JACC Vol . 23, No. 7June 1994J617-24

Improved Hemodynamic Function and Mechanical Efficiency inCongestive Heart Failure With Sodium Hichloroacetate

ROBERT M . BERSIN, MD, FACC,` CHRISTOPHER WOLFE, MD, FACC,MICHAEL KWASMAN, MD, DEBORAH LAU, RN . CINDY KLINSKI, RN, KEVIN TANAKA, BS,PAYMAN KHORRAMI, BS, GEORGE N . RENDER O)N, PHD,t TERESA DE MARCO, MD, FACC,KANU CHATTERJEE, MD. FACCSan Francisco, California and Gainesville . Florida

Objectives. The purpose of this study was to determine whethersodium dlcldoroaetate improves hemodynar, le performance andmechanical efficiency in congestive heart failure.

Background. Congestive heart failure is associated with im-paired hemodynamic performance and reduced mechanical eSi.fleecy, Dkhiomacetate stimulates pyruvate dehydrogenase activ-hy by i"tion of pyruvate dehydrogenase kinase, which resultsIn Inhibition of free fatty aid metabolism and stimulation of highresptnMrygndienl glucose and lactate consumption by the heart .Facilladon of glucose and lactate consumption with dichtoroace .toe should Improve mechanical e9lciency of the failing ventricle .

Merhedc. Ten patients with New York Heart Associationfunctional doss III to IV congestive heart failure were studied .DkWaoaefate (50 mg)kg body weight) was administered leara-venowly for 30 min, with measurements of thermodynamic vari-whim, coronary sinus blood How and blood gas, glucose and lactateleeve(a for2It. The same patients were also given dalmtamine (S to12 .5 pglkg per min) for comparison .

Resalks, Therapeutic levels of dichloroacetate were achieved(100 to 160 pgflter ofphuma) . Myocardial consumption of lactatewan dirmlakd from 29% to 37.4%. Forward stroke volumesIncreased (+5.3 upbeat, p < 0.02), as did left ventricular strokework (+IA g-mlms per beat, p < 0.02) and left ventricularminute work Moral 1.38 to 1.55 kg-m/mr per min, p < 0.01).Myocardial oxygen consumption decreased (from 19 .3 to

Congestive heart failure is associated with impaired mechan-ical performance of the left ventricle . Whether the cause ofheart failure is ischemic heart disease or previous myocar-dial infarction, congenital or valvular heart disease or aprimary disorder of cardiac muscle, interstitial fibrosis is

All editorial decisions for this ad ide, including selection of referees . weremade by a Guest Editor. This policy applies to all articles with authors fromthe University of Call is San Francium.

Fmm the Cardiology Division, University of California Medical Center,San Francisco, Celil'omia and ODepanment of Medicine, Univers :;, ofHodda, Gai-ilic, Read,,

Manusedpt received May 12, 1993 ; revised manuscript received January20, 1994, accepted January 26, 1994 .

*Present eddre :n m'uddre-, for rorresrondence : Dr . Robed M. Bersin,Cardiology Section, inn Singer Clinic . P.A . . 1001 Blythe Boulevard, Suite300, Chaelulte, North Carolina 28203 .

01994 by the American College of Cardiology

1617

HEART FAILURE

16 .5 mlfmin, p = 0.06) as left ventrkutnr minute work increased .Left ventricular mechanical efficiency thus improved from 15 .2%re ',0 .6% (p = 0 .03) .

Dobutamine adminishattioa ruled in the opposite trend withrespect to myocardial lactate extraction (from 34% to 15 .3%, p <0.02) . Stroke volume increased (+7 .4 rid/beat, p = NS vs .dicl•.loroacelle), as did left ventricular minute mark (from 1.29 to1 .59 g-m/ma per min, p < 0.01 vs . dlchtorooeemta) and myoar.dial oxygen consumption (from 18.6 to 21 .0 n llsda, p = 0 .06 vs .dichloreacetatv) . Left ventricular mechanical efficiency did nutchange with dohutamine administration them 16.4% to 15.8%,p = NS) .

Conclusions. Diehloroacetate administration stimulates myo-cardial lactate consumption and improves left ventricular me,chaniet efficiency. Forward stroke volume and left ventricularminute work increase significantly, with a shi ultanemss reductionin myocardial oxygen camiumption. Dahutandne ndmlnistratiaoresults in similar hemodyaamk improvements but with no changein left ventricular mechanical efficiency and with opposite effectson lactate metabolism . The opposing metabolic actions, yet similarhemodynamic responses, of dkhloroaeutate and dobutamine sug .gest that these agents may be complementary in the treatment ofcongestive heart failure.

(J Am Coll Candiol 1994;23d6d7-24)

generally present, which decreases the -number of contractileproteins available far myocardial contraction (I) . Two com-pensatory mechanisms for interstitial fibrosis and impairedcontractility are volume retention and myocyte hypertrophy(1,2). Volume retention improves the loading conditions ofthe left ventricle by means of the Frank-Starling mechanism(2). Myocyte hypertrophy improves contractility by increas-ing the total mass of myocardial tissue available for contrac-tion (3) . It also reduces wall stress by increasing the wall

thickness of an otherwise dilated chamber (4).Unfortunately, cavilary enlargement, myocyte hypertro-

phy and increased wall thickness combine to fibre thandouble left ventricular mass in chronic heart failure (4,5).Myocardial oxygen requirements increase commensuratelyeven though developed pressure and stroke and minute work

0735-1037194157.00

1618

aERSIN ET AL .DICHLOROACETATE IN CONGESTIVE HEART FAILURE

are generally subnormal (4-9) . Left ventricular mechanicalefficiency thus declines and is typically 15% in chronic heartfailure, whereas it is as high as 40% in normal subjects orpatients with coronary artery disease without heart failure(10).

Sodium dicholoroacetate is an investigational agent thatstimulates pyruvate dehydrogenase activity by inhibitingpyrovate dehydrogenase kinase (11). Stimulation of pyru-vate dehydrogenase with dichloroacetate occurs in mosttissues, particularly the myocardium (11,12) . Stimulation ofpyruvate dehydrogenase activity with dichloroacetate leadsto enhanced glycolysis of glucose and utilization of lactateby the myocardium for aerobic respiration (11,12) . Myocar-dial consumption of free fatty acids is simultaneously inhib-ited, with the overall effect of a change of substrate utiliza-tion from predominantly nonesterifled free fatty acids toglucose and lactate (11-16). The mechanical efficiency of theleft ventricle should thus improve with dichloroacetate to theextent that high respiratory quotient glucose and lactate areconsumed preferentially over low respiratory quotient non-esterified free fatty acids (17-22) . In preliminary studies inpatients with coronary artery disease and normal left ven-tricular function, improvements in left ventricular mechani-cal efficiency were achieved with dichloroacetate adminis-tration (23). Forward stroke volume and cardiac outputincreased at the same time that myocardial oxygen consumetion decreased . Myocardial mechanical efficiency improvedas a result (23) .

The present study was thus undertaken to assess whetherdichloroacetate improves left ventricular mechanical effi-ciency and hemodynamic performance of the failing leftventricle in patients with congestive heart failure .

MethodsPatients. Ten patients with New York Heart Association

functional class III to IV congestive heart failure werestudied . The underlying cause of congestive heart failurewas coronary artery disease in seven patients and idiopathicdilated cardiomyopathy in three. All patients were studiedwhile in hospital for the management of congestive heartfailure at the University of California, Moffitt Hospital, SanFrancisco, California . This study was approved by theInstitutional Review Board of the University of California,San Francisco, and informed consent was obtained for eachpatient .

Study protocol. Pulmonary artery and coronary sinuscatheters were placed using standard percutaneous tech-niques, and proper positioning was confirmed by fluoros-copy. Cardiac indexes (liters/min per m' body surface area)were determined by the thermodilution technique (24) withthe patient supine in the fasting state . A minimum of threedeterminations were made for each patient, and these wereaveraged to obtain a single value . Coronary blood flow(ml/min) was estimated by continuous thermodilution (25)using Wilton-Webster coronary sinus catheters and infusion

JACC VoL 23 . No. 7June 1590:1617-24

of room temperature 5% dextrose in water at a rate of46 ml/min . Oxygen saturation of arterial, pulmonary arterialand coronary sinus blood were measured directly with ahemoximeter (Radiometer, model OSM-2) . Blood gaseswere measured for arterial, mixed-venous and coronarysinus blood using automated blood gas analyzers (ComingMedical, models 168 and 178) . Hemoglobin content wasmeasured by the cyanmethemoglobin method. Blood lactateconcentration was measured using the NAD/NADH assaymethod for samples deproteinized in perchloric acid . Glu-cose concentration was measured using a glucose analyzer .

Analysis of dichloroacetate in plasma . Gas chromatogra-phy was used to analyze dichloroacetate . Briefly, plasma(100 pill is transferred to a 16- x 100-mm screw-cap culturetube (Kimax) . To this, 50 pl of a 50-4m1 solution oftrichloroacetic acid (internal standard) is added, followed byI ml of 14% of boron trifluoride in methanol . The culturetube is vortexed briefly and then heated to 100°C for 15 min .Cyclohexane (I nil) and water (I rap are added to the mixtureand agitated for 10 min (Lab-Line shaker), followed bycentrifugation for 10 min (-2,500 x g). The organic layer(top I )d) is injected into a Varian 3600 gas chromatographequipped with a 6-ft (1.8 m) x 2-mm inner diameter Chro-mosorb 101 column, a nitrogen-63 electron capture detectorand a Nelson Analytical 2600 chromatographic integrator .The column temperature is 160°C . The peak areas obtainedfrom the integrator are used in the determination of the ratioo(dichloroacetate to tcchlooacetate. Every batch of dichlo-roacelate analysis involved calibration with known mixturesof dichkKacetate in plasma (2 .5 to 200 pg/m) and trichkJ-roacetate . The calibration graph was linear .

Calcdadmr . Measured oxygen content was obtainedfrom the directly measured data using the formula: Oxygencontent (vol %) = Oxygen saturatiwJ100 x Hemoglobincontent (g/dl) x 1.34 (ml/g) + 0.0031 Partial pressure ofoxygen (Pot) (mm Hg) . Systemic Oxygen transport (mloxygen/min per m2 body surface area) was defined as Arte-rial oxygen content x Cardiac index x 10. Myocardialoxygen transport (ml oxygen/min) was established similarlyas (Arterial oxygen content x Coronary blood f ow)I100.Systemic oxygen consumption (VO 2) (ml oxygen/min per m'body surface area) was determined with the formula Vo2 =AVDO2 X Cardiac index x 10, where AVDO2 = arterial-venous oxygen difference. Myocardial oxygen consumption(ml oxygen/min) was estimated similarly, where Myocardialoxygen consumption = Coronary blood flow1100 x (Arterialoxygen content - Coronary sinus oxygen content).

Left ventricular stroke work index (g-m/m 2) was esti-mated as Stroke work index = (Mean systolic pressure -Pulmonary capillary wedge pressure) x Stroke volumeindex x 0.0136, where mean systolic and pulmonary capil-lary wedge pressures and stroke volume index were mea-sured, and 0.0136 is a conversion factor from ml x mmHg/m2 to g-m/m2 . Left ventricular minute work was deter-mined by factoring stroke work by heart rate. Left ventric-

:ACC Vol. 23. No. 7June 1994:1617-bi

ular mechanical efficiency was estimated by the ratio of leftventricular minute work to myocardial oxygen consumption .

Drag Infusions. Sodium dichloroacetate, 50 mg/kO, wasadministered intravenously for 30 min with hemodynamicmeasurements made every 15 mrn and blood measurementsevery 30 min for 2 h. After dichloroacetate infusions on thefirst day, the patient was allowed to recover overnight toallow metabolism of dichloroacetate before being givenintravenous dobutamine infusions on day 2 . Dobutaminewas then titrated to clinically optimal doses (5.0 to12.5 psg/kg per min) on the basis of the patient's clinical andhemodynamic response . Measurements of hemodynamicvariables, coronary blood flow, blood lactate levels andoxygen consumption were then performed both before andafter clinically optimal dosing of dobutamine . The 60-minpeak effects of dichloroacetate were compared with theeffects of dobutamine once the patient had achieved a steadystate t'uring a dobutansine infusion for at least 30 min. Bloodsampling and hemodynamic measurements were performedat baseline just before dobutamine infusion and after 30 minof continuous infusion at steady state. The steady-statehemodynamic and blood measurements obtained with do-butamine were then compared with the peak actions ofdichloroacetate ;u the same patients.

StatWial analyses. Statistical analyses were performedusing a repeated-measures analysis of variance. Compari-sons between the peak effects of sodium dichloroacetate andof intravenous dobutamine were performed using a singlepaired two-tailed r test . Differences between dichloroacetateand dobutamine infusions were considered statistically sig-nificant when p < 0.05. Power calculations were performedas follows: with an alpha level of 0.05, a power of 80%, asample size of 10, and a standard deviation of 0.26, onewould be able to detect a clinically important difference of0.23 between baseline and dichloroacetate for cardiac index .Similarly, for left ventricular mechanical efficiency, with astandard deviation of 5 .8, a clinical difference of 5 .2 could bedetected . All statistical analyses were performed using aMacintosh computer and a Statview Statistical Softwarepackage (Brain Power, Inc .). The data are presented as meanvalues t SE unless otherwise indicated.

ResultsPbarmaeodynric variables . Plasma dichloroacetate lev-

els peaked at the end of the 30-min infusion and decreased toplateau levels within 60 min. Plateau levels achieved at60 min were maintained at 120 min, consistent with atwo-compartment distribution with slow elimination kineticsafter an early rapid distribution phase (Fig . 1). There was nopreferential extraction of dichloroacetate by the myocar-dium, as arterial, mixed-venous and coronary sinus concen-trations of dichloroacetate were the same and showed thesame time kinetics (Fig . 1).

Heamadyaamle variables. Baseline hemodynamic vari-ables of the patients studied showed the usual hemodynamie

BERSIN ET AL .

1619DJCHLOROACETATE IN CONGESTIVE HEART FAILURE

z

m aa

Z2®

zOEaa irn

TIME(mtnutan)

Figure 1 . Plasma concentrations of dichloroacetate (DCA) after a30-win infusion of 50 mg/kg of dichloroacetate . Sclaores = arterialplasma samples ; etreles = coronary sinus plasma samples.

profile observed in dilated cardiomyopathy, with a de-creased rest cardiac index and stroke volume and elevationof the central venous and left atrial pressures (Table 1) . Withdichloroacetate, the heart rate, mean arterial, right atriai,pulmonary artery, pulmonary capillary wedge pressures andpulmonary and systemic vascular resistances did not changeappreciably . However, cardiac index increased by an aver-age of 15% (from 1 .95 to 2 .05 liters/min per mt, p < 0.02)(Fig. 2). Stroke volume increased and showed a peak cheerat 60 min (net change +53 mllbea', o < 0.02) (Fig. 3) . Leftventricular stroke work also increased and showed a similartime course, with a peak effect at 60 mia (from 14.7 to16 .5 g-m/mt per beat, p < 0 .02). Estimated left ventricularminute work increased from a baseline of 1 .38 to 1 .55kg-m/mt per min, p < 0.01, with a similar peak at 60 min .The peak hemodynamic actions of dichloroacetate thusoccurred at 60 min, coincident with the rapid distributionphase of the drug and establishment of plateau drug levels.

Coronary blood flow decreased with administration ofdichloroacetate and showed its nadir at 60 min, when for-ward stroke volume and left ventricular stroke work were attheir peak . The trend in coronary blood flow over lime

Table 1. Baseline Hemodynamic VariablesVuouble Mean ± s0

Head rate (beatslmin) 95 z 3Mean arterial pressure (mm Hg) 53 ± 5Cardiac index (lilershnin per m e) 1 .5

0.2Stroke volume index (mlbeal per wJ) 19.4 ± 2.0Pulmonary capillary wedge pre sure (mm Hg) 29 ? 1Central venous pressure (mm Bill 19 ± Isystemic vascularsesistanca(dynesscm ') 1,631 ± 235Pulmrnery vascular resistance (dyaesS- ) 420 c 82Coronary blood flow (ntforin) 163 . 29Myocardial oxygen consumption (ml oxygerdmin) 19.3 `2.8

JACC VO1. 23 . No. 7June 1994:1617-24

I=

Figure 6. Relation between myocardial work and oxygen consump-tion (ehelts) after dichloroacetate )DCA) infusion . Note that leftventricular mechanical efficiency improves after dichloroacetateinfusion at 60 min . Squares = left ventricular mechanical efficiency ;diamonds = left ventricular stroke work index ; n = change. °p <0 .05 .

concentrations decreased, consistent with stimulation ofmyocardial lactate consumption. Coronary sinus lactatelevels reached exceeding low levels, typically in the 0.2 to0.3.mmol4iter range, consistent with maximal stimulationof lactate consumption by the myocardium . Lactate levelsreached their nadir at 60 min of therapy and remained atextremely low levels for the remainderof the 2-h observationperiod. Conversely, blood glucose concentrations did notchange (172 mf/dl at baseline vs . 158 mg/dl at 60 min, p =NS). Coronary sinus glucose concentrations also did notchange. such that net myocardial extraction ratios for glu-cone were unaltered by dichloroacetate .

Myocardial aaygee consumption and mechanical efficiency .Baseline myocardial energetics were characterized by highmyocardial oxygen consumption, normal left ventricularminute work and low left ventricular mechanical efficiency(15.2%). With dichknoacetate, left ventricular minute workincreased as myocardial oxygen consumption decreased .Left ventricular mechanical efficiency increased to 20.6%(p - 0.03) (Fig. 6) . Approximately half of the improvementin myocardial mechanical efficiency was related to thedecrease in myocardial oxygen consumption (maximalchange of -18% at 60 min), with the balance of the improve-ment related to an increase in left ventricular stroke work ofsimilar magnitude (maximal change of +15% at 60 min) .

Comparison of dl tote with dobutaolloe. Cardiacoutput and forward stroke volume increased similarly withdichloroacetate and dobutamine administration (Fig. 7), butthe patterns of substrate utilization were markedly different .Whereas dichkBoacetate increased myocardial lactate con-sumption significantly, dobutattdtle decreased myocardiallactate extraction and consumption (from 34% to 15 .3%,p < 0.02). Net positive lactate production was seen in 3of die 10 patients who received dobutamine, consistentwith the development of myocardial ischemia . Myocardialoxygen consumption decreased with dichloroacetate andincreased with dobutamine (from 18.6 to 21 .0 ml/min,

BERBIN ET AL.

1621DICHLOROACETATE IN CONGESTIVE HEART FAILURE

P,Na P.NM

PEAK rain(om t Ni v. BASELa6)

Figure 7. Peak hemodynamie action of dichbmacetate (solid bras)at 60 min versus optimal effects of dobutamine (hatched bars).LVSWI = left ventricular stroke work index ; t1 = change .

p = 0 .06) . Left ventricular stroke work increased propor-tionately to the increase in myocadiai oxygen consumptionwith dobutamine such that left ventricular mechanical effi-ciency did not change (from 16.4% to 15 .8%, p = NS) (Fig .8) . to contrast, left ventricular mechanical efficiency im-proved significantly with dichloroacvtate, as noted earlier.

DiscussionMyocardial oxygen requirements with dkhloreaeetnle.

Congestive heart failure is associated with impaired mechan-ical performance of the left ventricle . Impairment of leftventricular systolic function leads to reduction in forward

Figure 8 . Peak metabolic actions of dichloroocetate (std bars)versus dobutondne %atebed bats). Note the opposite trends inmyocardial oxygen consumption (MV02) ad coronary blood flow(CBF) and the significant difference in left ventricular mechanicalefficiency (LVME) with the two drugs . MOT = myocardial oxygentransport .

4.7

copP.NS

-3.0

NOT14.05

MVOsP. .06

P.Ns

LVMEPc .05

1622

RERSINETAL .DICHLOROACETATE IN CONGESTIVE HEART FAILURE

stroke volume. Compensatory changes for reduced forwardstroke volume include an increase in left ventricular diastolicvolume and mass. Mechanical efficiency declines as a resultof the increased oxygen demands imposed by increased wallstress, cavitary enlargement and increased left ventricularmass (4-10,26), not because of fundamental changes in thecontractile economy of the left ventricle because the con-tractile economy of failing myocardium remains essentiallynormal (Ishihara H, personal communication, January 1993) .Oxygen requirements are greater than normal to perform anequivalent amount of cardiac work as a result of thisinefficiency (4-10). In the present study, the infusion of50 mg/kg of dichloroacetate for 30 min was associated withan improvement in stroke volume and mechanical efficiencyof the failing left ventricle in patients with congestive heartfailure . To our knowledge, this is the first report of apharmacologic agent that is capable of improving mechanicalefficiency in heart failure .

Although vasodilators reduce myocardial oxygen requ iHo-ments and the improved forward stroke volume, this islargely a result of reduced aflerload, lower peak systolicpressure and commensurate reduction in cardiac work, withno overall change in left ventricular mechanical efficiency(27-30) . Left ventricular volume also tends to decrease as aresult of reduced afterload, which further reduces oxygenrequirements (30). Inotropes, on the other hand, increaseforward stroke volume and left ventricular work by stimu-lation of myocardial contractility . Oxygen requirements gen-erally increase linearly with contractility, with no overallchange in left ventricular mechanical efficiency (31) . In thepresent study, dobutamire increased forward stroke vol-ume, left ventricular work and myocardial oxygen require-ments commensurately such that left ventricular mechanicalefficiency did not change .

Mechanisms of action. The manner in which dichloroac-etate augments myocardial mechanical performance is spec-ulative. Administration of 50 mglkg of dichloroacetate for30 min achieved therapeutic levels of x150 pg/liter ofdichloroacetate in most patients . Plasma levels of 100 to200 pg/liter of dichloroacetate are probably sufficient tostimulate pyruvate dehydrogenase activity in most tissues(11,12). Stimulation of the pyruvate dehydrogenase enzymecomplex by dichloroacetate has been shown to increaseglucose oxidation in cardiac muscle (12,13) . Pyruvate dehy-drogenase is the principal regulatory enzyme in the pathwaycontrolling entry of glycolytic intermediates (glucose andlactate) into the trichloroacetate cycle for aerobic respiration(11-14). Stimulation of pyruvate dehydrogenase thus accel-erates formation of acetyl coenzyme A from pyruvate .Conversion of lactate to pyruvate for aerobic respiration isalso facilitated by dichloroacetate in cardiac muscle(12,16,17,23) . Lactate consumption is increased generally,and marked reductions in blood lactate concentrations areseen with dichloroacetate administration (32,33) . At thesame time, oxidation of short- and medium-chain nonester-ified fatty acids is inhibited, particularly in the myocardium.

JACC Vol. 23. No. 7J .. 1994 :1617-74

A shift from consumption of predominantly nonestedfiedfatty acids to glucose and lactate results (12-16) .

The fact that myocardial lactate extraction was main-tained at extremely low blood lactate concentrations isconsistent with stimulation of myocardial lactate consump-tion by dichloroacetate . Although arterial and coronarysinus blood glucose concentrations did not change overall,the normal myocardium is known to extract only a smallfraction (19%) of delivered blood glucose under basal con-ditions (22). Increases in myocardial glucose consumptionwould not be expected to alter coronary sinus glucoseconcentrations appreciably because baseline extraction ra-tios are so low. Radiolabeled glucose uptake measurementswould be required to determine whether or not myocardialglucose uptake and metabolism were increased with dichlo-roacetate (16), which was beyond the scope of this study.

Enhancement of glucose and lactate oxidation by themyocardium could explain the increase in left ventricularmechanical efficiency seen with dichloroacetate administra-tion . Cardiac muscle, both normal and failing, preferentiallyconsumes aonesterified fatty acids over glucose and lactate .Nonesterified fatty acids, although a rich fuel source foraerobic respiration, are a relatively inefficient fuel source,yielding only 2.8 tool of adenosine triphospism per mole ofnonesterified fatty acids eonsuaxd (low respiratory quotientof0.7) (22). Glucose and lactate, on the other hand, are moreefficient fuels for aerobic respiration, yielding 3 .0 to 3.2 molof adenosine triphosphate per mole of glucose or lactateconsumed (high respiratory quaint of 1.0) (22) . Switchingmyocardial substrate utilization from low to high respiatoryquotient f era could theoretically yield a 16% to 26% im-provement in the oxygen comumption efficiency of themyocardium (12,32) . This is strikingly similar to the actualmeasured improvement of mechanical efficiency observedwith dichloroacetate, which occurred at 60 min (from 16 .1%at baseline to a peak of 20.6% at 60 min, a 28% increasecompared with baseline vahes) . Although not rigorouslyproved, it is suggested that the improvement seen in leftventricular mechanical efficiency in this study is related tothe metabolic effects of dichloroacetate, switching cardiacmuscle from primary nonesteri&d fatty acid consumption toglucose and lactate consumption .

Tore kinedta, Why left ventricular mechanical efficiencydeclined after 60 min is not known. Although plasma levelsof dichloroacetate peaked at 30 min at the end of theinfusion, they remained relatively constant in the 100. to120.mmol/iter range during the 2-h observation period (Fig .I). Despite therapeutic plasma levels of dichloroacetate forthe entire 2 h, left ventricular mechanical efficiency peakedat I h and declined toward baseline levels thereafter. Onepossible explanation may be that lactate levels were drivento such low levels by dichloroacetste that little remained forconsumption in the second hour, leading to "substrateburnout." Against this possibility is the fact that bloodglucose concentrations did not change . Whether the hero-dynamic and metabolic benefits of dichloroacetate could be

BERSIN ET AL.

1623June 1994:1617-2A

DICHLOROACETATE IN CONGESTIVE HEART FAILUREJACC Vol . 23, No . 7

extended beyond 60 min by fueling the myocardium with

ranted to assess whether dichloroacetate is useful as anmore substrate (i .e ., with supplemental glucose, lactate or

adjunct to beta-adrenergic stimulation in the treatment ofpyruvate infusions) is not known. Also, this small pilot study

congestive heart failure.may not be sufficiently powerful to detect small improve-ments in hemodynamic and metabolic function beyond60 min . The precise reason for the increase and decrease inleft ventricular mechanical efficiency during 2 h after dichlo-roacetate administration remains unexplained .

The hemodynamic actions of dichloroacetate demon-strated a similar increase and decrease during the 2-h obser-vation period and also peaked at 60 min . The increase inforward cardiac index observed appears to result from aprimary increase in forward stroke volume because strokevolume increased progressively at the same time that pul-monary capillary wedge pressure fell, with a peak at60-75 min (Fig. 3). The increase in cardiac output appears tobe unrelated to volume loading because the volume admin-istered was small (-50 ml), and pulmonary capillary wedgepressure actually declined during the dichloroacetate infu-sions .

Whether the increase in stroke volume was caused by adirect inotropic action, a primary reduction in afterload or acombination of these actions cannot be determined from thedata obtained in this study . Studies in patients with coronaryartery disease and in isolated perfused rat hearts withendotoxic stress or coronary occlusion have shown evidencefor primary stimulation of myocardial contractility withdichloroacetate (16,18,19,23,34,35). Although a primary va-

sodilating action of dichloroacetate has also been described,this appears to be due to the ganglionic blocking properties

of the diisopropyl ammonium salt of dichloroacetate, not to

dichloroacetate itself (it). The sodium salt of dichloroace-tate has not been associated with hypotension when used

clinically (11,32,33), but the direct effects of dichloroacetateon vascular smooth muscle have not been characterized.

Ccoectnlnin. Whether dichloroacetate exhibits inotro-pism or vasodilation or both, the magnitude of the effect of

dichloroacetate on forward stroke volume and stroke work issimilar to that of clinically optimal doses of intravenous

dobutamine . The striking aspect of dichloroacetate's actionsis that such hemodynamic improvements in the failing myo-cardium can be achieved with a net reduction in myocardialoxygen consumption, with a resultant improvement in leftventricular mechanical efficiency . In contrast, myocardialoxygen consumption increases commensurately with anyincrease seen in left ventricular work stimulated by dobut-amine, and there is no change in overall mechanical effi-ciency. Because beta-adrenergic stimulation is associatedwith increased myocardial extraction of nonesterified fattyacids and reduced consumption of glucose and lactate (the

opposite of what occurs with dichloroacetate), a metabolicbasis for opposing effects of dobutamine and dichloroacetateon left ventricular mechanical efficiency is suggested . One

would predict that inotropes and dichloroacetate might wellbe complementary and have additive effects on hemody-

aun ie performance (36) . Further studies thus appear war-

We am indebted to Vem B. Jackson for preparation of the manuscript andDr . James Sonar for assistance with the statistical analyses .

ReferencesI . Zimmer G . Zimmermann R, Hess OM, et al. Decreased concentration of

myofibrils and myofiber hypertrophy are structural dctermioarts of ion.paired lea ventricular function in patients with chronic heart disease : amultiple logistic regression analysis . J Am Cob Control 1992;20:1135-42.

i

2 . Eon AM Cordiomyopoohyofacrdond: nncj000slrooicaatafpcoosccicnin concesuve hear failure . N East I Med 1990922 :100-10.

3 . Kinyenbuehl HP, He . OM, Schneider J . Tluina M. Physiologic orpathologic hypertrophy . Eur Sean J 1983 ;4 (Sappl A):29-34.

4. Stmuer BE. Myocardial oxygen consumption in chronic heap disease :rote of wall stress, hypertrophy and coronary reserve . Am J Cordial1979 ;44 :730-4e.

5. Weiss MB. Ellis K . Sciacca RS, Johnson LL, Schmidt DH, Cannon PJ .Myocardial blood Bow in congestive and hypenrophte cardiomyopmhy .Circulation 197634:484-93 .

6 . Weber KT. Janicki JS. The metabolic demand and oxygen supply aideheart: physiologic and clinical considerations. Am J Cardiol 1979;44:722-9.

7. Brau^wald E . Control of myocardial oxygen consumption-physiologicand clinical considerations . Am J Candid 1971 :27 :416-32.

8. DeMureo T, Chattedee K. Rouleaa JL. Paeodey WW. Abnormal coro-nary hemodynamks and myocardial energetics in patients with chronicheart failure caused by ischemic heart disease and dilated cardiomyopa .,by . Am Heart J 1988 :115 :809-15.

9 . Hasenfuss G. Holuhmsh C, Heiss W, et al . Myocardial emrgehcs inpatients with dilated cardiomyopathy. Circulation 1959 ;80:51-64 .

10 . Takaoka H . Takeuchi MT, Odake M . et al . Comparison of hemodynamicdeterminants for myocardial oxygen consumption under different candeac-file slates in human ventricle . Circulation 1993;87 :59-69,

I I . Stacpoole PW . The pharmacology of dichloroacetate. Metabolism 1989;38:1124-44 .

12. McAllister A . ,Aaron SP. Randle PJ . Effects of diddoroacetate on themetabolism of glucose, pymvate, acetate, 3-hydmxybutymte and palmi-(ate in rat diaphragm and bean muscle in vilm and on extinction o`glucose, lactate, pymvate and free fatty acids by dog heart in vivo.Biochen 11973 ;134 :1067-8I.

13. Latipaa PM. Kolervo Hiltmlen J, Pcakurinen Kl, Hassinen E . Regulationof fatty acid oxidation in heart muscle: effects on pyruvale and diehloto-acetate. Biochim Biophys Acta 1953 ;752 :162-71 .

14 . Weiss RG . Chacko VP, Gerstenblith G . Fatty acid regulation of glucosemetabolism in the intact beating not heart assessed by carbon-13 NMRspectroscopy: the critical rote of pymvate dehydrogenate. J Mol CeaCardio11999:21:469 18.

15 . Nichoa TA. Lopaschuk GD, McNeill JH. Effects of free fatty acids anddichloroacetate on isolated working diabetic rat hearts . Am 1 Physiol1991 ;261:H1053-9.

16. Mjos OD . Miller NS, Riemelsma RA, Oliver MF. Effects of dichloroac-etate on myocardial substrate extraction, epicadial ST segment cleva.lion, and ventricular blood flaw following coronary occlusion in dogs .CaNiovasc Res 1976 ;10:427-M.

17. Salty M, Suedlow DB, Schaibte DH, Fleisher G . Effects of dichloroace-tate and naloxone m: hemodynmmic perform and substrate usefollowing endotoxin adminislntion . J Cril Cam 1985;3:82-8.

16. Matsuoka 5, Toshima X, Naito E . It al . Effects of dichloroacetate on themechanical function of the isolated schemic heap . Jpn Heart J 1987 ;25:531-7.

19. Bums AH, Giaimo ME, Summer WR. Dichlmoacelic acid improves invitro myocardial function following in viva endatozin administration.J Cot Care 1986 :1 :11-7.

1624

BERSIN ET AL .DICHLOROACETATE IN CONGESTIVE HEART FAILURE

20. Vik-Mo H, Mjos OD . Influence of free fatty acid on myocardial oxygenconsumption and ischemic injury . Am I Cardiol 1911 ;48:361-5.

s

21 . Liedtke JA, Nellis S, Neely JR . Effects of execs free fatty acids onmechanical and metabolic function in norrnaland iuhenuc myocardium inwine. Cioc Res 1978 ;43 :652-61 .

22 . :iahles H, Hellige G, Hunneman DH, Meager VA, Breluhneider HJ .Influence of myocardial substrate utilization on oxygen consumption ofthe been . Ckn Cardiff 1912 ;5 :216-93 .

23. uargovich TI, Macdonald RG, Hill JA, Feldman RL, Smcpook PW,Pepine CJ . Myocardial metabolic and hemodynamic effects of dichloro-acetate in coronary artery disease . Am 1 Cardiol 11 ;61 :65-70.

24. Ganz W, Donate R, Marcus HS, Folrester 1S, Swan HIC. A newtechnique for measurement of cardiac output by thermoddution in man .Am I Cardiol 197127:392.6.

23, Ganz W, Tam.. K, Marcus HS, Donose R, Yoshida S, Swan HIC .Meamrmnent of coronary sinus blood flow by cominuons theemodllutionin man . Circulation 1971 ;44:181-95.

26. Sup H, Igarashi Y . Yamada 0, Gob Y . Mechanical effmiency of the leftventricle as a function of prelond, afterk ad, and contractilay. HeartVessels 1985;1 :3-8 .

27. Rubin SA, Brown HV, Swan HJC. Arterial oxygenation and arterialoxygen transport in chronic myocardial failnreat real, during exerciseandafter hydrelatine treatment. Circulation 1982;66:143-8.

28. Cohn IN, Johnson G, Zkache S, cl al. A comparison of enalapril withhydralezine-isosorbide dinivate in the treatment of chronic congestivefailure . N Engl I Med 1991 .325 :303-t0.

JACC Vol. 23. No. 7June 1994 :1617-24

29. K.-Y .. T, Asami H . Ishiraka S . Sasayama S . Ventricular loadoptimvationbyunloadingtharapyinpatientswithheart failure . IAmCallCardiol 1991 :17 :199-207.

30 . Belle CD,KramerB,HoffieneistarHM,Muller-SebauenburgW,RElerT,Seipel L . Low-done emlapdl in severe chronic heart failure . CardiovascDmgsTher 1989 ;3:211-8 .

31, Sup H, Hisam R, Gnto Y, Yansrdo 0 . Igansbl Y. Effect of positiveimtropicagentsonthe relatimtbetween oxygen consumption nosystdkpressure volume area in canine left ventricle . C oc Res 1933:5'1 :306-I8 .

32 . SmcpoolePW,LoreneAC.ThomasRG,HarmonEM .DichknoncelateinThe treatment of lactic acidosis. Ann Intern Med 198&105:58-63 .

33 . Staepoole Flat, Weight EC, Baamgarmm TG, et al . A eontrc2ed clinitdvial of dkhlorwcelale for treatment of lactic acidosis in adults . N Foe! IMed 1993-327:1564-9.

34 . Sakai K . Ichihan K. Nasa Y, Kamigaki M, Abiko Y . Dichtoroacetareanemones myocardial acidosis and metabotk changes induced by partialocclusion of the coronary artery in dogs . Arch lot Pharmamdyn 1990 :307:92-108.

35. McVeigh 11. Lapauhuk GD. Dichtmoaceme stimulation of glucoseoxidation improves recovery of iuheetic rat hearts . Am 3 Physic]1990;259:HI079-85.

36. Bums AH, Summer WR. Ramy Runs LA. Shepherd RE. Imtropicinteractions of dichkroacetate with amrinone and ouabain in isolatedhearts from cndotoxbrshocked not. I Cardiovasc Phannacol 1985 ;11 :379-86.