the relationship the anatomic position the heart...

The Relationship between the Anatomic

Position of the Heart and theElectrocardiogram

A Criticism of "Unipolar" ElectrocardiographyBy ROBERT P. GRANT, M.D.

Using a method for identifying precisely the lie of the ventricular structures in the body at au-

topsy, the anatomic position of the heart was compared with the directions of the QRS electricalforces in a large number of subjects with and without heart disease. While the mean QRS axis variedthrough 180 degrees in these subjects, the anatomic position of the left ventricle varied less than45 degrees. No instances of significant rotation of the heart around its long axis were encountered,and it is shown that the unipolar electrocardiographic criteria for position and rotation of the hearthave little validity. An explanation for the axis deviations of ventricular hypertrophy is offered.

T HAT the form of the QRS complex isinfluenced by the position of the heartin the chest has been known since

Einthoven demonstrated changes in the direc-tion of the mean QRS axis with respiration.Later, the introduction of unipolar precordialleads made it possible to study the relation-ship between the electrocardiogram and posi-tion of the heart in three-dimensional space,and Wilson and co-workers devised a methodfor recognizing five different heart positionsfrom combinations of QRS complexes in theunipolar limb and precordial leads.' Subsequentworkers have greatly extended this method inorder to define in more detail the position ofthe heart and its components from the electro-cardiogram.

Wilson's method was the outgrowth of aparticular hypothesis regarding the factorswhich govern the form of the QRS complex.On the basis of experiments performed byLewis and in his own laboratory, Wilson con-cluded that the unipolar lead might be con-sidered to resemble a "direct lead," that is,one placed directly on the heart and recordingprincipally the electrical events of the portionof the heart underlying it.2, This hypothesis

From the Clinic of General Medicine and Experi-mental Therapeutics of the National Heart Institute,Section of Cardiodynamics, U. S. Public HealthService Hospital, Baltimore, Md.

890

suggested that when a unipolar electrode faceda particular region of the heart it would writea QRS complex with a specific contour for thatregion, and there would be specific QRSpatterns for many different regions of theheart. Thus, when a particular QRS contourwas seen in one or another unipolar lead, itcould be concluded that the region of the heartknown to yield that QRS pattern must be fac-ing that electrode location. This hypothesishas been extensively applied to clinical electro-cardiography and detailed systems have beendevised for detecting various types and de-grees of rotation of the heart, the lie of theseptum, anterior and posterior tilting of theapex, the location of sites of infarction in the leftventricle, and similar features from the contoursof QRS complexes in the various unipolar limband precordial leads.4-7

In its clinical application this approach hasbecome known as the "unipolar method" forelectrocardiographic interpretation. However,although widely accepted, it has some seriousshortcomings. The most important of these isas Wilson recognized, that, only an extremelybrief portion of the unipolar lead QRS com-plex, the "intrinsic deflection," can be con-sidered to be written by the underlying regionof the heart. And in a body surface lead eventhis is written to only a limited extent by theunderlying portion of the heart, which is why

Circulation, Volume VII, June, 1953

by guest on May 14, 2018

http://circ.ahajournals.org/D

ownloaded from

http://circ.ahajournals.org/

ROBERT P. GRANT

in clinical practice it is called the "iintrinisicoiddeflection." The portions of the complex beforeand after the instantaneous intrinsicoid de-flection are a much larger portion of the com-plex and represent the electrical activity of allthe other regions of the heart as they undergoactivation. To this extent a unipolar lead isnot a direct lead registering the electrical eventsof a single region of the heart, and it was forthese reasons that Wilson called it a "semi-direct" lead. The same caution was used indescribing the five positions of the heart, forhe termed them "electrical positions" of theheart and pointed out that they might or mightnot be related to the actual anatomic positionof the heart.These uncertainties regarding the validity

of the clinical application of this hypothesishave tended to be justified by subsequentexperimental studies. These have shown that,for clinical purposes, the contour of the uni-polar lead QRS complex is more accuratelyand more rationally treated by considering itwritten by all regions of the heart rather thanprincipally by the region facing the electrode.These studies, which are discussed else-where8' 9, 10 indicate that the validity of the"unipolar" method of electrocardiographicinterpretation is open to question.

There is also an important practical short-coming in the "unipolar" approach to theclinical electrocardiogram. The body surfaceQRS patterns for the various specific regionsof the heart and the electrocardiographiccriteria for position and rotation of the hearthave been based largely upon conjecture, andthere have been no careful anatomic studies inthe human subject to test the validity of thesecriteria. Even the notion that the heart rotateson its long axis with one or another type ofventricular hypertrophy, a basic concept inelectrocardiographic as well as x-ray andcardiologic literature, is unproved, for therehave been no systematic anatomic studiesdemonstrating that such rotation occurs. Thefollowing studies were undertaken to examinethe extent to which the electrocardiogram re-flects the anatomic properties of the heart.A method was devised for accurately measuringthe position and rotation of the heart at post-

mortem examination and this information wascompared with the electrocardiographic dataobtained shortly before death in a large numberof subjects with and without heart disease.

METHODS AND RESULTSTwenty-four consecutive cases were studied.

Seven of the patients died with clinically manifesthypertensive cardiovascular disease. Two subjectsdied of mitral stenosis and insufficiency, one ofcor pulmonale due to pulmonary emphysema, oneof aortic stenosis due to rheumatic heart disease,one of syphilitic aortic insufficiency, and two ofarteriosclerotic heart disease. The remaining 10 sub-jects had no recognized heart disease clinically or atautopsy. Patients with clinical, electrocardiographic,or pathologic evidence of myocardial infarction wereexcluded from this series as were also those with aQRS interval in excess of .12 second.

Bipolar and unipolar limb leads and precordialV leads were recorded on the Sanborn Viso-Cardiettewith the patient in the recumbent position. In allsubjects an electrocardiogram had been recordedwithin three days of death and at least three electro-cardiograms were obtained from all subjects duringthe terminal illness. The tracings were analyzed forthe direction of the mean spatial QRS vector, themean spatial vector for the first .02 second of theQRS interval, and the frontal plane projection ofthe QRS loop. The method for determining thedirections of the spatial vectors and the QRS loopsfrom the conventional electrocardiogram have beendescribed elsewhere.9

In accurately defining the anatomic locationof the various regions of the heart at autopsy, thefollowing method was used. After removal of thesternal plate and before intrathoracic contents weredisturbed, the pericardium was carefully opened anda thin, sharply pointed metal spear was gentlythrust anterior-posteriorly through the heart, per-pendicular to the autopsy table. Then, another spearwas passed from left to right through the chest walland through the ventricular mass to be parallelwith the frontal plane of the body and at right anglesto the anterior-posterior spear. The heart was nowremoved by severing the great vessels. The auriclesand the right ventricle were gently packed withformalin-soaked gauze to restore the contour ofthe heart, and a third spear was inserted perpendic-ular to the plane defined by the first two, to beparallel with the long axis of the body. The impaledheart was then mounted on a frame to support thespears in their proper rectangular relationships andthe whole placed in a jar containing 10 per centformalin. In certain instances the coronary vesselswere injected with formalin to improve fixation.

After 36 to 48 hours of fixation, photographswere made of the anterior, inferior (or diaphrag-matic) and lateral surfaces of the intact heart, using

891



ownloaded from


UNIPOLAR ELECTROCARDIOGRAPHY

the three spears as precise reference guides to de-fine exactly how the heart lay within the chest atthe time of death for each of these views. Afterthese photographs were taken and without disturb-ing the spears, the auricles, great vessels, epicardialfat and connective tissue, coronary vessels, andvalves were carefully removed leaving only theventricular muscle mass. Photographs were again

to determine the exact lie of the left ventricle andthe interventricular septum in the body at the timeof death. From these two measurements the typeand degree of rotation of the heart on its long axis,transverse axis and anterior-posterior axis could beprecisely determined. The various parts of the heartwere weighed and measured at each stage of dissec-tion. Eleven hearts were studied by this method,

yR VL~ Vf.4

;r jI

fi VI

2160- RT.65 IN. WT. 165 La AGECYVD g DECOMP.

FIG. 1. The method of study, showing the three stages of dissection, with three views of the heartat each stage. The region where the free wall of the right ventricle was dissected from the septum wasmarked with indelible pencil before photography in order to show more clearly the relation of theseptum to the remainder of the left ventricle. This line has been retouched with white ink in thelowest photograph. In this specimen, the aortic leaflet of the mitral valve has been preserved; insome of the later illustrations this valve was removed showing the mitral-aortic orifice of the leftventricle.

taken of the heart at this stage of dissection usingthe spears as reference axes for the three views.Next, the free wall of the right ventricle was cutaway where it joins the septum, leaving only theleft ventricle and septum; and photographs of thethree views were taken at this stage of dissection,using the spears to obtain exactly the same positionsas at the earlier stages of dissection. From the photo-graphs at the third stage of dissection it was possible

nine of which are illustrated. In the remainingcases the lie of the left ventricle and the septumwere determined by making serial sections throughthe heart parallel with the transverse plane as de-fined by the spears.

In some cases the dissection was greatly facili-tated by immersing the heart in slightly acidifiedwater and bringing it to a boil. The epicardial fatand connective tissue were much more easily removed

892



ownloaded from


ROBERT P. GRANT

following this treatment giving a much "cleaner"specimen and one which lent itself more satisfactorilyto dehydrating and clearing or "mummifying" for

SUP

~~~~ ~

the dissolution of only intracardiac fat and con-nective tissue is unlikely.

Figure 1 shows the photographs in one case. The

2184 Ht. 69 in., Wt. 165 lb., Age 50Emphysema E Cor Pulmonole

Total heart weightVentriclesLV. and septumR.V. (free wall )

500 gm.280 gm.190 gm.*90gm.

2167 Ht. 74in., Wt. 114 lb., Age 25Hodgkins Disease

Total heart weightVentriclesLV. and septumRV. (free wal)

280 gm.180gm.150gm.30gm...'

QR ~4SEPT

FIG. 2. Anterior view of the heart at the three stages of dissection. The right ventricular surfaceof the septum is shaded with oblique lines to facilitate identification. On the triaxial system beloweach series of drawings are plotted the mean spatial QRS vector and the frontal plane projectionof the QRS loop with .02 and .04 second markers. In addition, the spatial directions of the ana-tomic long axis of the left ventricle and the long axis of the septum (which was selected as theline from the septal leaflet of the tricuspid valve to the apex of the right ventricle) are drawn onthe triaxial system to show the relationship between the directions of the electrical forces and thelie of the heart. The duration of the QRS interval for the given case is indicated in parenthesisbeside the triaxial figure.

demonstration purposes. However, it reduced theweight of the ventricular mass by as much as 30per cent and reduced some of the measurements by10 to 20 per cent. That this weight loss represented

upper three photographs show anterior views of theheart at the three stages of dissection: first, of theintact heart; second, of the ventricular muscle mass;and, third, of the isolated left ventricle and septum.

893



ownloaded from



The second row of photographs show views of theinferior or diaphragmatic surface of the heart atthe same three stages of dissection; and the bottomrow shows views of the lateral surface (looked atfrom the patient's right side) at the same threestages of dissection.

was selected as the line from the septal leaflet ofthe tricuspid valve to the apex of the right ventricle)are drawn on the triaxial system to show the rela-tionship between the directions of the electricalforces and the lie of the heart. The duration of theQRS interval for the given case is indicated in paren-

2162 Ht. 69 in. Wt. over 300 lb.,

H.C.V. D.Total heart wt.VentriclesL.V. and septumR.V. (free wall)

950gm.45 0gm.340gm.

I Ogm.

(.09)" " QRS2166 Ht. 66 in. Wt. 100 lb. Age 24

Aortic Stenosis

Total heart wt. 650 gm.Ventricles 355 gm.

/# As i LV L.V. and septum 280 gm.

SEPT R.V. (free wall) 75 gm.FIG. 3. See text and legend for figure 2.

Only the anterior views of the heart at the threestages of dissection are shown in the illustrationsfor the other cases (figs. 2, 3, 4). Pen and ink tracingswere made from the photographs in order to facili-tate reproduction. Since the focal distance was notthe same for each heart, the scale of one inch isshown in the upper left drawing in each case. Thearrow drawn in the photograph of the left ventricleindicates the direction of the long axis of this ven-

tricle. On the triaxial system below each series ofdrawings are plotted the mean spatial QRS vectorand the frontal plane projection of the QRS loopwith .02 and .04 second markers. In addition, thespatial directions of the anatomic long axis of theleft ventricle and the long axis of the septum (which

thesis beside the triaxial figure. The age, sex, height,weight, postmortem diagnosis, and weights of theheart at various stages of dissection* are recorded

* The weight of the dissected ventricle is a muchmore accurate index of the presence and degree ofventricular hypertrophy than is the total weight ofthe heart. As has been known by pathologists sinceMuller's classic dissections in 1883, the nonventric-ular portions of the heart (auricles, epicardial fat,and connective tissue) may represent from 20 to 60per cent of the total weight of the heart. Averagenormal weights are: left ventricle and septum, 140to 180 Gm., free wall of the right ventricle 30 to 50Gm. 11, 12

894



ownloaded from


ROBERT P. GRANT

at the lower right side of each figure. Nine caseswere selected from the series for illustration. Of theseven cases shown in figures 1, 2, 3 and 4, cases2162, 2166, 2177, and 2180 represent left ventricularhypertrophy, 2184 represents right ventricular hy-pertrophy, 2187 is an instance of arterioscleroticheart failure with little or no ventricular hyper-trophy, and 2167 is a normal heart. A relativelylarge number of cases of left ventricular hypertrophy

QRS cycle, (3) the long axis of the left ven-tricle, and (4) the long axis of the septum areshown projected on the frontal plane of thebody. It can be seen that there were widevariations in direction of the mean QRS vectors,from marked left axis deviation to definiteright axis deviation. However, while the direc-

"I'.11

- 11..h

1,'1

ORS

2187 Ht.65 in. Wt. 165 lb., Age 52A.S.H.D. ' failure

Total heart weight 440gm.Ventricles 22 0gm.L.V. and septum 160gm.R.V. (free wall) 60gm.

are shown because of the wide variation in directionof the mean QRS vector in these cases.

In figure 5 are shown the relevant anatomicand electrocardiographic findings in all cases

studied. The range of variation in directionof (1) the mean QRS vector, (2) the mean .02second vector, which is the mean vector for theearliest electrical forces generated during the

1t. 126 lb., Age 44

weight 740 gm.360 gm.

)tum 280 gm.asll ) 80 gm.

tion of the mean QRS vector varied through180 degrees, the long axis of the left ventriclevaried in direction less than 45 degrees. The.02 vector varied less than the mean QRSvector in direction, and the long axis of theseptum varied less than 30 degrees in direction.The implications of these findings will be dis-cussed shortly.

2177 Ht. 68 in., WH.C.V.D.

rtf QRS Total heart aVentriclesL.V. and sep

LV R.Y. (free u[SEPTFIG. 4. See text and legend for figure 2.

895



ownloaded from



To measure the rotations of the heart aroundits long axis, the direction of the axis of theleft ventricle wvas compared with the directionof the septal axis in each case. It was foundthat these two axes had relatively the samerelationship to one another in all cases studied.Thus, although the series included instances ofright and left ventricular hypertrophy as wellas normal hearts, no instances of significantclockwise or counterclockwise rotation wereencountered. This can be seen in the illustra-tions, for it will be noted that the septum hasnearly exactly the same relationship to thebody of the left ventricle in all views in allsubjects.

BA

MEAN ORS

VECTOR

Fi(;. 5. (A) The range in direction of the mean

QRS vector and the mean .02 second vector in the

cases studied. (B) The range in direction of the longaxis of the left ventricle and of the septum in these

cases.

Before discussing the electrocardiographicand anatomic implications of the findings,it is necessary to examine the validity of themethod used. Obviously, the characteristicsof the heart at post mortem are not identicalwith those in the living subject. The dyinghuman heart is believed to cease beating insystole, yet the QRS electrical forces, which are

the basis of the present study, are writtenduring cardiac diastole. Furthermore, Friedmanhas shown that the heart is more completelyemptied at death than it is by systole in theliving subject, and that the volume of the heart(the tissue mass plus the contained blood) as

studied by three-dimensional x-ray methodsmay be as much as 50 per cent less at deaththan it was in the living state.'3

Accordingly, the over-all size of the heartsshown in the illustrations is not the same as isobtained in the living state when QRS forceswvere generated. However, the general position

of the heart in the chest postmortem examina-tion was probably very nearly the same duringlife, and the relationships of the various partsof the ventricles to one another shown in theillustrations were unquestionably the sameduring life. Furthermore, since in all hearts thepositions of the right and left ventricle relativeto one another and to the three body axeswere remarkably similar at postmortem ex-amination, in spite of wide variations in heartsize, it is probable that these positions werealso remarkably similar during life.

DIscussIoNThese studies indicate that the QRS axis

deviations commonly seen with ventricularhypertrophy are not due to anatomic rotationof the heart. Indeed, the lie of the left and rightventricles in the body was remarkably similarin all hearts, both normal and hypertrophied.To be sure, among normal subjects, those witha more horizontal left ventricle tended to havea QRS axis (or "mean QRS vector" in thenomenclature of vector electrocardiography)which was horizontal to about the same degree.However, in the presence of left ventricularhypertrophy the electrical axis was oftenmarkedly rotated leftward while the anatomicposition of the left ventricle in these cases wasessentially the same as in the normal heart.

Furthermore, in all hearts, both normal andhypertrophied, the long axes of the septum andleft ventricle had essentially the same direc-tions relative to one another and to the bodyaxes. Thus, there was no significant clockwiseor counterclockwise rotation of the heart inany of the 24 hearts studied. The septumdid not rotate with the development of ventric-ular hypertrophy, and there was little varia-tion in the degree of anterior tilt of the longaxis of the heart, as can be seen in the illustra-tions. The detection of rotation of the hearton its long axis has been a prominent part ofthe unipolar method of electrocardiographicinterpretation, with, in some instances, a rota-tion of over 90 degrees inferred from the elec-trocardiogram. From the present studies itis apparent that such rotation rarely if everoccurs, and the "unipolar" electrocardiographic

896



ownloaded from


ROBERT P. GRANT

criteria for position and rotation of the hearthave little validity.The notion that the heart rotates in one or

another direction in various heart diseasesseems to have arisen principally in electro-cardiographic literature, for no basis for it hasbeen found in anatomic literature. Evidentlyit arose as the simplest way to explain therather marked changes in direction of the meanQRS axis with ventricular hypertrophy andfrom an imperfect picture of the structure of theheart. Although anatomists had clearly de-fined the relationships of the various parts ofthe heart to one another, they had not beenconcerned with the relationship of these struc-tures to various regions of the body surface.The usual diagram of the ventricular heart

in electrocardiographic and other cardiologicliterature is of two chambers lying side by side,separated by an anterior-posterior partition,the septum, with the outflow orifices of thetwo chambers facing superiorly. Schematizedin this way as a vertically pedunculated struc-ture, it is easy to see how the idea could arisethat the heart might rotate with hypertrophyof one or the other ventricle.

Actually, however, as can be seen in theseillustrations, the left ventricle is a conicalstructure which in the supine subject lies rela-tively horizontally in the body. Its inflow andoutflow orifices face the right side of the body.The septum is structurally a part of the leftventricle, representing that part of the leftventricle to which the free wall of the rightventricle is attached. The septum does notextend anterior-posteriorly but instead is rela-tively parallel with the frontal plane, the rightventricle lying altogether anteriorly to theleft ventricle. The remainder of the left ven-tricle, the "free wall," which is considered inelectrocardiographic literature to be a lateralstructure facing the left side of the body,actually includes nearly 300 degrees of thecircumference of the left ventricle, and facessuperiorly, posteriorly and inferiorly, a fairlylarge portion of it resting on the diaphragm.The inflow and outflow orifices of both

ventricles communicate quite directly withrelatively fixed intrathoracic structures. Whilein the left ventricle the inflow and outflow

tracts are relatively parallel with one another,in the right ventricle they form a right angle(fig. 6). Thus, the right ventricle lies across theleft ventricle, embracing it at its base, andanchoring the heart by its superior and in-ferior attachments. When these anatomic

RIGHT VENTRICLE

LEFT VENTRICLE 1..FIG. 6. The directions of inflow and outflow tracts

of the right and left ventricles.

features are considered, it can be seen that theheart tends to be pyramidal with the base ofthe pyramid facing the right side of the bodyand the apices of this base firmly attached tothe other intrathoracic structures. Thus, whilethe heart may shift superiorly or inferiorly, itis prevented from rotation on its long axis by

897



ownloaded from


89 UNIOLAR ELECTROCAII1)IOGlRAPIIY

the fixation of the base. Then, as these studieshave shown, when right or left ventricularhypertrophy develops, the involved ventriclesimply increases in thickness and length in thissame relationship to other parts of the heart.

Since anatomic rotation does not accountfor the marked axis deviation of left ventricularhypertrophy, what is its explanation? In orderto answer this, it is first necessary to review thefactors governing the sequence in which thevarious parts of the heart are excited."4 15During the first .03 to .04 second of the QRSinterval, endocardial surfaces of the right andleft ventricles undergo activation. Since thesesurfaces are not changed in position by hyper-trophy, the vectors during this early part of theQRS interval should be relatively normallydirected in the presence of venitricular hyper-trophy. This proved to be the case in, thepresent series, for the direction of the mean.02 vector varied mtuch less than did that ofthe mean QRS vector and had a more fixedrelationship with the lie of the heart in all cases.The forces after the first .03 second of the

QRS interval are effectively generated fromthe epicardial surfaces of the ventricles andtend to be directed perpendicularly to thegenerating surface. Since the epicardial area isgreater than the endocardial area, these vec-tors are much greater in magnitude than theendocardial vectors. Furthermore, the epi-cardial forces from the right ventricle arenormally smaller in magnitude and are gener-ated earlier than those of the left ventricle.Accordingly, the direction of the mean QRSvector in the normal subject and in the subjectwith left ventricular hypertrophy dependsprincipally upon the characteristics of epi-cardial depolarization of the left ventricle.There are two factors which determine thesecharacteristics: (1) The conduction system,that is the sequence in which the excitationprocess is distributed to the various regionsof the left ventricle, and (2) the thickness ofthe wall of the left ventricle, for it takes ap-proximately 10 times longer for excitation totraverse the wall thans to spread an equal dis-tance along the conduction pathways of theendocardium.The characteristics of left ventricular con-

duction have been recently reviewed byMurray.16 For present purposes it need onlybe pointed out that left ventricular excitationextends first down the endocardial surface ofthe anterior and diaphragmatic walls of theleft ventricle toward the apex, and thenspreads toward the base on the superior andposterior endocardial surfaces, ending usuallyat the remotest posterobasal region of the freewall of this ventricle. With normal wall thick-ness, epicardial excitation tends to follow thissame sequence, but more slowly. The terminalQRS vectors in the normal subject tend to beresultants of vectors from the last epicardialregions to be depolarized, and therefore theseterminal forces tend to be directed posteriorlyand somewhat leftward. This is the explanationfor the direction of the mean QRS vector inthe normal subject.To be sure minor variations in this sequence

of left ventricular depolarization are extremelycommon and, indeed, the fact that everyonehas a slightly different QRS loop is probablyprimarily due to the fact that this sequence ofendocardial and epicardial depolarization variesslightly from person to person. When the loopbecomes prolonged in duration or abnormalin contour it is often possible to determine thelocation of the region of disturbed intra-ventricular conduction by studying the direc-tion of the initial QRS forces.16 And, since theepicardial vectors tend to be directed perpen-dicularly to the surface where they are gener-ated, the direction of terminal vectors willalso reveal the delayed region, for, whenplotted from the electrocardiogram as vectors,they tend to point toward the blocked region.9

Increases in left ventricular wall thickness asin left ventricular hypertrophy delay thetransmyocardial progression of excitation. Thistends to extend the period of epicardial ex-citation'7 and more clearly separate in time thevectors generated from the very last regionof the left ventricle from those generated frommore intermediate regions. Because of this theterminal vectors are more exclusively con-tributed from the superior and posterior re-gions of the left ventricle in the hypertrophiedthan in the normal heart, and hence have amore superior and posterior direction than

898



ownloaded from


ROBERT P. GRANT

normally. In addition, because of the increasedsurface area which accompanies the hyper-trophy, the vectors generated from the hyper-trophied region are increased in magnitude.Accordingly, in left ventricular hypertrophy theterminal QRS vectors are more leftward andposterior in direction and are greater in magni-tude than in the normal subject, even thoughthe heart has essentially the same anatomicposition in the two instances. This then is theexplanation for the marked leftward QRS axisin left ventricular hypertrophy, with presentviews of ventricular excitation.There are other factors which, n o doubt,

contribute to the leftward direction of meanQRS vector in left ventricular hypertrophy.The anatomic studies of the hearts in thisseries of cases have shown that with hyper-trophy the diameter and length of the wallof the left ventricle are increased, but thediameter of the mitral-aortic orifice usuallyremains normal. This causes the contour of thefree wall of the left ventricle to become con-siderably more bowed superiorly and pos-teriorly than normally (fig. 7). As a conse-quence, the epicardial surface of the base of theleft ventricle comes to face more superiorly,and, in some cases, even rightward. This alsocauses the terminal vectors of the left ventricleto be directed more superiorly and posteriorlythan in the normal subject.

Still another factor has been suggested byWilson.18 The wave of excitation travels muchfaster along the endocardial surface than it doesacross the wall of the left ventricle. When thewall is increased in thickness, this differencein rate of movement causes the wave of ex-citation advancing across the myocardial wallto become more and more oblique to the surfaceof the heart. As a consequence, the epicardialQRS vectors tend to be oblique to the surfacerather than perpendicular as in the normalsubject. This would cause the direction ofmean QRS vectors to be more leftward andposterior in the presence of left ventricularhypertrophy than in the normal heart.Not all hearts with left ventricular hyper-

trophy show a left axis deviation. For example,in case 2180 the mean QRS vector is somewhatvertical in direction in the frontal plane, in

2170 and 2206 it is horizontal, while in 2166and 2162 it is markedly leftward. Yet the leftventricle had essentially the same position inall these cases, and no systematic feature ofwall thickness, length of inflow or outflow tract,or left ventricular weight could be identifiedto account for this variation. It has been sug-gested that a normally directed QRS axis inthe presence of ventricular hypertrophy imdi-

FIG. 7. Comparison of a normal and an hyper-trophied left ventricle, right lateral view above,inferior or diaphragmatic view below. The free wallof the right ventricle has been removed. Largerheart: case 2227, mitral stenosis and aortic insuf-ficiency due to chronic rheumatic fever; weight ofthe left ventricle and septum: 360 Gm. Smaller heart:case 2209, bronchogenic carcinoma; weight of the leftventricle and septum: 130 Gm.

cates the presence of combined right and leftventricular hypertrophy."9 This was not thecase in the present series as can be seen bycomparing the right ventricular weights. Itmust be concluded that the mean QRS vectoris vertical or horizontal in these cases because ofanomalous intraventricular conduction. Per-haps delay in the posterior subdivisions of theleft bundle branch is responsible for the rela-tively vertical axis in some of these cases.16

89'9



ownloaded from



When left ventricular hypertrophy is marked,the directions of the early QRS forces are oftenchanged, becoming more or less parallel withthe mean QRS vector (and manifested byabsence of an initial R wave in V1 and V2and lead III). Although no instance of thistype with a QRS interval measuring less than.12 second wxas encountered in the presentseries, there are two possible explanations forthis abnormality, either of which may beresponsible in a given case: (1) it may representselective delay in the anterior divisions of theleft bundle branch,'6 or, (2) it may representthe presence of significant dilatation in addi-tion to the hypertrophy. The anatomic studiesdone in connection with this series of caseshave confirmed the observation of Kirch thatin left ventricular dilatation the inflow tractand superior wvall of the left ventricle may oftenbecome more markedly stretched and bowedthan the outflow tract and inferior wall. Underthese circumstances the vectors generated fromthe endocardium of the inflow tract tend to beincreased in magnitude. The resultant vectorsduring the first .03 second of the QRS loopwould be dominated by these larger posteriorlydirected vectors and point leftward and pos-teriorly, thus producing Q waves in V1, V2and V13.

There was a single instance of right ven-tricular hypertrophy in the present series, acase of cor pulmonale due to pulmonary em-physema (case 2184), and the mean QRS vectorwas relatively vertical in this case. Althoughthe descent of the diaphragms in pulmonaryemphysema may often cause the heart torotate somewhat rightward producing rightaxis deviation, this proved not to be the casein the present instance. As can be seen in theillustration, the right ventricle became moreprominent but did not produce significant rota-tion of the heart as a whole. Accordingly, thevertical axis in this case must be due, not torotation of the heart, but to the larger epi-cardial surface area and increased wall thick-ness of the right ventricle. This would cause theepicardial vectors from the right ventricle tobe increased in magnitude and slightly delayedin time, causing the mean QRS vector to have

a more rightward direction than normally.That the mean QRS vector xvas directed in-feriorly rather than anteriorly may be due tothe fact that the inflowx tract of the right ven-tricle was more hypertrophied than the outflowvtract.2" The measurements of the right ventriclein this heart indicated that while the lengthof the inflowv and outflow tracts were aboutequally increased, the xvall of the infloxv tractwas significantly more increased in thicknessthan was the wall of the outflowx tract. As canbe seen in the protocol in the illustration, thefree +Xall of the right ventricle xvas nearly threetimes normal weight in this case.There are txvo other aspects of these studies

which are of electrocardiographic interest. Oneof these concerns the electrical forces generatedby the septum. It has been widely acceptedthat the earliest region of the ventricular myo-cardium from which QRS forces are generatedis the interventricular septum, producing aforce traveling from the left to the right sideof the septum.7' 22 If this force were perpen-dicular to the septum it would be significantlydifferent in direction from the remainder of theresultant forces during the QRS interval. Onthis supposition, the appearance of Q waves inleads on the left lateral chest have been at-tributed to a septal electrical force, and vari-ations in the distribution of these Q wvaves inthe various unipolar leads have been used toidentify rotation of the septum.

In the present study, it was found that theseptum has very nearly the same position inthe chest in all hearts, both normal and ab-normal as can be seen in the illustrations. Itis relatively parallel with the frontal plane,its apex somewhat anterior to its base, andin none of the 24 cases studied was significantrotation of the septum encountered. Whenthe lie of the septum was compared with thedirection of the mean force for the first .02second of the QRS interval, it was found thatthis vector tended to be more tangential thanperpendicular to the septum (fig. 5). Thus,there appears to be little validity in the notionof a "septal Q wave." As others have suggested,if there is a measurable QRS force due to ex-clusive septal activation, it is likely that it is

900



ownloaded from


ROBERT P. GRANT

of considerably shorter duration than .02 andmay be too small in magnitude to be recordedfrom body surface leads.14

Another implication of these anatomicstudies concerns the nomenclature of myo-cardial infarctions. In electrocardiography,"posterior" myocardial infarction was sodesignated because of the notion that theseptum extended anterior-posteriorly, definingan anterior and posterior region of the leftventricle. From this reasoning, infarcts of theleft ventricle adjacent to the "posterior" at-tachment of the septum were called "posterior"infarcts. However, as can be seen in the il-lustrations and as has been long recognized byanatomists, the interventricular septum is notan anterior-posterior structure, but more nearlylies parallel with the frontal plane. The sulcusfor the descending branch of the right coronaryartery and the adj acent region of the leftventricle where "posterior" infarcts occur,rests on the diaphragm. Hence infarcts in thisregion are more accurately termed "dia-phragmatic" or "inferior" than posterior.The notion that "posterior" infarcts lie on

the anatomically posterior wall of the leftventricle has been a source of confusion inclinical electrocardiography as well. The QRSand T wave deformities of myocardial in-farction are due to abnormal QRS and T forcespointing away from the location of the in-farct.9 18 Therefore, if a "posterior" infarct wereactually due to a lesion on the anatomicallyposterior wall of the heart, the QRS and T ab-normalities should be most conspicuous in theprecordial leads, for these leads effectivelymeasure anterior-posterior forces, while thelimb leads measure forces parallel to the frontalplane of the body. Actually, however, the mostconspicuous QRS and T abnormalities of whatis called "posterior" infarction are seen in thelimb leads (Q, T3, with Q waves in leads IIand Vf, and relatively normal precordialleads in most cases). When the directions ofthe abnormal QRS and T forces of such "pos-terior" myocardial infarctions are measured,they prove to point away from the diaphrag-matic and not the posterior region of the heart.Thus, the electrocardiographic evidence con-

firms the point of view expressed in the previousparagraph, that "posterior" myocardial in-farctions actually lie in the diaphragmatic andnot in the anatomically posterior wall of theleft ventricle.

SUMMARY1. A method was devised for accurately

defining the postmortem position of the leftventricle and the interventricular septum interms of three body axes. These anatomicfindings were compared with the characteristicsof the mean QRS and mean .02 second vectorsof the premortem electrocardiogram in 24 sub-jects, 14 of whom had varying degrees of rightor left ventricular hypertrophy and/or dilata-tion, and 10 of whom had no heart disease.

2. It was found that the left ventricle andthe interventricular septum have remarkablysimilar positions in the body in all subjectsregardless of the presence or type of heartdisease, and no instances of significant clock-wise or counterclockwise rotation of the heartwere encountered. On the basis of these find-ings it is concluded that the criteria in "uni-polar" electrocardiography for identifying theposition and rotation of the heart and its com-ponents have little validity.

3. In view of the fact that rotation of theheart is not responsible for the QRS axis de-viation of the ventricular hypertrophy analternative explanation is offered which isconsistent with prevailing concepts of ventricu-lar excitation.

ACKNOWLEDGMENTThe author wishes to express his gratitude to

Drs. Llewellvn Ashburn, Ronald A. Welsh andFrank London, Department of Pathology of theU. S. Public Health Service Hospital, Baltimore,for their generous cooperation in these studies.

SUMARLO EsPAROL

La posici6n anat6mica del coraz6n se com-par6 con las direcciones de las fuerzas el6ctricasQRS en uni nlimero de sujetos con y sin en-fermedad cardiaca, usando un metodo paraprecisar la localizaci6n de las estructurasventriculares en el cuerpo durante la autopsia.Aunque el QRS promedio vari6 por 180° en

901



ownloaded from



estos sujetos, la posiciOln anatomica del ventri-culo izquierdo vario menios de 45 grados. Enningun instante se encontro rotacion significa-tiva del corazo'n alrededor de su eje, y se de-muestra que el criterio electrocardiogralficounipolar para posicion y rotacion del corazontiene poco valor. Una explicacion para lasdesviacioiies del eje en hipertrofia ventricularse ofrece.

REFERENCES

1 WILSON, F. N., ROSENBAUM, F. F., ERLANGER,H., KOSSMAN, C. E., HECHT, H., COTRIM, N.,MENZES D}E OLIVERI, R., SCARSI, R., AND BAR-KER, P. S.: Precordial electrocardiogram. Am.Heart J. 27: 18, 1944.

2 LEWIS, T., AND ROTHSCHILD, M. A.: The excita-tory process in the dog's heart, the ventricles.Phil. Tr. Roy. Soc. London, s.B. 206: 181, 1915.

3 WILSON, F. N., ROSENBAUM, F. F., AND JOHNSTON,F. D.: The interpretation of the ventricularcomplex of the electrocardiogram. AdvancesInt. Med. 2: 1, 1947.

4 GOLDBERGER, E.: Unipolar Lead Electrocardiog-raphy, ed. 2. Philadelphia, Lea & Febiger, 1949.

5LIPMAN, B. S., AND AIASSIE, E.: Clinical UnipolarElectrocardiography. Chicago, Year Book Pub-lishers, 1951.

6 BARKER, J. Al.: Interpretation of the UnipolarElectrocardiogram. New York, Appleton-Cen-tury-Crofts, 1952.

LEPESHKIN, E.: Modern Electrocardiography. Bal-timore, Williams & Wilkins Co., 1951. Vol. I.

GRANT, R. P.: Spatial vector electrocardiography.Circulation 2: 676, 1950.

9 AND ESTES, E. H.: Spatial Vector Electro-cardiography. Philadelphia, Blakiston, 1951.

10 GRISHMAN, A., AND SHERLIS, L.: Spatial Vector-cardiography. Philadelphia, Saunders, 1952.

11 DULL, M.: Gewichtsbestimmungen der reinenMuskelmasse beider herzkammern bei normalerund pathologischer Herzbelastung. Beitr. path.Anat. 105: 337, 1941.

12 MERKEL, H., AND NADOLNY, G.: Das verhalten derMusklemasse des rechten und linken Ventrikelsbei Hypertonie. Ztschr. Kreislaufforsch. 40: 341,1951.

13 FRIEDMAN, E. C.: The residual blood of the heart.Am. Heart J. 39: 397, 1950.

14 WILSON, F. N., HILL, I. G. W., AND JOHNSTON,F. D.: Form of the electrocardiogram in experi-mental myocardial infaretion. Am. Heart J. 9:596, 1934.

11 BAYLEY, R. H.: On certain applications of modernelectrocardiographic theory to interpretationof electrocardiograms which indicate myocardialdisease. Am. Heart J. 26: 769, 1943.

16 MURRAY, R. H.: To be published.17 WILSON, F. N., AND HERRMANN, G.: Relation of

QRS interval to ventricular weight. Heart 15:135, 1930.

18 , MACLEOD, A. G., AND BARKER, P. S.: Inter-pretation of initial deflections of ventricularcomplex of electrocardiogram. Am. Heart J. 6:637, 1931.

19 LANGENDORF, R., HURWITZ, AI., AND KATZ, L. N.:Electrocardiographic patterns of combinedstrain. Brit. Heart J. 5: 27, 1943.

20 KIRCH, E.: Pathogenese und Folgen der Dilata-tion und der Hypertrophie des Herzens. Klin.Wchnschr. 9: 767, 817, 1930.

21 PODKAMINSKY, N. A.: Entwicklung der Hyper-trophie und der Dilatation des Herzens inAbhangigkeit von seinem functionellen Bezeich-ungen. Virchow's Arch. path. Anat. 284: 92,1932.

22 GARDBERG, AM., AND ASHMAN, R.: The QRS com-plex of the electrocardiogram. Arch. Int. Med.72: 210, 1943.

902



ownloaded from


ROBERT P. GRANTElectrocardiogram: A Criticism of "Unipolar" ElectrocardiographyThe Relationship between the Anatomic Position of the Heart and the

Print ISSN: 0009-7322. Online ISSN: 1524-4539 Copyright © 1953 American Heart Association, Inc. All rights reserved.

is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231Circulation doi: 10.1161/01.CIR.7.6.890

1953;7:890-902Circulation.

http://circ.ahajournals.org/content/7/6/890the World Wide Web at:

The online version of this article, along with updated information and services, is located on

http://circ.ahajournals.org//subscriptions/

is online at: Circulation Information about subscribing to Subscriptions:

http://www.lww.com/reprints Information about reprints can be found online at: Reprints:

document.

Permissions and Rights Question and Answer Further information about this process is available in therequested is located, click Request Permissions in the middle column of the Web page under Services.the Editorial Office. Once the online version of the published article for which permission is being

can be obtained via RightsLink, a service of the Copyright Clearance Center, notCirculationpublished in Requests for permissions to reproduce figures, tables, or portions of articles originallyPermissions:



ownloaded from

http://circ.ahajournals.org/content/7/6/890

http://www.ahajournals.org/site/rights/

http://www.lww.com/reprints

http://circ.ahajournals.org//subscriptions/


the relationship the anatomic position the heart...

Documents