Hemodynamic and metabolic correlates of dipyridamole-induced myocardial thallium-201 perfusion abnormalities in multivessel coronary artery disease

Hemodynamic and Metabolic Correlates of Dipyridamolelnduced Myocardial Thallium-201 Perfusion Abnormalities in Multivessel Coronary Artery Disease David P. McLaughlin, MD, George A. Beller, MD, Joel Linden, PhD, Carlos R. Ayers, MD, Marcia L. Ripley, Heidi Taylor, Denny D. Watson, PhD, and Marc D. Feldman, MD The mechanisms responsible for the develop merit of reversible thallium201 (V-201) defects with dipyridamole stress in patients with coronary artery disease (CAD) is not well understood. Previous experimental animal studies have demonstrated coronary steal characterized by an absolute m in subendocardial flow distal to a stenosis in response to dipyridamole incUsion. Accordingly, the purpose of this study was to determine if reversible D-201 defects in response to dipyridamole infusion are reflective of myocardial ischemia or secondary to regional differences in flow reserve. Dipyridamole (0.55 mg/kg) D-201 imng was performed in 23 patients in whom serial electrocardiographic, hemodynamic, aortic and coronary sinus lactate, and comnary sinus adenosine measurements were obtained. All patients with CAD had D-201 redistribution (3.8 -c 2.0 defects/patient), and all ps tients without CAD had normal scans. Mean aor tic pressure was similar in both groups and did not change in response to dipyridamole (nouCAD 103 f 11 vs CAD 99 + 15 mm Hg, p = NS). Pulmonary capillary wedge pressure was similar at baseline (nonCAD 11 +- 4 vs CAD 13 f 5 mm Hg, p = NS) and did not change in response to the drug (nor&AD 14 f 3 vs CAD 15 f 7 mm Hg, p q NS). Lactate extraction fraction was similar at baseline (non-CAD 0.22 f 0.09 vs CAD 0.17 f 0.14, p = NS) and decreased similarly in both groups (no&GAD 0.08 f 0.05 vs CAD 0.05 f 0.12, p q NS). Coronary sinus adenosine concentration was si@Micantly higher at baseline in patients with CAD (nouCAD 15 -C 4 vs CAD 35 + 13 rig/ml, p = 0.034), presumably to maintain resting coronary blood flow, and increased significantly with a comparable percent increase in From the Cardiovascular Division, Department of Medicint, University of Virginia Health Sciences Center, Charlottesville, Virginia. This study was supported in part by American Heart Association Virginia Affiliate Grant-in-Aid G880010; the Bayer Fund for Cardiovascular Research, Sterling Drug, New York; and a FIRST Award R29HL47046 01 from the National Heart, Lung, and Blood Institute, Bethesda, Maryland. Manuscript received July 8, 1993; revised manuscript received December 3, 1993, and accepted December 5. Address for reprints: Marc D. Feldman, MD, Director, Cardiac Catheterization Laboratory, University of Pittsburgh Medical Center, Room 392, 200 Lothrop Street, Pittsburgh, Pennsylvania 15213.

both groups after dipyridamole (nouCAD 35 + 10 vs CAD 59 + 35 rig/ml, p q 0.0001). In this group of patients with multivessel CAD, dlpyritk amole-induced D-201 perfusion abnounalnies were not associated with significant myocardiil ischemia by hemodynamic and metabolic criteria and appeared to be more indicative of abnormal flow reserve. (Am J Cardiol1994;74:1159-1194)

P

harmacologic stress perfusion imaging has increasingly been used as an alternative to exercise imaging for the detection of physiologically significant coronary artery disease (CAD) and for prognostication.1-3The rationale for vasodilator stressimaging is basedon the concept of coronary flow reserve.On dilating the peripheral coronary bed with dipyridamole, coronary flow can increase by as much as fourfold.4 When increasing blood flow with a vasodilator, however, an impairment in flow reserve in a stenotic bed is seen by failure to augment flow comparedwith hyperemia in the normally perfused bed. This results in a relative disparity of regional myocardial perfusion between normal and stenotic beds, which can be detected in patients by thallium-201 (Tl-201) scintigraphy. Intravenous dipyridamole can produce a significant endocardial-to-epicardial flow gradient distal to a critical coronary stenosisP7 This transmural “coronary steal” may cause subendocardial ischemia, rather than simply abnormal flow reserve. Intravenous dipyridamole can also cause a significant decreasein mean arterial pressurein the setting of a critical coronary stenosis.5JjThis decreasein perfusion pressuremay also cause“true” myocardial ischemia. In these experimental studies,4,6,7anesthetizedanimal models with resting tachycardia were most often used,and dipyridamole causeda substantialreduction in arterial blood pressure.The objective of this study was to determine whether dipyridamole-induced Tl-201 perfusion abnormalities in patients with CAD are predominantly secondaryto regional differencesin flow reserve, or are due primarily to myocardial ischemia. If induction of myocardial ischemia plays a significant role in producing Tl-201 redistribution defects after dipyridamole infusion, we hypothesized that hemodynamic and metabolic abnormalities indicative of ischemia should be manifest.

CORRELATESOF DIPYRIDAMOLE THALLIUM-201 DEFECTS 1159

METHODS Patient group: Twenty-three patients were enrolled

in the study; 17 had CAD, and 6 had angiographic normal coronary arteries(non-CAD). All patients with CAD had typical angina, and all non-CAD patients had atypical chest pain syndromes.Each patient gave written informed consent, and the protocol was approved by the Human Investigation Committee of the University of Virginia. Protocol: All medicationswere discontinued24 hours before cardiac catheterization.Thirty minutes passedbetween the routine diagnostic catheterization and the researchprotocol to minimize any hemodynamic effectsof the contrast agent used. A dual-lumen adenosinecatheter (8Fr) was passedfrom the right internal jugular vein to the coronary sinus, with the tip positioned at least 3.0 cm into the sinus. Details of this catheterand solution to arrest adenosinemetabolism at the cathetertip have previously been described.8A 7Fr Swan-Ganzcatheter was positioned in the pulmonary artery. A 7Fr Judkins right catheter was positioned in the ascendingaorta. Baseline hemodynamic measurementswere obtained, including heart rate, mean central aortic pressure,and mean pulmonary capillary wedge pressure. Baseline metabolic measurementsincluded coronary sinus adenosineas well as aortic and coronary sinus lactate. Other baseline measurementsincluded a 1Zlead electrocardiogram. Dipyridamole was infused intravenously at the approved dose (0.56 mg/kg) over 4 minutes.9 At 7 minutes, all baseline measurementswere repeated, and 2.0 mCi of Tl-201 was administered intravenously. Angiography and ventriculography: Coronary angiograms were interpreted by 2 attending cardiologists who had no knowledge of patients’ clinical data. The number, location, and percent diameter narrowing (by calipers) were recorded for each patient. Left ventricular ejection fraction was calculated from the right anterior oblique ventriculogram using the technique described by Sandler and Dodge.loThe number of regional wall motion abnormalities were obtained using an 8-segment model by interpretation of right anterior oblique and left anterior oblique ventriculograms.l1 Coronary sinus adenosine measurement: In this study,coronary sinus adenosinewas used as an index of interstitial adenosine concentration. There are 2 possible sourcesof coronary sinus adenosine: (1) interstitial adenosine,which is releasedlocally in responseto a decrease in the oxygen supply/demand ratio12*‘3;and (2) an increase secondary to dipyridamole which exerts its physiologic effects by blocking cellular uptake and degradation of adenosine, thereby increasing the halflife of the nucleoside.13v14 We have previously demonstratedconsistent ischemia-induced increasesin coronary sinus adenosinein patients with CAD during atria1pacing stress.*The adenosine catheter used in this study allows 1:1 mixing of blood and stop solution at the catheter tip. The stop solution used to arrest adenosine metabolism has undergone extensive development and validation.8,*5The stop solution includes the 5’ nucleotidaseinhibitors ethylenediaminetetraacetic acid (10 n&I), ethylene glycol-bis (beta-aminoethyl ether)-N,N’-tetraacetic acid (10 mM), 1160

THE AMERICANJOURNALOF CARDIOLOGY VOLUME73

and AOPCP (0.1 n&I); it also includes a nonspecific phosphatase inhibitor, DL-glycerophosphate (10 mM, IND #35,253), erythro-9(2- hydroxyl-3-nonyl)-adenine (5 p,M), an adenosinedeaminaseinhibitor, and dipyridamole, an inhibitor of adenosineuptake. Immediately after collection, the coronary sinus blood sample was centrifuged at 4,800 g for 2 minutes. Each supematantwas filtered through a 0.2 p filter (Acrodisc, Gelman Sciences,Ann Arbor, Michigan). The filtrate was partially purified and concentrated as previously described8by both boronate affinity chromatography (AffiGel 601, Bio-Rad Laboratories, Richmond, California) andreverse-phasechromatography(C l8Sep-Pak,Waters, Milford, Massachusetts).The amount of adenosine in each samplewas determinedby high-performance liquid chromatography as previously described.8 Lact&e measurement and analysis: Blood lactate levels were determinedby a modification of the Marbach and Weil method, which employs the oxidation of lactate to pyruvate (du Pont automatic clinical analyzer,E.I. du Pont de Nemours and Co., Wilmington, Delaware). Lactate extraction is detined as aortic lactate minus coronary sinus lactate, divided by aortic lactate. A negative lactate extraction was delined as lactate production. Thallium201 imagmg analysis: PlanarTl-201 scintigrams were assessedusing a previously described computer-assistedapproach for quantitation of regional Tl201uptake and washout.l6After backgroundinterpolative subtraction,peakTl-201 activity was measuredin 11myocardial scan segmentsin 3 views. The initial and delayed imageswere assessedby 2 independentobserversfor the presenceor absenceof defects and the presenceof redistribution, respectively. Differences in interpretation were resolvedby a third interpreter.The lung-to-heart ratio was considered increasedif the p value was 9.51. Electrocardiographic analysis: Diagnostic electrocardiographic changeswere defined as 21 mm of horizontal or downsloping ST-segment depression 280 ms past the J point. Statistical methods: Data were expressedas mean f SD. Mean values were comparedusing a 2-tailed Student’s t test. A p value co.05 was considered significant. Linear regression was calculated using the statistical package,RS- I (BBN Software, Cambridge, Massachusetts). RESULTS Angiographic data: All 17patients in the CAD group had >70% narrowing of the left anterior descending or left circumflex coronary arteries. Ten of these patients had >70% narrowing of both the left anterior descending and left circumflex arteries. Five of the patients with left anterior descending stenoses also had significant right coronary artery stenosis. The mean number of stenotic vesselswith >70% lesions was 2.2 f 0.8 in the CAD group. The mean number of regional wall motion abnormalities was 0.5 _+2.0 in the CAD group and 0 f 0 in the non-CAD group (p = NS). Ejection fraction was comparable in both groups (0.64 f 0.11 vs 0.71 f 0.04, CAD vs non-CAD, p = NS). Hemodynamic data: Hemodynamic results are summarized in Figure 1 and Table I. Heart rate was similar

JUNE 15,1994

TABLE I Hemodynamic and Metabolic Data Baseline

Heart rate (beats/mln) Mean aortlc pressure (mm Hg) Rate-pressure product Pulmonary capillary wedge pressure (mm Hg) Lactate extraction fraction Coronary smus adenosrne (rig/ml) Coronary smus adenosine (% increase)

Dipyridamole

CAD (n = 17)

Non-CAD (n = 6)

CAD (n = 17)

Non-CAD (n = 6)

75 r 12 102 T 14

73k 11 105 2 16

86 2 11* 992 15

104 t 16*t 103 t 11

11.2 f 2.4

12.0 rt 2.2

11.1 2 2.6

13.9 2 2.5t

13 + 5

11 -t4

15 2 7

14 2 3

0.05 r 0.12

0.08 A 0.06

0.17 2 0.14

0.22 + 0.09

35 2 13t

16 + 4

69 + 35*t

35 + 10*

45 k 4

68k

15

*p < 0.05 basellneversus dipyrldamole; tp < 0.05 non-CAD versus CAD. CAD = coronary artery disease.

at baseline in CAD and non-CAD groups (p = NS) and increased significantly in both groups in response to dipyridamole infusion. Heart rate increase was significantly greater in the non-CAD (104 f 16) than CAD (86 f 11) patients (p = 0.005). Aortic mean pressure was comparableat baseline in both groups and did not change significantly in responseto dipyridamole infusion (nonCAD 103 f 11vs CAD 99 + 15 mm Hg, p = NS). The rate-pressureproduct in the 2 groups was comparable at baseline, but increasedto a greater extent in responseto dipyridamole infusion in the non-CAD group. Pulmonary capillary wedge pressurewas similar at baseline between non-CAD (11 + 4) and CAD (13 + 5 mm Hg, p = NS) groups and did not change in response to dipyridamole infusion (non-CAD 14 f 3 vs CAD 15 + 7 mm Hg, p = NS). The same statistical comparisons between the 10 CAD patients with >70% stenosesof both the left anterior descendingand left circumflex coronary arteries and the non-CAD patients yielded identical results. Metabolic deta: The metabolic data are summarized in Figure 2. Lactate extraction fraction was not significantly different at baseline in the 2 groups. The decrease

0

Non-CAD h6)

in lactate was not different in the non-CAD (0.076 f 0.064) versus the CAD (0.049 + 0.124) group (p = NS). Coronary sinus adenosine concentration was significantly higher at baseline in patients with (35 f 13r&ml) than without (16 f 4 rig/ml, p = 0.004) CAD (Figures 3 and 4). Adenosine concentration was greaterin the CAD (69 f 35 rig/ml) than the non-CAD (35 f 10) group in responseto dipyridamole infusion (p = 0.034), and increased significantly (p = 0.0001) in both groups after drug administration (Figure 3). However, the percent increasein coronary sinus adenosine concentration in responseto dipyridamole infusion was comparablein both groups (p = NS, Table I). The same statistical comparisons between the nonCAD patients and the 10 patients with >70% stenoses of both the left anterior descending and left circumflex coronary arteries yielded identical results. We found no correlation between the increase in coronary sinus adenosineconcentration and the following variables for both patients with and without CAD: number of diseased vesselson angiography (r = .37, p = NS), change in pulmonary capillary wedge pressure (r = .27, p = NS), or number of defects with redistribution (r = .06, p = NS).

0

CAD (n- I71

Non-CAD III4

CAD

(n.17)

Non-CAD (II=61

CAD (n=17)

FIGURE 1. The hsmodynamic effects of intravenous dipyridamole. Heart rate, aortic mean, and pulmonary capillary wedge pressures are shown on the left l~affof each panel for patients without coronary artery disease (no&AD) (n q 6) and the rigm ha/f for those with coronary artery disease (CAD) (n = 17). Baseline values are imlicated by black bars, and dipyridamole values by hatched bars. Data are expressed as mean f SD. *p SO.05 baseline versus dipyridamole. PCWP = pulmonary capillary wedge pressure.

CORRELATESOF DIPYRIDAMOLE THALLIUM-201 DEFECTS 1161

Thallium201

and electrocanliographic

data: The

mean number of Tl-201 defects/patientwas significantly greater in the group with than without CAD (4.0 + 1.9 vs 0.0, p = 0.0001).The number of myocardial segments showing delayed redistribution was also greater in the CAD group (3.8 + 2.0 vs 0.0, p = 0.0001). Eleven of the 17 patients with CAD had Tl-201 defects in the left anterior descending distribution. Four of these 11 patients also had defects in the distribution of the right coronary artery. The remaining 6 had defects in the left circumflex distribution. All 17 showed delayed redistribution. The lung-to-heart Tl-201 ratio after dipyridamole was comparable in both groups (0.53 + 0.11 vs 0.43 It 0.15, CAD vs non-CAD, p = 0.12). Five patients in the CAD group had diagnostic ST-segment depression. No patient in the non-CAD group had ischemic ST-segment changes. DISCUSSION Our results suggestlittle metabolic or hemodynamic evidence of significant ischemia in this cohort of CAD patientswith predominantly multivessel CAD despitedevelopment of Tl-201 redistribution defectsafter dipyridamole infusion. Compared with non-CAD patients, patients with CAD had no greater increase in pulmonary capillary wedge pressure,a hemodynamic marker of ischemia, during dipyridamole infusion. Similarly, patients with CAD had no lactate production. Although there was a decreasein the lactate extraction fraction after dipyridamole, the magnitude of the decreasewas comparable in patients with and without CAD. Further comparison of only the 10CAD patients with >70% stenosesof both the left anterior descendingand left circumflex coronary arteries with the non-CAD group did not change these

H Baseline H Dipyrldamole

results. Because the coronary sinus catheter was sampling venous return from both of theseterritories, our results in the subset of patients with diseaseof both major branchesof the left coronary systemwere less likely to be due to mixture of lactate from stenotic and normal coronary territories, as would be observed in patients with 1-vessel disease. Local myocardial ischemia is associatedwith release which is a sensitive indiof endogenousadenosine,12,13 cator of ischemia. In this study, an increase in coronary sinus adenosinein both groups after dipyridamole infusion was expected, in keeping with the known effect of dipyridamole, which inhibits cellular uptake of adenosine.14If significant ischemia had been induced by a transmural coronary steal mechanism by dipyridamole, a further increasein adenosinein the CAD group should have been demonstrated.An additional increasein coronary sinus adenosineover and above that anticipated as a result of the dipyridamole effect alone, was not observed. The percent increase in coronary sinus adenosine concentration was the samein CAD and non-CAD groups. The percent increase in coronary sinus adenosine concentration was also the samein the 10CAD patients with >70% stenosesinvolving both the left anterior descendingand left circumflex coronary arteries and the non-CAD group, arguing that dilution of coronary sinus blood by myocardium not jeopardized by an epicardial coronary stenosis should not be a factor. The increasein adenosineconcentration might have been anticipated to correlate with the number of stress-induced perfusion abnormalities. This also was not observed. Thus, thesedata suggestthat Tl-201 defectswith delayed redistribution, a scintigraphic marker of reversible re-

loo* W Baseline 0” 0 Dipyridamole 3

* t

-r

$60 .i g 40 I 20

0

Non-CAD

Non-CAD (n=6) FIGURE 2. The effect of intravenous dipyridamole on lactate extraction fraction. Patients without coronary artery disease (non-CAD) (n q 6) are shown on the left M/f of the panel and those with coronary artery die ease (CAD) (n = 17) on the right ha/f, Baseline values are indicated by Mack bars, and dipyridamole values by hatched biws. Data are expressed as mean f BD. *p 50.05 baseline versus dipyridamole.

II62

THE AMERICANJOURNALOF CARDIOLOGY VOLUME73

(n=6) FIGURE 3. The effect of itivenous dipyridamole on cofmary sinus adenosine concentration. Patients witI+ out coronary artery disease (no&AD) (n = 6) are shown on the left ha/f of the panel and those with coronary artery disease (CAD) (n = 17) on the tight AaM Baseline values are indicated by black bars, and dipyridamde values by hatched hers Data are expressed as mean f SD. *p SO.05 baseline versus dlpyrklamole; tp 10.05 no* CAD versus CAD.

JUNE 15,1994

gional hypoperfusion, were not accompaniedby significant myocardial ischemia in this patient group. A major difference between our study and data from published experimental reports on significant endocardial-to-epicardial steal in myocardium distal to a coronary stenosis, was the absence of a decreasein aortic perfusion pressure in these patients during infusion of dipyridamole. During intravenous infusion of dipyridamole in the experimental laboratory, we5 and others6 have shown a consistent reduction in aortic mean pressure and, hence, coronary perfusion pressure. Interestingly, the stenosis gradient was unchanged in these experimental studies, implying that the major contribution to transmural ischemia was the decreasein mean arterial pressurein responseto the vasodilator. Furthermore, in 1 experiment, dipyridamole was administered directly through the intracoronary route with no consequentreduction in mean arterial pressure; only a slight decrease in subendocardial flow was seenconsequentto drug infusion.i7 The open-chest dog model is, by nature, an ischemic preparation. The dogs are tachycardiac due to the pentobarbital anesthesia,and therefore much more likely to develop subendocardialischemia in responseto intravenous vasodilators. There are great similarities between the current study and another patient study performed by Feldman et al.‘s These investigators demonstratedischemia (defined as a decreasefrom baseline in coronary blood flow) in only 2 of 12regions of the myocardium supplied by a stenotic epicardial coronary artery in response to intravenous dipyridamole during the same time frame examined in the current study. The majority of myocardial regions supplied by stenotic epicardial vessels did not show a decreasein coronary blood flow but only an absenceof an increase. Similar to the current study, there was no change in left ventricular end-diastolic pressure. There was regional lactate production in 5 of 12myocardial regions supplied by stenotic epicardial coronary arteries; 3 patients had lactate production consequentto dipyridamole infusion in the present study. Perhapsany minor differencesbetween the 2 studies reflect a significant decreasein mean aortic pressure in the study of Feldman et al ‘* but none in the current study. This decreasein perfusion pressuremay explain the 2 myocardial regions with ischemia documented by a decrease in coronary blood flow. With high-dose dipyridamole (0.84 mg/kg), regional wall motion abnormalities perhapsreflective of ischemia can be observed. This is often associatedwith a significant decreasein systemic arterial pressure.i9-**This reduction in systemic pressuremay be the mechanism by which ischemia-induced wall motion abnormalities are seen. Zoghbi and co-workers,23administering intravenous adenosine, showed a high incidence (85%) of new adenosine-induced wall motion abnormalities on echocardiography. Ogilby et a124reported a greater increase in pulmonary capillary wedge pressurein responseto intravenous adenosineinfusion in patients with CAD than in control subjects, which was interpreted as reflecting diastolic dysfunction secondary to vasodilator-induced ischemia. Although adenosine infusion may result in

more severe hemodynamic and echocardiographic abnormalities than dipyridamole, the exact mechanism for this has not yet been entirely determined. For example, the increasein pulmonary capillary wedge pressureconsequentto the vasodilator responseto adenosinemay be the result of the Gregg phenomenon, also known as the “garden hose effect.“25-28This effect would be consistent with hyperemia-induced diastolic dysfunction, with a decreasein compliance and an increase in pulmonary capillary wedge pressure. Ogilby et a124demonstrated that left ventricular ejection fraction did not decreasein response to adenosine infusion, which should occur if ischemia were the causeof the elevation in wedge pressure. In the present study,even when patients with ST-segment depression (n = 5) were separatelyanalyzed, there was no hemodynamic or metabolic evidence of “true” myocardial ischemia. When these 5 patients were compared with the 12 patients without these symptoms or electrocardiographic changes, the hemodynamic and metabolic markers of ischemia were not different. However, these patients may have had very mild ischemia, which might not have been seenby detectableincreases in wedge pressure or lactate production. In the present study, coronary sinus adenosine concentration at baseline was higher in patients with than control subjects without CAD (Figure 4). This finding demonstratesfor the first time that flow reserve in myocardial segmentssupplied by stenotic coronary arteries may be limited not only during provocation with pharmacologic stress,but also to some extent in the resting state.Increasedcoronary sinus adenosinelevels in these patients with CAD suggest that some vasodilatory reserve is used simply to maintain resting coronary blood flow. The current study provides evidence that release of endogenous adenosine may be an intrinsic homeo-

. .

60-

.

; 8 .

10

I

1

o-

Non-CAD (n=6)

CAD (rk17)

FIGURE 4. Individual basal patient coronary sinus adenosine concentrations. Patients without coronary artery disease (non-CAD) (n q 0) are shown on the left half of the panel and those with coronary artery disease (CAD) (n = 17) on the right half. Adenosine corn centration is higher in patients with coronary artery disease (p = 0.004).

CORRELATESOF DIPYRIDAMOLE THALLIUM-201 DEFECTS 1163

J

static mechanism to maintain resting flow distal to a stenotic coronary artery, even in the absenceof clinical symptoms or signs of ischemia. None of the 17 patients with CAD in this study had angina or electrocardiographic changes indicative of ischemia at rest. In addition, only 4 patients exhibited resting wall motion abnormalities, and these were the patients who had prior infarction. Whether the increased basal coronary sinus adenosineconcentration in patients with CAD has a preconditioning effect on the myocardium29*30 cannot be answeredby the current study,and it is an observation that will need conlirmation in larger numbers of patients. Study limitations: We used more indirect metabolic and rather gross hemodynamic markers of ischemia as criteria. It is possible that markers such as an increased pulmonary capillary wedge pressure or lactate production may simply not be sensitive enough to detect mild subendocardial ischemia. Regional systolic thickening was not assessedduring dipyridamole infusion. Serial measurementof coronary sinus adenosine was considered to be a rather sensitive approach for the detection of such mild ischemic responses.Although a two- to threefold increase in coronary sinus adenosine was detected in patients with CAD in this study, this surely is an underestimation of the true increase in tissue adenosine production. The absolute increasein coronary blood flow in responseto dipyridamole infusion was not measured.The concentration of coronary sinus adenosinereported in this study could not, therefore, be normalized for the diluting effect of increasing coronary blood flow. Acknowledgment: We thank the staff in the cardiac catheterization and nuclear laboratories for the technical assistancethat was provided, as well as Jerry Curtis for his help in preparing this manuscript. We also thank the Cordis Corporation for supplying components to build the adenosinecatheter.

1. Leppo J, Boucher CA, Okada RD, Newell JB, Straw W, Pohost GM. Serial thallium-210 myocardial imaging after dipyridamole infusion. Diagnostic utility in detecting coronary stenoses and relationship to regional wall motion. Circularion 1982;66:649d56. 2. Leppo JA, O’Brien I, Rothendler JA, Getchell JD, Lee VW. Dipyrid amole-thallium-201 scintigraphy in the prediction of future cardiac events after acute myocardial infarction. N En@ J Med 1984;310:101&1018, 3. Boucher CA, Brewster DC, Darling RC, Okada RD, Strauss HW, Pohost GM. Determination of cardiac risk by dipyridamole-thallium imaging before peripheral vascular surgery. N En@ J Med 1985;312:384-394. 4. Beller GA, Holzgrefe HH, Watson DD. Effects of dipyridamole-induced vasodilation on myocardial uptake and clearance kinetics of thallium-201. Circu[ution 1983;68:1328-1338. S. Beller GA, Holzgrefe HH, Watson DD. Intrinsic washout rates of thallium-201 in normal and ischemic myocardium after dipytidamole-Induced vasodilation. Circularion 1985:71:37X-3X6. 6. Okada RD, Leppo JA, Boucher CA, Pohost GM. Myocardial kinetics of thallium-201 after dipyridamole infusion in normal canine myocardium and in myo-

1164

THE AMERICANJOURNALOF CARDIOLOGY VOLUME73

cardium distal to a stenosis. J C/in Invest 1982;69: 199-209. 7. Fung AY, Gallagher KP, Buda AJ. The physiologic basis of dobutamine as compared with dipyridamole stress interventions in the assessment of critical cororuy stenosis. Circulalion 1987:76:943-95 I 8. Feldman MD, Ayers CR, Lehman MR, Taylor HE, Gordon VL, Sabia PJ, Ras D, Sk&k TC, Linden J. Improved detection of ischemia-induced increases in con nary sinus adenosine in patients with coronary artery disease. C/in Chum 1992; 38:2X-262. 9. Ranhosky A, Kempthome-Rawson J, and the Intravenous Dipyridamole Thallium Imaging Study Group. The safety of intravenous dipyridamole thallium myocardial perfusion imaging. Cinuludon 1990:X1: 1205-1209. 10. Sand& H, Dodge HT. The use of single plane angiocardiograms for the calculation of left ventricular volume in man. Am Heart J 1968;75:325-334. 11. Palacios IF, Miller SW. Coronary arteriography and left ventticulography. In: Eagle KM, Haber E, DeSanctis RW, Austen WG, eds. The Practice of Cardiology. The Medical and Surgical Cardiac Units at the Massachusetts General Hospital. 2d ed. Vol II. Boston: Little, Brown, 1989;1637-1655. 12. Beme RM. Cardiac nucleotides in hypoxia: possible role in regulation of corenary blood flow. Am .J Phrsiol 1963;204:317-322. 13. Belardinelli L, Linden J, Beme RM. The cardiac effects of adenosine. P,-og Cardiormc Dis 1989;32:73-97. 14. German DC, Kredich NM, Bjomsson TD. Oral dipyridamole increases plasma adenosine levels in human beings. C/in Phurmacol Ther 1989;45:8&84. 15. Linden J, Taylor H, Feldman MD, Woodward EB, Ayers CR, Ripley ML, Patel A. The precise radioimmtoo assay of adenosine: minimization of sample collection artifacts and immunocrossreactivity. Anul Biochem 1992;210:24&254. 16. Watson DD, Campbell NP, Read EK, Gibson RS, Teates CD, Belier GA. Spatial and temporal quantitation of plane thallium myocardial images. J Nucl Med 1981;22:577-584. 17. Belier GA, Granato JE, Cannon JM, Belardinelli L, Watson DD. Effects of intracoronary dipyridamole infusion on regional myocardial blood flow and intrinsic thallium-201 washout in dogs with a critical coronary stenosis. Am Hearr J 1992: l24:5&64 18. Feldman RL, Nichols WW, Pepine CJ, Conti CR. Acute effect of intravenous dipyridamole on regional coronary hemcdynamics and metabolism. Cwcukufion 1981;64:333-344. 19. Picano E, Lattanzi F, Masini M, Distante A, L’Abbate A. High dose dipyridamole echocardiography test in effort angina pectoris. J Am Co// Card& 1986;8: 84X-854. 20. Picano E, Pirelli S, Marzilli M. Faletra F, Lattanzi F. Campolo L, Massa D, Albeni A, Gara E, Distante A, L’Abbate A. Usefulness of high-dose dipyridamole echocardiography test in coronary angioplasty. Cwcukfion 1989;80:807-815. 21. Bolognese L, Sarasso G, Aralada D, Bongo A, Rossi L, Rossi P. High dose dipyridamole echocardtography early after uncomplicated acute myocardial infarction. Correlation with exercise testing and coronary angiography. J Am Coil Cardial 19X%14:357-363. 22. Martin TW, Seaworth IF, Johns JP, Pupa LE. Condos WR. Comparison of adenosine, dipyridamole, and dobutamine in stress echocardiography. Ann Inrern Med 1992:116:19&196. 23. Zoghbi WA, Cheitif J, Kleiman NS, Verani MS, Trakhtenbroit A. Diagnosis of ischemic heart disease with adenosine echocardiography. J Am Co// Card& 1991;18:1271-1279. 24. Ogilby JD, Irkandrian AS, Untereker WJ, Heo J, Nguyen TN, Mercuro J. Effect of intravenous adenoaine infusion of myocardial perfusion and function. Circularion 1992:86:887-895. 25. Gdbert JC, Glantz SA. Determinants of left ventricular tilling and of the diastolic pressure-volume relation. Circ RES 1989;64:827-852. 26. Vogel WM, Apstein CS, Btiggs LL, Gaasch WH, Ahn J. Acute altera tions in left ventricular diastolic chamber stiffness. Role of the erectile effect of coronary arterial pressure and flow in normal and damaged hearts. Circ Res 1982;5 I :465478. 27. Verrier ED, B&tow JD, Hoffman JIE. Coronary vasodilation shifts the diastolic pressure-dimension curve of the left ventricle. .I Mel Cell Cardiol 1986; 18: 579-594. 28. Farhi ER, Canty JM, Kocke FJ. Dissociation of diastolic pressure-segment length and pressure-wall thickness relations during vascdilation in the conscious dog. J Am Co/l Cardiol 1991;18:85&857. 29. Liu GS, Thornton J, Van Winkle DM, Stanley AW, Olsson RA, Downey JM. Protection against infarction afforded by preconditioning is mediated by A, adenosine receptors in rabbit heart. Cirudcrrion 1991:84:35&356. 30. Thornton JD, Guang SL, Olsson RA, Downey JM. intravenous pretreatment with A,-selective adenosine analogues protects the heart against infarction. Circu/ation 1992;85:659-665.

JUNE 15, 1994