Left ventricular area correlates

systolic pressure-volume with oxygen consumption

FARAH KHALAFBEIGUI, HIROYUKI Department of Biomedical Engineering, Baltimore, Maryland 21205







SUGA, AND KIICHI SAGAWA The Johns Hopkins University, School


GAWA. Left ventricular systolic pressure-volume area correlates with oxygen consumption. Am. J. Physiol. 237(5): H566H569, 1979 or Am. J. Physiol.: Heart Circ. Physiol. 6(5): H566H569, 1979.-In 13 excised, cross-circulated canine hearts, we studied the correlation between left ventricular oxygen consumption per beat (MVO~) and the magnitude of a specific pressure-volume (P-V) area circumscribed by the end-systolic and end-diastolic P-V relationship curves and the systolic segment of the P-V trajectory of a left ventricular contraction. The pressure and volume load of the ventricle were changed with a volume servo pump in order to alter the P-V area, and MVo2 was measured (after each change in the pressure and volume load). In the data collected from both isovolumic and ejecting contractions of each left ventricle contracting with a stable inotropic background, we found a linear correlation between MVoz and the P-V area. The average correlation coefficient was 0.92 t 0.016 (SE). Linear regression analysis yielded the formula: Mv02 (ml/beat) = a[P-V area (mmHgsml/beat)] + b, where a, the slope coefficient, was (1.53 t, 0.14) X lOA” and b, which probably represents the basal 02 consumption, was 0.019 t 0.003 ml/beat. We propose that the P-V area as defined above may be a good index of ventricular oxygen consumption under a given inotropic background. heart; cardiac energetics;



stroke work

MECHANICAL FACTORS Of left ventricular COntraCtion that determine cardiac oxygen consumption have been the subject of extensive research for many years (1, 5). Various factors reported to predict myocardial oxygen consumption per beat include ventricular pressure and volume (7)) active tension developed (1,4), and “tensiontime index” (8, 14), “cardiac effort index” (product of peak ventricular pressure and heart rate) (6), contractile element work (2), external work (3, 4), shortening velocity, and inotropic state (1,3,4). However, agreement has not been reached as to which or what combination of them correlates best with Mv02 under a variety of loading conditions. Particularly, there has been no method proposed to predict MT02 of a variably loaded heart directly from its pressure and volume data. The present paper describes a new index of myocardial oxygen consumption per beat (Mvoz) which is a pressure-volume (P-V) area defined in the P-V plane as the area bounded by the end-systolic and end-diastolic P-V relation curves and the systolic segment of the P-V loop trajectory of a contraction. A highly significant correlation was found between MVoz and the P-V area, sugTHE


of Me&c&

gesting that the latter can be used to predict the former under a stable contractile state. METHODS

Preparation. Thirteen pairs of mongrel dogs (body wt 15-22 kg) were anesthetized with sodium pentobarbital (30 mg/kg iv). The heart excised from one of the pair was perfused with arterial blood provided by the other dog with a constant perfusion pressure between 75 and 120 mmHg at a temperature of 37 t 1°C. The details of the surgical preparation of the ventricle and the servopump system to control and alter the pressure-volume load have been reported elsewhere (11). Briefly, a thin latex balloon (Qualatex no. 9 helium balloon) was placed in the left ventricle through the mitral annulus and connected to the servo pump (Fig. 1). The balloon and the water housing of the pump were primed with tap water without leaving any air bubbles. The water reciprocated between the heart and the pump in a programmed manner but in synchrony with the spontaneous ventricular contraction. Left ventricular thebesian flow was drained out by suction via a cannula with multiple side holes placed between the balloon and the endocardium so that the balloon would fit well on the endocardial surface. A miniature pressure transducer (Konigsberg, P-21) was placed inside the balloon for measurement of intraventricular pressure. The right ventricle was maintained totally unloaded by draining coronary venous return and right ventricular thebesian flow through a large-bore cannula that was placed in the right ventricle and connected to a negative hydrostatic pressure source. To determine coronary blood flow we measured the total mixed coronary venous outflow from the right ventricle by a rotameter (type 4S, Clifford Wilson). The heart beat at a regular sinus rhythm (115 t 3.8 (SE) beats/min) for 4-6 h with the cross perfusion. We attempted to minimize a change in the inotropic background of the ventricle by maintaining the support dog in a stable condition. We recorded instantaneous pressure and volume of the left ventricle while controlling primarily the instantaneous ventricular volume by the servo-pump system. The pressure and volume signals were fed into the x and y variable inputs of a storage oscilloscope in order to obtain the pressure-volume loop trajectory of contraction in the P-V plane. From the P-V trajectories of steady-state contractions under various loading conditions we determined a specific P-V area, which is defined in the next section.



0 1979 the American


Downloaded from www.physiology.org/journal/ajpheart at Glasgow Univ Lib ( on February 15, 2019.













FIG. 1. Schematic illustration of the servo-pump heart preparation under cross circulation. A, coronary V, coronary venous return tube; LV, left ventricle.

system and the perfusion tube;

Total mixed coronary venous blood was sampled from the right ventricular drain tube, and arterial blood was sampled from the arterial perfusion line. The blood samples were analyzed for hemoglobin concentration (Hb) and oxyhemoglobin percent saturation (%HbOz) by an Instrumentation Laboratory model 182 Co-oximeter, and also for pH, Pco~, POTby pH-blood gas analyzer (Instrumentation Laboratory model 113-01). From [Hb] and %HbOz, oxygen content of blood samples was calculated. Mi702 per minute was calculated as the product of the measured coronary blood flow per minute and the arteriovenous difference of oxygen content. MVo2 per beat was obtained by dividing Mv02 per minute by heart rate per minute. Definition of P-V area. Our previous finding on the instantaneous P-V relationship of left ventricle (12) indicated that ventricular contraction could be modeled by a time-varying elasticity regardless of loading conditions. On the basis of the experimental finding, Suga (9) has advanced a theory that the P-V area, as defined below, represents the total mechanical energy required to change the elastic state of the ventricle from the enddiastolic state to the end-systolic state. To examine the validity of this theory, we studied the correlation of MVoz with a specific area on the pressure-volume plane that is circumscribed by three sides. These are 1) the end-diastolic P-V relation curve, 2) the end-systolic P-V relation line (Line A0 in Fig. 2 and explained at the beginning of RESULTS), and 3) the systolic portion of the P-V loop trajectory, namely, Line CBA in the lower left panel of Fig. 2, which schematically represents an ejecting contraction, or Line BA in the lower right panel of Fig. 2 representing an isovolumic contraction. These areas (shaded area AODCB or AOB in Fig. 2) will be simply called the P-V area in this paper.

superimposed P-V loops of isovolumic and ejecting contractions of one left ventricle is shown in Fig. 2. The P-V data point in an ejecting contraction moves in a rectangular fashion with time during a cardiac cycle in the P-V plane. The position and the size of the P-V loop were varied extensively by changing the programmed time course of V through the servo-pump system. The arrowheads indicate the end-systolic P-V data points. To determine the end-systolic P-V relationship we drew a straight line through these end-systolic P-V points regardless of end-diastolic volume and mode of contraction and linearly extrapolated it to the volume axis. The volume-axis intercept (point 0) is Vd, which represents the end-systolic dead space in the ventricular lumen. This point was determined at the beginning of each experiment. The P-V areas, as defined in METHODS, were determined on the pictures taken on the storage oscilloscope while ventricles were contracting steadily for at least 35 min under one of several different loading conditions. The P-V areas were correlated with the values of MVoz measured simultaneously. A plot of the correlation between MVo2 (ml Ox/beat) and the P-V area (mmHg .ml/beat) in a ventricle is shown in Fig. 3. In this and every other ventricle the P-V area and Mvoz were determined in isovolumic contractions at several volumes (solid circles) and in several ejecting contractions from different end-diastolic volumes and against different systolic pressures (open circles). The order of contraction mode as well as change in load was randomized. Statistical analysis of this set of data indicated a highly significant correlation between MVoz and the P-V area (P < 0.001) with a correlation coefficient of 0.96. Linear regression analysis of the MVO~ data on the



I” E E


# z


2 E


0 I








2. Top: representative example of superimposed P-V loops of isovolumic and ejecting contractions of the left ventricle in one heart. Arrowheads indicate the end-systolic P-V data points. Bottom: schematic diagrams to explain definition of P-V area for isovolumic contraction (r&#) and ejecting contraction (left). FIG.


Instantaneous P-V loops were obtained under a variety of loading conditions in each ventricle. An example of

Downloaded from www.physiology.org/journal/ajpheart at Glasgow Univ Lib ( on February 15, 2019.



k zi .



6 -z.025 0







3. Correlation between isovolumic contractions;




3000 l



and P-V area in one heart: cZosed circles, ejecting contractions.

P-V area from the same ventricle

yielded the formula

MVoz = 1.55 x 10-” [P-V area (mmHg=ml/beat)] + 0.0082 (ml/beat) The regression coefficient represent the oxygen cost of unit P-V area and the last term is considered to represent the basal M~oQ of the ventricle muscle. A similarly high correlation between Mvoz and P-V area was found in all other hearts. The mean correlation coefficient averaged from the 13 hearts was 0.92 2 0.016 (SE). Figure 4 depicts all the regression lines for 13 hearts. Each of the regression lines was calculated from 5-15 data points from both isovolumic and ejecting contractions of each ventricle under a stable contractile state. The average regression coefficient was [1.53 t 0.1431 x lo-“, and the mean value of the intercept with the MVo2 axis (basal MVoz) was 0.0194 t 0.003. As can be seen in Fig. 4, there are variations of the intercept and the slope of the regression line among the hearts. In an attempt to identify the source of these variations, we examined the correlation of the slope and intercept with a contractility index E,,, (12), heart weight, heart rate, coronary blood flow, and blood gas parameters mentioned above. None of these were statistically significantly correlated with the magnitude of the slope or the MVoz-axis intercept.




and 2) we only need to measure the ventricular pressure and volume and do not need to convert pressure to force, and volume to length of muscle, assuming a simple shape of the ventricle, a homogeneous structure of the ventricular wall, and a mechanical model of the cardiac muscle. Mv02 measured in this study is not just for the left ventricle because the coronary venous oxygen content reflects MVoz of the right ventricle and both atria as well. However, because right ventricle and right and left atria remained entirely unloaded throughout the experiment we consider that Mv02 by these three chambers added only a minute amount to the value of b in the regression equation, contributing little to the variation in MVoz with changes in the P-V area. In this experiment, MVO~ and P-V area were measured only from ventricles contracting under a stable heart rate and contractile state without any inotropic intervention. Therefore, it remains to be known whether or not the regression coefficient and the MVoz-axis intercept of the regression equation vary with change in contractile state or heart rate. According to Suga’s theory the total mechanical energy associated with an ejecting contraction (lower right panel of Fig. 2) consists of one part which is expended to perform the external work (area ABCD) and another part which exists as a potential (elastic) energy stored in the ventricular wall at the end of systole (area ADO). The latter concept is not yet popular and can be disputed. However, Suga (10) has also shown that by unloading the ventricle at an appropriate speed, as much as ‘70% of the potential energy represented by area ADO could actually be retrieved as the external work done by the ventricle to the external system. Based on the theoretical consideration and the experimental finding cited above, we believe that the high correlation of MiTo2 with the P-V area reported here is not coincidental but has a direct mechanistic relation. In turn, the presently observed high correlation certainly supports, though does not prove, Suga’s theory. Lastly, a question must be raised: If the P-V area ADO in Fig. 2 represents the mechanical potential energy stored at the end of systole as proposed by Suga (9), what happens to it when the ventricle relaxes in the physiological manner? We would consider that it is eventually dissipated as heat. At present there is no experimental evidence to support this notion. A definite evi-


A highly significant correlation was found between the myocardial oxygen consumption and the specific P-V area of the canine left ventricle in the present study. The mean value of the correlation coefficient we obtained seems convincingly higher than the correlation coefficient of MVoz with contractile element work (2), with peak developed tension times heart rate (6), or with the time integral of systolic force (14). Thus, on a purely empirical basis, the P-V area seems to be a better index than those previously proposed. Even if the P-V area correlates with MVog to the same degree as the other indices, it would still have the major advantages over most of the other indices. The reasons are that 1) P-V area has a dimension of work (or energy)




1000 P-V

4. Regression in 13 hearts. FIG.


ines of MAO:!








on P-V area obtained

Downloaded from www.physiology.org/journal/ajpheart at Glasgow Univ Lib ( on February 15, 2019.






dence remains to be gained by careful measurement of heat generated by isovolumically contracting ventricle. This study was partly HL- 14903. F. Khalafbeigui Association award.

supported by Public Health Service was a recipient of the American

Grant Heart

H. Suga’s present address is National search Institute, Suita, Osaka 565, Japan. An abstract of this study was presented Meeting (Federation Proc. 37: 824, 1978). Received

9 March

1979; accepted



at the 62nd Annual

in final form

10 July




REFERENCES 1. BRAUNWALD, E. The determinants of myocardial oxygen consumption. Physiologist 12: 65-93, 1969. 2. BRITMAN, N. A., AND H. J. LEVINE. Contractile element work: a major determinant of myocardial oxygen consumption. J. CZin. Inuest. 43: 1397-1408, 1964. 3. BURNS, J. W., AND J. W. COVELL. Myocardial oxygen consumption during isometric and isovolumic contractions in the cat heart. Am. J. PhysioZ. 223: 1491-1497, 1972. 4. COLEMAN, H. N. III, E. H. SONNENBLICK, AND E. BRAUNWALD. Myocardial oxygen consumption associated with external work. The Fenn effect. Am. J. PhysioZ. 217: 291-296, 1969. 5. GIBBS, C. L. Cardiac energetics. PhysioZ. Reti. 58: 174-254, 1978. 6. MCDONALD, R. H., R. R. TAYLOR, AND H. F. CINGALANI. Measurement of myocardial developed tension and its relation to oxygen consumption. Am. J. PhysioZ. 211: 667-673, 1966. 7. MONROE, R. G., AND G. N. FRENCH. Left ventricular pressurevolume relationships and myocardial oxygen consumption in the isolated heart. Circ. Res. 9: 362-374, 1961. 8. SARNOFF, S. G., E. BRAUNWALD, G. H. WELCH, R. B. CASE, W. N. STAINSBY, AND R. MACRUZ. Hemodynamic determinants of oxygen consumption of the heart with special reference to the tension-time

index. Am. J. Physiol. 192: 148-156, 1958. 9. SUGA, H. Total mechanical energy of a ventricular model and cardiac oxygen consumption. Am. J. Physiol. 236: H498-H505,1979 or Am. J. Physiol.: Heart Circ. Physiol. 5: H498-H505, 1979. 10. SUGA, H. External mechanical work from relaxing ventricle. Am. J. Physiol. 236: H494-H497, 1979 or Am. J. Physiol.: Heart Circ. Physiol. 5: H494-H497, 1979. 11. SUGA, H., AND K. SAGAWA. End-diastolic and end-systolic ventricular volume clamper for isolated canine heart. Am. J. PhysioZ. 233: H718-H722,1977 or Am. J. Physiol.: Heart Circ. PhysioZ. 2: H718H722, 1977, 12. SUGA, H., AND K. SAGAWA. Instantaneous pressure-volume relationship and their ratio in the excised, supported canine left ventricle. Circ. Res. 35: 117-126, 1974. 13. TAYLOR, R. R., H. E. CINGOLANI, T. P. GRAHAM, AND R. L. CLANCY. Myocardial oxygen consumption, left ventricular fiber shortening and wall tension. Cardiovasc. Res. 1: 219-238, 1967. 14. WEBER, K. T., AND J. JANICKI. Myocardial oxygen consumption: The role of wall force and shortening. Am. J. PhysioZ. 233: H421H430, 1977 or Am. J. PhysioZ.: Heart Circ. Physiol. 2: H421-H430, 1977.

Downloaded from www.physiology.org/journal/ajpheart at Glasgow Univ Lib ( on February 15, 2019.

Left ventricular systolic pressure-volume area correlates with oxygen consumption.

Left ventricular area correlates systolic pressure-volume with oxygen consumption FARAH KHALAFBEIGUI, HIROYUKI Department of Biomedical Engineering,...
898KB Sizes 0 Downloads 0 Views