control myocardial oxygen consumption:...

Control of Myocardial Oxygen Consumption: Relative

Influence of Contractile State and Tension Development

THOMASP. GRAHAM,JR., JAMESW. CovELL, EDMUNDH. SONNENBLICK,JOHNRoss, JR., and EUGENEBRAuNwALDFrom the Cardiology Branch, National Heart Institute,Bethesda, Maryland

A B S T R A C T Myocardial oxygen consumptionwas measured in 11 anesthetized, open-chest dogsin order to compare in the same heart the relativeinfluence on oxygen usage of tension developmentand the contractile or inotropic state, as reflectedin Vmax, the maximum velocity of shortening of theunloaded contractile elements. The isovolumetri-cally contracting left ventricle was studied withleft ventricular volume, heart rate, and systemicperfusion rate controlled. Wall tension, contractileelement velocity, and Vmax were calculated. Peakdeveloped tension was increased at a constantVma. by increasing ventricular volume, and theeffect Onl oxygen consumption was determined.Oxygen utilization was then redetermined at anincreased Vmax but at a constant peak developedtension by infusing norepinephrine (0.76 to 7.6jug/min) and decreasing ventricular volume tomatch the tension existing before norepinephrineinfusion. Oxygen consumption consistently in-creased with increases in both developed tensionand VnaX With the following multiple regressionequation relating these variables: myocardial oxy-gen consumption (dl/beat per 100 g in LV) = K+ 0.25 peak developed tension (g/cm2) + 1.43Vmx,,, (cm/sec). These data indicate that the oxy-gen cost of augmentation of contractility is sub-

This work was presented in part before the Society forClinical Investigation, Atlantic City, N. J., 30 April 1967.J. Clin. Invest. 46: 1063. (Abstr.), and the AmericanPhysiological Society, Washington, D. C., 25 August1967.

Address requests for reprints to Dr. Eugene Braun-wald, National Heart Institute, Bethesda, Md.

Receized for putblication 12 July 1967 and in revisedform 2 October 1967.

stantial, can be independent of any change in fibershortening, and is similar in order of magnitudeto the effect of alterations in tension development

INTRODUCTIONThe delineation of the factors that regulate

myocardial oxygen consumption (MVo2) has beenthe -objective of extensive research. It was demon-strated by Rohde in 1912 that MVo2 varied di-rectly as a function of the product of developedpressure and heart rate in the cat isovolumetricleft ventricular preparation (1). Shortly thereafter,Evans and Matsuoka (2) concluded from studieson the dog heart-lung preparation.that "there isa relation between the tension set up on contrac-tion and the metabolism of the contractile tissue."These demonstrations, that tension 2 developmentis an important determinant of MVo2, have beenverified repeatedly (3-8). It has also been pro-posed that the heart's oxygen utilization is deter-mined primarily by a closely related variable, thework performed by the contractile elements(CEW) (9). Several subsequent studies, how-ever, have. demonstrated that CEWdoes notinvariably correlate with MVo2 (10-12).

Recently, it has also become apparent that walltension is not the' sole determinant of the heart'soxygen utilization, and there is' substantial evi-dence that changes in myocardial contractility, orinotropic state, as reflected in Vmax, the maximumvelocity of shortening of the unloaded-contractileelements, can exert considerable influence on

1 Defined herein as either total tensile force or asforce/unit cross-sectional area (stress).

The Journal of Clinical Investigation Volume 47 1968 375

MVo2 (11, 13-18). This effect has been demon-strated both for positive inotropic interventions,which increase MVo2 when myocardial tension isheld constant or falls (13-18), and also for nega-tive inotropic interventions which reduce MVo2 atconstant levels of tension (11). It has been sug-gested that this effect of the contractile or inotro-pic state on MVo2 might be related to changes in(a) the velocity of shortening of the contractileelements (CE) and myocardial fibers, (b) theextent of shortening of the CE and of the myo-cardial fibers, and (c) the contractile state asreflected in the extrapolated velocity of shorteningof the unloaded CE (Vmax). This last-namedvariable has the property of being largely inde-pendent of changes in myocardial fiber length( 19), and for this reason it is the best indexcurrently available for evaluating the contractilestate of the heart (20-21).

Thus, while both the contractile state of themyocardium and the developed tension have beenshown to affect MVo2, no direct comparisons ofthe relative quantitative influences of these twovariables have been made. Furthermore, the effecton MVo2 of increasing contractility without alter-ing either the extent of fiber shortening or theCEWhas not been defined. The present investi-gation, therefore, was undertaken to assess, inthe same heart, the relative effects on MVo2 ofchanges in tension development and in the con-tractile state of the myocardium. This assessmentwas accomplished by utilizing an isovolumetric,left ventricular preparation in which wall tensioncould be altered at a constant contractility, and,conversely, the contractile state could be increasedat a relatively constant level of peak developedtension and CEW.

METHODSExperiviental preparation. 11 mongrel dogs weighing

from 17.3 to 24.6 kg were anesthetized with pentobarbi-tal sodium (35 mg/kg), and a bilateral thoracotomy wasperformed. Ventilation was provided by a positive pres-sure respirator. The basic experimental design was firstto investigate the effect on MVo2 of changing peak de-veloped tension (PDT) by altering intraventricularvolume at a constant contractility. Thereafter, the effecton MVo1, of changing contractility at a relatively con-stant PDT was investigated by infusing norepinephrineand decreasing the intraventricular volume in order toachieve levels of PDT similar to those observed beforenorepinephrine.

The experimental preparation is depicted schematicallyin Fig. 1. The right heart was bypassed, and the sys-temic and coronary venous blood were drained sepa-rately into a large reservoir. The blood, which was oxy-genated with either 100%o 02 or with 97%o 02 and 3%oC02, was passed through a heat exchanger and was thenpumped back into a femoral artery to provide retrogradesystemic perfusion at flow rates which were held es-sentialy constant in each experiment and averaged 110ml/min per kg (range 73-146). Heart rate was main-tained constant at an average of 141 beats/min (range100-160) by right atrial stimulation after crushing thesinoatrial node.

In order to control ventricular volume and to achieveisovolumic contractions, we inserted a latex balloon intothe left ventricle through the mitral valve and securedit at the apex to a large bore (0.5 mm) metal cannulaused to measure left ventricular pressure within theballoon (Fig. 1). The opposite end of the balloon wastied to a small plastic disc which fitted snugly into theleft ventricular outflow tract just below the aortic valveand prevented herniation of the balloon through theaortic valve. Stabilization of the disc in the outflow tractwas achieved by attaching it to a polyethylene-coveredguide wire which was anchored at its opposite end tothe metal cannula in the balloon. A separate plasticgrooved disc was sutured into the mitral annulus toprevent balloon protrusion at this point. Left ventricu-lar Thebesian flow was drained through perforations inthe mitral disc.

The intraventricular balloon was filled with saline,and volumes were varied using a calibrated syringe. Thevolume of the nondistended balloon was large relativeto that of the left ventricular cavity, and the pressure-volume relationship of the balloon alone was determinedat the conclusion of each experiment. In 7 of the 11 ex-periments, there was a detectable pressure (average =2.8 mmHg, range 1-5) attributable to the balloon aloneat the largest volumes used during the experiments.Whenever it occurred, the pressure contributed by theballoon was substracted from the recorded left ventricularpressure.

M1easurements. Aortic pressure (0-200 mmHg) andleft ventricular pressure (LVP, 0-40 mmHg and 0-200mmHg) were measured with cannulae attached directlyto Statham P23Db transducers (Statham InstrumentsInc., Los Angeles, Calif.). The first derivative of LVPwith respect to time (LV dp/dt) was measured withan electronic differentiator.2 The dynamic characteristicsof this instrument have been described previously (21).

Coronary blood flow minus left Thebesian flow (CBF)was measured by timed collection of the right heartdrainage. The latter was sampled to provide the coronaryvenous 02 content. The coronary arteriovenous oxygendifference (A-V 02) was monitored continuously witha Guyton analyzer (22) to indicate steady-state condi-tions, and arterial and coronary venous blood samples

2Electronic Gear, Inc., Valley Stream, N. Y., ModelNo. 5602.

376 Graham, Covell, Sonnenblick, Ross, and Braunwald

CANNULAINI NTRAVENTRICULAR

BALL ON

FIGURE 1 Isovolunmetric left ven-tricular preparation with right heartbypass and retrograde systemicperfusion. Stim., stimulator withelectrodes on the right atrium; SG,strain gauge.

were analyzed manometrically for 02 content (23).MVo2 was calculated as the product of CBF and themanometrically determined A-V 02. Experiments -wereexcluded from analysis if arterial unsaturation occurred.The right ventricle was drained by gravity and main-tained in a collapsed state. Because right ventricularmechanical activity was minimal, MVo2 was expressedas ml/min per 100 g of LV or as Al/beat per 100 g ofLV. In two experiments the right coronary artery wasligated approximately 2 cm from its origin in order toinvestigate the possibility that the collapsed right ven-tricle made a variable contribution to the MVo2 and im-portantly influenced the results. However, the changesin M¶7o2 in these experiments were similar to thosewithout right coronary ligation.

Myocardial wall tension (Purr) 8 was monitored withan analogue computer 4 by a method described previously(11) to facilitate the matching of this variable before

3 P = LV pressure in g/cm', r = internal radius incentimeters.

4 Electronic Associates, Inc., Long Branch, N. J.,Model TR 20.

and after norepinephrine. The tension, electrocardiogram,A-V 02, LV dp/dt, LVP, and aortic pressure were re-corded on a multichannel oscillograph5 at paper speedsof 100 mm/sec. A more definitive estimate of wall forcenormalized for wall thickness (stress) was calculatedat the conclusion of the experiment, as described below.

Experimental protocol. MVo2 was first measured atvarying levels of PDT achieved by varying ventricularvolume at the same contractile state. Thereafter nor-epinephrine (0.76-7.6 ,g/min) wvas infused at a constantrate into the ascending aorta just above the aortic valve(Fig. 1). When a steady state of increased contractilitywas attained, as evidenced by a stable peak LVP andLV dp/dt, the volume of the LV balloon was decreasedin order to match the monitored PDT of each of theprenorepinephrine controls, and MVo2 was redetermined.Whenever PDT was not returned to the exact controllevel, the effect on MVo2 of changing contractility wasdetermined by extrapolation to the appropriate value ofPDT, as shown in Fig. 4. In one experiment three sepa-rate states of contractility were produced by varying

5 Sanborn Co., Waltham, Mass., Model No. 350.

Control of Myocardial Oxygen Consumption 377

norepinephrine concentrations while maintaining PDTrelatively constant. This procedure was then repeatedat a different level of PDT.

In 6 of the 11 dogs, MVo2 was determined at the con-clusion of the experiment after the intraventricular bal-loon had been emptied and the heart had been arrestedwith 15%o potassium chloride. After the conclusion of allexperiments, the weight of the LV including the intra-ventricular septum was determined after the right ven-tricle and atria had been removed. This value was usedto express MXVo2/100 g of LV and to derive LV musclevolume for use in the calculations.

CalculationsThe following calculations were performed utilizing

a digital computer with manual conversion of LVP, LVdp/dt, end-diastolic volume, and LV muscle volume fromanalog to digital form. Stress, velocity, and power werederived at 10-msec intervals throughout three representa-tive contractions for each experimental state. The threecontractions were analyzed in order to obtain a largenumber of data points on the stress-velocity curves andthus achieve a more accurate estimation of Vmax.

Mean wall stress. Wall stress in g/cm' was calcu-lated as Pr/2h where P = intraventricular pressure ing/cm', r = internal radius in centimeters, assuming aspherical ventricle, and h = wall thickness in centimetersderived from internal volume and muscle volume. Thedetails of this calculation have been discussed previously(11). Stress was chosen as the force parameter foranalysis in an attempt to normalize for the cross-sectionalarea of the muscle producing the force. PDT was definedas peak tension minus one-half of the resting tension(RT) (11). Total developed tension (TDT) was definedas the area under the developed tension-time curve andwas calculated as bf. (T) dt - (RT) /2 (n), where a =onset of systole, b = the point at which tension falls tothe level of resting tension, and n =the number of 10-msec intervals over which the integration was performed.The definition of peak developed tension was designed toprovide some consideration of the progressive unloadingof the parallel elastic components (PEC) during fibershortening that is predicted by the three-componentmuscle model. Since in this model resting tension isborne partially or entirely by the PEC, during shortening,resting tension must be transferred progressively to thecontractile elements (CE) and series elastic components(SEG). In cardiac muscle, when resting tensions arehigh, the magnitude of this transfer of load during sys-tole can be quite large. Since in the intact heart thePEC stiffness and the distribution of resting tension be-tween PEC and SEC are unknown, the assumption wasmade that!at peak tension one-half of the resting tensionhad been transferred to the contractile system.

Contractile element velocity (VcE). In an isovolumiccontraction, Vcz is equal to the rate of lengthening ofthe. series 'elastic components (SEC) of the myocardium.This rate of series elastic lengthening (VsE) is directlyproportional to the rate of stress development (dT/dt)and inversely related to the stiffness of the SEC, (dT/

dl). This relationship then can be written as: VCE= V8E= dl/dt = (dT/dt) /(dT/dl), where dT/dt = the derivativeof Pr/2h with respect to time and dT/dl = 28T. In thisform Vcu is given in muscle lengths or circumferencesper second and can be converted to centimeters per sec-ond by multiplying by the instantaneous circumference(24). Vmax was determined by extrapolation of VcE tozero load when VCE (cm/sec) was plotted as a functionof stress, Pr/2h. Also, V.,.x was determined by the Hillequation (25); the values of tension and velocity atthree separate points on the force-velocity curve, ex-cluding PO, were used to solve the Hill equation (P + a)(V + b) = K, where P = force and V = velocity, andthe simultaneous equations were solved for Vmax, i.e., thepoint on the curve at which P equals zero. Althoughsimilar results were obtained by extrapolation and alge-braic solution, it is recognized that neither approachprovides an absolute measurement of V.,x.

The extent of contractile element (CE) shortening perbeat was calculated as Cf. (VrEp) dt where c = p)eak ten-sion, the point at which CE shortening ceases in an iso-volumic contraction.

Contractile element work (CEW). During isovolu-mic contraction, CEWwas calculated as bfa (T) (VCE)dt. CEWwas calculated both with T determined asPr/2h and as (Pirr2) (14). The results were directionallysimilar with either calculation, and only the CEWusingPr/2h is presented in the results section.

RESULTS

MVo2 always increased, both when PDT wasincreased at a constant Vmax and when Vm.x wasincreased at a constant PDT. In 15 observationson 11 dogs MVo2 increased from an average of41.5 ± 2.2 (SE) to 49.0 ± 2.4 ul/beat per 100 gof LV (+18%o) when PDT was increased froman average of 49.7 ± 5.0 to 80.2 ± 5.7 g/cm2(+60%) by increasing ventricular volume (TableI). With this increase in PDT, total developedtension increased to an even greater extent, from9.0 + 1.1 to 15.9 ± 2.9 g/sec per cm2 (+76%).The increase in MVo2 was characterized by asignificant increase in both coronary blood flowand coronary A-V 02 difference.

In 22 observations on 11 dogs MVo2 increasedfrom an average of 40.7 + 2.7 to 57.0 + 3.5 ~J/beat per 100 g of LV (+ 40%o ) when Vmax was in-creased from an average of 41.8 ± 2.4 to 54.6 ±2.9 cm/sec (+31%o) (Table II). This increasein MVo2 occurred despite a constant level ofPDT and a 21%o decrease in total developed ten-sion. Again, this change in MVo2 was achievedwith both an increase in coronary flow and awidening of the A-V 02 difference.


TABLE I

Effect of MVo2 of Increasing Tensionat a Constant Vm.,

Control Increased volume

LVELDP, mmHg 3.3±(0.6 5.84066*LV vol. ml 194±1 24±41*MAP, mmHg 81±4 81±4$4Flow, ml/min per kg 114±5 114+5tPeak LVP, mmHg 6946 97±6*Max dp/dt, mmHg per sec 894±84 1189±78*PDT, g/cm2 per beat 49.7 ±5.0 80.2 ±t5.7*TDT, g-sec/cm2 per beat 9.0 ±41.1 15.9 ±2.9*TPT, sec 0.130±+0.003 0.135±+0.0034CEW, g-m/beat 27.0±-2.9 44.3 +3.0*ALCE, cm/beat 1.50±0.03 1.48±0.024CIBF, ml/min 134±-10 142 +10*A-V 02, vol % 5.1 ±0.02 5.7 ±0.2*MVo2, -l/beat per 100 g of LV 41.5 ±2.2 49.0 ±2.4*

LVEDP, left ventricular end-diastolic pressure; LV vol, left ventricu-lar volume; MAP, mean aortic pressure; flow, systemic flow; peak LVPpeak left ventricular pressure; max dp/dt, maximum rate of rise ofLVP; PDT, peak developed tension; TDT, total developed tension;'[PT, time from onset of contraction to peak tension; CEW,contractileelement work; ALCE, extent of contractile element shortening; CBFcoronary blood flow; A-V 02, coronary arteriovenous oxygen difference;M'~o5, myocardial oxygen consumption. Data represent mean values± si. for 15 observations in 11 dogs.

* P < 0.001 Student's t test for paried differences.4 Not significant.

When PDT was increased at a constant Vmax,the calculated extent of Contractile element short-ening (ALCE) did not change significantly. How-ever, when Vmax was increased at a constant PDT,

A B

Vokwo -1. mI %bkm* 24 mlM1V,0 - 7. ml/men / 1O q LV MV)O *&Owml Ain AW10 LV

TABLE I IEffect on MVo2 of Increasing Vmax at a Constant Peak

Developed Tension

Control N'orepinephrine

LVEDP, mmHg 4.9±0.6 1.1 ±0.4*LV vol. ml 22+1 15+1*MAP, mmHg 78±5 78 ±54Flow, ml/min per kg 105±410 100 ±+104Peak LVP, mmHg 80±6 95 ±8*Max dp/dt, mmHg/sec 1006±76 1622 +143*TDT, g-sec/cm2 12.94±2.2 10.2 +2.2*TPT, sec 0.123±0.006 0.106±0.006*Vmax, cm/sec 41.842.4 54.6+2.9*CEW,g-cm/beat 38.0+2.5 37.5 +2.2:ALCE, cm/beat 1.43 ±0.06 1.68 +0.06*CBF, ml/min 133±9 156+10*A-V 02, vol % 4.8 ±0.3 5.8 +0.3*MVo2, ul/beat per 100 g of LV 40.74±2.7 57.0±3.5*

LVEDP, left ventricular end-diastolic pressure; LV vol. left ventricu-lar volume; MAP, mean aortic pressure; flow, systemic flow; peak LVP,peak left ventricular pressure; max dp/dt, maximum rate of rise ofLVP; TDT, total developed tension; TPT, time from onset of contrac-tion to peak tension; Vmax, extrapolated velocity at zero tension; CEW-contractile element work; ALCE, extent of contractile element shorten-ing; CBF, coronary blood flow; A-V 02, coronary arteriovenous oxygendifference; MVo2, myocardial oxygen consumption. Data representmean value ± SE for 22 observations in 11 dogs.

* P < 0.001 Student's t test for paired differences.4 Not significant.

the calculated ALCE showed an average increasefrom 1.43 + 0.06 to 1.68 + 0.06 cm/beat. (TableII). CEWincreased when PDT was raised at aconstant Vmax (Table I) but was not significantly

c

Vdwm* 27mI valunm * 7mlM90t1 * 8.9nm* L!OO/AtV * 6.9, WAin A00/ LV

FIGURE 2 The effects of increasing ventricular volume and developed tension (A, B, and C) andthe infusion of norepinephrine (D) on hemodynamics and MVo2 in experiment No. 4. ECG, electro-cardiogram; tension= P7rr; A-V 02, coronary arteriovenous oxygen difference; LV dp/dt, the rateof left ventricular pressure development; LV Pr, left ventricular pressure; A Pr, aortic pressure.


volume VMot Pd- 0(ml) (cm/sec) (ml/min/IOOgLV)

A /2 44 5.7Control * 17 44 6.5

* 23 44 7.9

Norepinephrine A 15 58 11.8

I-

-j

LJ

0 50 100

FpRCE (g/cm2)

altered when V11.,. was increased at a constantPDT due to the decrease in total developed ten-sion (Table II).

In Fig. 2, tracings from one experiment areillustrated. In panels A, B, and C, ventricularvolume was increased at a constant Vma., and asa consequence LV end-diastolic pressure, peakLVP, peak LV dp/dt, and wall tension (Pirr2)all increased progressively, along with MVo2.During norepinephrine inusion, contractility in-creased, and in order to achieve a level of PDTin the range of that observed during the controlperiod, ventricular volume was decreased by re-

FIGURE 3 Left ventricular contractileelement force-velocity relationships inexperiment 8. Velocity = contractile ele-ment velocity; force = force/unit area(Pr/2h). V.,x = intercept on the ordi-nate. Closed symbols = effects of vary-ing ventricular volume at a constantVmax; open symbols = effect of increasingVma. with norepinephrine. Values of VmaXare estimates and not absolute values.(See Results).

moving saline from the intraventricular balloon.At this point (panel D), MVo2 was 8.9 ml/minper 100 g of LV, and, when compared to thecontrol state at a similar PDT (a tension midwaybetween that shown in panels A and B), MVo2had increased by 1.6 ml/min per 100 g of LV(+21 %).

The results from another experiment are shownin Fig. 3 in the form of the left ventricular con-tractile element force-velocity relationships ofrepresentative contractions. The three curves withsolid symbols represent three control states inwhich PDT was increased by increasing ventricu-

*-@ Con/ro/S--LN Noreovnephrine Vmax-I76 -,

-Vmox r 63

in

50PEAK DEVELOPEDTENSION (g/cm2)

100

FIGURE 4 MVo3 as a function of peak de-veloped tension in experiment 7. Closedcircles = increasing tension at a constantVmax; open triangles = observations dur-ing norepinephrine infusion; open square= MVo2 with potassium chloride arrest.The change in MVo2 (AMVo2) with nor-epinephrine at a constant tension was ob-tained by interpolation when tension wasnot precisely matched, as illustrated hereby the dotted vertical line.


10

-J

00

E 5

-

To2

0

Peak Developed Tension0 82.3±1.8A 70.2± 2 2

(00

0.24 ml/min per 100 g of LV and was alwayslower than the MVo2 extrapolated to a PDT ofzero, as illustrated here.

In Fig. 5, MVo2 is plotted as a function ofVmax at two distinct levels of PDT in the experi-ment in which Vmaz was changed by varying therate of norepinephrine infusion while maintainingPDT constant. A linear increase in MVo2 withincreasing contractility at constant levels of walltension is apparent.

The data are summarized in Fig. 6 both graphi-cally and by the equation: MVo2= K + 0.25PDT+ 1.43 Vmax. This equation was obtained bymultiple regression analyses of the data obtainedin 7 of the 11 experiments. These seven experi-ments were chosen for these analyses because theyhad at least two values of MVo2for changes in PDTat a constant Vm.. and at least two values ofMVo2 for changes in Vma. at a constant PDT.

50

Vmax (cm/sec)

FIGURE 5 MVo2 as a function of Vmax in experiment 11.

lar volume. The extrapolated velocity intercept ofthese curves at zero tension (Vma.) was un-

changed with this intervention, indicating that thecontractile state was unchanged. The intercept on

the abscissa, i.e. peak tension, however, was in-creased, and MVo. changed directionally with thisvariable. In contrast, Vmi1 was clearly increasedduring norepinephrine infusion (open triangles),signifying an augmentation of contractility, andLV volume was reduced to match the highesttension existing before norepinephrine infusion.With this increase in Vmai, MVo2 rose substan-tially.

MVo2, plotted as a function of PDT, for an-

other experiment, is shown in Fig. 4. The solidline illustrates the linear increase in MVo2 as

PDT was increased at a constant Vmin1 Thebroken line demonstrates the upward shift of thePDT-MVo2 relationship when Vin. was' in-creased. The open triangle lying above the brokenline was obtained with further increases in Vmaxand MVo2produced by increasing the norepineph-rine infusion rate. The open square on the ordi-nate represents MVo2 after cardiac arrest withpotassium chloride. MVo2 determined after potas-sium chloride arrest in six dogs averaged 1.43 +

60MV02 = k + 0.25 PDT + 1.43 Vmox

4O_

Q

20>E

20

x

0

hr

N,

0

r-

'20,v 120 240

PEAK DEVELOPEDTENSION (g/cm2)

FIGuRE 6 MVo2 isopleths as a function of V..x and peakdeveloped tension. Isopleths were calculated from theequation at the top of the figure which was derived bymultiple regression analysis. Coefficient of PDT 0.25 +

0.04 (SE), coefficient of Vma. = 1.43 0.15, = 0.99.Broken lines = effect on MVot of hypothetical increasesin PDT at a constant V... (horizontal line) and inV.a. at a constant PDT (vertical line).


10r

51

-J

00

0

25

The coefficients of PDT and Vma. in this equationare average values for the multiple regressionanalysis performed in each of the seven animals.The three diagonal parallel lines are MVo2 iso-pleths calculated from the multiple regressionequation. The broken lines in the center of thefigure illustrate the increases in PDT and Vmaxrequired to increase MVo2 by 50%o, from 40 to60 ixl/beat per 100 g. These particular changes inPDT, Vmax, and MVo2 were chosen for illustra-tive purposes because they are in the physiologicalrange.

DISCUSSION

The major objective of the present investigationwas to determine, in the same heart, the relativeinfluences of changes in the contractile state andin tension development on MVo2. It was observedthat elevations of both of these variables are asso-ciated with comparable increases in energy utili-zation. Moreover, the increase in MVos with in-creased contractility occurred under conditions inwhich peak developed tension and contractileelement work remained constant, and total devel-oped tension actually decreased. These resultsserve to quantify and to place into perspectiveprevious investigations showing increases inMVo2 with increases in contractility produced bysympathetic nerve stimulation or excitement (13),isoproterenol (14), norepinephrine (15, 17), car-diac glycosides (16-18), Ca++ (15), and pairedelectrical stimulation (15). It has also been re-ported that increases in contractility produced bycardiac glycosides (26), norepinephrine (27), andCa++ (28) do not always augment MVo2. How-ever, in those investigations ventricular volume al-ways declined as cardiac contractility was aug-mented, and tension development undoubtedly fellas well. From the present investigation in whichventricular volume and tension could be con-trolled, it seems clear that simultaneous, oppositechanges in tension and contractility would tendto cancel out any alteration of MVo2, and thesefindings can explain the reported absence of eleva-tions of MVo2 with augmentation of contractility(26-28).

Although the assumptions involved in the cal-culations of stress and velocity utilized in thisstudy have been discussed previously (11), cer-tain aspects deserve consideration here. The as-sumption of a spherical left ventricle with varying

ventricular volume undoubtedly involves someerror, since the larger the chamber the morespherical its shape. A spherical model, in contrastto an ellipsoidal one, underestimates total wallforce (29), and tension may actually have beengreater with norepinephrine in the smaller, moreellipsoidal heart than was calculated. We haveinvestigated this problem by means of cineangio-cardiography in dogs in which ventricular volumeswere varied by blood infusion. With the largestchange in the left ventricular major/minor semi-axis ratio which occurred (2.1: 1-1.4: 1 ), the er-ror in tension estimation using a spherical modelwas found to approximate 5 (11 ). Thus, even ifthis large change in the shape of the left ventricledid occur in the course of the present experiments,the resultant error in the calculation of tensioncould not conceivably account for the reported re-lations among developed tension, contractility,and MVo2.

Peak developed tension was defined as the dif-ference between peak tension and one-half of theresting tension in order to take into account theprogressive unloading of the parallel elastic ele-ments during contraction (30). The precise dis-tribution of resting tension between the paralleland series elastic components is unknown, andthus the exact contribution of the parallel elasticcomponent to the development of tension cannotbe calculated. However, the possible error intro-duced by this consideration is small, since restingtension constituted such a small fraction of peaktension.

The value used for series elastic stiffness inthese experiments (28T) was originally derivedfrom experiments in isolated cat papillary muscles(31), and a similar value has been found recentlyin the intact dog heart by the quick-release method(32). Furthermore, norepinephrine does not ap-pear to alter series elasticity either in papillarymuscle (33) or in the intact heart (32), a con-sideration of importance in the interpretation ofthe present results.

This investigation was carried out in an isovolu-metric preparation in order to obviate any large al-terations in the extent of shortening of the myo-cardial fibers and contractile elements. It is widelyrecognized, however, that isovolumetric contrac-tions are not isometric, since shape changes do oc-cur in the course of contraction. However, since thechanges in ventricular volume in the course of any


experiment were relatively small, it is unlikely thatchanges in the extent of fiber shortening couldhave affected the results materially. The calcu-lated extent of contractile element shortening didincrease slightly during norepinephrine infusion.However, since large changes in fiber shorteningand therefore contractile element shortening ex-ert only a small effect on MVo2 (10), the smallincrease in contractile element shortening whichoccurred in these experiments could not haveaffected the results substantially.

The possibility must also be considered that theincrease in MVo2 observed when Vm.x was ele-vated by norepinephrine resulted largely from adirect metabolic effect of the catecholamine, in-depeindent of any effect on the mechanics of con-traction. However, this possibility is unlikelysince it has been shown that much larger dosesof norepinephrine administered to potassium-ar-rested hearts resulted in far smaller incrementsof MVo2 than those observed in the present study(34). Furthermore, it has been shown that whenvelocity of left ventricular ejection and of pres-sure development were elevated to comparablelevels by norepinephrine and sustained postextra-systolic potentiation, the elevations of MVo2 weresimilar (15). This finding suggests that the al-terations in myocardial contractility produced bynorepinephrine were principally responsible forthe stimulation of MVo2.

In a comprehensive evaluation of the determi-nants of the MVo2, factors other than wall ten-sion and contractility must be considered. A for-mulation which considers the available experi-mental findings would indicate that at least fivefactors influence MVo2. These factors includebasal 02, activation 02, tension development 02,external work, i.e. (tension X shortening) 02,and contractile state 02. The first term, the basal029 is known and makes up a small proportion ofthe total 'MVo2 of the in situ heart. In the presentexperiments basal MVo2 was estimated in heartsarrested with potassium chloride, and it averaged1.43 ml/min per 100 g, a value similar to thosewhich have been found previously, both withhearts arrested with potassium excess (34-37)and with calcium deficiency (37, 38). The resting,or basal 02, may be increased slightly with in-creasing levels of catecholamines (34, 39-42).

The, activation energy is -a complex term -andwas defined by Hill in skeletal muscle as resulting

from "a 'triggered' reaction setting the muscle ina state in which it can shorten and do work" (43).The activation 02 of cardiac muscle might beconsidered to be made up of three components:electrical activation, contractile site activation anddeactivation, and the maintenance of the activestate. The 02 cost of electrical activation has beenmeasured in the nonbeating, dog heart and wasfound to be quite small (38), probably less than1%o of the total MVo2 of the resting unanesthe-tized dog (44). Since no mechanical activity tookplace in the nonbeating heart (38), this electricalactivation presumably did not result in the activa-tion of contractile sites on the myofilaments. Inpresent theories of muscle contraction (45, 46),repetitive reactions at these sites occur when Ca"becomes available to them, and it is likely that af-ter each contraction energy is required for the re-binding of Ca++, which is necessary for relaxation(46). It has recently been found that activationheat in frog sartorius muscle is independent oflength and temperature but is increased by agentsthat prolong the duration of the active state (47).However, the energy associated with increasingthe duration of the active state in cardiac muscleis not known. Although activation energy has notbeen quantified in the intact heart, it has been esti-mated, in isometrically contracting, isolated papil-lary muscle, from heat measurements (48) as wellas from high energy phosphate determinations(49) and is probably only a small fraction of thetotal energy requirements. Whether or not thisfraction is altered by inotropic influences in cardiacmuscle remains unknown.

It seems established that the major determinantsof MVo2 relate to the mechanical aspects of con-traction. The 02 utilization associated with stressor tension development in these experiments agreedclosely with that found previously. By recalculatingthe data of Monroe 'and French (6) and of Mc-Donald, Taylor, and Cingolani. (8), coefficientsof PDT (MVo2 = k PDT) of 0.26 and 0.21, re-spectively, were found, values similar to thosefound in this study (average 0.25, Fig. 6). Thework of Monroe (50), demonstrating that over90%o of the MVo2 was accounted for by the timepeak tension was- reached, serves to emphasizethe major energetic importance of tension develop-ment in contrast to the total developed tension(TDT), i.e., the area under the tension-timecurve. Accordingly, in the present study, peak

Control of Myocardial Oxygen Consumprion 383

developed tension rather than total developedtension was held constant when Vmax was altered.

With the Hill muscle model (25), tension de-velopment results from the work performed bythe contractile elements in stretching the serieselastic component, and in isometric contractions,CEWand PDT are directly proportional to oneanother and to MVo2. The work performed by thecontractile elements in shortening the myocardialfibers has recently been found to be associatedwith considerably lower energy costs than thework performed in developing tension, i.e., instretching the series elastic component (11, 12).Further evidence that CEW, as presently calcu-lated (9), is not the major factor determiningMVo2 is provided by the present experiments inwhich Vmax was altered at a constant PDT, CEWremained unchanged, and MVo2 was alteredmarkedly.

In conclusion, the data presented herein indi-cate that both developed tension and contractilestate are significant factors in the regulation ofMVo2. Furthermore, a comparative analysis ofthe effect of altering these two factors demon-strates that the effect on oxygen utilization ofchanges in contractility is substantial and similarin magnitude to the effect of altering tension de-velopment. From the current state of knowledge ofmyocardial energetics, the major determinants ofthe MVo2 may best be expressed in terms of ten-sion development and contractile state. The basal02 requirements, the activation energy, and thecost of contractile element shortening against aload would appear to make relatively smaller con-tributions than do tension development and con-tractile state.

ACKNOWLEDGMENTSThe authors wish to express their gratitude to Mr.Robert M. Lewis, Mr. Richard D. McGill, and Mr.Noel Roane for their technical assistance; to Mrs. HopeCook for the manometric determinations of oxygen con-tent; and to Miss Joan Gurian for her help in statisticalanalysis of the data.

REFERENCES

1. Rohde, E. 1912. Uber den Einfluss der mechanischenBedingungen auf die Tdtigkeit und den Sauerstoffver-brauch des Warmbluterherzens. Arch. Exptl. Pathol.Pharnacol. 68: 401.

2. Evans, C. L., and Y. Matsuoka. 1915. The effect ofvarious mechanical conditions on the gaseous metab-

olism and efficiency of the mammalian heart. J.Physiol. 49: 378.

3. Sarnoff, S. J., E. Braunwald, G. H. Welch, Jr., R. B.Case, WV. N. Stainsby, and R. Macruz. 1958. Hemo-dynamic determinants of oxygen consumption of theheart with special reference to the tension-time index.Anl. J. Physiol. 192: 148.

4. Katz, L. N., and H. Feinberg. 1958. The relation ofcardiac effort to myocardial oxygen consumption andcoronary flow. Circulation Res. 6: 656.

5. Rodbard, S., F. Williams, and C. Williams. 1959. Thespherical dynamics of the heart (myocardial tension,oxygen consumption, coronary blood flow, and effi-ciency). Am. Heart J. 57: 348.

6. Monroe, R. G., and G. N. French. 1961. Left ven-tricular pressure-volume relationships and myocardialoxygen consumption in the isolated heart. CirculationRes. 9: 362.

7. Rolett, E. L., P. M. Yurchak, W. B. Hood, Jr., and R.Gorlin. 1965. Pressure-volume correlates of left ven-tricular oxygen consumption in the hypervolemic dog.Circulation Res. 17: 499.

8. McDonald, R. H., Jr., R. R. Taylor, and H. E. Cingo-lani. 1966. Measurement of myocardial developed ten-sion and its relation to oxygen consumption. Am. J.Physiol. 211: 667.

9. Britman, N. A., and H. J. Levine. 1964. Contractileelement work: A major determinant of myocardial oxy-gen consumption. J. Clin. Invest. 43: 1397.

10. Taylor, R. R., H. E. Cingolani, T. P. Graham, Jr.,and R. L. Clancy. 1967. Myocardial oxygen consump-tion, left ventricular fiber shortening, and wall tension.Cardiovascular Res. 1: 219.

11. Graham, T. P., Jr., J. Ross, Jr., J. W. Covell, E.Sonnenblick, and R. L. Clancy. 1967. Myocardial oxy-gen consumption in acute experimental cardiac de-pression. Circulation Res. 21: 123.

12. Coleman, H. N. Effect of alterations in shorteningand external work on oxygen consumption in cat papil-lary muscle. Am. J. Physiol. In press.

13. Gregg, D. E. 1963. Physiology of the coronary circula-tion. Circutlation. 27: 1128.

14. Krasnow, N., E. L. Rolett, P. M. Yurchak, W. B.Hood, Jr., and R. Gorlin. 1964. Isoproterenol andcardiovascular performance. Am. J. Med. 37: 514.

15. Sonnenblick, E. H., J. Ross, Jr., J. W. Covell, G. A.Kaiser, and E. Braunwald. 1965. Velocity of contrac-tion as a determinant of myocardial oxygen consump-tion. Ant. J. Physiol. 209: 919.

16. Covell, J. W., E. Braunwald, J. Ross, Jr., and E. H.Sonnenblick. 1966. Studies on digitalis. XVI. Effectson myocardial oxygen consumption. J. Clin. Invest. 45:1535.

17. Clancy, R. L., T. P. Graham, Jr., W. J. Powell, Jr.,and J. P. Gilmore. 1967. Inotropic augmentation ofmyocardial oxygen consumption. Am. J. Physiol. 212:1055.

18. Coleman, H. N., III. 1967. Role of acetylstrophanthidinin augmenting myocardial oxygen consumption: Rela-


tion of increased MV02 to change in velocity of con-traction. Am. J. Cardiol. 21: 487.

19. Sonnenblick, E. H. 1962. Force-velocity relations inmammalian heart muscle. Am. J. Physiol. 202: 931.

20. Abbott, B. C., and W. F. H. M. Mommaerts. 1959.A study of inotropic mechanisms in the papillary mus-cle preparation. J. Gen. Physiol. 42: 533.

21. Covell, J. XV., J. Ross, Jr., E. H. Sonnenblick, and E.Braunwald. 1966. Comparison of the force-velocityrelation and the ventricular function curve as measuresof the contractile state of the heart. Circulation Res.19: 364.

22. Guyton, A. C., C. A. Farish, and J. W. Williams.1959. An improved arteriovenous oxygen differencerecorder. J. .4ppl. Physiol. 14: 145.

23. Van Slyke, D. D., and J. M. Neill. 1924. The de-termination of gases in blood and other solutions byvacuum extraction and manometric measurement. I.J. Biol. Chem. 61: 523.

24. Ross, J., Jr., J. W. Covell, E. H. Sonnenblick, and E.Braunwald. 1966. Contractile state of the heart charac-terized by force-velocity relations in variably after-loaded and isovolumic beats. Circulation Res. 18: 149.

25. Hill, A. V. 1938. The heat of shortening and thedynamic constants of muscle. Proc. Roy. Soc. (Lon-don), Ser. B. 126: 136.

?6. Sarnoff, S. J., J. P. Gilmore, A. G. Wallace, N. S.Skinner, Jr., J. H. Mitchell, and W. M. Daggett. 1964.Effect of acetyl strophanthidin therapy on cardiac dy-namics, oxygen consumption, and efficiency in theisolated heart with and without hypoxia. Am. J. Med.37: 3.

27. Sarnoff, S. J., J. P. Gilmore, M. S. Weisfeldt, W. M.Daggett, and P. B. Mansfield. 1965. Influence ofnorepinephrine on myocardial oxygen consumption un-der controlled hemodynamic conditions. Am. J.Cardiol. 16: 217.

28. Gilmore, J. P., W. M. Daggett, R. H. McDonald,and S. J. Sarnoff. Influence of calcium on myocardialpotassium balance, oxygen consumption, and perform-ance. Am. Heart J. In press.

29. Gorlin, R., E. L. Rolett, P. M. Yurchak, and W. C.Elliot. 1964. Left ventricular volume in man measuredby thermodilution. J. Clin. Invest. 43: 1203.

30. Parmley, WN. WV., and E. H. Sonnenblick. 1967. Serieselastic in heart muscle: Its relation to contractile ele-ment velocity and proposed muscle models. CirculationRes. 20: 112.

31. Sonnenblick, E. H. 1964. Series elastic and contractileelements in heart muscle: changes in muscle length.Am. J. Phvsiol. 207:1330.

32. Covell, J. W.. R. R., Taylor, and J. Ross, Jr. 1967.Series elasticity in the intact left ventricle determinedby a quick release technique. Federation Proc. 26: 382.(Abstr.)

33. Sonnenblick, E. H. 1962. Implications of musclemechanics in the heart. Federation Proc. 21: 975.

34. Klocke, F. J., G. A. Kaiser, J. Ross, Jr., and E.Braunwald. 1965. Mechanism of increase of myocardialoxygen uptake produced by catecholamines. Am. J.Physiol. 209: 913.

35. McKeever, W. P., D. E. Gregg, and P. C. Canney.1958. Oxygen uptake of the nonworking left ventricle.Circulation Res. 6: 612.

36. Monroe, R. G., and G. French. 1960. Ventricular pres-sure-volume relationships and oxygen consumption infibrillation and arrest. Circulation Res. 8: 260.

37. Kohn, R. M. 1963. Myocardial oxygen uptake duringventricular fibrillation and electromechanical dissocia-tion. Am. J. Cardiol. 11: 483.

38. Klocke, F. J., E. Braunwald, and J. Ross, Jr. 1966.Oxygen cost of electrical activation of the heart.Circulation Res 18: 357.

39. Whalen, W. J. 1957. Oxygen consumption and tensionof isolated heart muscle during rest and activity usinga newv technic. Circulation Res. 5: 556.

40. Berne, R. M. 1958. Effect of epinephrine and norepi-nephrine on coronary circulation. Circulation Res. 6:644.

41. Challoner, D., and D. Steinberg. 1965. Metabolic effectof epinephrine on the QO2of the arrested isolated per-fused rat heart. Nature. 205: 602.

42. Hauge, A., and I. Dye. 1966. The action of adrenalinein cardiac muscle II. Effect on oxygen consumptionin the asystolic perfused rat heart. Acta. Physiol.Scand. 68: 295.

43. Hill, A. V. 1949. The heat of activation and the heatof shortening in a muscle twitch. Proc. Roy. Soc.(London), Ser. B. 136: 195.

44. Gregg, D. E., E. M. Khouri, and C. R. Rayford. 1965.Systemic and coronary energetics in the resting unan-esthetized dog. Circulation Res. 16: 102.

45. Huxley, A. F. 1957. Muscle structure and theories ofcontraction. Progr. Biophys. Biophys. Chem. 7: 255.

46. Davies, R. E. 1963. A molecular theory of muscle con-traction Calcium-dependent contractions with hydro-gen bond formation plus ATP-dependent extensionsof part of the myosin-actin cross-bridges. Nature. 199:1068.

47. Gibbs, C. L., N. V. Ricchiuti, and W. F. H. M.Mommaerts. 1966. Activation heat in frog sartoriusmuscle J. Gen. Physiol. 49: 517.

48. Ricchiuti, N. V., and C. L. Gibbs. 1965. Heat produc-tion in a cardiac contraction. Nature. 208: 897.

49. Pool, P. E., and E. H. Sonnenblick. 1967. Themechanochemistry of cardiac muscle. I. The isometriccontraction. J. Gen. Physiol. 50: 951.

50. Monroe, R. G. 1964. Myocardial oxygen consumptionduring ventricular contraction and relaxation. Circula-tion Res. 14: 294.


control myocardial oxygen consumption:...

Documents