myocardial reference vasodilatory effects humans · haemoglobin, c02,.,, is arterial oxygen...

Heart 1997;78:1 17-126

Myocardial oxygen supply:demand ratio as

reference for coronary vasodilatory drug effects inhumans

Isabelle Vergroesen, Jasper E Kal, Jos A E Spaan, Harry B Van Wezel

AbstractObjective-Introduction and measure-ment of human myocardial oxygen sup-ply:demand ratio as a reference forquantification of coronary microvascularvasodilating drug effects in clinical stud-ies. Myocardial oxygen consumption isthe major determinant of coronary bloodflow; therefore, the true vasodilatingproperties of coronary vasodilating drugsthat may have an effect on oxygen con-sumption cannot be correctly assessedfrom blood flow changes alone.Design-Prospective, controlled trial.Setting-Academic hospital.Patients-12 patients with multivesselcoronary artery disease (CAD) under-going coronary artery bypass grafting.Interventions-Cardiac pacing at 30beatslmin above sinus rhythm in awakeand anaesthetised patients (fentanylpan-curonium bromide).Main outcome measures-Myocardialoxygen supply, defined as coronary sinusblood flow multiplied by arterial oxygencontent; myocardial oxygen demand,defined as coronary sinus blood flow mul-tiplied by arteriovenous oxygen contentdifference. The change in oxygen demandinduced by pacing was related to thechange in myocardial oxygen supply inawake and anaesthetised patients. Thismyocardial oxygen supply:demand ratiodetermined in the reference study wascompared with that induced by intra-venous and intracoronary drugs (nifedip-ine, felodipine, urapidil, and sodiumnitroprusside) in two pharmacologicalstudies: patients with CAD undergoingcardiac surgery (45 treated with sodiumnitroprusside, 27 with nifedipine, and 27with urapidil to manage arterial bloodpressure); and patients with unstableangina (and a similar degree of CAD)undergoing cardiac catheterisation fordiagnostic purposes (10 treated withintracoronary nifedipine and 10 withintracoronary felodipine).Results-When awake, the ratio of pacinginduced oxygen supply:demand changesin the 12 reference study patients was 1 50(95% confidence intervals (CI), 1*41-1.58), similar to the 1-45 (1.35-1.56) mea-sured in the same patients after inductionof anaesthesia. Anaesthesia per se did notincrease coronary oxygen supply abovethe expected increase related to demandchanges. The only significant change in

the oxygen supply:demand ratio wasinduced by intracoronary bolus adminis-tration of nifedipine and felodipine (10.6(SE 1.9) and 13-9 (1.9) mllmin, respec-tively, above the demand related supply).Conclusions-Quantification of coronaryvasoactive properties in relation to thephysiological reference ratio betweenmyocardial oxygen supply and demandmay be a powerful tool to differentiatebetween true and apparent coronaryvasoactive drugs.

(Heart 1997;78:1 17-126)

Keywords: coronary circulation; vasodilatation; meta-bolic flow regulation; vasoactive drugs; pacing

Normal regulation of coronary blood flow is amultifactorial process possibly depending onmetabolic, myogenic, neurohumoral, andendothelial responses. 1-6 The overall effect ofthese combined responses leads to continuousmatching of myocardial oxygen supply anddemand.7-9 In humans the myocardial oxygensupply:demand ratio should be consideredwhen the effects of coronary vasodilatory stim-uli are described. Although there is a widebody of literature dealing with the effect ofcoronary vasodilators on myocardial bloodflow and oxygen consumption, the potentialinvolvement of the normal regulation processof coronary flow is rarely taken into account.'0 11The purpose of the present study was toredress this inadequacy in the assessment ofthe capacity of drugs to vasodilate the resis-tance vessels of the coronary circulation.

MethodsA physiological reference study gathered datafrom 12 patients (both awake and anaes-thetised) with multivessel coronary arterydisease (CAD) undergoing coronary arterybypass surgery. These data were comparedwith those from 45 anaesthetised patientstreated with sodium nitroprusside, 27 anaes-thetised patients treated with urapidil, and 27anaesthetised patients treated with nifedipine.These patients had been studied to evaluatethe effectiveness of the drugs in the manage-ment of arterial blood pressure during cardiacsurgery.' 2-14 They were also compared with astudy of awake patients undergoing cardiaccatheterisation for diagnostic purposes; 10treated with intracoronary nifedipine and 10treated with intracoronary felodipine.'5

Department ofMedical Physics,Academic MedicalCenter, University ofAmsterdam,Meibergdreef 15, 1105AZ Amsterdam, TheNetherlandsI VergroesenJ A E SpaanDepartment ofAnaesthesiologyJ E KalH B van WezelCorrespondence to:Dr I Vergroesen, email:I.Vergroesengamcx.uva.nlAccepted for publication10 April 1997

117 on D

ecember 26, 2020 by guest. P

rotected by copyright.http://heart.bm

j.com/

Heart: first published as 10.1136/hrt.78.2.117 on 1 A

ugust 1997. Dow

nloaded from

http://heart.bmj.com/

Vergroesen, Kal, Spaan, Van Wezel

PATIENT CHARACTERISTICSAll patients gave informed consent to partici-pate in one of the studies that all had institu-tional approval. For the physiologicalreference study and for the drug studies underanaesthesia, the patients with stable coronaryartery disease (CAD) were scheduled for elec-tive coronary artery surgery. Excluded fromthe study were patients with left ventricularend diastolic pressure higher than 18 mm Hg,left ventricular hypertrophy, ejection fractionless than 45%, atrioventricular conductiondefects, main stem stenosis or unstable angina.Patients undergoing additional surgical proce-dures-for example, valve replacement oraneurysmectomy, were also excluded. In thepharmacological study of awake patients aftercardiac catheterisation for diagnostic pur-poses, patients were confirmed to meet similarcriteria of CAD as the other groups.'5

INSTRUMENTATIONFor patients undergoing surgery, ECG leadswere connected and leads II, III, and V5 werecontinuously monitored (HP Merlin System,Hewlett-Packard, Boblingen, Germany). Awide bore peripheral infusion and a 20 gaugeradial artery canula were inserted under localanalgesia.

For the awake patients, a triple lumenpulmonary artery catheter (Baxter HealthCare Corporation, Irvine, California, USA)and a coronary sinus thermodilution catheter(Wilton-Webster Laboratories, Altadena,California, USA, type CCS-7U-90B) wereintroduced via the left subclavian vein asdescribed before.'2"3'6 The absence of rightatrial admixture in coronary sinus blood waschecked by injection of cold saline in the rightatrium, while coronary sinus temperaturecurves were recorded simultaneously.'7 18Under fluoroscopy, pacing (via coronary sinuscatheter) was practised during 10 to 30 sec-onds to ascertain the stability of the position ofthe tip of the coronary sinus catheter in rela-tion to the surrounding anatomical structuresand fluoroscopic landmarks. In addition, theelectrical threshold for pacing was determined.For the measurement of coronary sinus bloodflow, normal saline at room temperature wasused as indicator and infused into the coro-nary sinus at a rate of 45 ml/min via a Mark IVinfusion pump (Medrad Technology forPeople, Pittsburgh, USA).'9 Infusion rateswere verified by timed volume collection andflow calculations reflected the indicator infu-sion rate used.

ANAESTHESIA TECHNIQUECalcium channel blockers and long actingnitrates were given until the evening beforesurgery. adrenoreceptor blocking agentswere continued until the morning of surgery.Lorazepam 4-5 mg was given for premedica-tion two hours before surgery.

At the start of induction of anaesthesia thepatients were pre-oxygenated. Pancuroniumbromide (2 mg) was given followed by fen-tanyl (100 ug/kg) injected over five minutes.When the patient became unresponsive to

commands an additional 6 mg dose of pan-curonium was given, ventilation was assistedand then controlled manually. After intuba-tion of the trachea the lungs were ventilatedwith air and oxygen (FIO2 = 0 5). Ventilationwas adjusted to maintain the end-tidal carbondioxide concentration between 4% and 4-5%.In the first 15 minutes following induction ofanaesthesia 250-500 ml of gelofuscine (ViforMedical SA, Crisier, Switzerland) was infusedto maintain a stable haemodynamic situation.Gelofuscine is a gelatine solution containing(per 500 ml) 20 g of modified gelatin,77 mmol Na+, 63 mmol Cl-, pH 7d1-7-7,279 mOsm/l.

PHYSIOLOGICAL REFERENCE STUDYThe measurement protocol and data analysisof the reference study are shown in figure 1.After adequate instrumentation and a restingperiod of 20 minutes a measurement serieswas obtained in 12 awake patients. Twentyminutes after induction of anaesthesia andendotracheal intubation, the complete proto-col was repeated. Each set of measurementsconsisted of a steady state recording period ofcoronary sinus blood flow, arterial blood pres-sure, pulmonary artery pressure, right atrialpressure, and ECG lead II for 20 seconds.These recordings were stored on a computerdisk for off-line analysis. Furthermore, pul-monary capillary wedge pressure and singlebolus cardiac outputs were obtained. Cardiacoutput was calculated as the average of at leastthree measurements and reported as cardiacindex (CI) (cardiac output/body surface areain 1/min/m2). Blood samples from the radialartery and coronary sinus were drawn to deter-mine plasma haemoglobin, partial pressure ofoxygen (P02), and oxygen saturation (SatO2).

SatO2 was measured by an OSM-II haem-oxymeter (Radiometer, Copenhagen) and P02by an ABL-III (Radiometer).

Myocardial metabolic indices were calcu-lated according to standard formulae.'6

c02 = 2-24 - Hb * SatO2+ 0 003 * P02 (ml02/100 ml)MV02 = CSBF (cO2,ar, - c02,cJ)/1 00 (ml02/min)myocardial oxygen supply = CSBF -

c02,/100 (ml 02/min)where c02 is oxygen content, Hb is plasmahaemoglobin, c02,.,, is arterial oxygen con-tent, c02,c, is coronary sinus oxygen content,and MV02 is myocardial oxygen consump-tion.

PHARMACOLOGICAL STUDIESIn the pharmacological studies in patientsundergoing coronary artery bypass grafting(CABG) the protocol consisted of a series ofbaseline measurements 10 minutes afterinduction of anaesthesia, followed by onset ofcontinuous infusion of the study drug.Infusion rates of each drug were adjusted tomaintain systolic blood pressure between 80%and 120% of baseline values. Additional mea-surements were obtained 10 minutes after theonset of infusion of the study drugs (beforesurgery started) and after sternotomy when the

118 on D



j.com/


ugust 1997. Dow

nloaded from


Myocardial oxygen supply:demand ratio as reference for coronary vasodilatory drug effects in humans

Data analysis and presentation

A Left A Right t

Left vI

Right A

Figure 1 The upper part shows the protocol. Each measurement series includes 20 seconds recording ofcoronary sinusbloodflow, arterial blood pressure, pulmonary artery pressure, right atrial pressure, andECG lead II, followed bymeasurement ofpulmonary capillary wedge pressure, cardiac output, and withdrawal of arterial and coronary sinus bloodsamples. Data analysis and presentation: the pacing induced changes in MVO, and oxygen supply were calculatedfromthe difference between values obtained in measurement series 2 and 1 (awake), and 5 and 4 (anaesthesia). Changesinduced by cessation ofpacing were calculatedffrom the differences between measurement series 3 and 2 (awake), and 6and 5 (anaesthesia). The results of this analysis are presented infigure 3. The effect of anaesthesia was calculated by thechanges in myocardial oxygen supply and demand between measurement series 4 and 3 (sinus rhythm), and 5 and 2(pacing). The results of the latter analysis are presented in figure 4.

pericardium was opened, representing maxi-mal surgical stress. The changes in oxygensupply and demand were calculated for bothmeasurement periods (start dose and follow-ing stemotomy) in relation to baseline valuesas for the reference study.

In the pharmacological study in awakepatients after cardiac catheterisation for diag-nostic purposes, a baseline measurement wasobtained just before intracoronary bolus infu-sion of the study drug. Additional measure-ments were obtained 1, 3, 5, 7 5, and 10minutes after the bolus injection. Changes inoxygen demand and supply were measuredone minute after bolus injection, when the

A02 supply 20(ml 02/min)

t 0

-20 20

*A02 demand(ml 02/min)

-20

Figure 2 Theoretical demonstration of calculation ofcoronary microvascular vasodilating potency in a virtualdataset. 0, supposed measurements; solid line, humanphysiological oxygen supply:demand ratio (reference line);dotted lines, individual differences between measuredoxygen supply and expected oxygen supply. The meancoronary microvascular vasodilating potency is the meanvertical distance to the reference line according to the dottedlines, where points below the reference line yield negativevalues and points above the line positive values.

change in myocardial oxygen demand wasmaximal in all cases.'5

STATISTICAL ANALYSISData were analysed using paired standard ttests. A value of P < 0 05 was considered sig-nificant. Results are reported as mean (SD)unless stated otherwise. The change in coro-nary oxygen supply was related to the changein coronary oxygen demand by standardregression analysis. In the physiological refer-ence study using the pacing protocol, a changein myocardial oxygen supply cannot occurwithout change in myocardial oxygen demand,therefore, this regression line was only calcu-lated without intercept. (Standard formula:AO, supply = ratio * AO, demand.) For thequantification of the drug induced effects(including the effect of anaesthesia) onmyocardial supply:demand ratios two regres-sion analyses were performed, one with inter-cept (AO, supply = ratio - AO, demand) andone without (AO, supply = ratio * AO,demand + intercept). The estimated ratios(slopes of regression lines) and intercepts arereported with their 95% confidence intervals(CI) only if the regression analysis resulted ina significant fit.The coronary microvascular vasodilating

potency of each drug was calculated as themean difference between the expected changein oxygen supply (based on the physiologicalreference relation and the measured change inoxygen demand) and the actual change in oxy-gen supply found in each patient. The coro-nary microvascular vasodilating potency is themean difference between oxygen suppliesfound and oxygen supplies expected on thebasis of the measured oxygen demand and thereference human oxygen supply:demand ratio

119 on D



j.com/


ugust 1997. Dow

nloaded from



Table 1 Patient characteristics and clinical data ofphysiological reference study

Age Height Weight Extent ofPatient (years) Sex (cm) (kg) CAD PriorMI Hypertension

1 67 M 168 68 RCA,LAD,LCX YES NO2 62 M 172 85 RCA,LAD,LCX NO NO3 66 F 157 58 RCA,LAD YES YES4 69 M 175 73 RCA,LAD,LCX NO NO5 67 M 176 70 RCA,LAD,LCX NO YES6 51 M 171 94 RCA,LAD,LCX NO NO7 72 M 182 78 RCA,LAD,LCX YES NO8 47 M 185 86 RCA,LAD YES NO9 55 M 174 70 LAD,LCX YES NO10 71 M 172 80 RCA,LAD,LCX NO NO11 54 M 172 97 RCA,LAD,LCX YES YES12 61 M 180 86 RCA,LAD,LCX NO NO

RCA, right coronary artery; LAD, left anterior descending coronary artery; LCX, left circumflex coronary artery; CAD, coro-nary artery disease; MI, myocardial infarction.

Table 2 Haemodynamic characteristics ofphysiological reference study (n = 12)

Awake Fentanyl

Baseline Pacing Post pacing Baseline Pacing Post pacing

HR 61 (8) 86 (6)* 61 (7) 58 (7) 87 (7)* 59 (8)MAP 102 (21) 108 (22) 105 (22) 79 (17)t 91 (16) 80 (16)RAP 4-8 (3 5) 4-2 (3 7) 5-1 (3-7) 7-3 (4 2)t 6-7 (4-1) 7-5 (4 0)PAP 23 (4) 26 (5) 24 (4) 20 (6) 22 (5) 20 (5)PCWP 13 (4) 11 (6) 12 (5) 11 (3) 11 (3) 10 (3)CI 2-5 ( 7) 3-2 (.6)* 2-7 ( 7) 2-4 ( 7) 3-1 (.8)* 2 6 ( 7)C02,=,t 17-9 (2-2) 17-9 (2 2) 17-9 (2-2) 19-4 (2 2)t 19-2 (2 2) 19-2 (2 2)C02 CS 6-3 (1-3) 6-5 (1-3) 6-3 (1-3) 7-4 (1-3)t 7-2 (1-1) 7-2 (1-1)MVO2 13 (4) 18 (7)* 13 (5) 12 (5) 18 (7)* 13 (60)CSBF 108 (23) 154 (35)* 111 (24) 99 (32)t 143 (46)* 100 (400)

All data are mean (SD); HR, heart rate (beats/min); MAP, mean arterial pressure (mm Hg); RAP, right atrial pressure(mm Hg); PAP, mean pulmonary artery pressure (mm Hg); PCWP, pulmonary capillary wedge pressure (mm Hg); CI, cardiacindex (1/min/m2); C02,.,,, arterial oxygen content (m/l/00 ml); cO2,C1 coronary venous oxygen content (ml/I00 ml); MVO2,myocardial oxygen consumption (ml/min); CSBF, coronary sinus blood flow (ml/min).*Significantly different from baseline, t significantly different from awake state.

(fig 2, solid line). Data points below theexpected oxygen supply give a negative contri-bution to the mean coronary microvascularvasodilating potency and data points above theline of expectancy a positive. The physiologi-cal reference relation was obtained from theregression line fitted to the pacing inducedchanges in oxygen supply and demand.

ResultsREFERENCE STUDYPatient charactenisticsThe 12 patients participating in the referencestudy were comparable with respect to age,weight, and height (table 1). Although sixpatients had suffered a previous myocardial

infarction and three had (treated) hyperten-sion, there was no evidence of either impairedleft ventricular function and dilatation, or leftventricular hypertrophy. All patients wereusing long acting nitrates, ,B blockers, and cal-cium channel blockers for the chronic man-agement of ischaemic heart disease. Nopatient was excluded from the data analysisduring the course of the study for any reason.

Haemodynamic and metabolic characteristicsHaemodynamic results obtained in both theawake and anaesthetised condition arereported at baseline, after onset of pacing, andafter discontinuation of pacing (table 2).Pacing resulted in a small change in meanarterial pressure in the direction of the heart

Figure 3 Physiologicalmetabolic flow regulation;the effect ofpacing (A) at30 beatslmin above sinusrhythm, and cessation ofpacing (V) on myocardialoxygen supply and demandtn patients with coronaryartery disease. Left panel,awake; Right panel, afterinduction of anaesthesia byfentanyl and pancuroniumbromide. The linesrepresent the results ofregeession analysis to thedata presented here andwere used as reference infigures 4-7. Details areshown in table 3.

-20

20

20 -20

A02 demand(ml 02/min)

w

_A02 supply(ml 02/min)

//A

It20


-20 _

Anaesthesia

120

Awake

on Decem

ber 26, 2020 by guest. Protected by copyright.

http://heart.bmj.com

/H

eart: first published as 10.1136/hrt.78.2.117 on 1 August 1997. D

ownloaded from



Table 3 Regression analysis: AO2 supply = ratio AO2 demand

Coronaryvasodilatating

Intervention n Ratio 95% CI 72 potency

Pacing, awake 24 1-50 1-41-1-58 0-98 005 (1-20)Pacing, anaesthesia 24 1-45 1 35-1-56 0 97 -0-25 (1-35)Anaesthesia, sinus rhythm 12 1-41 1 29-1-52 0 93 0 35 (1-26)Anaesthesia, pacing 12 1 35 1-22-1-48 0 97 0-69 (1-40)SNP, start dose (iv) 45 1-29 1-14-144 0-87 0 55 (1-47)*SNP, sternotomy (iv) 45 1-50 1 26-1-74 0-72 1-78 (2 75)*Urapidil, start dose (iv) 27 1-57 1-38-1-75 0 90 0-38 (1-50)Urapidil, stemotomy (iv) 27 1-43 1-301-56 0-92 0-85 (2 50)Nifedipine, start dose (iv) 27 1-37 1-08-1-65 0 73 0 35 (2 25)Nifedipine, sternotomy (iv) 27 1 48 1-24-1 73 0-41 2-45 (5 12)*Nifedipine, awake (ic) 10 NS NS NS 10-57 (5.87)*Felodipine, awake (ic) 10 NS NS NS 13-85 (5.96)*

Coronary microvascular vasodilating potency (mI02/min) is calculated as explained in figure 2, presented as mean (SD). All ratiospresented were significantly different from zero.95% CI, 95% confidence interval; r2, regression coefficient; n, number of observations used to estimate the regression linespresented in tables 3 and 4; SNP, sodium nitroprusside; iv, intravenous; ic, intracoronary; NS, no significant fit.*Significantly different from zero.

rate change; during anaesthesia changes inmean arterial pressure were significantly largercompared with the awake state (12 mm Hg v6 mm Hg). Similarly, the discontinuation ofpacing resulted in a blood pressure decrease of3 mm Hg in the awake state, and a blood pres-sure decrease of 11 mm Hg in the anaes-thetised state. Compared with the awakesituation, mean arterial pressure decreased inall patients significantly after induction ofanaesthesia.

Right atrial pressure, pulmonary arterypressure, and pulmonary capillary wedge pres-sure did not change significantly during thestudy. CI increased significantly as a result ofpacing in both awake and anaesthetisedpatients.

Arterial and coronary sinus oxygen content(cO2,art and cO2,cs), and oxygen supply anddemand were calculated at baseline, duringpacing (after coronary sinus blood flow hadreached a steady state), and after cessation ofpacing. In awake and anaesthetised patients,cO,,2c did not change in response to heart ratechanges, although cO2,cs increased 1.1 ml/ml(P < 005) and cO2,,a increased 1-5 ml/ml afterinduction of anaesthesia because of ventilationof the patient. MVO2 increased significantlyduring pacing and this change was similar inboth the awake and anaesthetised state.

MYOCARDIAL SUPPLY:DEMAND RATIOSReference studyFigure 3 shows the pacing induced change inoxygen demand related to the subsequentchange in oxygen supply, awake and afterinduction of anaesthesia. In each patient (bothawake and anaesthetised) two supply:demandratios are shown in each plot, one depictingthe effect of pacing versus sinus rhythm, andthe second showing the effect of cessation ofpacing. The regression lines calculated forboth conditions were not significantly differ-ent: awake, AO,supply = 1-50 * AMVO2 (95%CI, 1-41-1 59); anaesthetised, AO2supply =1-45 * AMVO2 (1-35-1-56) (table 3). Thecoronary microvascular vasodilating potenciescalculated for the pacing induced results givean impression of the natural variation that maybe expected in data of this kind.

Figure 4 shows the effect of anaesthesiainduced changes in myocardial oxygendemand in relation to the pacing inducedregression line in the awake condition (solidline, identical to solid line in figure 3, leftpanel). Compared with the awake situation,MVO2 is significantly reduced by anaesthesia(table 2); however, this is associated witha matching decrease in coronary oxygensupply (r = 096). The calculated regressionline is characterised as AO2supply = 1 41

Figure 4 Effect ofanaesthesia on myocardialoxygen supply and demandin patients with coronaryartery disease Left panel,at sinus rhythm; Rightpanel, during pacing at 30beatslmmn above baseline.The lines presented in thesefigures are the regressionlines of the physiologicalreference data obtainedawake and presented infigure 3 (left panel).Results of regressionanalysis of the data andthe calculated coronarymicrovascular vasodilatingpotency relative to the lineshown are presented intable 3.

20


-20

A -20

Sinus rhythm

-20 20


121

Pacing

on Decem



/H


ownloaded from



A02 supplY(ml 02/min)

/

A02 d4(ml 0

A02 SU

407 (m 02

b/O

40 -20 *P.

emand -2/min) -20

A02 supply40 r (ml 02/min)

Af, die

i40 -20 a

4mand A(lA 2 u I IO' lu

(ml 02Jmin)

A02 SL40 F (ml 02

0

0 /

-20 '

40

I~~~~~~~~~~~~~~~~~~~~~~40 -20

Lmand /

(ml 02/min)-20 _

_ A02 SL

(ml 02

* .7"

SternotonFigure 5 Effects ofpotentialUy vasoactive drugs (A) urapidil, (B) sodiunand (C) nifedipine administered intravenously, on myocardial oxygen sujin anaesthetised patients with coronary artery disease. Left panels, 10 miudrug infusion before surgery was started; Right panels, after sternotomy, &infused at a dose sufficient to keep arterial blood pressure between 80% anpreinfusion values. Dashed lines in each panel represent the regression linephysiological reference data shown in figure 3 (right panel) obtained duriand were notfitted to the data shown in this figure. These lines were usedcalculate the coronary microvascular vasodilating potency of each drug. i

regression analysis of the data shown in this figure are presented in tables

AO,demand (95% CI, 1-29-1-5no significant intercept and tlines calculated at baseline hearting pacing were not significa(table 3). The coronary microvalating potency of anaesthesia wcantly different from zero.

ipply Pharmacological studies/min) The changes in myocardial oxygen demand and

supply measured during infusion of urapidil,sodium nitroprusside, and nifedipine are pre-sented in figure 5 for individual patients duringanaesthesia. In all panels, the dashed lines rep-resent the physiological supply:demand ratioobtained by pacing during anaesthesia in the

40 physiological reference study (fig 3).A02 demand Left panels of figure 5 show the effect of a(ml 02/min) start dose of the different study drugs

compared with baseline values. Mean oxygensupply in the control condition was 21

uipply (6) ml/min. There was a small but significant,/min) mean difference in oxygen supply for sodium-mm) nitroprusside in relation to the oxygen supply

based on the dashed line in figure 5 of0-55 ml/min (SE 0-21). For both urapidil andnifedipine the coronary microvascular vasodi-lating potency was not significantly differentfrom zero under these conditions.The results of regression analysis of these

40 data are presented with (table 3) and withoutA02 demand (table 4) intercept. Sodium nitroprusside, ura-(ml 02/min) pidil, or intravenous nifedipine did not change

the ratio between myocardial oxygen supplyand demand significantly, compared with the

upply physiological ratio obtained by pacing during,/min) anaesthesia (table 3). There was no significant* increase in coronary oxygen supply, without

change in myocardial oxygen demand, for anyof the drugs studied at start dose; therefore, sono significant intercept was found (table 4).

Right panels of figure 5 show the effect ofurapidil, sodium nitroprusside, and nifedipine

40 on myocardial oxygen supply and demand atA02 demand higher infusion rates after sternotomy and(ml 02/min) opening of the pericardium. Infusion rates were

adjusted to keep mean arterial pressurebetween 120% and 80% of baseline values.

ny The mean (SD) coronary microvascular vasodi-n nitroprusside, lating potency was calculated and was signifi-pply and demand cant for sodium nitroprusside (1-78 (2-75) ml/nutes after start of min) and nifedipine (2-45 (5-12) ml/min) butvhile drugs are nofruapdlid 120% of not for urapidil.e of the Table 4 shows that regression analysis of thensganaesthesia effect of sodium nitroprusside and nifedipine

aesults of the revealed a small but significant intercept, indi-3 and 4. cating an increase in coronary oxygen supply

independent of myocardial oxygen demand.2). There was However, for nifedipine only, the oxygen sup-:he regression ply:demand ratio was significantly decreased,rate and dur- using regression analysis with intercept estima-ntly different tion. All effects of nifedipine (significant coro-scular vasodi- nary microvascular vasodilating potency,as not signifi- significant intercept with the oxygen supply

axis, and decreased supply:demand ratio)

Table 4 Regression analysis: AO2 supply = ratio * AO2 demand + intercept

95% CI 95% CI ofIntervention n Ratio of ratio Intercept intercept r2

Anaesthesia, sinus rhythm 12 1-53 1-341-73 0-64 -0-16-1-44 0.94Anaesthesia, pacing 12 1-36 1 17-1-54 007 -1-17-1-31 097SNP, start dose (iv) 45 1-33 1-18-1-49 045 -0-01-089 0-88SNP, sternotomy (iv) 45 1-40 1 19-1 60 1-83* 1 01-2 65 0-81Urapidil, start dose (iv) 27 1-53 1-33-1-74 0-25 -0 45-0{96 0 90Urapidil, sternotomy (iv) 27 1-34 1-201-49 1-31 -0-03-2-46 0 94Nifedipine, start dose (iv) 27 1-41 1-07-1-76 0-28 -0-85-1-41 0 74Nifedipine, stemotomy (iv) 27 1 01f 0-70-1-32 5-73* 2-75-8-7 0-64

All ratios presented were significantly different from zero. All intercepts were on the myocardial oxygen supply axis in mlO,/min.95% CI, 95% confidence interval; r', regression coefficient; SNP, sodium nitroprusside.*Significantly different from zero; tsignificantly different from regression without intercept; tsignificantly different from thereference regression obtained under similar conditions.

A40

-20

-20 _

B

/

-20 f

-20

C Anr_ A02 supply(ml 02/min)

/e

AO, de

-20

-20

Start dose

122

4u

on Decem



/H


ownloaded from



A02 SUpply20 (ml 02/min)

0* /

20 -20


Felodipine NifedipineFigure 6 Effect of intracoronary bolus infusion offelodipine or nifedipine on myocardial oxygen supply and demand inawake patients with coronary artery disease one minute after bolus administration. Left panel, effect offelodipine bolusinjection of 0 1 mg. Right panel, effect of nifedipine bolus injection of 0-2 mg. The solid lines represent the regression line tothe physiological reference data obtained in awake patients and presented in figure 3 and was used to calculate thecoronary microvascular vasodilating potency of these drugs under these conditions (see table 3).

resulted from an increase in myocardial oxygensupply larger than expected from the physiolog-ical supply:demand ratio (dashed line) in onlysix of 27 patients.

Figure 6 shows the changes in coronaryoxygen supply induced by intracoronary bolusinjection of felodipine or nifedipine in awakepatients after cardiac catheterisation for diag-nostic purposes in relation to the concomi-tantly induced changes in myocardial oxygendemand. The solid lines represent the physio-logical supply:demand ratio obtained by pac-ing in the awake patients in the physiologicalreference study and was used as reference forthe calculation of the coronary microvascularvasodilating potency of these drugs underthese conditions.

Regression analysis of the changes in oxygensupply to the induced changes in oxygendemand did not yield a significant correlationwith or without intercept for either intracoro-nary felodipine or intracoronary nifedipine.

For nifedipine nine of 10 patients, and forfelodipine all 10 patients showed an increasein oxygen supply compared with the expectedphysiological values (solid line). The coronarymicrovascular vasodilating potency was 13-85(SE 1 92) ml/min for felodipine and 1057(1 89) ml/min for nifedipine. Both coronarymicrovascular vasodilating potencies were sig-nificantly different from zero and indicate sub-stantial changes in oxygen supply. (Mean(SD) baseline oxygen supply was 22-7(4 3) ml/min for intracoronary nifedipine and21-7 (6 3) for intracoronary felodipine, bothwere not significantly different from eachother or from the baseline condition in theawake patients of the reference study.)

DiscussionBy establishing the linear relation betweenmyocardial oxygen supply and demand bypacing, the effect of putative vasodilatingdrugs can be divided between those in which

this relation is unaffected (urapidil and sodiumnitroprusside) and those which increase theoxygen supply more than expected (felodipineand intracoronary nifedipine). We interpretedthe latter effect as a true opening of the resis-tance vessels of the coronary circulation. Theformer drugs may dilate the proximal epicar-dial vessels or the coronary veins, but not theresistance vessels that determine coronaryblood flow and thus oxygen supply. Any effectof these drugs on oxygen supply results fromthe concomitant effect on metabolic rate-thatis, myocardial oxygen consumption or oxygendemand.The aim of the present study was to quan-

tify the physiological response of the coronaryflow regulation process to pacing inducedchanges in myocardial oxygen demand. Thesedata were used to describe the effect of vasoac-tive compounds on the coronary circulation inrelation to this physiological regulationprocess, as these drugs may have separateeffects on MVO,. As the effects of these coro-nary vasoactive drugs were previously studiedin awake and anaesthetised patients, we had todetermine the possible effect of anaesthesia onthe oxygen supply:demand ratio.The results indicate that the ratio between

pacing induced changes in myocardial oxygensupply and demand was 1-50 (0-1) in awakepatients with CAD and 1-45 (0 1) after induc-tion of anaesthesia by fentanyl/pancuroniumbromide; therefore, this type of anaesthesiadoes not seem to affect the physiological coro-nary metabolic regulation at steady state.

Subsequently it was shown that vasoactivityof drugs was judged differently compared withnormal physiological regulation of coronaryblood flow. Intravenous administration ofsodium nitroprusside, urapidil, and nifedipineshowed rather similar behaviour, comparedwith the regression line obtained during pac-ing (fig 5). In the majority of patients studied,the induced changes in myocardial oxygensupply were related to changes in myocardial

0

-20 20

A02 demand(ml O2/min)

123

-20 _ -20

on Decem



/H


ownloaded from



4 A02 SUPPiY40 F (ml 02/min)Type 1

a

-20

-20 L

Type 2

*0

*0//. .

40 -20 /

/A02 demand(ml 02/min)

4 A02 supply40 (ml 02/min)

Type 3

0

-20 0 /

0/

-20

40 r

Type 5

-20 A

/--20

00

0* ///

AO,, dF

A02 supply40 r (ml 02/min)

/

-20 '

407

Type 4

40 -20

emand/(lx2 u0 mIO' lu(ml 02Jmin)

A02 supply(ml 02/min)

00

09

40

Type 6

40 -20

A02 demand I(ml 02/min) -20

A02 SUpI(ml 02/rr

00.x/ $

-20 _

A02 SUP(ml 02/rn

0

/A

Figure 7 ilustration of the potential of this new analysis. Theoretical vasc(@) are presented in relation to physiologicalflow regulation (dotted line inAll panels show an average increase in myocardial oxygen supply as resultinfusion; however, with the present analysis not all drugs wil be deemed cormicrovascular vasodilating drugs. Type 1: classic coronary vasodilator affeccoronary oxygen supply and not myocardial oxygen demand. The mean chzsupply wiUl be equal to the mean vasodilating potency of this type of drug. 1vasodilator underestimated by its change in oxygen supply alone. The inducoxygen demand enhances the vasodilating potency of this type of drug. Typ,vasodilator characterised by a consistent moderate increase in coronary oxygabove the expected physiological supply. Coronary microvascular vasodilatiintercept of regression line are equally effective in quantifying its action. Theoxygen supply and demand is unchanged. Type 4: this coronary vasodilatoithe same result in each patient. The intercept of the regression line indicatespossibilities at zero change in oxygen demand, but the ratio decreases compcphysiological line. The coronary microvascular vasodilating potency wiUl re)change in myocardial oxygen supply at the actual changes in myocardial onType 5: this type of vasodilator shows an increased oxygen supply relative tphysiologicaly expected supply if myocardial oxygen demand is increased. ,

oxygen demand a larger decrease in supply is expected as well. The intercepregression line would be equal to zero, but the ratio wiU increase. The coronimicrovascular vasodilating potency may be significantly greater than zero cthe change in MVO. Type 6: this drug is not a true vasodilator. The incremyocardial oxygen supply can be fully explained by the physiological coronzregulation. 7he range ofoxygen supply and demand changes shown here isthan induced by pacing in the reference study.

felodipine in the coronary vascular wall andthe myocardium. In addition, the intracoro-nary route of bolus administration is not asso-ciated with baroreflex mediated sympatheticactivation, as is the case with intravenousadministration of vasodilators.

PHYSIOLOGY OF CORONARY FLOW REGULATION40 Coronary flow is determined by myocardial

A02 demand oxygen consumption and coronary perfusion(ml 02/min) pressure.920 These parameters may vary con-

tinuously over time; therefore, coronary vascu-lar resistance is adjusted in a continuous

ply dynamic process. This process is mediated bynin) several mechanisms including myogenic,

metabolic, neurohumoral, and endothelialresponses. The relative influence of each sepa-rate response on the overall regulation processis unknown.'021

It is well documented that myocardial oxy-gen supply and demand are tightly coupled so

40 that each change in demand is matched by acorresponding change in supply.79 This phe-

(ml d0eman) nomenon is the result of physiological coro-(ml 02/mln) nary flow regulation and, in the case of drug

administration, this should not be referred toas drug induced vasodilatation or vasocon-

ply) striction. We propose to reserve the term "true,ii,) vasodilatation" for increases in coronary blood

flow above the physiological regulation level.This implies that the physiological coronaryflow regulation should be known and used as areference when potential vasodilating agentsare studied.

In humans, this physiological regulation40 process has not been studied extensively. In

A02 demand 1972, Kitamura et al measured coronary blood(ml 02/min) flow and MV02 in awake humans (young,

male volunteers) during rest and at several lev-vdilator effects els of exercise.22 From the average data thei each panel). myocardial supply:demand ratio was esti-ronauy mated to be approximately 1-4. In 1978cts only Mohrman and Feigl used myocardialange in oxygen supply:demand ratios to characterise the effected reduction in of a adrenoreceptor blockade on coronarye 3: coronary flow regulation in dogs.8 They found a base-gen supply line ratio of 0 85 and vasodilatation caused byngpotency and a adrenoreceptor blockade increased this ratioeratio ofrdoes not yield to 1-23. In 1987, Vergroesen et al reported avasodilating myocardial oxygen supply:demand ratio of 1-4ared with the (SE 0 2) in dogs and 1-5 (0'2) in goats withqect the average (E02tygen demand. and without a separate coronary perfusion sys-p the tem.9 The discrepancy between the resultst a decreased reported by Mohrman and Feigl, and

ary Vergroesen et al's study is possibly because oflepending on differences in the stimulus used to increasease in MVO2. However, the data of Vergroesen et alary flowonly larger and Kitamura et al are in good agreement with

the present findings in patients with CAD (figs3 and 4, table 3).

oxygen demand. Only a small additionalincrease in coronary oxygen supply could beattributed to the administration of nifedipineand sodium nitroprusside, but not urapidil. Incontrast, intracoronary administration ofnifedipine or felodipine showed markedvasodilatation in spite of the decrease inMVO2. This was probably caused by the veryhigh local concentration of nifedipine and

USE OF AO2 SUPPLY:AO2 DEMAND RATIO TOEVALUATE VASODILATING PROPERTIES OFDRUGSA large number of peripheral and coronaryarterial vasodilators are currently available forthe management of hypertension andischaemic heart disease. As in most studiesdescribing coronary phannacodynamic effectsonly mean changes in coronary blood flow and

124 on D



j.com/


ugust 1997. Dow

nloaded from



MVO, are reported (as was done previouslywith the present data,""13 '5), the relativevasodilating potency of these different agentsis not well defined. A number of parametersthat can be used to differentiate between theeffects of coronary vasodilators are shown intable 3. These include: the ratio of myocardialoxygen supply and demand at different infu-sion rates; the intercept with the Y axis (supplyaxis); and intrinsic coronary microvascularvasodilating potency, defined as the mean dif-ference between the measured increase in oxy-gen supply and the expected increase inoxygen supply calculated from the referencesupply:demand ratio.

Figure 7 shows the theoretical effects ofputative coronary vasodilators on myocardialoxygen supply in relation to the physiologicalreference supply:demand ratio as measured inthe present study. Six different types of drugsare shown that would have been deemed coro-nary vasodilators according to current clinicalpractice, but with this new method more dif-ferentiation is made.Comparing theoretical effects of potential

coronary vasodilators (fig 7) with the data pre-sented in figures 5 and 6, we see that the rela-tion between changes in oxygen supply anddemand calculated on the basis of the physio-logical reference study in anaesthetisedpatients (dashed lines in figure 5) predictedthe expected changes in coronary oxygen sup-ply to a very large extent for urapidil andnifedipine at low starting dose (type 6 vasodi-latation, fig 7). Only a few patients treatedwith high dose intravenous nifedipine had sig-nificant vasodilatation; therefore, nifedipine isa type 4 vasodilator as shown from the regres-sion line fit with intercept estimation pre-sented in table 4. Sodium nitroprussideappeared to induce a small but consistentupwards shift of the oxygen supply:demandratio, apparent in the small but significantcoronary microvascular vasodilating potency(type 3 vasodilator). The fact that the coro-nary microvascular vasodilating potency forsodium nitroprusside was significantly differ-ent from zero shows that this new method ofquantification is highly sensitive. Clinical sig-nificance of this small but consistent improve-ment has yet to be shown. However, themajority of patients treated with sodium nitro-prusside showed very little change in eitheroxygen supply or oxygen demand. Measure-ment of the vasodilating properties of thesedrugs by the percentage increases in oxygensupply alone would have overestimated thevasodilating properties of urapidil in bothawake and anaesthetised patients, and nifedip-ine at high dose intravenous infusion.

Figure 6 shows the strength of this analysis.The increase in oxygen supply induced byintracoronary administration of nifedipine andfelodipine relative to the reference curve ismuch larger then the change in oxygen supplyper se (type 2 vasodilatation, fig 7). Thereduction in MVO, following intracoronarybolus injection of high dose felodipine ornifedipine probably results from the potentnegative inotropic properties of these agents,

separate from the relaxing effect on the coro-nary smooth muscle cells. Using physiologicaloxygen supply:demand ratios as a reference,intracoronary administration of nifedipineclearly has more local effects on the coronaryresistance vessels than intravenous administra-tion of the same agent (figs 5 and 6).

CONTRIBUTION OF COLLATERAL FLOW TOOXYGEN SUPPLY:DEMAND RATIOIn the prospectively studied group of patientsthe contribution of potential collaterals(including those that cannot be seen by coro-nary angiography) to coronary sinus bloodflow was unknown. However, theoreticallythere is no reason to assume a different oxygenextraction ratio (and thus supply:demandratio) in myocardium perfused via collateralsunless these collateral vessels induce signifi-cant arteriovenous shunting. As coronaryvenous oxygen tention remained constant dur-ing all interventions in the present study, therewas no indication of significant intracoronaryshunting.

CONCLUSIONPhysiological regulation of coronary bloodflow in awake and anaesthetised patients withCAD can be characterised by similar myocar-dial oxygen supply:demand ratios.The myocardial oxygen supply:demand

ratio and coronary microvascular vasodilatingpotency are potential tools for comparingcoronary vasodilating drugs in clinical prac-tice.

Intravenous administration of sodium nitro-prusside, urapidil, and nifedipine in sufficientdoses to control arterial blood pressure duringsurgical stress has limited coronary vasoactiveeffects in the majority of patients with CAD.The physiological coronary flow control sys-tem is still the major determinant of coronaryflow in these patients.

Intracoronary high dose bolus injection ofnifedipine and felodipine results in significanttrue vasodilatation.

The authors are grateful to Marjolein Porsius, research nurse,for her assistance and technical support during the experimentsdescribed in this study.

1 Broten TP, Feigl EO. Role of myocardial oxygen and car-bon dioxide in coronary autoregulation. Am 3 Physiol1992;262:H1231-7.

2 Kuo L, Davis MJ, Chilian WM. Endothelium-dependent,flow-induced dilation of isolated coronary arterioles. Am J7Physiol 1990;259:H1063-70.

3 Feigl EO, Van Winkle DM, Miyashiro JK. Cholinergicvasodilation of coronary resistance vessels in dogs,baboons and goats. Blood Vessels 1990;27:94-105.

4 Meininger GA, Davis MJ. Cellular mechanisms involved inthe vascular myogenic response. Am 37 Physiol 1992;263:H647-59.

5 Drake-Holland AJ, Laird JD, Noble MIM, Spaan JAE,Vergroesen I. Oxygen and coronary vascular resistanceduring autoregulation and metabolic vasodilatation in thedog. 7 Physiol 1984;348:285-99.

6 De Fily DV, Chilian WM. Coronary microcirculation:autoregulation and metabolic control. Basic Res Cardiol1995;90:112-18.

7 Eckenhoff JE, Hafkenschiel JH, Landmesser CM, HarmelM. Cardiac oxygen metabolism and control of the coro-nary circulation. AmJPhysiol 1947;149:634-49.

8 Mohrman DE, Feigl EO. Competition between sympa-thetic vasoconstriction and metabolic vasodilation in thecanine coronary circulation. Circulation 1978;42:79-86.

9 Vergroesen I, Noble MIM, Wieringa PA, Spaan JAE.Quantification of °, consumption and arterial pressure

125 on D



j.com/


ugust 1997. Dow

nloaded from



as independent determinants of coronary flow. Am JfPhysiol 1987;252:H545-53.

10 Moffitt EA, Sethna DH. The coronary circulation andmyocardial oxygenation in coronary artery disease: effectsof anesthesia. Anesth Analg 1986;65:395-410.

11 Harrison DG, Kurz MA, Quillen JE, Sellke FW, Mugge A.Normal and pathophysiologic considerations of endothe-lial regulation of vascular tone and their relevance tonitrate therapy. Am J7 Cardiol 1992;70: 1 lB-17B.

12 Van Wezel HB, Bovill JG, Koolen JJ, Barendse GAM,Fiolet JWT, Dijkhuis JP. Myocardial metabolism andcoronary sinus blood flow during coronary artery surgery:effects of nitroprusside and nifedipine. Am Heart 1987;113:266-73.

13 Van Wezel HB, Bovill JG, Visser CA, Koolen JJ, Janse MJ,Meijne NG, et al. Myocardial metabolism, catecholaminebalance, and left ventricular function during coronaryartery surgery: effects of nitroprusside and nifedipine.Cardiothorac Anesthesiol 1987;1:408-17.

14 Van Der Stroom JG, Van Wezel HB, Vergroesen I, Kal JE,Koolen JJ, Dijkhuis JP, et al. Comparison of the effects ofurapidil and sodium nitroprusside on haemodynamicstate, myocardial metabolism and function in patientsduring coronary artery surgery. Br Anaesth 1996;76:645-51.

15 Koolen JJ, Van Wezel HB, Piek Jn, van Liebergen R, SwaanA, Visser CA. Effects of intracoronary felodipine versus

nifedipine on left ventricular contractility and coronarysinus blood flow in stable angina pectoris. Am Cardiol1994;74:730-2.

16 Van Wezel HB, Bovill JG, Koolen JJ, Patrick MR, FioletJWT, Van Der Stroom JG. Influence of glyceryl trinitrateand nifedipine on coronary sinus blood flow and globalmyocardial metabolism during coronary artery operation.Br HeartJ 1986;56:272-7.

17 Mathey DG, Chatterjee K, Tyberg JV, Lekven J, BrundageB, Parmley WW. Coronary sinus reflux. A source of errorin the measurement of thermodilution coronary sinusflow. Circulation 1978;57:778-86.

18 Koberstein RC, Pittman DE, Klocke lJ. Right atrialadmixture in coronary venous blood. Am Jf Physiol 1969;216:513-35.

19 Ganz W, Tamura K, Marcus HS, Donoso R, Yoshida S,Swan HJC. Measurement of coronary sinus blood flowby continuous thermodilution in man. Circulation 1971;44:181-95.

20 Feigl EO. Coronary physiology. Physiol Rev 1983;63:1-205.21 Jones CJH, Kuo L, Davis MJ, Chilian WM. Myogenic and

flow-dependent control mechanisms in the coronarymicrocirculation. Basic Res Cardiol 1993;88:2-10.

22 Kitamura K, Jorgensen CR, Gobel FL, Taylor HL, WangY. Hemodynamic correlates of myocardial oxygen con-sumption during upright exercise. Appl Physiol 1972;32:516-22.

126 on D



j.com/


ugust 1997. Dow

nloaded from


myocardial reference vasodilatory effects humans · haemoglobin, c02,.,, is arterial oxygen...

Documents