Spatial Nonspatial Influences on...

Spatial and Nonspatial Influences on the TQ-ST SegmentDeflection of Ischemia

THEORETICALANDEXPERIMENTALANALYSIS IN THE PIG

ROGERP. HOLLANDand HAROLDBROOKSwith the technical assistance of BARBARALIDL

From the Hecht Experimental Hemodynamics Laboratory, Department of Medicine-Cardiology,The University of Chicago Pritzker School of Medicine, Chicago, Illinois 60637

A B S T RA CT Spatial and nonspatial aspects of TQ-STsegment mapping were studied with the solid angletheorem and randomly coded data from 15,000 electro-grams of 160 anterior descending artery occlusions eachof 100-s duration performed in 18 pigs. Factors ana-lyzed included electrode location, ischemic area andshape, wall thickness, and increases in plasma potas-sium (K+). Change from control in the TQ-ST recordedat 60 s (ATQ-ST) was measured at 22 ischemic (IS)and nonischemic (NIS) epicardial sites overlying right(RV) and left (LV) ventricles. In IS regions, ATQ-STdecreased according to LV > septum > RVand LV base> LV apex. In NIS regions, LV sites had negative(Neg) ATQ-ST which increased as LV IS border wasapproached. However, RV NIS had positive (Pos)ATQ-ST which again increased as RV IS border wasapproached. With large artery occlusion IS areaincreased 123+18%, ATQ-ST at IS sites decreased(-38.1+3.6%), and sum of ATQ-ST at IS sites in-creased by only 67.3±10.3%. In RV NIS Pos ATQ-ST became Neg. With increased K+, ATQ-ST decreasedproportionately to log K+ (r = 0.97±0.01) at IS and NISsites on the epicardium and precordium. TQ-STat 60 s was obliterated when K+ = 8.7+0.2 mM. Allfindings were significant (P < 0.005) and agreed withthe solid angle theorem. Thus, a transmembranepotential difference and current flow at the IS boundaryalone are responsible for the TQ-ST. Nonspatial fac-tors affect the magnitude of transmembrane potentialdifference, while spatial factors alter the position ofthe boundary to the electrode site.

Presented, in part, at the 49th Annual Scientific Session ofthe American Heart Association, Miami, Fla., November 1976.

Received for publication 16 June 1976 and in revised form3 January 1977.

INTRODUCTION

"No other part of the electrocardiogram is subject toso many theories, interpretations, and even more mis-interpretations as the ST interval," Shaefer and Haas(1). "The nature of these electrocardiographic changesis in no way mysterious," Wilson et al. (2).

Techniques currently used to assess ischemic/infarcted myocardium include hemodynamic moni-toring, creatine phosphokinase (CPK)l curve analysis,myocardial radionuclide imaging, and precordial TQ-ST segment mapping. The latter has enjoyed consider-able popularity due primarily to its low cost, nonin-vasive approach, and simply applied rules of inter-pretation. Controversy has arisen recently, however,for despite all the time, effort, and resources in-vested in TQ-ST segment mapping studies, basicappreciation of the complex manner by which theTQ-ST segment deflection relates to the underlyingischemic region has been lacking. Inevitably, con-fusion and disagreement over the specificity and quan-titative value of this electrocardiographic measure of

IAbbreviations used in this paper: CPK, creatine phos-phokinase; ECG, electrocardiogram; [K+]0, plasma potassium;LAD, left anterior descending coronary artery; LV, leftventricle; RV, right ventricle; SA, average solid angle; ISA;summed solid angle; SUM(+), SUM(-), sum of all positiveand negative TQ-ST deflections, respectively; TQ-STO,preocclusion TQ-ST segment deflection; ATQ-ST, individualTQ-ST segment deflection changes; ATQ-ST, average changein the TQ-ST segment; £TQ-ST, sum of individual TQ-STsegment deflection changes from the ischemic region;V.., transmembrane potential of the ischemic region;VmN, transmembrane potential of the normal region; AVm,transmembrane potential difference; e, TQ-ST segment de-flection; Q, solid angle subtended at recording site by theischemic boundary.

The Journal of Clinical Investigation Volume 60 July 1977-197-214 197

RECORDINGELECTRODE

Q-ST= Sl(Vm VmImKNIr

Vm

FIGURE 1 Mathematic and pictorial characterization of thesolid angle theory. The solid angle fl is defined as the areaof spherical surface cut off a unit sphere (inscribed aboutthe recording electrode) by a cone formed by drawinglines from the recording electrode to every point along theischemic boundary. The ischemic boundary is a source ofcurrent flow established by differences in the transmembranepotentials of the normal (VmN) and ischemic (Vm.) cellsduring diastole and systole. The TQ-ST segment voltage erecorded at the electrode is given by the above equation.K is a term correcting for differences in intra- and extra-cellular conductivity and the occupancy of much of the heartmuscle by interstitial tissue and space. See text for additionaldiscussion. Adapted from Holland and Arnsdorf (19).

myocardial injury was to be expected (3-11). Recogniz-ing the current renewed interest in this area and theprofound clinical value of being able to quantifyischemic damage, we have in this study examined ina theoretical and experimental manner some basicspatial and nonspatial factors influencing the magni-tude and polarity of TQ-ST deflection during myo-

cardial ischemia.

METHODS

NomenclatureMany studies have shown that the injury deflection, con-

ventionally referred to as the "ST segment deflection,"has a diastolic as well as a systolic component. For thisreason we have here, as in other studies (12, 13), re-ferred to this deflection as the "TQ-ST segment deflection."So-called "ST segment elevation," then, is actually the sum-mation of TQ depression and ST elevation (14-17). Due tocapacitive coupling of most electrocardiogram (ECG) record-ing apparatus, the fact that the TQ segment is often not atzero potential (isoelectric) may not be obvious. The rela-tive contribution of the diastolic (TQ segment) and systolic(ST segment) components to the total TQ-ST segmentdeflection has not yet been clearly established (14-17).

Mathematical formulations from the solidangle theoremThe solid angle theorem was used in this study to mathe-

matically describe the behavior of the TQ-ST segment dur-

198 R. P. Holland and H. Brooks

ing ischemia. According to the theorem (Fig. 1) the magni-tude and polarity of the TQ-ST segment deflection (e)recorded at an electrode site P is equal to the productof three parameters. Q is the solid angle subtended atP by the ischemic boundary; VmNand Vm, denote the trans-membrane potentials of the normal and ischemic regionduring either diastole (TQ segment) or systole (ST seg-ment); and K is a term correcting for differences in intra-and extracellular conductivity and the occupancy of much ofthe heart muscle by interstitial tissue and space (13, 18-21). Since the solid angle depends only upon the positionof the recording electrode and the geometry of the ischemicboundary, it reflects spatial influences on the magnitudeof the TQ-ST deflection. The difference in transmembranepotential between the two regions does not, however,depend upon geometry and thus reflects nonspatial influenceson the TQ-ST deflection.

Spatial influences were studied with a three-dimensionalspherical model of the heart containing a well-definedischemic region (12). Calculations of solid angle values inthis model then permitted analysis of various spatial rela-tionships existing between the recording electrode and thegeometry of the ischemic region. These spatial relationshipswere simulated in the model with the help of a sphericalcoordinate system (see Appendix), FORTRANIV program-ming language, and a Sigma V digital computer (SigmaInstruments, Inc., Braintree, Mass.). The ischemic heartmodel used in this study, unless otherwise stated, had outerand inner wall radii of 3.0 and 2.0 cm, respectively.The shape of the ischemic region was assumed to be trans-mural.

Experiments in the intact heartAnimal model. Experiments were performed in 18 open-

chest, domestic, neutered male pigs weighing from 35 to 60kg. The pig was chosen because of similarities betweenthe gross coronary artery architecture (22) and collateralcirculation (23) of this animal and man. In addition, the highdegree of reproducibility of the coronary architecture (22)among individual animals, as well as the less extensivedistribution of collateral vessels in the pig ventricle ascompared to the dog ventricle, insures that occlusion of acertain length of a coronary artery (24) results in a visually(cyanotic discoloration) and electrically identifiable area ofischemic damage. The degree of reproducibility is docu-mented in the Results section of this paper.

Surgery. After induction of anesthesia with a small intra-venous injection of thiopental, the animals were anes-thetized with an intravenous infusion of a warmed solu-tion of alpha-chloralose (60 mg/kg). During the study, sup-plementary doses of chloralose were given to maintain arelatively uniform state of anesthesia. The hemodynamicactions of chloralose are minimal and transient in duration(25). Respiration was maintained by a volume respirator(Harvard Apparatus Co., Inc., Millis, Mass.), regulated to main-tain an arterial pH of 7.45±0.05 throughout the experiment.The pump was connected to a tracheostomy tube and sup-plemental oxygen was administered to maintain arterial oxy-gen saturation at 95%. The heart was exposed by a mid-sternal thoracotomy, a pericardiotomy was performed, and apericardial cradle was created to support the exposed heart.

Hemodynamic mwasurements. Heart rate was kept con-stant at 125 beats/min. This was accomplished by plac-ing a bipolar electrode in the wall of the right auricularappendage and stimulating with a Grass stimulator (GrassInstrument Co., Quincy, Mass.) and an optical isolationunit. Stimulus duration was 5 ms. Animals whose resting

RV

Ij SEPT<<AxZLV Boundary

SEPT-

1.0 cm

FIGURE 2 Left. Location of epicardial electrodes on the anterior surface of the pig heart.Right. Anatomic classification of electrode sites. Ischemic sites (1, 2, 3, 4) lie completelyinside ischemic (cyanotic) region (shaded area). Border sites (5, 6) overlay both normal andcyanotic tissue. Septal (SEPT) sites (3) were within 0.5 cm (dashed parallel lines) of the maintrunk of the LADartery and separated the RVand LV. LV ischemic sites were further subdividedinto those overlying apical (1) and more basally (2) located portions of the ischemic regionby a line (vertical dashed line) drawn from the site of the small artery occlusion to theapex of the heart. Nonischemic sites (7, 8, 9, 10) were labeled as either "near" (7, 9) forthose within 2.0 cm of the boundary (curved dashed line) or "distant" (8, 10) for thoselocated further away. The position of the large artery occlusion is also illustrated. Thelength of LAD occluded (LAD) was measured in millimeters from the site of occlusion tothe apical termination of the LAD (SEPT arrow).

heart rate exceeded 125 beats/min were excluded fromthis study. Pressure determinations were made through 6-inch Teflon 16T gauge catheters connected directly toStatham P23 Db pressure transducers (Statham InstrumentsDiv., Gould, Inc., Oxnard, Calif.) without intervening tubing.Systemic pressure was obtained by placing the cathe-ter in the left carotid artery and left ventricular pressureby placing it in the left ventricular cavity at the apex.The left ventricular pressure was differentiated electronicallyto obtain the time derivative of left ventricular pressure(LV dP/dt). Left ventricular end-diastolic pressure measure-ments were made from high-gain tracings via an additionalamplifier (1 mmHg/division).

In those animals in which the pattern of TQ-ST segmentdeflections was studied at different levels of plasma potas-sium ([K+I.), [K+]0 was increased from control values by aslow intravenous infusion of KCI with an infusion pump.[K+]o levels were obtained from heparinized carotid arterialsamples taken immediately after the beginning of each coro-nary occlusion and were measured with a flame photometer(Instrumentation Laboratory, Inc., Lexington, Mass.).

Ischemic area. The anterior left ventricular free wall sup-plied by branches of the left anterior descending (LAD)artery was selected for study. Reproducible areas of ischemicinvolvement in all animals were obtained by placing a silkligature (00) around the LAD artery at a point approximately

one-third to one-half the distance from the apical terminationof the LAD to its origin from the left main coronary artery.Reversible occlusion was obtained with a polyethylenecollar. In those experiments in which large and small areasof ischemic involvement were compared, the larger areawas produced by occluding the anterior descending arteryat a point approximately 1.0 cm from its origin with the leftmain coronary artery. This insured that the large area com-pletely enclosed the small area. Occlusions were limited to100-s duration. Recent studies have demonstrated that al-though ischemic TQ-ST segment deflections return towardsnormal after release of a brief occlusion(5- 10 min), return ofcontractile activity is delayed and may exhibit a permanentdeficit (26-28). Thus, to insure complete functional (bio-chemical, electrical, mechanical, etc.) recovery ofthe ischemicsegment, it would appear necessary to limit the occlusionsto very short periods of time, instead of the 15-20-minocclusions of most studies (5, 29). Although the TQ-STsegment deflection has not obtained a steady-state value bythis time (100 s), there is no indication from earlier studiesthat steady-state values are reached at any time within the1st h after occlusion. Finally, the incidence of conductionabnormalities (QRS complex, loss of the S wave, ectopicbeats, etc.) which may either obliterate, mimic, or alter theTQ-ST segment deflection characteristic of ischemia increaseswith the duration of the occlusion (5, 30). In the pig, such

Influences on the TQ-ST Segment Deflection 199

PRECORDIAL* 5IOmm* = lOOmmA . 50mm* 25mm

l0 20 30 40 50 60Isdcmic Area (cm')

FIGuRE 3 Effect of changes in wall thickness on the magni-tude of the solid angle at precordial and epicardial elec-trode locations. Curves were derived with the electrodecentrally positioned over the ischemic region. The outerwall radius used in this analysis was 3.5 cm. At either loca-tion increases in wall thickness increase the solid angle for agiven area of ischemic involvement.

abnormalities are clearly in evidence beyond the first 100 s

of the occlusion (30).The area of ischemic involvement was estimated in all

animals by the following methods: (a) The area of cyanoticdiscoloration after the first occlusion was drawn to scale ona previously sketched anterolateral view of the porcineheart which included the distribution of the LAD arteryand its primary branches (31); (b) At postmortem, the lengthof LAD artery occluded was perfused with methylene bluedye and the stained epicardial surface area along with variousanatomic arterial landmarks was measured with a pair ofcalipers. The two areas (cyanotic and stained) were thencalculated by planimetry and the caliper values; (c) The massof stained tissue (ischemic weight) after the above dye in-fusion was then cut out and weighed (24). Agreementin the two methods of estimating ischemic area is documentedin the Results section of this article.

Electrical measurements. Epicardial electrical potentialswere recorded from the heart's surface with 22 atraumatic,firmly attaching electrodes designed in this laboratory.The electrodes are a modification of the original design(32) and are constructed from polished brass screws havinga surface area of 14 mm2(radius = 2.2 mm). The electrodeswere spaced equidistant from one another over the an-terior surface of the left and right ventricles in the distribu-tion of the LAD artery (Fig. 2, left).

The need for near-simultaneous recordings of the electro-grams from 22 sites required that an automated rapid elec-tromechanical switching circuit be designed (32). Basicallythe system consists of a pulse generator and two steppingrelays. Upon receiving a pulse the stepping relay is advancedand connects a particular electrode site to the ECGampli-fier for a period of 0.8 s. The heart rate of 125 beats/minused in this study permits approximately 1% heart beats tobe recorded at each site. By using a pair of calibratedbioelectric amplifiers (No. 8811A, Hewlett-Packard Co., PaloAlto, Calif.) with a frequency response of DC to 10 kHz,all 22 electrode sites could be sampled within the spaceof 10 s. The signals were simultaneously recorded on FMmagnetic tape (73Y4 ips) and displayed on a high speed chartrecorder along with the hemodynamic signals. The FMtape recorder had a frequency response of DC to 2.5 kHz,a signal-to-noise ratio of 44 dB, total harmonic distortionof 2%, and flutter of 0.4%. The resulting signals were then


analyzed at a gain of 1.0 mV/div and an equivalent paperspeed of 50 mm/s with a frequency response of DCto 400Hz. With this system TQ-ST segment deflections after acutecoronary occlusion were characterized in the followingmanner. TQ-ST deflections were measured from all sitesimmediately before occlusion (control) and at exactly 60 safter coronary occlusion. The total TQ-ST deflection wasmeasured from that portion of the TQ segment occurringimmediately before the inscription of the QRScomplex tothe systolic endpoint, a point on the ST segment occurring100 ms after onset of the QRS complex (33). Track widthwas approximately V5 of a division and TQ-ST values weremeasured to Y%. of a division (0.1 mV). The control values(TQ-STO) were obtained from the mean of multiple (2-4)complexes recorded during the 30 s immediately precedingthe occlusion. The difference between this value and themagnitude of the deflection at 60 s after occlusion was la-beled ATQ-ST. To obtain values for TQ-ST at 60 s fromsites which were sampled at times other than 60 s (e.g.,56 and 66 s), linear interpolation was employed.

TQ-ST segment deflection changes at multiple ischemicand nonischemic sites were categorized solely on the basisof the position of the various recording sites to the ischemicregion. This was accomplished by dividing the anterosep-tal region of the heart into 10 regions as illustrated in Fig.2 (right). All ischemic sites were required to be completelyenclosed by the region of ischemic (cyanotic) myocardium.Border sites were required to overlie both normal and cya-notic tissue and thus the width of the border was definedby the electrode diameter (4.4 mm). Septal ischemic siteswere within 0.5 cm of the main trunk of the LAD. Leftventricular ischemic sites were further subdivided into siteslying in thin, apical portions of the ischemic region andthose in the thicker-walled areas nearer the base of the heart.This was again accomplished in an unbiased fashion bydrawing an imaginary line from the site of the occlusionto the apex of heart (see Fig. 2). All electrode locationswere placed in 1 of the 10 categories before measurementof their electrogram tracings.

Data analysis. To avoid any prejudicial measurement ofthe TQ-ST segment deflections, all electrogram tracings wererandomly coded and evaluated without knowledge as to elec-trode locations. As a result of the requirement for multiplemeasurements of the TQ-ST segment deflection an averageof 100 determinations were made during each occlusion andthus in the course of this study in excess of 15,000 electro-gram tracings were evaluated. Statistical analysis was confinedto paired and unpaired t-tests and linear regression analysis(34) carried out with the help of a Wang 700 desk cal-culator (Wang Laboratories, Inc., Lowell, Mass.).

RESULTS

Theoretical analysis

In Fig. 3 the effect of different wall thicknessesand areas of ischemic involvement on the solid anglesubtended at precordial and epicardial locations isillustrated. Although increases in ischemic size in-crease the magnitude of the solid angle at precordiallocations, and decrease it at epicardial sites (12),at either location increases in wall thickness increasethe solid angle. This finding immediately suggestedthat for equivalent areas of ischemic involvement,TQ-ST deflections recorded at thick-walled portions

0.000 10 20 30 40 50 60

Ischemic Area (cm')

EPICARDIAL

PRECORDIAL

o, * - 0.25( 3.5cm2 )

OD. z 0.50(13.8cm2)A, A - 0.75 (30.8cm")O. * a 1.00 (52.0cmn)

Prwecm

I ,I

Epirdiuml

IX

FIGURE 4 Influence of electrode position with respect to the ischemic boundary (dashedvertical line) on the polarity and magnitude of the solid angle computed at precordial andepicardial locations for different ischemic areas. Electrodes overlying ischemic regions (solidsymbols) have positive solid angle values at either location. On the precordium, negativesolid angle values occur at sites distant from the boundary while at the epicardium negativevalues are recorded from all nonischemic regions (clear symbols). Dashed lines indicateuncertainty of solid angle values at the epicardium and in the vicinity of the ischemicboundary. Arrows identify solid angle values at epicardial sites within 0.5 cm of ischemic boun-dary. Note that at the epicardium the negative solid angle values in the nonischemicregion increase in magnitude with increases in ischemic area despite a reduction in themagnitude of the positive solid angle values of the ischemic region.

(base) of the ischemic left ventricle should exceedthose recorded at the thinner-walled apex. Similarly,the TQ-ST deflection at left ventricular sites shouldexceed the magnitude of the deflections from rightventricular ischemic sites.

The solid angle values calculated above were forelectrodes centrally positioned over a circular area ofischemic involvement. However, in mapping studies,electrodes overlie all regions of ischemic and non-

ischemic tissue. The effect of moving the electrodefrom the center of the ischemic area (13/P = 0.0),towards the ischemic boundary (13/P = 1.0), and beyond(fl/P> 1.0) was computed and is shown in Fig. 4.The solid angle at precordial sites is maximal whenthe electrodes directly overlie the ischemic region(//P = 0.0) and decreases at sites distant from the center(f3/P > 0.0). The solid angle does not become negativeuntil electrodes are positioned at a considerable dis-tance from the ischemic boundary (13/P > 1.0). Thesolid angle at epicardial sites, in comparison, increasesin positivity as the boundary is approached. At theboundary a discontinuity in the curve occurs and be-

yond this point, in the nonischemic regions, the epi-cardial solid angle is negative, rapidly decreasing inmagnitude as the electrode moves further away fromthe ischemic boundary. Although the solid angle can

assume values of either positive or negative polarityin the immediate vicinity of the boundary, on theaverage the solid angle values should have a positivepolarity with a magnitude of approximately 40% thatrecorded by electrodes directly overlying the center ofthe ischemic region.

Having determined how the magnitude and polarityof the solid angle varies depending upon its locationwithin the ischemic region, the summed solid angle(I;SA) for different areas of ischemic involvement may

be computed. ISA is equal to the sum of the individualsolid angles subtended at each electrode site overlyingthe ischemic region. In Fig. 5 :SA and the average

solid angle SA were plotted as a function of ischemicarea. At the precordium both SA and ISA increasewith ischemic area. The curve relating SA to ischemicarea is remarkably linear over the range of areas from5 to 50 cm2. At the epicardium the situation is


w-JCDz

0-J0C')

EPICARDIAL

0.10w1

0.05w

0.00

20 30 40 5

ISCHEMIC AREA (cm2)

FIGuRE 5 Influence of changes in ischemicaverage (SA) and summed (ISA) solid angle vaelectrodes overlying the ischemic area. Xupon the electrode density, E, the area ofvolvement, and the SA recorded at each sitdepends in turn upon whether the electrodes;the precordium or epicardium. Although ISA ii

increases in ischemic area at either position,centage change in ischemic area results insmaller changes in YSA at the epicardium. Arelationship between ;SA and ischemic areaonly when SA is constant for all ischemic area

tion never exists but is approximated at pr(and large ischemic areas whereupon changesarea are accompanied by nearly equivalent cha

different. Although SA decreases with iischemic area, ;SA increases, but in a not

ner. At this location a given percentageischemic area results in a substantially sm

in ISA. In the range of areas from 10a doubling in ischemic area is met bycrease in SA and only a 60% increaseepicardial sites; at precordial sites theYESA (+ 153%) is greater than the char(+ 100%).

Experimental verificationSpatial influences. In Table I anatomi

dynamic characteristics of the experimentithe ischemic region are tabulated. Occspecific length (38.1+1.1 mm)of the antering coronary artery in animals of equivaclass resulted in the production of standischemic involvement (13.2+0.8 cm2). 7tissue becoming ischemic (cyanotic) clos


2.00-w 0 mated that area normally perfused by the occluded1< artery (stained).

100 Q In Table II TQ-ST segment deflection changes at2 multiple ischemic and nonischemic sites have beena. tabulated from a total of 77 occlusions performed in

-0.0 IiJ 16 animals. As was suggested by the theoreticalfindings, different TQ-ST potential populations can be

40.OEvcm2 successfully categorized simply on the basis of spatialrelationships existing between the various electrodesites and the geometry of the ischemic region. In

30.OEwcm' a the ischemic region TQ-ST deflections were always< observed to increase in magnitude as a thicker-N walled ischemic boundary was approached. Thus, the

20.0Ercmt o magnitudes of the individual ischemic populations de-X creased accordingly: left ventricle (LV) > SEPT*c > right ventricle (RV) and LV base > LV apex. The

10.0E.OEm2W relative magnitudes of the deflections at these ana-tomic locations exhibited even smaller variability when

00.0 deflection populations were compared in each animal(Fraction column). In this case the LV and RVischemicsites were respectively 1.47±0.06 and 0.65±0.04 the

area on the magnitude recorded at the septal locations, while thealues seen by sites closer to the LV base were 1.25±0.06 (P < 0.001)SA depends those nearer the apex.ischemic in- Although there was no significant difference in the

e. The latter means of the positive potential deflections recordedare located at at the RV and LV ischemic borders, a distinctionncreases with

between these two populations may be noted. At thea given per-substantially

purely linearl is expected TABLE ILs. This situa- Characteristics of the Experimental Modelecordial sites

in ischemic

mges in I5SA.

increases in

alinear man-

change inaller change

to 50 cm2a 22% de-in I SA atchange in

rige in area

c and hemo-il model and2lusion of a

ior descend-dlent weight

lard areas ofThe area ofely approxi-

AnatomicAnimal weight, kgHeart weight, g

Heart/animal weight, %Ventricle weight, g

Length of artery occluded, mm

Ischemic area-cyanotic, cm2Ischemic area-stained, cm2Ischemic weight, g

Ischemic/ventricle weight, %Number of ischemic sites, n

HemodynamicMAP, mmHgLVP, mmHgLVEDP, mmHgLV dP/dt, mmHgls

46.6±1.7184.3±6.1

4.2±0.2152.9±4.5

38.1±1.113.2±0.814.6±0.919.9±0.813.1±0.59.1±0.5

102.5±5.4109.1±5.2

6.4±0.41,649+106

Abbreviations: MAP, mean arterial pressure; LVP, leftventricular peak systolic pressure; LVEDP, left ventricularend-diastolic pressure; LV dP/dt, time derivative of leftventricular pressure.

Values shown are means_SEM obtained from experimentsperformed in 17 animals. Heart rate was kept constant at 125beats/min.

0

W-

.Q

8c00.

right ischemic border the potential deflections were

always positive, while at the left negative deflectionswere measured in a few instances.

At electrode sites overlying nonischemic portions ofthe LV reciprocal negative deflections were recorded,their magnitudes decreasing at locations more distantfrom the LV ischemic boundary. These reciprocaldeflections averaged -0.25±0.03 (near) and -0.09+0.01 (distant) the magnitude at the ischemic septalsites.

In contrast, at nonischemic positions on the RVthe polarity of the TQ-ST deflections was positivedespite the fact that the sites from where they were

recorded neither overlay cyanotic tissue nor were inthe apparent distribution of the occluded artery.The magnitude of these deflections maintained itspositivity but diminished in intensity at locationsmore distant from the right ventricular ischemicboundary.

In Table III the effects of small and large coronary

artery occlusion on the area of ischemic involvementand the magnitude and polarity of the deflections re-

corded from sites overlying ischemic and nonischemicportions of the heart are tabulated. With the porcineheart model, production and subsequent measurementof a standardized area of ischemic involvement was

easily accomplished. The ischemic area after largeartery occlusion was approximately twice that observedafter the smaller occlusion. The percentage increase inischemic weight, however, was greater than this duepresumably to the involvement of the thicker, more

basally located portions of the anterior wall onlyduring large artery occlusion.

The change in ischemic area had a predictableinfluence on the magnitude and polarity of the TQ-STdeflections recorded from all areas. Sites originallyoverlying the smaller ischemic area (LV apex) main-tained their positive polarity but decreased in magni-tude as the LV ischemic boundary was removed to a

more distant location nearer the base of the heart.New sites closer to the base of the heart, whichafter the small artery occlusion were in nonischemicor border regions, now registered positive deflectionswhich exceeded in magnitude those deflections re-

corded at more apical portions of the ventricle. Simi-lar though less dramatic changes were observed in theRVischemic regions, for after the large artery occlusionthe RVboundary extended itself upwards towards thepulmonary outflow tract but still maintained a positionparallel to and a fixed distance away from the septum(Fig. 6).

Although the area of ischemic involvement more thandoubled after large artery occlusion, the sum of theindividual TQ-ST segment deflections from theischemic region, YTQ-ST, increased by only 67.3+ 10.3%. This was predictable on the basis of the

TABLE IITQ-ST Segment Deflections at Multiple Ischemic

and Nonischemic Sites

Magnitude Fraction (septum)

mV

Ischemic regionLeft ventricle 4.70±0.30 1.47±0.06

Apex 4.25+0.29 1.32±0.06Base 5.26+0.36 1.64±0.07Difference 1.00±0.25* 0.32±0.07*

Septum 3.21±0.18 1.00Right ventricle 2.03±0.15 0.65±0.04

Border regionLeft ventricle 1.66±0.32 0.56±0.11Right ventricle 1.14+0.15 0.38±0.08Difference 0.52±0.45 0.18±0.16

Nonischemic regionLeft ventricle

Near -0.80±0.09 -0.25±0.03Distant -0.28±0.05 -0.09±0.01Difference -0.51±+0.09* -0.16+0.03*

Right ventricleNear +0.57+0.06 +0.19±0.02Distant +0.19±0.05 +0.05±0.01Difference +0.38±0.06* +0.14±0.03*

Values shown are meanstSEM obtained from experiments in17 animals. Average number of occlusions performed in eachanimal was 4.5±0.5.* P < O.O0; paired t test.

theoretical findings and is a reasonable expectationwhen one considers that although the number ofischemic sites increased to the same degree as the is-chemic area, the average potential deflection recordedat these sites decreased by 20%. Indeed, if the thick-ness of the LV ischemic boundary had not increasedafter large artery occlusion, the average ischemicTQ-ST deflection might have experienced an evengreater decrease (and YTQ-ST a smaller increase)and thus might have yielded results in even closeragreement with the theoretical predictions.

In the nonischemic portions of the RV, the normalpositive deflections recorded with a small area ofischemia declined in magnitude and in many instanceseven became negative after the large artery occlusion(P < 0.005). In contrast, however, the normally nega-tive deflections recorded from the nonischemic por-tions of the LV appeared to increase in magnitudeafter large artery occlusion. These results suggestedthat the position of both right and left ischemicventricular boundaries helped to determine both thepolarity and magnitude of the deflections recorded atall sites.

This hypothesis was tested further as is illustrated


TABLE IIIEffect of Changes in Ischemic Area on the TQ-ST Segment Deflection

Small Large Difference Change

AnatomicLAO, mm 39.2±1.3 79.8±1.8 40.6±2.5* 105.4±9.5*IA, cm2 13.0±0.6 28.5±1.8 15.6±1.9* 123.3±17.9*IW, g 19.6± 1.0 45.3± 1.6 25.6± 1.0* 133.4±9.6*#IS 8.1±0.3 16.6±0.5 8.5±0.5* 106.3±9.2*

ElectricalIschemic region

LV + SEPT (ATQ-ST)Original 3.77±0.21 2.28±0.11 - 1.49±0.22* -38.5±3.6*New 3.86±0.26§Total 3.77±0.21 3.01±0.14 -0.76±0.15* - 19.4±3.0*

RV (ATQ-ST)Original 1.73±0.15 1.32±0.17 -0.41±0.12t -24.9±8.2tNew 2.62±0.28§Total 1.73±0.15 1.95±0.20 +0.23±0.23 +19.1±15.7

IATQ-STOriginal 28.5±2.1 17.4±1.1 -11.1±1.7* -38.1±3.6*New 29.8±2.95Total 28.5±2.1 47.2±3.3 + 18.7±2.5* 67.3±10.3*

Nonischemic regionLV + SEPT (ATQ-ST) -0.59±0.12 -1.14±0.50 -0.55±0.35 -93±59RV (ATQ-ST) +0.44±0.08 -0.25±0.12 -0.69±0.15* - 160±26*

Abbreviations: Difference, large-small; LAO, length of artery occluded; IA, area of cyanotic discoloration; IW, ischemicweight; #IS, number of electrode sites completely overlying cyanotic region; LV + SEPT, left ventricle and septal sites; RV,right ventricle sites; :ATQ-ST, sum of ATQ-ST at all ischemic sites; Original, sites overlying small ischemic area; New,additional sites becoming ischemic after large artery occlusion.Values shown are means±SEMobtained from experiments performed in eight animals. The average number of small and largeartery occlusions performed in each animal was 5.0 and 2.1, respectively. ATQ-ST at nonischemic LV + SEPT sites afterlarge occlusions were obtained in only three animals.* P < 0.005; paired t test (difference and percent change).4 P < 0.02; paired t test (difference and percent change).§ P < 0.005; paired t test (original-new).

in Fig. 6. In this experiment multiple occlusions ofthe small coronary artery in an animal were made andthe average TQ-ST (ATQ-ST) changes at all sites meas-ured at 30, 60, and 90 s. During every other occlusionat precisely 65s after the small artery occlusion,the larger artery was then occluded. Again despiteincreases in the amount of ischemic damage, the de-flection magnitude at sites originally overlying thesmaller area of ischemia began to decline, being signif-icantly lower 25 s later than if the smaller occlusionhad been maintained the full 90 s. In this exampleSUM(+) and SUM(-) refer to the sum, respectively,of all positive and negative ATQ-ST deflections meas-ured from all 22 sites. With a sudden increase inischemic area at 65s, sites originally overlying non-ischemic regions of the LV now overlay ischemicregions and began to develop positive ATQ-ST. Thusthese sites were no longer categorized in SUM(-) and


this parameter dramtically decreased in value. SUM(+)on the other hand had still not exhibited any change at90s; for although it had begun to receive positivecontributions from previously nonischemic sites(shaded) it also lost contributions from (a) sites fromthe originally ischemic area whose average deflectionbegan to decrease (solid) and (b) nonischemic sitesoverlying the RV which were initially positive butwhose magnitudes declined towards zero after thesubsequent large occlusion (see Table III).

Preocclusion TQ-ST segment deflection. While pre-occlusion control values for the TQ-ST deflection (TQ-STO) closely approached an isoelectric value for the ani-mal population as a whole (TQ-STo = +0.05+0.10 mV),the values in individual animals and at different elec-trode sites showed greater variation. This is illustratedin Fig. 7. The SEMfor TQ-STo at all sites in a singleanimal was regularly 0.10 mVwhile the mean was

often significantly different from zero (either positiveor negative) depending upon the animal studied (darkcircles).

Nonspatial influences. The effect of progressivehyperkalemia (3.0 to 8.0 mM) on the evolution of theTQ-ST segment deflection was studied in six animals.Results in a representative animal are shown in Fig. 8.The average ATQ-ST recorded from all ischemic andnonischemic regions decreased with steadily increas-ing plasma potassium levels varying in a linear fashionwith the logarithm of the [K+]o level (r = 0.974+0.013).The value of [K+]o at which no changes in TQ-STsegment are expected was found by extrapolation tobe 8.72±+0.15 mM(n = 5). Changes in whole ventricu-

40--10- sum,

SUM(-

-20 -

5.0

4.00 LV + SEPT33.01

~2.01l1.0 RV

a,_

U)

E tZ

. EZ it

&-

lar dynamics did not change significantly in thisstudy until [K+]o exceeded 8.0 mM. By this time theTQ-ST segment changes were nearly nonexistent. Toinsure that the effect of K+ on the TQ-ST deflec-tion was nonspatially mediated, in the sixth animalthe polyethylene occluder was brought out of the chestcavity, saline-soaked gauze pads were placed aroundthe heart (to insure proper conductive pathways tothe precordium), and the chest was closed. Needleelectrodes were then placed under the skin at variouspoints on the right and left precordium. With 11repeated coronary occlusions at varying [K+]0 levelsthe ATQ-ST again declined in a logarithmic manner(r = 0.98).

20

10

0~

I I It0 30 60

Seconds Post Occlusion

-9

FIGuRE 6 Effect of a sudden increase in ischemic area on the TQ-ST segment deflection.Average change in the TQ-ST segment (ATQ-ST) was measured at all sites at 30 and 60 safter seven repeated small artery occlusions (one star; solid symbols) in one animal. Duringocclusions 1, 3, 5, and 7 the ischemia was maintained for 100 s and the ATQ-ST at90 s also measured. During occlusions 2, 4, and 6 the large artery was occluded at 65 s andATQ-ST again measured at 90 s (two stars; clear symbols). Despite the increase in ischemicarea as evidenced by the doubling in the number of ischemic sites, the ATQ-ST at all ischemicsites in the distribution of the occluded small artery (dark circles) declined in magnitudewithin the following 25 s. SUM+ (0), the sum of all positive ATQ-ST at all sites remainedunchanged after the large artery occlusion while the sum of the negative ATQ-ST, SUM-(-),rapidly declined. Dashed lines indicate ischemic boundary after small and large arteryocclusions. *P < 0.001 (unpaired t-test). **P < 0.01 and ***P < 0.05 (paired t test at individualischemic sites). See text for additional discussion.


%f

etry of the anteroseptal region produced experi-80- - *mentally. A closer look at this geometry suggested

that three different boundaries were present: right,0 60 - Lot 1 ^ left, and septal ventricular boundaries. Each of these

Ri t boundaries would make its own contribution to the40 total TQ-ST segment deflection recorded at each elec-

S. 20 1 l ll 1trode site. In Fig. 9, the positions of the boundaries2 are depicted in a theoretic heart model containing an

__2 ischemic region of 18 cm2 in area. Also shown are

-20X, .0- 0_.0T-- -2.0 solid angle values calculated for a number of ischemicTO -ST. (mv) and nonischemic sites. It is observed that electrodes

positionled over the ischemic region subtend positiveFIGURE 7 Deflections from the isoelectric line in the sones, the itudes of are influeepicardial electrogram recorded before coronary artery oc- sold angles, the magnitudes of which are influencedclusion (TQ-STO). TQ-STO values varied among different ani- by the electrodes proximity to the different ischemicmals and electrode sites. Histogram constructed from meas- boundaries and the boundaries' thicknesses. Thus, theurements of the TQ-ST deflection at all electrode sites and solid angle increases as the thicker LV boundary isin all animals used in the compilation of the data in Table approached (LV > septum > RV) as did the TQ-ST seg-II. The main trunk of the anterior descending artery divided ment deflections in the experiment model. Elec-right from left ventricular sites, however, neither spatialnor hemodynamic factors were thought to be responsible trodes positioned over nonischemic regions, however,for these variations. Although the mean of this population sense current flow from different directions and thusis close to zero, the mean of the deflections in individual the polarity and magnitudes of their solid anglesanimals (0) did, in some instances, assume values which weresignificantly different from zero.

DISCUSSION

Spatial influences and epicardial mapping. Afteracute coronary artery occlusion in the pig TQ-STsegment deflection changes (ATQ-ST) during earlyischemia may be recorded from all epicardial regionsof the heart. This and earlier studies (12, 35-37)have demonstrated that the relative magnitudes andpolarities of these deflections in each animal are de-termined primarily by (a) the position of the record-ing electrode with respect to the ischemic tissueboundaries; (b) the area of ischemic involvement;(c) the transmural shape of ischemic involvement;and (d) the thickness of the ventricular wall at theboundary.

Spatial factors which affect the TQ-ST segmentdeflection may be investigated in a physiologic andquantitative fashion by way of solid angle analysis.Predictions obtained from this analysis may then betested and validated in the intact pig heart, a modelwhich permits production and measurement of astandardized and well-defined area of ischemic in-volvement.

Boundary location in anterseptal ischemia. Al-though it was assumed that negative TQ-ST segmentdeflections should be recorded from both right and leftventricular nonischemic regions, the finding of positivedeflections on the RV came initially as a surprise.This finding indicated that the spherical heart modelwith an ischemic region of uniform wall thicknessfrom which the theoretical curves were derived (seeAppendix) rather incompletely characterized the geom-


ISCHEMIC *NON-ISCHEMIC °

3.5

3.0

0.0

-100 -90 -80 -70Ek(mv)

3.0 6.0 12.0Plasma K (mM)

FIGURE 8 Plot of changes in the TQ-ST segment deflec-tion recorded during multiple occlusions of the anteriordescending artery in the pig at steadily increasing plasmapotassium (K+) levels. The ATQ-ST recorded from bothischemic (cyanotic and nonischemic (left ventricle only) sitesappeared to vary in a linear faishion with the transmembranepotential (Ek calculated from the Nermst equation at 37°C,assuming intracellular [K+] = 140 mM)and with the log of theK+ level. Values are means+SEM. Number of electrodesoverying ischemic and schemic areas were 10 and 7, re-spectively.

reflect the difference (as opposed to the sum whenthe electrodes overlie the ischemic region) of contribu-tions from the different boundaries. At nonischemicRVsites, despite the nearby contribution from the thin-walled RV boundary, the more distant, but thicker,septal and LV boundaries subtend the larger solidangles and therefore, as seen experimentally, positiveTQ-ST segment deflections may be recorded from thisnonischemic region (electrode sites B and C).

Since spatial influences on the TQ-ST deflectiondepend upon the position of the electrode to the variousboundaries, movement of the LV boundary to a moredistant location should result in the following. First,the positive voltage at the RVnonischemic sites shouldeventually become negative if the boundary is movedfar enough. Second, the positive potentials of theischemic regions should decrease in magnitude. Fi-nally, the LV nonischemic sites should becomeslightly more negative. In the experimental model sucha relocation of the LV boundary was accomplished byproximally occluding the LAD artery, therby creating alarger area of ischemia. The results (Table III and Fig.6) fulfilled these expectations.

Boundary vs. local influences on the TQ-ST seg-ment deflection. Recognizing that it is the rela-tionship of the electrode to the ischemic boundarywhich determines the relative magnitude and polarityof the TQ-ST segment deflection, only then can it beappreciated that although sites of TQ-ST segment ele-vation are usually found to overlie tissue exhibitinglactate production (38), ATP and CPK depletion (29,38, 39), histologic abnormalities (31, 40), QRS com-plex alterations (5, 30, 41) and lowered PO2 (42-44), contractile activity (45), and coronary blood flow(46-48), the frequent lack of quantitative relationshipsexisting between the magnitude of the TQ-ST segmentdeflection at a particular site and these alternatedmarkers should not be surprising. Although the magni-tude of the TQ-ST deflection varies depending uponthe electrode's orientation to the boundary, there is noreason to expect CPK levels, histology, etc., to varyin the same manner. Recent studies in which TQ-STsegment magnitudes at individual sites on the epi-cardium were compared to later changes (24 h) seenin tissue blood flow (49), histologic abnormalities (40),and CPK levels (49) at the same site, demonstratethe absence of such quantitative relationships.

Electrode location and ischemic shape. Further ap-praisal of the solid angle concept reveals additionalspatial factors such as heart size and ischemic shapewhich may influence the polarity as well as the magni-tude of the TQ-ST deflection. Differences in ischemicshape, first considered by Prinzmetal et al. (50), pro-vide a good example of how solid angle analysis ismost useful when employed in a quantitative manner.Some of the most difficult conceptual problems with

-RIGHT + LEFT-

FIGURE 9 Solid angle values computed at ischemic (shadedarea) and nonischemic sites during anteroseptal ischemia ina theoretical model of the heart. Area of ischemic involve-ment equaled 18 cm2. Distribution of solid angle valuesapproximates TQ-ST deflections observed experimentally(Fig. 1 and Table II). Note that positive solid angle valuesmay be obtained from electrodes overlying nonischemic por-tions of the right ventricle (sites B and C). Appreciation ofthe significant contribution of left ventricular sites may notbe gained from this two dimensional (lower) view of ischemicsegment. Much of the left ventricular boundary is considera-bly closer and right ventricular boundary farther away fromthese sites as the upper view may suggest.

the solid angle or boundary concept developed be-cause of experimental data which did not appear toagree with predictions obtained from a rather super-ficial solid angle analysis. The following were held bysome to be unreconcilable with the boundary con-cept: (a) TQ-ST segment elevation was observed inintramyocardial electrode leads when no correspond-ing change was recorded from epicardial leads (51);(b) An absence of TQ-ST depression was noted inintracavitary leads despite the finding of TQ-ST ele-vation at epicardial sites (52, 53); and (c) TQ-STelevation was observed at both endocardial and epi-cardial locations when equal inner and outer wall


A0

0

0

c

+ +

N

E

FIGURE 10 Polarity and relative magnment deflections computed at epicar(endocardial, and intracavitary locations;ischemic involvement. + (-)TQ-ST sedicates negative (positive) TQsegment alST segment displacement. Dark circlhpositions.

(transmural) ischemic involvementThe last of these observations was dat the time (1942), including of;(54), but Pruitt and Valencia (35) shca few years later that there was "aelusive, explanation" (32) providedplication of the solid angle concepttion is illustrated in Fig. 10, wh4relative magnitudes of solid anglerdeflections) were computed for clocations and ischemic shapes. Thetioned above can be explained by pischemic involvement (Fig. IOA), a cing in normal dogs (55) and patielcoronary artery disease (56). Themerely require a transmural ischemD, and E).

Well-delineated and diffuse isc)As shown in Fig. 4, the solid angepicardial electrode is moved fromperiphery of the ischemic region,boundary is well delineated (zero wness of the ventricular wall is const

B baboon (57), and normal human (58, 59), there existed+++++ few functionally significant collateral vessels and one

expects the boundary to be narrow and well defined(24). However, in the dog, with its richer collateralblood supply (23), evidence suggests that the boundarylacks such definition (60). In this species a wide gra-

-- dient of transmembrane potentials most likely exists atthe periphery of the ischemic region and the boundaryor separation between ischemic and nonischemic tis-

D sue is wider and much less distinct (18).+++ The influence of boundary width on the distribution

of solid angle values at ischemic and nonischemicW \ sites is shown in Fig. 11. Although different boundaryr+++> \ widths do not alter the precordial distribution, changes/ \ in the epicardial distribution in the vicinity of the

°) boundary are seen. When the boundary is wide,solid angle values now decrease, instead of increasing,as electrodes are moved from the center to the

F periphery of the ischemic region.One would expect this difference in the solid angle

distribution from the center to the periphery of theischemic region to be reflected by differences in theTQ-ST segment voltages in species with (e.g., pig and

/ ++++ a \ baboon) and without (e.g., dog) a well-defined+++ boundary. Studies in the dog have shown that the TQ-

ST segment deflection is greater in the center of theLitude of TQ-ST seg- ischemic region than at the periphery (61). Studies indial intramyocardial,and various shapes of the baboon, on the other hand, show only a rapidgment deflection in- transition in polarity as electrodes are moved from thend positive (negative) ischemic to the nonischemic region (15). The lack ofes refer to electrode uniform wall thickness of the anteroseptal ischemic

region and the size of the electrodes did not permitus to investigate the possibility that TQ-ST segment

was present (32). deflections might actually increase as the boundarylistressing to many is approached in the pig. The suggestion that the bound-all people Bayley ary is both diffuse and dependent upon collateral)wed rather clearly blood flow in the dog raises a number of interestingsimple, if initially possibilities. First, precise localization of the bound-by the proper ap- aries of the ischemic region is less likely to be

t. Such an applica- accomplished in the dog than in the pig. Second,ere polarities and one would expect the width of this boundary to depends (TQ-ST segment upon collateral vessel patency and perfusion pressureiifferent electrode as suggested by Bayley (18). Finally, one would alsofirst situation men- expect that the area of ischemic involvement might betostulating a patchy modified in the dog by interventions which eitherluite common find- alter the level of collateral blood supply to this widents with advanced boundary or the oxygen demands of the tissue at thesecond and third boundary. In comparison, in the pig such modification

ic shape (Fig. 10C, might not be possible. Here the amount of tissuebecoming ischemic closely approaches that normally

hemic boundaries. perfused by the occluded artery (Table I). Modifica-rle increases as an tion of eventual infarct size in humans may similarly

the center to the depend upon available collateral flow.provided that the Spatial influences and precordial mapping. Al-

,idth) and the thick- though solid angle analysis was carried out for bothant. In the pig (23), precordial and epicardial locations, experimental veri-


/I/ , , , ""\

PRECORDIAL

BOUNDARYTHICKNESS (T) 1.57r0 z 0.0 mmA z 7.5mm

OP a 15.0 mm

-0.5vL

Precordi

EI

l~

Epicardium

t'FIGURE 11 Influence of boundary thickness (T) on the polarity and magnitude of the solidangle computed at precordial and epicardial locations. Area of ischemic involvement equaled13.8 cm2 (P = 0.50 radians). At the precordium, whether the boundary is well defined(T = 0) or diffuse (T = 15 mm), the solid angle values are essentially unchanged. At the epi-cardium, differences in solid angle values for the well-defined and diffuse boundariesoccur only in the region of the boundary. Although solid angle values in the ischemic regionincrease in magnitude as a distinct boundary is approached, they decrease when theboundary is wide and diffuse. See text for additional discussion.

fication was limited to the latter. In this instancethe polarity and magnitude of the TQ-ST segment ateach site could be associated with the position theelectrode maintained with respect to the ischemicregion. Similar verification at precordial sites thoughdesirable will be difficult to accomplish. Not only doesthe distance from the electrode to the heart's surfacevary depending upon the electrodes' position on theprecordium, but whether changes in ischemic area andshape, heart size, etc., decrease or increase the magni-tude of the TQ-ST deflection depends in turn upon

this distance. This matter has received little attentiondespite its significant influence on the magnitude of theTQ-ST deflection. For instance, a change in the pre-

cordial-to-heart distance of only 1 cm (i.e., from 5 to 6cm) may reduce the magnitude by over 40%. Similarvariations in this distance may occur in a variety ofphysiologic (posture, pregnancy, etc.) and pathologic(pulmonary congestion, pneumothorax, cardiac dilata-tion, etc.) conditions.

TQ-ST segment data obtained at the precordiumwhich does suggest general agreement with solidangle analysis include observations that: (a) the magni-tude of the precordial TQ-ST segment deflection in-creases with increases in ischemic area (12, 62);and (b) maximal precordial TQ-ST segment deflections

are recorded over the center of the ischemic region,decreasing at more peripheral locations (5, 63).

Nonspatial influences on the TQ-ST segment de-flection. According to the solid angle theorem (Fig.

1), the TQ-ST segment deflection is also a functionof the difference in transmembrane voltage existingbetween the normal (VmN) and ischemic (Vm.) cellsduring both diastole (TQ segment) and systole (STsegment). During early ischemia this difference is be-lieved to arise primarily as a consequence of anoxiaand potassium leakage out of the ischemic cells,and its accumulation in the surrounding extracellularspace (13, 32, 64, 65). If the difference in trans-membrane voltage between the normal and ischemiccells widens, the TQ-ST deflection magnitude in-creases. If, for any reason, these differences disappear,then so must the TQ-ST deflection.

The number of potential interventions and mech-anisms by which the TQ-ST segment deflection may bemodified in a nonspatial manner is large and diverse.These would include factors which directly modify thetransmembrane potentials of either the normal or

ischemic cells during diastole or systole (e.g., elec-trolytes, temperature, catecholamines, heart rate,metabolic inhibitors, and anti-arrhythmic therapy).Also included are agents which alter the transmem-


0.0371

EPICARDIAL

brane voltage of the ischemic tissue by changing therate and degree of potassium leakage out of thesecells. In the presence of ischemia-inhibited Na+-K+ATPase pump, chronotropic (66), inotropic (catechola-mines [30, 67], ouabain [68], calcium [69]), andmetabolic (hyperthermia [70], acidosis [711) factorsserve to accelerate potassium efflux. This facilitatespotassium accumulation in the extracellular space ofthe ischemic tissue and causes the diastolic restingpotential to decline and the action potential durationto shorten. Factors which slow potassium loss (e.g.,hypothermia, potassium solutions, and myocardial de-pressants) may help to restore the transmembranepotential of the ischemic cells back to more normalvalues.

The possibility, therefore, that a change in theischemic TQ-ST segment deflection elicited by a givenintervention can be reliably and exclusively attributedto a change in the spatial geometry of the ischemicregion (ischemic area and shape) and not to a non-spatial change in the transmembrane potential dif-ference (AVm) between the two tissues would seem tobe unlikely. For example, one indication that propran-olol may benefit patients with angina pectoris is theobservation that it decreases the magnitude of theischemic TQ-ST segment deflection (29, 72). Hemo-dynamically, propranolol decreases heart rate and theforce and velocity of contraction (73). In this way itmay reduce demands of the ischemic tissue, andmay, as shown in the dog (74), help to decreaseischemic area. As shown earlier (Figs. 3-5), at the pre-cordium a decrease in ischemic area will decreasethe solid angle and the TQ-ST deflection magnitude(12, 62). There are other, nonspatial ways, however,in which propranolol may decrease the deflection.As suggested by the work of Sarnoff et al. (75),Parker et al. (66), and Polimeni and Vassalle (76),decreases in oxygen consumption will also decreasethe rate of potassium leakage and accumulation in theischemic region. This in turn will limit the develop-ment of AVm and the TQ-ST deflection magnitudewill decline. Specific electrophysiologic actions of pro-pranolol may also affect AVm. Kupersmith et al. (77)and Wittig and Williams (17) found that propranololreturns the action potential duration of ischemic cellsback towards normal. The systolic AVm, therefore, isreduced, resulting in a decline in the ST and totalTQ-ST segment deflection. This effect of propranololmay be related to its ability to antagonize catechol-amine effects on calcium currents across the myo-cardial membrane (78). The mechanism by whichverapamil, a specific calcium current blocker, reducesthe ischemic TQ-ST segment may be similar (79, 80).

Potassium ion infusion was chosen for this studybecause it has minimal effects on oxygen consump-tion and contractile activity of the heart while having


significant electrophysiologic actions. Potassium can(a) directly alter the transmembrane potential (de-polarizes cells during diastole and shortens their ac-tion potential duration during systole [81]); (b) in-fluence the rate of potassium leakage and accumula-tion in the extracellular apace (stimulates uptake bythe cell of potassium ion [82]); and (c) increasepotassium conductance (83). The precise mech-anism by which potassium infusion reduced the magni-tude of the TQ-ST segment deflections in this studycannot be ascertained without making simultaneousmeasurements of transmembrane potentials before andduring coronary artery occlusion in the normal andischemic regions. The possibility that some of thechanges in the TQ-ST segment may be due to hyper-kalemic changes in conduction velocity (30) should alsobe investigated. The finding that the decline in theTQ-ST deflection appeared to vary in a linear fashionwith the logarithm of K+ level is interesting, however,for both EK and the action potential duration (whichcorrespond to the TQ and ST segments of the elec-trogram) both vary with the log of K+. The latterdiscovery was made by examining data from Surawicz'original study (84) and then plotting action potentialduration versus the log of K+; the resultant curvewas remarkably linear over the range of K+ valuesfrom 2.0 to 10.00 mM(r = 0.992). The finding thatchanges in the TQ-ST segment were similar at eithernonischemic or ischemic regions on the epicardiumand precordium suggests that potassium exerts a pri-marily nonspatial influence on the TQ-ST segment andthus is not dependent upon either electrode locationor ischemic geometry. This nonspatial influence of po-tassium may help to account for any previously ob-served absences of TQ-ST deflections in the ECGfromthe acute myocardial infarction patient who is also inrenal failure (85), as well as suggest an alternatemechanism by which glucose-insulin-potassium im-proves the ECGpicture of patients with an acutemyocardial infarction (86).

Preocclusion TQ-ST segment deflections. Thecause of control (preocclusion) TQ-ST deflectionsabove or below zero in a normal animal or man isnot known (87, 88). In this study heart rate was keptconstant (89) and electrode artifact, if present, wouldpresumably account only for positive deflections. Sinceneither spatial (right vs. left ventricular sites) nor hemo-dynamic correlation apparently accounted for this vari-ation, it is assumed that slight changes in the tempera-ture at the epicardial surface, because of the open-chest preparation, might be responsible (90, 91).The conventional view, therefore, that the normal TQ-ST segment is isoelectric, is not supported by thisstudy.

Solid angle theory and the experimental model.The solid angle theorem employed in this study is a

z[M Electrode Site

i a

x8.1

FIGURE 12 Left. Relationship between spherical and rectangular coordinates of a point P inspace. Right. Model of transmural ischemic region in the ventricular wall during electricalsystole used in solid angle computations.

mathematical expression which relates the potentialrecorded at an electrode site to the flow of currentsin homogeneous volume conductors of infinite extent.Its conception can be traced at least as far back asHelmholtz in 1853 (92) and was utilized extensivelyin the formulation of modem electrocardiographictheory by Wilson and Bayley (93) and McFee and Johns-ton (94). It requires that the recording electrodeand excitable tissue both be located in a homogeneousconductor of infinite extent. However, in both clini-cal (precordium) and experimental (epicardium) situa-tions the boundaries are finite and the substancessurrounding the heart (skin, bones, lung, blood)of somewhat varying conductivity. These factors, how-ever, may tend more to alter the absolute magnitudeof the electrocardiographically recorded potentialsthan their relative distribution (18, 95). For example,Bayley showed that the potential at a point on a volumeconductor of arbitrary shape (as in the case of thethorax) is approximately twice the value obtained atthe same point were the volume conductor infinitein extent (18). A review of the advantages andlimitations of solid angle analysis as related toECGinterpretation has recently been written (19).

Summary and conclusions. The TQ-ST segmentdeflection arises due to differences in transmembranepotential at the boundary between normal and ischemictissue during diastole and systole. Nonspatial factorsinfluence the magnitude of this potential gradient,while spatial factors alter the relative position of theboundary to the electrode site. Spatial factors includingelectrode location, ischemic area and shape, heartsize, and wall thickness can be quantitatively analyzed

with the solid angle theorem and then experimentallyverified in the intact porcine heart, a model whichpermits production and measurement of a well-definedarea of ischemic involvement. Nonspatial influencesare of equal importance, though less easily analyzedin a theoretical fashion. One such influence, the eleva-tion of plasma K+ levels, results in a decrease in themagnitude of the TQ-ST deflections recorded from bothischemic and nonischemic portions of the heart aftercoronary artery occlusion. The majority of pharmaco-logic agents previously shown to influence the TQ-STsegment deflection during ischemia probably do so viaboth spatial and nonspatial mechanisms.

APPENDIXSpherical coordinate transform. The spherical coordin-

ates (R,*,6) of a point P in space are illustrated in Fig. 12(left). Rectangular coordinates (X,Y,Z) are related to sphericalcoordinates by the equations:

X = R sinm cosO; Y = R sinm sinO; Z = R cos*.The distance AB between two points in space A(X,, Y1, Z,)and B(X2, Y2, Z2) can be obtained by the equation:

AB = [(XI,- X2),2+ (Yl- Y2)2 + (Zl - 2)21i.

Solid angle computations. In Fig. 12 (right) a smallsegment of ventricular wall is diagrammed illustrating thenormal and ischemic (shaded) myocardial tissue with theboundary between them. The solid angle constructed at theelectrode site N by the ischemic boundary may be computedfrom:

2wra = f2| [cosa - cosy] dO,


where a and y are obtained from the Cosine Law or:

a = cos-l[(KL2 - N-2 - LN2) -K2 LN]and

y = cos-I[(LM2 - MNTN2 - KN2)i/2-M-KNI.

ACKNOWLEDGMENTS

This work was supported by U. S. Public Health Servicegrants 5P50 HL 17648 and HL 7013202, the Louis BlockFund, and Sinton Charitable Associates of the Universityof Chicago.

REFERENCES1. Shaefer, H., and H. G. Haas. 1962. Electrocardiography.

Handb. Physiol. 1 (Sect. 2. Circulation): 380.2. Wilson, F. N., F. D. Johnston, I. G. W. Hill, and G. C.

Grout. 1933. The electrocardiogram in the later stagesof experimental myocardial infarction. Trans. Assoc. Am.Physicians Phila. 48: 154-163.

3. Nielsen, B. L. 1973. ST-segment elevation in acute myo-cardial infarction. prognostic importance. Circulation.48: 338-345.

4. Braunwald, E., and P. R. Maroko. 1974. The reduction ofinfarct size-an idea whose time (for testing) has come.Circulation. 50: 206-209.

5. Muller, J. E., P. R. Maroko, and E. Braunwald. 1975.Evaluation of precordial electrocardiographic mapping asa means of assessing changes in myocardial ischemicinjury. Circulation. 52: 16-27.

6. Flaherty, J. T. 1976. Clinical uses of precordial ST-segment mapping and the pathophysiology of ST-segmentvoltage changes. Circulation. 53 (Suppl. I): 85-87.

7. Ross, J., Jr. 1976. Electrocardiographic ST-segment anal-ysis in the characterization of myocardial ischemia andinfarction. Circulation. 53 (Suppl. I): 73-81.

8. Holland, R. P., and H. Brooks. 1975. ST segment mapping:Fact and fallacy. Circulation. 52 (Suppl. II): 6. (Abstr.)

9. Norris, R. M., C. Barratt-Boyes, M. K. Heng, and B. N.Singh. 1976. Failure of ST segment elevation to predictseverity of acute myocardial infarction. Br. Heart J. 38:85-92.

10. Fozzard, H. A., and D. S. DasGupta. 1977. ST seg-ment potentials and mapping: Theory and experiment.Circulation. 54: 533-537.

11. Reese, L., S. Scheidt, and T. Killip. 1973. Variabilityof precordial ST segment maps after acute myocardialinfarction in man. Circulation. 48 (Suppl. IV): 38.(Abstr.)

12. Holland, R. P., and H. Brooks. 1975. Precordial andepicardial surface potentials during myocardial ischemiain the pig. A theoretical and experimental analysis of theTQ and ST segments. Circ. Res. 37: 471-480.

13. Holland, R. P., and H. Brooks. 1977. TQ-ST segmentmapping: A critical review and analysis of current con-cepts. Am. J. Cardiol. In press.

14. Samson, W. E., and A. M. Scher. 1960. Mechanism of S-Tsegment alteration during acute myocardial injury. Circ.Res. 8: 780-787.

15. Bruyneel, K. J. J. 1975. Use of moving epicardial elec-trodes in defining ST-segment changes after acutecoronary occlusion in the baboon. Relation to primaryventricular fibrillation. Am. Heart J. 89: 731-741.

16. Cohen, D., and L. A. Kaufman. 1975. Magnetic deter-mination of the relationship between the S-T segment

shift and the injury current produced by coronary arteryocclusion. Circ. Res. 36: 414-424.

17. Wittig, J. H., and E. M. V. Williams. 1976. Mechanismof decrease of ST elevation by propranolol in ischemicmyocardium. Circulation. 54 (Suppl. II): 15. (Abstr.)

18. Bayley, R. H. 1958. Electrocardiographic analysis. Bio-physical Principles of Electrocardiography. Paul B.Hoeber, Inc., New York. 19-34, 184-196, and 207-222.

19. Holland, R. P., and M. F. Amsdorf. 1977. Solid angletheory and the electrocardiogram. Physiologic and quanti-tative interpretations. Prog. Cardiovasc. Dis. In press.

20. Plonsey, R. 1969. Bioelectric Phenomena. McGraw-HillBook Co., New York. 202-275.

21. Plonsey, R. 1974. An evaluation of several cardiac ac-tivation models. J. Electrocardiol. (San Diego). 7: 237-244.

22. Howe, B. B., P. A. Fehn, and R. R. Pensinger. 1968.Comparative anatomical studies of the coronary arteriesof canine and porcine heart. I. Free ventricular walls.Acta Anat. 71: 13-21.

23. Schaper, W. 1971. Comparative Arteriography of the Col-lateral Circulation of the Heart. North Holland Publish-ing Company, Amsterdam. 29-50.

24. Brooks, H., J. Al-Sadir, J. Schwartz, B. Rich, P. Harper,and L. Resnekov. 1975. Biventricular dynamics duringquantitated anteroseptal infarction in the porcine heart.Am. J. Cardiol. 36: 765-775.

25. Van Citters, R. L., D. L. Franklin, and R. F. Rush-mer. 1964. Left ventricular dynamics in dogs duringanesthesia with alpha-chloralose and sodium pentobar-bital. Am. J. Cardiol. 13: 349-354.

26. Lichtig, C., and H. Brooks. 1975. Myocardial ultrastruc-ture and function during progressive early ischemiain the intact heart. J. Thorac Cardiovasc. Surg. 70: 309-315.

27. Weiner, J. M., C. S. Apstein, J. H. Arthur, and W. B. Hood,Jr. 1974. Persistence of myocardial injury following briefperiods of coronary occlusion. Am. J. Cardiol. 33: 177.(Abstr.)

28. Swain, J. A., H. G. Hanley, K. R. Genth, R. M. Lewis,C. J. Hartley, and A. Schartz. 1976. Prolonged depres-sion of regional myocardial function following transientischemia in conscious dogs. Physiologist. 19: 383. (Abstr.)

29. Maroko, P. R., J. K. Kjekshus, B. E. Sobel, T. Watanabe,J. W. Covell, J. Ross, Jr., and E. Braunwald. 1971.Factors influencing infarct size following coronary arteryocclusions. Circulation. 43: 67-82.

30. Holland, R. P., and H. Brooks. 1976. The QRS com-plex during myocardial ischemia. An experimentalanalysis in the porcine heart. J. Clin. Invest. 57: 541-550.

31. Ginks, W. R., H. D. Sybers, P. R. Maroko, J. W. Covell,B. E. Sobel, and J. Ross, Jr. 1972. Coronary arteryreperfusion. II. Reduction of myocardial infarct size at1 week after coronary occlusion. J. Clin. Invest. 51:2717-2723.

32. Holland, R., F. Pashkow, and H. Brooks. 1974. Atraumaticepicardial electrode and rapid sampling switch for cardiacsurface mapping. J. Appl. Physiol. 37: 424-427.

33. Redwood, D. R., E. R. Smith, and S. E. Epstein. 1972.Coronary artery occlusion in the conscious dog. Effectsof alterations in heart rate and arterial pressure on thedegree of myocardial ischemia. Circulation. 46: 323-332.

34. Snedecor, G. W., and W. G. Cochran. 1967. StatisticalMethods. Iowa State University Press, Ames, Iowa.6th edition. 593 pp.

35. Pruitt, R. D., and F. Valencia. 1948. The immediateelectrocardiographic effects of circumscribed myocardial


injuries: An experimental study. Am. Heart J. 35: 161-197.

36. Hellerstein, H. K., and L. N. Katz. 1958. The electricaleffects of injury at various myocardial locations.Am. Heart

J. 36: 184-220.37. Craib, W. H. 1927-29. A study of the electrical field

surrounding active heart muscle. Heart. 14: 71-109.38. Karlsson, J., G. H. Templeton, and J. T. Willerson.

1973. Relationship between epicardial S-T segmentchanges and myocardial metabolism during acutecoronary insufficiency. Circ. Res. 32: 725-730.

39. Griggs, D. M., Jr., V. V. Tchokoev, and C. C. Chen.1972. Transmural differences in ventricular tissue sub-strate levels due to coronary constriction. Am. J. Physiol.222: 705-709.

40. Irwin, R. G., and F. R. Cobb. 1976. Relationship be-tween epicardial (Ep) ST segment changes and extent ofhistologic myocardial infarction (MI) in awake dogs.Circulation 54 (Suppl. II): 15. (Abstr.)

41. Hamlin, R. L., C. R. Smith, H. K. Pellerstein, and R. S.Pipers. 1969. Alterations in QRSduring ischemia of theleft ventricular free-wall in goats. J. Electrocardiol.(San Diego). 2: 223-228.

42. Sayen, J. J., W. F. Sheldon, G. Peirce, and P. T. Kuo.1958. Polarographic oxygen, the epicardial electrocardio-gram and muscle contraction in experimental acuteregional ischemia of the left ventricle. Circ. Res. 6: 779-798.

43. Sayen, J. J., G. Peirce, A. H. Katcher, and W. F. Sheldon.1961. Correlation of intramyocardial electrocardiogramswith polarographic oxygen and contractility in the non-ischemic and regionally ischemic left ventricle. Circ.Res. 9: 1268-1279.

44. Khuri, S. F., J. T. Flaherty, J. B. O'Riordan, B. Pitt, R. K.Brawley, J. S. Donohoo, and V. L. Gott. 1975. Changesin intramyocardial ST segment voltage and gas tensionswith regional myocardial ischemia in the dog. Circ. Res.37: 455-463.

45. Ross, J., Jr., and D. Franklin. 1976. Analysis of regionalmyocardial function, dimensions, and wall thickness inthe characterization of myocardial ischemia and infarc-tion. Circulation. 53 (Suppl. I): 88-92.

46. Wegria, R., M. Segers, R. P. Keating, and H. P. Ward.1949. Relationship between the reduction in coronaryflow and the appearance of electrocardiographic changes.Am. Heart J. 38: 90-96.

47. Timogiannakis, G., I. Amende, E. Martinez, and M.Thomas. 1974. ST segment dediation and regional myo-cardial blood flow during experimental partial coronaryartery occlusion. Cardiovasc. Res. 8: 469-477.

48. Smith, H. J., B. N. Singh, R. M. Norris, M. B. John, andP. J. Hurley. 1975. Changes in myocardial blood flowand S-T segment elevation following coronary artery oc-clusion in dogs. Circ. Res. 36: 697-705.

49. Heng, M. K., B. N. Singh, and R. M. Norris. 1975. Rela-tionship between epicardial ST segment elevation andexperimental infarct size. Circulation. 52 (Suppl. II):6. (Abstr.)

50. Prinzmetal, M., H. C. Bergman, H. E. Kruger, L. L.Schwartz, B. Simkin, and S. S. Sobin. 1948. Studies onthe coronary circulation. III. Collateral circulation of beat-ing human and dog hearts with coronary occlusion.Am. HeartJ. 35: 689-717.

51. Durrer, D., P. Formijne, R. T. Van Dam, J. Buller,A. A. W. Van Lier, and F. L. Meyler. 1961. In revivedhuman fetal heart and in acute and chronic coronaryocclusion. The electrocardiogram in normal and someabnormal conditions. Am. Heart J. 61: 303-316.

52. Wolferth, C. C., S. Bellet, M. M. Livezey, and F. D.Murphy. 1945. Negative displacement of the RS-T seg-ment in the electrocardiogram and its relationships topositive displacement; an experimental study. Am. Heart

J. 29: 220-245.53. Ekmekci, A., H. Toyoshima, J. K. Kwoczynski, T. Nagaya,

and M. Prinzmetal. 1961. Angina pectoris. IV. Clinicaland experimental difference between ischemia with S-Televation and ischemia with S-T depression. Am. J.Cardiol. 7: 412-426.

54. Bayley, R. H. 1942. An interpretation of the injury andthe ischemic effects of myocardial infarction in ac-cordance with the laws which determine the flow ofelectric currents in homogeneous volume conductors,and in accordance with relevant pathologic changes.Am. HeartJ. 24: 514-528.

55. Cohn, P. F., E. S. Kirk, J. M. Downey, E. H. Sonnen-blick, and R. Gorlin. 1973. Autoradiographic evaluationof myocardial collateral circulation in the canine heart.Cardiovasc. Res. 7: 181-185.

56. Fulton, W. F. M. 1956. Chronic generalized myocardialischaemia with advanced coronary artery disease. Br.HeartJ. 18: 341-354.

57. Grayson, J., and M. Irvine. 1968. Myocardial infarction inthe monkey: Studies on the collateral circulation afteracute coronary occlusion. Cardiovasc. Res. 2: 170-178.

58. Pitt, B. 1959. Interarterial coronary anastomoses. Occur-rence in normal hearts and in certain pathologic condi-tions. Circulation. 20: 816-822.

59. Schlesinger, M. J. 1938. An injection plus dissection studyof coronary artery occlusions and anastomoses. Am. Heart

J. 15: 528-568.60. Becker, L. C., R. Ferreira, and M. Thomas. 1973.

Mapping of left ventricular blood flow with radioactivemicrospheres in experimental coronary artery occlusion.Cardiovasc. Res. 7: 391-400.

61. Ekmekci, A., H. Toyoshima, J. K. Kwoczynski, T. Nagaya,and M. Prinzmetal. 1961. Angina pectoris. V. Giant Rand receding S wave in myocardial ischemia and certainnonischemic conditions. Am. J. Cardiol. 7: 521-532.

62. Capone, R. J., A. S. Most, and P. A. Sydlik. 1975. Pre-cordial ST segment mapping. A sensitive technique forthe evaluation of myocardial injury. Chest. 67: 577-582.

63. Reid, D. S., L. J. Pelides, and J. P. Shillingford.1971. Surface mapping of RS-T segment in acute myo-cardial infarction. Br. Heart J. 33: 370-374.

64. Johnson, E. A. 1976. First electrocardiographic sign ofmyocardial ischemia: An electrophysiological conjecture.Circulation. 53 (Suppl. I): 82-84.

65. Case, R. B., M. G. Nasser, and R. S. Crampton. 1969.Biochemical aspects of early myocardial ischemia. Am. J.Cardiol. 24: 766-775.

66. Parker, J. O., M. A. Chiong, R. 0. West, and R. B. Case.1970. The effect of ischemia and alterations of heartrate on myocardial potassium in man. Circulation.42: 205-217.

67. Vassalle, M., and 0. Bamabei. 1971. Norepinephrineand potassium fluxes in cardiac Purkinje fibers. PfluegersArch. Eur. J. Physiol. 322: 287-303.

68. Hajdu, S., and E. Leonard. 1959. The cellular basis ofcardiac glycoside action. Pharmacol. Rev. 11: 173-209.

69. Nayler, W. G. 1967. Calcium exchange in cardiac muscle:A basic mechanism of drug action. Am. Heart J. 73: 379-394.

70. Page, E., R. J. Goerke, and S. R. Storm. 1964. Catheart muscle in vitro. IV. Inhibition of transport inquiescent muscles.J. Gen. Physiol. 47: 531-543.

71. Brown, E. B., Jr., and B. Goott. 1963. Intracellular hy-


drogen ion changes and potassium movement. Am. J.Physiol. 204: 765-770.

72. Becker, L. D., R. Ferreira, and M. Thomas. 1975. Effectof propranolol and isoprenaline on regional left ventricu-lar blood flow in experimental myocardial ischemia.Cardiovasc. Res. 9: 178-186.

73. Graham, T. P., Jr., J. Ross, Jr., J. W. Covell, E. H. Sonnen-blick, and R. L. Clancy. 1967. Myocardial oxygen con-sumption in acute experimental cardiac depression.Circ. Res. 21: 123-138.

74. Reimer, K. A., M. M. Rasmussen, and R. B. Jennings.1973. Reduction by propranolol of myocardial necrosisfollowing temporary coronary artery occlusion in dogs.Circ. Res. 33: 353-363.

75. Sarnoff, S. J., J. P. Gilmore, R. H. McDonald, Jr., W. M.Daggett, M. L. Weisfeldt, and P. B. Mansfield. 1966. Rela-tionship between myocardial K+ balance, 02 consump-tion, and contractility. Am. J. Physiol. 211: 361-375.

76. Polimeni, P. I., and M. Vassalle. 1970. Potassium fluxesin Purkinje and ventricular muscle fibers during restand activity. Am. J. Physiol. 218: 1381-1388.

77. Kupersmith, J., H. Shiang, R. S. Litwak, and M. V. Her-man. 1976. Electrophysiological and antiarrhythmic ef-fects of propranolol in canine acute myocardial ischemia.Circ. Res. 38: 302-307.

78. Wit, A. L., B. F. Hoffman, and M. R. Rosen. 1975. Elec-trophysiology and pharmacology of cardiac arrhythmias.IX. Cardiac electrophysiologic effects of beta adrenergicreceptor stimulation and blockade. Part A. Am. Heart J.90: 521-533.

79. Belardinelli, L. 1975. Reduction of acute myocardialischemia with verapamil. Evaluation by epicardial map-ping. Arq. Bras. Cardiol. 28: 599-607.

80. Shigenobu, K., J. A. Schneider, and N. Sperelakis. 1974.Verapamil blockade of slow Na+ and Ca++ responsesin myocardial cells. J. Pharmacol. Exp. Ther. 190:280-288.

81. Gettes, L. S., B. Surawicz, and J. C. Shiue. 1962. Effectof high K, low K, and quinidine on QRSduration andventricular action potential. Am. J. Physiol. 203: 1135-1140.

82. Goerke, J., and E. Page. 1965. Cat heart muscle invitro. VI. Potassium exchange in papillary muscles. J.Gen. Physiol. 48: 933-948.

83. Carmeliet, E. E., 1961. Chloride and Potassium Perme-ability in Cardiac Purldnje Fibers. Presses Acad6miquesEuropeennes, Brussels. 152 pp.

84. Surawicz, B. 1967. Relationship between electrcardio-gram and electrolytes. Am. HeartJ. 73: 814-834.

85. Bellet, S., P. C. Gazes, and W. A. Steiger. 1950. Theeffect of potassium on the electrocardiogram in the normaldog and in dogs with myocardial infarction. Am. J. Med.Sci. 220: 237-246.

86. Sodi-Pallares, D., A. Bisteni, G. A. Medrano, A. DeMicheli, J. Ponce De Leon, E. Calva, B. L. Fishleder, M. R.Testelli, and B. L. Miller. 1966. The polarizing treat-ment in cardiovascular conditions. Experimental basisand clinical applications. In Electrolytes and Cardio-vascular Diseases. E. Bajusz, editor. S. Karger, NewYork. 2: 198-238.

87. Edeiken, J. 1954. Elevation of the RS-T segment, ap-parent or real, in the right precordial leads as a probablenormal variant. Am. Heart . 48: 331-339.

88. Goldman, M. J. 1953. RS-T segment elevation inmid- and left precordial leads as a normal variant.Am. HeartJ. 46: 817-820.

89. Sjostrand, T. 1950. The relationship between the heartfrequency and the S-T level of the electrocardiogram.Acta Med. Scand. 138: 201-210.

90. Hoffman, B. F. 1956. Temperature effects on cardiactransmembrane potentials. In The Physiology of InducedHypothennia. N. A. S. N. R. C. Pubi. 451: 302-324.

91. Toyoshima, H., M. Prinzmetal, M. Horiba, Y. Kobayashi,Y. Mizuno, R. Nakayama, and K. Yamada. 1965. Thenature of normal and abnormal electrcardiograms. VIII.Relation of ST segment and T wave changes to intra-cellular potentials. Arch. Intern. Med. 115: 4-16.

92. Helmholtz, H. von. 1853. Uber einige Gesetze derVertheilung elektrischer Strome in korperlichen Leitern,mit Answendung auf die thier-elektrischen Versuche.Ann. Phys. Chem. 89: 211, 353.

93. Wilson, F. N., and R. H. Bayley. 1950. The electricfield of an eccentric dipole in a homogeneous sphericalconducting medium. Circulation 1: 84-92.

94. McFee, R., and F. D. Johnston. 1953. Electrocardio-graphic leads. I. Introduction. Circulation. 8: 554-568.

95. Boineau, J. P., and M. S. Spach. 1968. 'Te relationshipbetween the electrocardiogram and the electrical activityof the heart. J. Electrocardiol. (San Diego). 1: 117-124.


Spatial Nonspatial Influences on...

Documents

Transcript of Spatial Nonspatial Influences on...