mal electrograms that weregreater than 1 SD from the mean values of theabnormalelectrograms as a group. Again, ifonly one parameter of amplitude or duration met the criteria for

fractionation, then the amplitude/duration ratio was used to define fractionation.

Lateelectrogramsweredefined as those in which a compo-nentwasrecorded after theinscriptionof thelatest surface QRS

activity.Thiswasatimingcharacteristic andhence it could be

applied to any of the three types of electrograms described above.

CIRCULATION

TIME

FIGURE 2. Three surface electrocardiographic recordings (I, aVF, and

V,)accompanied by three local bipolar electrograms (normal, abnor-mal, and fractionated and late) recorded from different left ventricular endocardial sites. Thedashed vertical line denotes the end of surface QRS activity. Thearrowsshow theonsetand offset (characterized by theamplification signal decay artifact) of local electrical activity. Each electrogram isaccompanied by 1 mV calibration signals.

The longest electrogram recorded in each patientwas also

analyzed and was defined as the electrogram of the greatest duration.

The site oforigin of ventriculartachycardia wasarbitrarily defined as the earliestrecordedelectrical activity occurring in

late diastolebefore the surface QRS duringventricular tachycar-dia. Thiswasdetermined by catheter and/or operative mapping.

Whenaventriculartachycardiawaslocalized betweentwo stan-dardsites, the electrogram from each of these siteswas consid-ered in theanalysis of site of origin. Induction of tachycardia

wasperformed by standard previously described methods.7 All

other sites wereconsideredtobe non-sites oforigin.

Dataanalysis. Electrographic amplitudes, durations, and ac-tivation times wereassessed ateach siteby atleasttwo inde-pendent observers. A minimum of fourtosixcomplexes were assessedat each site. The datawere analyzed with theuse of

everyelectrogram recorded but, becausewebelievedthat this

might introduce asampling bias, the datawere also evaluated

with theuseof themeanof all theelectrogramsatasitewhen

more than one electrogram was obtained at agiven site. An electrographic signal thatcamefrombetweentwostandardsites

wasdesignatedtoonesitebasedonits localactivation time. By this technique a maximum of 12 electrograms from the left ventricle could beobtained. It is themeandatathatare present-ed since analysis ofboth methodsyielded similar results.

All resultsareexpressedintermsofmean + SD. Student'st

testforunpaireddatawasusedwhencomparing electrographic

characteristics from sites oforigin andthosefrom areas that

werenotsites oforigin. Calculation of sensitivity, specificity, andpositive predictive valueweredonebystandardmethods,as

summarized in table 1.

Results

General. The 52 patients exhibited 102

morpholog-ically

distinctventricular

tachycardias

for

analysis:

20

patients had only one morphology, 17 had two mor-phologies, 13 had three mormor-phologies, one had four morphologies, and one patient had five ventricular tachycardia morphologies. The total number of elec-trograms available foranalysis was 546 (10.5 per pa-tient). Of these

electrograms,

102 were from sites of origin while the remaining 444 were from non-sites of origin. The number of electrograms that met the nor-malcriteria was 234 (42%); the remaining 312 (58%) electrograms were considered abnormal.

Abnormal electrograms - quantitative analysis. The

mean abnormalelectrographic amplitude was 1.4 ± 0.9 mVand the mean abnormal electrogram duration was93 ± 40 msec.The mean abnormal electrographic amplitude/duration ratio was 0.017 ± 0.012.

Thegeneration of these data allowedaquantitative definition of fractionated electrograms. The quantita-tive criteria used to define fractionated electrograms basedonvalues 1 SDfrom the mean becamean ampli-tudeof 0.5 mVorless,aduration of 133msec ormore, and/or an amplitude/duration ratio 0.005 or less.

Classification ofelectrograms. A distribution of each type of electrogram is seen in table 2. Of the 312 abnormal electrograms, 47 were fractionated by the described criteria. This represents 9% of all electro-grams. Eighty of the abnormalelectrogramswere clas-sifiedaslate,representing 15% of allelectrogramsand 26 (5%) were considered both fractionated and late. Examples of the various types of electrograms are

shown in figure 2.

Analysis of sites oforigin. Results of the quantitative comparative analysis ofsite oforigin andnon-site of originisshown in table 3. The sites of

origin

had lower amplitudes (2.0 ± 2 vs 3.6 + 2.8 mV; p < .001), longer durations (88 ± 33 vs 75 ± 27 msec; p <

TABLE 1

Definitions ofsensitivity,specificity, andpositivepredictive value Truepositive ElectrographictypeAatsiteoforigin

Falsepositive Electrogramtype,notsite oforigin

Truenegative - Nonelectrogramtype,Bnotsiteoforigin

Falsenegative - Nonelectrogramtypeatsite oforigin

Truepositive Sensitivity =

Truepositive + Falsenegative

Truenegative Specificity =

Truenegative + Falsepositive Truepositive

Positivepredictivevalue =

--True positive + False positive

ADefinedbyoneof thefollowing: normal,abnormal, fractionated,

abnormal late, fractionatedlate, orlongest.

BDefinedbyallelectrograms notexhibitingparticularcharacteristic beinganalyzed.

TABLE 2

Distribution of the 546electrograms

Total (%) Site of Non-site of all

Electrogram type origin oforigin electrograms Normal 14 220 234 (42) Abnormal 88 224 312 (58) Fractionated 10 37 47 (9) Abnormallate 26 54 80 (15) Fractionated late 8 18 26 (5) Longest 16 36 52 (10)

.001), and loweramplitude/duration ratios (0.027 ± 0.03 vs 0.058 + 0.055;p < .001) than did the

non-sites oforigin. Although all values reached statistical significance, much overlap between the groups was

noted.

Value of electrographic types.Thesensitivity, specific-ity, andpositive predictivevalueof each type of elec-trogram is summarized in table 4.

Ofall the types, only abnormal electrograms pos-sessed goodsensitivity, at86%. Normal electrograms

were notsensitive, beingpresent14%of the time. Late abnormalelectrogramshadasensitivityof 26%. Like-wise,the longest electrogram recordedonly possessed

asensitivity for the siteoforiginofventricular tachy-cardia of 16%. Theleast sensitive type of electrogram

was the fractionated, whether or not it waslate (10% and 8%, respectively).

Thespecificityofeach type ofelectrogramwasalso examined. Both normal and abnormal electrograms had a specificity of 50%. The fractionated electro-grams had a 92% specificity. The characteristic of "lateness" also led to increased specificity, i.e., ab-normally late electrograms hadan88%specificity and fractionated late electrograms had a 96% specificity. The longestelectrogram recordedwas alsoquite spe-cific, at 92%.

Theabilitytopositively predict the site of origin of ventricular tachycardia was also examined for each electrographic subtype. The positive predictive value

TABLE 3

Comparison ofelectrogramsof siteoforigin with thosenotof siteof

origin

Site of Non-site

origin oforigin pvalue Amplitude (mV) 2.0±2A 3.6+2.8 <.001 Duration (msec) 88±33 75+27 <.001 Amplitude/duration ratio(mV/msec) 0.027±0.030 0.058±0.055 <.001 AMean + SD. 1106

of normal electrograms was 6%, and for abnormal electrograms it was 29%. Fractionated electrograms only predicted site of origin 21% of the time. The characteristic of lateness added little to thepredictive value; abnormal late electrograms had a 33% predic-tive value while fractionated late electrograms posi-tively predicted the origin oftachycardia 31% ofthe time. Similarly, the longest electrogram hadapositive predictive value of only 31%. Therefore, no electro-graphic type manifested a good positive predictive value.

Single ventricular tachycardia morphology analysis. A

separate analysis was performed on data from the 20 patients who had only one distinct ventricular tachy-cardia morphology. These results are summarized in table 5. Thedistributionand value ofeach ofthe types of electrograms were similar to those in the entire population. Abnormal electrograms werefairly sensi-tive (80%) but were not specific (65%) and were of poor positive predictive value (21%) for the site of origin ofventriculartachycardia. The only difference noted in this analysis was the lack of normal electro-grams at thesite oforigin of ventricular tachycardia. This mostlikelyreflects thefewer number of morpho-logicallydistinctventriculartachycardias analyzed (20 vs 82).

An analog map from one patient is shown in figure 3. This map shows the dispersionof various types of electrograms throughout the entire endocardial

sur-face, regardless ofsite oforigin.

Discussion

Mapping during ventricular tachycardia has been establishedtobeboth safe anduseful;however, it may be associated with several potential problems.8 The time required to map during ventricular tachycardia

can result in unfavorable hemodynamics. Some

ven-tricular tachycardias lead to rapid hemodynamic col-lapse andpatientswith thesetachycardiasarenot suit-able candidates for endocardialmapping. Occasionally tachycardias are totally noninducible, are of different morphologies than spontaneously documented ventric-ulartachycardias, orhaverapidly changing morpholo-giesthat make them unsuitable for activation mapping. Finally, operative mapping leads to increased cardio-pulmonary bypass time and its inherent problems. It wouldtherefore be beneficial to be able to predict the site oforigin of ventricular tachycardia from mapping during sinus rhythm.

It has been demonstrated with both epicardial and endocardialmapping during sinus rhythm that patients withrecurrentsustainedventricular tachycardia have a

TABLE 4

Sensitivity, specificity, and positive predictive value of each electrogram type(entire population)

Positive predictive Specificity (for site value (for site of Sensitivity other than site origin of Electrogram type (forsite of origin) oforigin) electrogram type) Normal 14/102 (14%) 224/444 (50%) 14/234 (6%) Abnormal 88/102 (86%) 220/444 (50%) 88/312 (29%) Fractionated 10/102 (10%) 407/444 (92%) 10/47 (21%) Abnormal late 26/102 (26%) 390/444 (88%) 26/80 (33%) Fractionated late 8/102 (8%) 426/444 (96%) 8/26 (31%) Longest 11/102 (16%) 408/444 (92%) 16/52 (31%)

variety of abnormal electrographic findings that are not present in patients without this arrhythmia.4' 9 This

may well define an electrophysiologic substrate for ventricular tachycardia. However, the mere presence of suchelectrograms does not allow the conclusion that mapping during sinus rhythm can be used to guide surgical therapy of ventricular tachycardia. This hypothesiscanonlybe proved if such electrograms are shown to besensitive, specificand, mostimportantly, to possess positive predictive value for the site of origin of ventricular tachycardia. The goal of this study was to evaluate the hypothesis that endocardial map-ping during sinusrhythm inpatientsis usefulin guid-ing localized surgical therapy of ventricular tachy-cardia.

Classification of electrograms. The literature dealing with mapping techniques is confusing with regard to electrographic descriptions. The confusion has arisen because of a lack ofa standardization with regardto

amplifierandfiltersettingsandinterelectrode distance ofrecording apparatus and mostimportantly because ofa lack of

quantitative electrographic

description

of characteristics of amplitude and duration. This last problemcanbe

partially

attributedtothe fact thatthere has beennoquantitativedescriptionofbipolar electro-graphic characteristics ofnormal left ventricles.

In a prior study6 wedefined quantitative standards fornormalbipolar electrograms basedon thefindings in 10 subjects with normal left ventricles from whom 112 bipolar electrograms were obtained for analysis. Thedescription of abnormal electrograms is the result ofcomparison with this data base.

Of the electrograms recorded in patients in sinus rhythm, the ones that are fractionated are the most abnormal. Only recently have microelectrode tech-niques been used in attempts to define fractionated activity.'0Thequalitative descriptionthatis in general agreementwithresults of clinicalmappingis thatof an electrogram withhigh frequency,lowamplitude, mul-ticomponent signalsofprolongedduration. The quan-titative descriptions in the literature, however, have varied from none to amplitudes less than 1 mV and durations greater than 50

msec,

11-1 without anydetails about howthese values wereobtained. Our definition offractionation encompasses all previous definitions andis in fact morerigid. Webelieved that the signifi-canceofthistypeofelectrogramwouldbebestjudged by analyzingthe mostabnormal(morethan 1 SD from themean) ofthe abnormalelectrograms. To makesure

thesecriteriawerenot toorigidabnormalelectrograms were also analyzed as a group.

Electrograms having late activation or exhibiting

TABLE 5

Analysisofelectrographic distribution,sensitivity,specificity,andpositivepredictivevalue inpatientswithone

morpho-logically distinct ventriculartachycardia

Positivepredictive Specificity (forsite value (forsite of

Sensitivity other than site originof

Electrogramtype Number(%) (for site oforigin) oforigin) electrogram type)

Normal 97 (49%) 0/20 (0%) 80/177 (45%) 0/97 (0%) Abnormal 78 (51%) 16/20 (80%) 115/177 (65%) 16/78 (21%) Fractionated 22 (11%) 4/20 (20%) 159/177 (90%) 4/22 (18%) Abnormal late 13 (7%) 4/20 (20%) 168/177 (95%) 4/13 (31%) Fractionated late 12 (6%) 2/20 (10%) 167/177 (94%) 2/12 (17%) Longest 20 (10%) 2/20 (10%) 159/177 (90%) 2120 (10%) 6,

NSR 1 aVF VJ RVA 1 ,- FL 2- .A AL 3 g AA 5 ~~~~~N 6 it N 7 t AL 9 NN 9 ~~~~~~~FL VT 1 oVF VA RVA h+ 8 A

10-

10

11 11 12 12 TIME. tIME

FIGURE3.Analog map fromapatientin sinusrhythm(left)andduring

ventriculartachycardia (right). Three surface leads(I, aVF,andVj)are

accompanied byonerightventricularapical (RVA) recordingand 12(1 to 12) left ventricular endocardialrecordings.Time line isonthe

bot-tom.Thedotted vertical line denotes theonsetof surfaceQRS activity.

Theasteriskatsite 2during sinusrhythmdenotes the site oforiginof ventricular tachycardia(arrow, site 2ventriculartachycardia).The

let-ters at theend ofthe sinusrhythmelectrograms denote the typeof

electrogram (N= normal;A = abnormal;F = fractionated;and L = late).

latepotentialsarebelievedbysomeinvestigatorstobe

significant.5'

16-1 Previous work from our laboratory has demonstratedthat normal endocardialbipolar elec-trograms do not extend beyond the first 50% of the

surface

QRS complex.6Wetherefore electedtodefine

as late any electrogram thatextendedbeyond the

sur-face QRS complex. Againthe intentwas tobest deter-mine the significance of this finding.

The last type ofelectrogramanalyzed was the long-est electrogram recorded by endocardial mapping. Long electrograms may represent slow conduction, which is one of the requirements for reentry and reentry is presumed to be the mechanism of sustained ventricular tachycardia. It is conceivable that the site oforiginexists near this area, but an obvious limitation exists in theuseof thiselectrographiccharacteristic in patientswho havetachycardias ofmorethanone

mor-phologic type originating from nonadjacent sites.

1108

Anotherelectrogram reported to be auseful marker for ventricular tachycardia is the "split" or "double potential"5 electrogram.3 1'17,18 This electrographic finding was observed so rarely during endocardial cathetermapping that it was not suitable for use in the analysis. This may be a function of the interelectrode distance used in this investigation or of the number of sampling sites available.

Analysis of sites of origin. The electrograms from the sites oforigin were characterized by lower amplitude and longer duration. However, considerable overlap existedwith the other electrograms so that this finding alone would not be helpful in distinguishing between recordings from sites of origin and those from other sites.

The electrograms from the sites of origin were gen-erally abnormal (86%). This is in accord with what previous investigatorshave characterized the left ven-tricular electrograms in this patient population.4'9

Twofindings werestriking in the analysis of the site oforigin. First, a completely normal electrogram was obtained from the site oforigin 14% of the time (14 patients, 16 morphologically distinct ventricular tachycardias). Thisfindingisnotsurprising since

ven-triculartachycardia can arise from normal left ventri-cles. Electrograms recorded from these normal left ventricles shownodifference between the siteof origin

vs non-site oforigin.* The second important finding

was thatthe electrograms of the leftventricular endo-cardium thatwerenotfrom sitesof originwere abnor-mal 50% of the time. Thesetwofactshave to be con-sidered if theapplication oflocalized directed surgery is being considered.

Predictive value. Mapping in patients during sinus rhythm, if it istobe clinically useful, must be able to aid inthelocalization of the site oforigin of ventricular tachycardia. Although abnormal electrograms were

sensitive (86%) they lacked specificity (50%). The various other types of abnormal electrograms had higherspecificitybut very low sensitivity. None of the types ofelectrograms demonstrated a positive predic-tive value of greaterthan 33% and an attempt to com-bine thevarioustypeswould still fail to give a positive predictive value of greater than 40%.

Thereareanumber ofpossibleexplanationsfor the fact that the mapping, during sinus rhythm, of endo-cardial electrograms does not predict site oforigin of ventricular tachycardia. Our patient population had de-pressed left ventricularfunction (mean- ejection frac-tion of <30%) with diffuse mycocardial disease,

*Cassidy DM, etal:Unpublished observations.

which typifies patients with recurrent sustained ven-tricular tachycardia. The electrograms recorded, re-gardlessof type, may simply reflect this diffuse myo-cardial disease. The size and depth of the actual anatomical circuit has not been ascertained with the currently available techniques. Recordings obtained during endocardial catheter mapping may reflect activ-ity outside this circuit, thereby rendering such infor-mation less useful. Finally, conduction velocity and activation patterns during sinus rhythm at the recorded site of origin of ventricular tachycardia may not be linear, so that the assumption that such electrograms are a marker of the slow conduction necessary for reentry may not be valid.

Limitations. A major limitation of this study is related to the number of points sampled. Although a maxi-mum of 12 sites were used for this analysis, these values were derived from up to 32 samples in some patients. This, however, may still not be enough. It is conceivable that if a much larger sample could be obtained the results would be different. However, a limited number of points can besampled with endocar-dial catheter mapping during sinus rhythm, and it is this method that must be usedclinically. Also, prelimi-nary data obtained with intraoperative mapping under direct visualization support our findings.'3

A second limitation related tothetechnique of cath-eter mapping involves preciselocalization of the cath-eter position, which can be difficult. Eachlocation was confirmedin at least threefluoroscopic positions in the presence of two independent physicians experienced with theendocardial mapping scheme. When possible the catheter was maintained in the same position for recording ofthesinus rhythm andventricular tachycar-dia electrograms; this was accomplished by starting and stopping the tachycardia. Otherwise the catheter was respositioned each time to the area from which a recording had previously been made. An earlier study from this laboratory showed a good correlation be-tweenresults ofcatheter and operative mapping, sug-gesting reasonable reproducibility ofthis technique.'9 Finally, when more than oneelectrogram was obtained froman areathe average value was used in the analy-sis. We therefore believe that adherence to these guide-lines has minimized the chance of an error that could result from catheter mispositioning.

Itmight be argued that the criteria selected for some of the types of electrogram were

tbo

rigid, but if sig-nificant differences are to be found, they should be found with the use of the most abnormalelectrograms. Itshould benoted that ifintermediate criteria had been chosen some gain in sensitivity would have resulted,

Vol.

but at the expenseof specificity and with no significant improvement in positive predictive value.

This analysis was performed on a selected popula-tion. All patients had coronary artery disease and left ventricularaneurysms.The application of theseresults to other patientpopulations with ventricular tachycar-dia may notbe justified.

We conclude that surgical therapy of ventricular tachycardia based on results of endocardial catheter mapping during sinus rhythm would lead to excision of a larger area of endocardiumthan that based on results of mapping during ventricular tachycardia. In addi-tion, not all sites of origin would be excised if only those showing abnormality on the map during sinus rhythm were selectively removed.

References

1. Josephson ME, Harken AH, Horowitz LN: Endocardial excision: a new surgical technique for the treatment of recurrentventricular tachycardia. Circulation 60: 1430, 1979

2. Josephson ME, Horowitz LN, Farshidi A, Spear JF, Kastor JA, Moore EN:Recurrent sustained ventricular tachycardia. 2. Endo-cardial mapping. Circulation57: 440, 1978

3. JosephsonME, Horowitz LN, Spielman SR, Waxman HL, Green-span AM: Roleof catheter mapping in the preoperative evaluation ofventricular tachycardia. Am J Cardiol 49: 207, 1982 4. WeinerI, Mindich B, Pitcon R: Determinants of ventricular

tachy-cardia in patients with ventricular aneurysms: results of intraopera-tiveepicardial and endocardial mapping. Circulation65: 856, 1982 5. Klein H, Kard RB, Kouchoukos NT, Zorn GL Jr, James TN, Waldo AL: Intraoperative electrophysiologic mapping of the ven-tricles during sinus rhythm in patients withaprevious myocardial infarction. Circulation66: 847, 1982

6. Cassidy DM, Vassallo JA,MarchlinskiFE, BuxtonAE, Untereker WJ, Josephson ME:Endocardial mapping of sinusrhythm in hu-mans withnormalleft ventricles: activation patterns and character-istics ofelectrograms. Circulation (in press)

7. Josephson ME, Horowitz LN, Farshidi A, Kastor JA: Recurrent sustainedventricular tachycardia. I. Mechanisms. Circulation 57: 431, 1978

8. Mason JW, Stinson EB, WinkleRA, Oyer PE, GriffinJC, Ross DL: Relativeefficacyof blind left ventricular aneurysm resection for the treatment ofrecurrentventriculartachycardia. AmJCardiol

49: 241, 1982

9. Untereker WJ, Spielman SR, LaskeyW, GreenspanAM, Hirsh-feld JW, Goldberg S, HorowitzLN, KastorJA, Josephson ME: Theendocardial electrogramas apredictorof wall motion. Circula-tion 62 (supplII):11-249, 1980

10. Gardner PI, Ursell PC, Fenoglio JJ, Wit AL: Anatomical and

electrophysiological basis forelectrograms showing fractionated activity. Circulation 66 (supplII): II-78, 1982

11. Waldo AL, Kaiser SGA:Astudy of ventriculararrhythmias associ-ated withacutemyocardialinfarction in the canine heart. Circula-tion42: 1222, 1973

12. Boineau JP, Cox JL: Slow ventricular activation inacute

myocardi-al infarction: a sourceof reentrant premature ventricular

contrac-tions. Circulation43: 702, 1973

13. Kienzle MG, Falcone RA, KempfFC, Miller JM, Harken AH,

Josephson ME: Intraoperative endocardial mapping: relation of fractionatedelectrograms in sinusrhythmtoendocardial activation in ventricular tachycardia -surgical implications. J Aol Coll Cardiol 1: 582, 1983

14. Waldo AL, Arciniegas JG, Klein H: Surgical treatmentof life-threateningventriculararrhythmias: the role ofintraoperative map-ping and consideration of the presently available surgical tech-niques. Prog Cardiovasc Dis23: 247, 1981

15. Gallagher JJ, KasellJH, CoxJL, Smith WM, IdekerRE, Smith 1109

WM: Techniques of intraoperative electrophysiologic mapping. Am JCardiol 49: 221, 1982

16. Josephson ME, Simson MB, Harken AH,Horowitz LN, Falcone RA:The incidence and clinicalsignificance of epicardial late po-tentials in patientswith recurrent sustained ventricular tachycardia and coronary artery disease. Circulation 66: 1199, 1982 17. Fontaine G, Guiraudon G, FrankR:Doubled epicardial potentials

inidiopathic ventriculartachycardia. Circulation54 (supplII): II-78, 1976

18. Fontaine G, Guiraudon G, Frank R, VedelJ, Grosgogeat Y, Cabrol C, FacquetJ: Stimulation studies and epicardial mapping in ven-triculartachycardia. Study of mechanisms and selections for sur-gery.In KulbertusHE, editor: Reentrant arrhythmias. Baltimore, 1976, University Park Press, pp 334-350

19. Josephson ME, Horowitz LN, Spielman SR, Greenspan AM, Van-dePolC, Harken AH: Comparison of endocardial catheter mapping with intraoperative mapping of ventricular tachycardia.Circulation 61: 395, 1980

CIRCULATION