European Heart Journal (1990) 11 {Supplement C), 8-21

Abnormal intracellular calcium handling in acute and chronic heart failure: role in systolic and diastolic dysfunction C. L. PERREAULT, A . J. MEUSE, L. A . BENTTVEGNA AND J. P. MORGAN

Charles A. Dana Research Institute and the Harvard-Thomdike Laboratory, Department of Medicine (Cardiovascular Division), Beth Israel Hospital and Harvard Medical School, Boston, Massachusetts, USA

KEY WORDS: Calcium, heart failure, systolic dysfunction, diastolic dysfunction. Acute or chronic heart failure may be caused by one or more of a variety of abnormalities including changes in excitation-contraction coupling processes (i.e. decreased availability of activator Ca2+ or a change in myofilament Ca2+ responsiveness), a change in myocardial energetics, or a change in extracellular factors, such as connective tissue content. Most of the animal and human models of acute cardiac failure that we have studied in our laboratory (i.e. negative inotropic responses to drugs, hypoxia, acidosis and ischaemia) appear to involve changes in excitation-contraction coupling as the predominant cause of dysfunction. On the other hand, the models of chronic cardiac dysfunction that we have studied (i.e. chronic right ventricular pressure overload in ferrets, hypertrophic cardiomyopathy in Syrian hamsters, hypertensive cardiomyopathy in rats, hypothyroidism in ferrets, end-stage dilated and hypertrophic cardiomyopathy in man) predominantly appear to reflect a combination of changes involving abnormalities in both excitation-contraction coupling and extracellular factors involving myocyte drop-out and increases in connective tissue content. However. In most of these models of acute and chronic heart failure, abnormal intracellular Co2* handling appears to be a major cause of both systolic and diastolic dysfunction. Introduction Although the clinical spectrum of heart failure includes cases of predominant systolic or diastolic dysfunction, most patients with this diagnosis have a combination of both abnormalities. Since changes in intracellular levels of free ionised calcium (i.e. [Ca 2+ D have been shown to play a central role in the regulation of both systolic and diastolic performance of the heart, it is reasonable to hypothesize that abnonnal [Ca 2+ ] ; handling may be an important subcellular mechanism in at least some cases of heart failure'1"41. Studying the effects of drugs and disease states on excitation-contraction coupling processes in the mammalian heart has been a major thrust of our laboratory; the purpose of this report is to summarize our work concerning the role of abnormal [Ca2+], handling in animal and human models of acute and chronic heart failure The cellular control of cardiac myocyte function will be briefly summarized; excitation^ontraction

coupling and the effects of drugs on failing myocard^ b e e n ** t 0 P i c o f r e c e n t r e v i e w s ^ . Tbets « « **>** myofilaments to initiate cardiac contraction; relaxation occurs when Ca dissociates

SrnS'K^S.IS^ASSSSt'SJ:

from

02215, U.S.A. 0195-668X/90/0CO08 + 14 $03.00/0

ium

« * contractile apparatus and is resequestered

by the sarcoplasmic reticulum. Intracellular Ca © 1990 The European Society of Cardiology

Abnormal intracellular calcium handling

Action potential

Isoprotcrenol |

Na - K

Figure 5 Influence of nifedipine on the aequorin signals (noisy traces) and isometric contractions of a cat papillary muscle. Records were made at steady state. Stimuli delivered at four-second intervals. Lower trace shows stimulus artifact. Numbers in panels show molar concentration of nifedipine. Reproduced from Morgan et al.pi\ with permission.

Abnormal intracellular calcium handling

case of ischaemia, or to changes in extra-myocyte factors (i.e. connective tissue content) in the ferret model of right ventricular dysfunction that develops with chronic pressure overload. As discussed below, changes in myofilament Ca2+ responsiveness and/or connective tissue content appear to be important in some of the other models of chronic heart failure included in Tables 1 and 2. However, it is important that decreased availability of Ca2+ is the predominant factor in the majority of these models. DIASTOLIC FORCE AND [ C a 2 + ] ,

Under the experimental conditions used in our studies, most of the models listed in Table 2 also demonstrate evidence of diastolic dysfunction. This may be reflected by a marked delay in relaxation of the isometric twitch compared with controls, or by an elevation in end-diastolic Ca2+ and tension. The diastolic changes which occur in the models of acute

heart failure tend to be relatively modest with the exception of ryanodine (Fig. 6), which produces a marked prolongation of the Ca2+ transient and isometric contraction in papillary muscle preparations'2'; digitalis toxicity, which is associated with marked Ca2+ overload and elevated end-diastolic Ca2+ with incomplete relaxation'2-3117', and hypoxia and ischaemia in the whole heart preparation'13'. In contrast to hypoxia induced in an isolated papillary muscle preparation, hypoxia of the whole heart is associated with an increase in end-diastolic [Ca2+]i and left ventricular pressure (see Grossman, Fig. 2, this issue). In ischaemia, although systolic and diastolic pressures both fall below control levels, enddiastolic Ca2+ concentrations actually rise towards the 1 fim range (Fig. 7). This dramatic increase in [Ca2+]i is masked by a coincident decrease in myofilament Ca2+ responsiveness which prevents tension from rising.

Cot Control 3-3 k

Drug 0-6 k

ControI 2-4 mN

Drug 0-8 mN

Rabbit Control 2-1 mN Control 1-4 k

Drug 0-2 k

15

Drug I I mN

0-5 •

Figure 6 Influence of ryanodine, 1 x 10"6 M, on the time course of light and tension responses in cat and rabbit papillary muscles. Tracing recorded before and after (arrows) administration of the drug are superimposed with vertical scales adjusted to give the same peak. True amplitudes are indicated to the right. Note that although the time courses of light and tension are prolonged, ryanodine has a negative inotropic effect. Tension is expressed in mN; k = 1000 cycles s"1. Reproduced from Morgan and Morgan'2', with permission.

16

C. L. Perreault et al.

Itchatmio

.

Reperfusion

.„ _.,.nrmwri#'fPi'rPlT"1

100 n

0-1

100 -i

0-1

Figure 7 Recordings of [Ca2+]j transients (top), isovolumic left ventricular pressure (middle) and coronary perfusion pressure (bottom) during 3 min ischaemia followed by reperfusion. Pacing rate wa£ 130 beats min"1. Reproduced from Kihara et a/.[13), with permission.

The hypothyroid model of chronic heart failure in the ferret is associated with a marked prolongation of the isometric contraction; however, end-diastolic [Ca2+]i and tension returned to control values in our experiments'12'. Similar changes were observed in the hypertrophic cardiomyopathy of the Syrian hamster'30', hypertensive cardiomyopathy of the rat'34', and right ventricular dysfunction induced by pressure-overload in the ferret'29]. The human models of end-stage heart failure that we have studied also show marked prolongation in the time course of the [Ca2+]j transient and isometric relaxation (Fig. 8); moreover, at physiologic rates of pacing, end-diastolic [Ca2+]i rises and this increase is associated with an elevation in end-diastolic tension (i.e. incomplete relaxation; Fig. 9)'1'3-4'91. These changes can be exacerbated by agents that increase [Ca2+]j further, such as digitalis and increased extracellular concentrations of Ca 2+ ; they can be ameliorated by drugs that enhance the ability of the myopathic myocytes to maintain Ca2+ homeostasis, such as forskolin and noradrenaline (both of which increase intracellular cAMP levels)'3'4-9"36'. In addition to the changes discussed above, which

occur in end-diastolic Ca2+ and tension under isometric conditions, decreased diastolic compliance can also be reflected (in the whole heart preparations) by a shift of the pressure-volume relationship in a direction suggesting decreased ventricular compliance or (in papillary muscle preparations) by an increase in passive tension noted during length-tension determinations'1'. Such abnormalities have been described by a number of investigators, including ourselves, in the same or similar models to those listed in Table 1. Therefore, although changes in excitationcontraction coupling are of central importance with regard to the aetiologies of failure in each of these models, it must be kept in mind that, particularly with regard to the chronic models of heart failure, other calcium-independent factors may be playing an important role. MY0FILAMENT C a 2 + RESPONSIVENESS

With regard to the models of acute and chronic heart failure described in Tables 1 and 2, changes in myofilament Ca2+ responsiveness appear to play an important role. Perhaps the most striking example of

Abnormal intracellular calcium handling Dilated cardlomyopathy

Control 6nA

17

Hyptrtrophic cardlomyopathy

\yL'

53 nA

0-9 s

- 8 0 mV —

- 7 8 mV —

200 m»

Figure 8 Representative light tracing (upper noisy trace) and tension tracing (middle trace) in control and myopathic human muscles; lower trace is stimulus artifact, light expressed in nanoamperes (nA) of anode current; tension in g. The cross-sectional areas of the muscles are (in mm2) control, 1-3; dilated cardiomyopathy, 0-5; hypertrophic cardiomyopathy, 0-5. Action potentials recorded in muscle from these same hearts are also shown. Reproduced from Gwathmey el a/.'4', with permission.

JUUUUJUUUUUJUUJ^ Stimulus Interval

3s

2s

Is

Figure 9 Response of hypertrophied trabecula from patient with end-stage heart failure to increased rates of stimulation at 30°C; [Ca2+]i = 16 mM. Upper trace, aequorin signals; middle trace, tension development; lower trace, stimulus artifact. Interval between stimuli noted at bottom of Figure. Reproduced from Gwathmey et alS*\ with permission.

the importance of these mechanisms in acute heart failure is provided by ischaemia. In this condition, as shown in Fig. 7, systolic and diastolic tension actually fall at a time when intracellular Ca2+ concentrations are rising to high levels. This apparent discrepancy is due to the dramatic increases in intracellular inorganic phosphate and hydrogen ion concentrations which occur during ischaemia as ATP is broken down. Increased inorganic phosphate and acidosis have both been shown to decrease markedly the myofilament Ca2+ sensitivity producing a rightward shift in the force versus pCa relationship (as shown in panel B of Fig. 3 ^ ) . The effects of acidosis are shown in Fig. 10 where the disproportionate decrease in tension relative to light is readily apparent. A similar response is illustrated

in Fig. 11 for BDM under the conditions of our experiments. We have also shown that hypoxia is associated with the development of a decrease in myofilament Ca2+ sensitivity*261; although the subcellular mechanisms are less well understood'251. The chronic models of heart failure listed in Table 2 have not been completely characterized with regard to changes in Ca 2+ sensitivity. On the basis of information currently available, it appears that the right ventricular pressure-overload model in ferrets'29"38', and the hypertensive cardiomyopathy model in rats'32"33 •39-40\ are not associated with significant changes in left ventricular myofilament Ca2+ sensitivity. In addition, end-stage heart failure in man does not show a significant difference from control with regard to the force vs pCa relationship of

18

C. L. Perreault et al.

Control

Acidoiu

A—

48 nA

0-36 g

100 ms

Figure 10 Effects of acidosis on light and tension responses of a ferret right ventricular papillary muscle, paced at 0-33 Hz, 30°C, [Ca 2+ ] o = 1 DIM. Note preservation of Ca2+ transient in face of marked reduction of isometric tension development.

BDM (mM) ; Ferret papillary mujcle

^fj

^**H);»"E">-

120 nA

0-5g

100 ms

Figure 11 Effects of BDM on light and tension responses of a ferret rightventricularpapillarymuscle,pacedat0-33Hz,30°C,[Ca2+]o = 1 mM. Note relative preservation of Ca2+ transient in face of marked reduction of isometric tension development. W/O = wash-out.

Abnormal intracellular calcium handling skinned ventricular fibres'18'. However, of particular interest is the finding that the force vs pCa relationship may be differentially altered by pharmacological agents. For example, it has recently been reported that the experimental inotropic agent, DPI 201-106, can sensitize to calcium the myofilaments of myopathic but not control human working myocardium'18'. Therefore, a change in myofilament Ca2+ responsiveness appears to be an important mechanism by which cardiac function can be altered in heart failure.

19

connective tissue content is markedly increased'4' Therefore, extracardiac factors must be kept in mind when considering the complex aetiology of most cases of heart failure. It should also be noted that in vitro studies such as those described in Table 2, have not considered the important in vivo role of neuroautonomic dysfunction or the effects of acute and chronic changes in cardiac preload and afterload due to pathophysiological changes in the vasculature'44'.

Therapeutic implications EXTRA-MYOCYTE FACTORS AND MYOCARDIAL ENERGETICS

It must be kept in mind that most of the steps in the excitation-contraction coupling process require the expenditure of energy. .Relaxation abnormalities often play a major role in the pathogenesis of the clinical syndrome of heart failure and recent studies have indicated that relaxation is more readily impaired by ischaemia and pressure-overload than is contraction. As noted by Katz'41', a deficit in chemical energy has a greater effect in slowing the Ca2+ fluxes that occur during diastole than during systole. This is largely because delivery of activator Ca2+ to the contractile apparatus is an intrinsically rapid downhill process, whereas relaxation involves the much slower transport of Ca2+ against a concentration gradient as calcium is taken up again by the sarcoplasmic reticulum. A fall in myocardial highenergy phosphate concentrations can slow the Ca2+ pump of the sarcoplasmic reticulum and thereby impair resequestration. It is also clear that extra-myocyte factors, involving myocyte necrosis and increased connective tissue in the heart, may play an important role in the aetiology of heart failure''42'43]. Increased connective tissue content can be associated with a decrease in systolic and diastolic compliance of the left ventricle and ventricular papillary muscle preparations. This may translate into decreased active systolic force generation and, in some cases, a decreased rate of diastolic relaxation. Striking examples of these types of changes are provided by several of the chronic models of heart failure listed in Tables 1 and 2. Chronicrightventricular dysfunction in ferrets due to chronic pressure-overload appears to be caused predominantly by the increased connective tissue content which develops, rather than by changes in intracellular Ca2+ handling1291. Although marked changes in excitation-contraction coupling processes occur in the human models of heart failure, the

In summary, acute or chrome heart failure may be caused by one or more of a variety of abnormalities including changes in excitation-contraction coupling processes (i.e. decreased availability of activator Ca2+ or a change in myofilament Ca 2+ responsiveness), a change in myocardial energetics, or a change in extracardiac factors such as connective tissue content. Most of the cases of acute and chronic heart failure discussed above appear to be associated with changes in excitation-contraction coupling, although other mechanisms also may be involved. Cardiac dysfunction may be due to abnormalities in excitation-contraction coupling at multiple levels in the cell; this fact is well illustrated by the pharmacological models of acute heart failure illustrated in Tables 1 and 2. The importance of sarcolemmal mechanisms is shown by the response to decreased extracellular calcium or calcium channel blockers (Fig. 5). Both of these manoeuvres decrease the amount of calcium available for excitationcontraction coupling processes in the myoplasm and result in lesser degrees of activation. Local anaesthetics produce a similar effect by impairing sodiumcalcium exchange. Ryanodine, a drug that acts selectively on the sarcoplasmic reticulum to decrease the release of Ca 2+ , produces profound degrees of heart failure by altering this mechanism (Fig. 6). Finally, BDM has been shown to have an as yet poorly characterized effect on the cardiac contractile apparatus to decrease its sensitivity to activator calcium. When administered in appropriate concentrations under the proper experimental conditions, this agent can produce uncoupling of excitation from contraction although intracellular calcium handling, as reflected by the aequorin signal, appears to persist relatively unchanged (Fig. 11). A similar response occurs with acidosis (Fig. 10). The other models of acute and chronic heart failure included in Tables 1 and 2 appear to reflect a combination of factors

20

C. L. Perreault et al.

involving both excitation-contraction coupling and the extra-cardiac or metabolic factors discussed above. Although it is difficult to extrapolate the results of in vitro experiments to man, in vivo, rational therapy of the patient with heart failure clearly depends upon identifying the specific functional abnormalities that are present and selecting specific modalities or drugs targeted to correct them. Supported in part by grant HL31117 and a Research Career Development Award from the National Institutes of Health, Bethesda Maryland, U.S.A. (HL01611); DAO5171 from the National Institute of Drug Abuse, Bethesda, Maryland, USA and a Grant-in-Aid from the American Heart Association. References [1] Grossman W, Lorell B, eds. Diastolic relaxation of the heart. Boston: Martinus Nijhoff Publishing, 1988: 1-305. [2] Morgan JP, Morgan KG. Calcium and cardiovascular function: intracellular calcium levels during contraction and relaxation of mammalian cardiac and vascular smooth muscle as detected with aequorin. Am J Med 1984; 77 (Suppl 5A) 33-^6. [3] Morgan JP, Morgan KG. IntraceUular calcium and cardiovascular function in heart failure: effects of pharmacologic agents. Cardiovasc Drugs Ther 1990; 3 (Suppl 3): 959-70. [4] Gwathmey JK, Copelas L, MacKinnon R et al. Abnormal intraceUular calcium handling in myocardium from patients with end-stage heart failure. Ore Res 1987; 61: 70-6. [5] Ruegg JC, ed. Calcium in muscle activation, New York: Springer-Verlag, 1988: 1-285. [6] Reiter CA. Calcium mobilization and cardiac inotropic mechanisms. Pharmacol Rev 1988; 40: 189-218. [7] Fleischer S, Inui M. Regulation of muscle contraction and relaxation in heart. PTog Clin Biol Res 1988; 273: 435-50. [8] Blinks JR, Endoh M. Modification of myofibrillar responsiveness to Ca + + as an inotropic mechanism. Circulation 1986; 73 (Suppl III): 85-96. [9] Feldman MD, Copelas L, Gwathmey JK et al. Deficient production of cyclic AMP: pharmacologic evidence of an important cause of contractile dysfunction in patients with end-stage heart failure. Circulation 1987; 75: 331-9. [10] Blinks JR. IntraceUular [Ca ++ ] measurements. In: Fozzard HA, Haber E, Jennings RB, eds. The heart and cardiovascular system. New York: Scientific Foundations, 1986: 671-701. [11] Kihara Y, Morgan JP. A comparative study of three methods for intraceUular loading of the calcium indicator aequorin in ferret papillary muscles. Biochem Biophys Res Comm 1989; 162: 402-7. [12] MacKinnon R, Gwathmey JK, Allen PD, Briggs GM, Morgan JP. Modulation by thyroid state of intraceUular calcium and contractility in ferret ventricular muscle. Circ Res 1988; 63: 1080-9.

[13] Kihara Y,.Grossman W, Morgan JP. Direct measurement of changes in intraceUular calcium transients during hypoxia, ischemia, and reperfusion of the intact mammalian heart. Circ Res 1989; 65: 1029-44. [14] Levine M, Meuse AJ, Watanabe J, Bentivegna L, Morgan JP. IntraceUular [Ca2+]i during ischemia in the blood perfused dog heart. Biophys J 1990; 57: 173a. [15] Meuse AJ, Perreault CL, Grossman W, Morgan JP. An experimental procedure for obtaining aequorinloaded isolated mammalian cardiac myocytes. J Gen Physiol 1989; 94: 46. [16] Morgan JP. Mechanism of action of inotropic drugs. In: Braunwald E, ed. Heart disease update. Philadelphia: W.B. Saunders Company, 1989: 136-44. : [17] Morgan JP, Erny RE, Allen PD, Grossman .W, Gwathmey JK. Abnormal intraceUular calcium handling: a major cause of systolic and diastoUc dysfunction in ventricular myocardium from patients with heart failure. Circulation 1990; 81 (Suppl HI): 21-32. . [18] Hajjar RJ, Gwathmey JK, Briggs GM, Morgan JP. Differential effects of DPI 201-106 on the sensitivity of myofilaments to Ca2+ in intact and skinned trabeculae from control and myopathic human hearts. J Clin Invest 1988; 82: 1578-84. [19] Perreault CL, Brozovich FV, Ransil BJ, Morgan JP. Effects of MCI-154 on Ca2+ activation of skinned human myocardium. Eur J Pharmacol 1989; 165: 305-8. [20] Gwathmey JK, Slawsky MT, Briggs GM, Morgan JP. The role of intraceUular sodium in the regulation of intraceUular calcium and contractility. Effects of DPI 201-106 on excitation-contraction coupling in human ventricular myocardium. J Clin Invest 1988; 82: 1592605. [21] Morgan JP, Wier WG, Hess P, Blinks JR. Influence of Ca2+ channel blocking agents on calcium transients and tension development in isolated mammalian heart muscle. Circ Res 1983; 52 (Suppl I) 47-52. [22] Perreault CL, Hague N, Morgan JP. Cocaine's cardiac actions are modified by caffeine. Circulation 1988; 78 (Suppl II) 359. [23] Blanchard EM, Alpert NR, AUen DG, Smith GL. The effect of 2,3-butanedione monoxime on the initial heat-tension integral relation and aequorin Ught output from ferret papillary muscles. Biophys J 1988; 53: 605a. [24] Ohkusa T, Gwathmey JK. The effects of 2,3butanedione monoxime on twitch force and steady state force relationship in ferret papUlary muscle. Biophys J 1989; 55: 100a. [25] Allen DG, Orchard CH. IntraceUular calcium concentration during hypoxia and metaboUc inhibition in mammaUan ventricular muscle. J Physiol 1983; 339: 107-22. [26] MacKinnon R, Gwathmey JK, Morgan JP. Differential effects of reoxygenation on intraceUular calcium and isometric tension. Pfiugers Arch 1987; 409: 448-53. [27] Allen DG, Orchard CH. The effects of changes of pH on intraceUular calcium transients in mammalian cardiac muscle. J Physiol (Lond) 1983; 335: 555-67. [28] AUen DG, Orchard CH. Myocardial contractile function during ischemia and hypoxia. Circ Res 1987; 60: 153-68.

Abnormal intracellular calcium handling

[29] Gwathmcy JK, Morgan JP. Altered calcium handling in experimental pressure-overload hypertrophy in the ferret. Ore Res 1985; 57: 836-43. [30] Warren SE, Hague NL, Morgan JP. Normal intracellular calcium availability in the hypertrophic Syrian hamster. Clin Res 1989; 37: 3O5A. [31] Strobeck JF, Factor SM, Blau A et al. Hereditary and acquired cardiomyopathy in experimental animals. Mechanical, biochemical and structural features. Ann NY Acad Sci 1979; 317: 59-88. [32] Bing OHL, Wiegner AW, Brooks WW, Fishbein MC, Pfeffer JM. Papillary muscle structure-function relations in the aging spontaneously hypertensive rat. din Exp Hypertens 1988; 10: 37-58. [33] Bing OHL, Sen S, Conrad CH, Brooks WW. Myocardial function structure and collagen in the spontaneously hypertensive rat: progression from compensated hypertrophy to hemodynamic impairment. Eur Heart J 1984; 5 Suppl F 43-52. [34] Bing OHL, Brooks WW, Perreault CL, Morgan JP. Myocardial function and calcium transients in hypertrophied and failing LV of spontaneously hypertensive rats. Circulation 1989; 80 (Suppl II): 506. [35] Perreault CL, Bing OHL, Brooks WW, Ransil BJ, Morgan JP. Differential effects of hypertrophy and failure on right versus left ventricular Ca2+ activation. Circulation 1989; 80 (Suppl II): 100. [36] Phillips PJ, Gwatiuney JK, Feldman MD, Schoen FJ, Grossman W, Morgan JP. Postextrasystolic potentiation and the force-frequency relationship; differential augmentation of myocardial contractility in working myocardium from patients with end-stage heart failure. J Mol Cell Cardiol 1990; 22 (in press). [37] Feldman MD, Gwathmey JK, Phillips P, Schoen F, Morgan JP. Reversal of the force-frequency

21

relationship in working myocardium from patients with end-stage heart failure. J Appl Cardiol 1988; 3: 272-83. [38] Hajjar RJ, Morgan JP. Calcium-activation in myopathic human and pressure overload hypertrophied ferret hearts. In: Proceedings of the 1987 HST Forum, Cambridge and Boston: The Harvard-MIT Division of Health Sciences and Technology, 1987: 87-97. [39] Perreault CL, Bing OHL, Brooks WW, Ransil BJ, Morgan JP. Mechanisms responsible for mechanical abnormalities in hypertrophied and failing myocardium. J Gen Physiol 1989; 94: 19. [40] Bing OHL, Brooks WW, Perreault CL, Morgan JP. Calcium transients and inotropy in hypertrophied and failing myocardium from the spontaneously hypertensive rat. Circulation 1989; 80 (Suppl II): 502. [41] Katz AM. Potential deleterious effects of inotropic agents in therapy of chronic heart failure. Circulation 1986; 73 (Suppl III) 184-90. [42] Weber KT, Janicki JS, Shroff S, Pearlman ES. Shape and structure of the normal and failing human heart. In: Alpert NR, ed. Perspectives in cardiovascular research, Vol. 7, Myocardial hypertrophy and failure. New York: Raven Press, 1983: 85-108. [43] Strobeck JE, Sonnenblick EN. Pathophysiology of heart failure: deficiency in cardiac contraction. In: Cohn J, ed. Drug treatment of heart failure. New Jersey: Advanced Therapeutics Communications International, 1988: 49-78. [44] Goldsmith SR, Kubo SH. Pathophysiology of heart failure: peripheral vascular factors and neurohormonal mechanisms. In: Cohn J, ed. Drug treatment of heart failure, New Jersey: Advanced Therapeutics Communications International, 1988: 49-78.