Effect of lntracoronary Nitroglycerin on Myocardial Blood Flow and Distribution in Pacing-Induced Angina Pectoris

Quantitative Assessment By Single-Photon Emission Tomography

PETER LIU, MD, SYLVAIN HOULE, MD, PhD, ROBERT J. BURNS, MD,

BRIAN KIMBALL, MD, ANN WARBICK-CERONE, BScPhm,

LINDA JOHNSTON, BScPhm, DAVID GILDAY, MD, RICHARD D. WEISEL, MD,

and PETER R. MCLAUGHLIN, MD, with the kc/&a/ assistance of

JOAN IVANOV, RN, ati MARIA GREEN, RTNM

Changes in regional coronary flow after adminis- tration of intracorgnary nitroglycerin were assessed by measuring total coronary blood flow (using cor- onary sinus flow catheters) and it+ regional distri- bution (by quantitative single-photon emission to- mography of injected radioactive microspheres). After pacing io angina, 10 patients with coronary artery disease received serial selective lefl coronary injections of technetium-99m microspheres, 40 pg of nitroglycerin, and indium-11 1 microspheres. Significant changes in coronary flow distribution were determined by subtracting prenitroglycerin from postnitroglycerin tomographic profiles. Perfu- sion of each myocardial segment was classified as

normal mildly, moderately or severely compromised, based on upstream coronary anatomy. The overall increase in coronary flow was 23% in the normal territories and 33 %, 44% and 15 % (p <0.05), in the mildly, moderately and severely compromised territories, respectively, compared with control values. Thus, intracoronary nitroglycerin increased coronary blood flow to all perfusion territories. The increase in distribution of coronary flow was greatest in the mildly and moderately compromised regions and the least in the most severely compromised regions; this is probably a reflection of the underlying coronary reserve.

(Am J Cardiol 1985;55:1270-1278)

Previous studies have established that the predominant were either done at rest without ischemia or failed to beneficial effect of nitroglycerin in the relief of angina measure local flow distribution.4 This study was de- is in peripheral vasodilatation, which decreases myo- signed to examine the direct effect of intracoronary cardial oxygen demandl; but whether nitroglycerin has nitroglycerin on regional myocardial blood flow distri- a direct beneficial effect on myocardial oxygen supply bution during pacing-induced angina in patients with by redistributing coronary flow to acutely ischemic coronary artery disease. This was done by combining myocardial zones is controversiaL2-s Some data have the measurements of both total coronary blood flow shown that nitroglycerin increases flow to ischemic re- through coronary sinus catheters and regional distri- gions,4 other data have not.3 However, these studies bution by the dual microsphere technique.2lg

From the Division of Cardiology, Toronto General Hospital, and the Divisions of Nuclear Medicine, Toronto General Hospital and Hospital For Sick Children, Toronto, Canada. This study was supported by Grant Tl-7, The Heart and Stroke Foundation of Ontario, Toronto, Ontario, Canada. Manuscript received October 8, 1984; revised manuscript received February 7, 1985, accepted February 12, 1985.

Address for reprints: Peter R. McLaughlin, MD, Cardiovascular Unit, Toronto General Hospital, l-424, Eaton Wing, 200 Elizabeth Street, Toronto, Ontario, M5G lL7, Canada.

Methods Twenty patients with clinically suspected coronary artery

disease scheduled for cardiac catheterization were studied. Patients with valvular heart disease, cardiomyopathy, a his- tory of myocardial infarction, or collaterals on coronary ar- teriography were excluded. Cardiac medications were with- held for 24 hours before the study, and sublingual

1270

May 1, 1985 THE AMERICAN JOURNAL OF CARDIOLOGY Volume 55 1271

nitroglycerin was used for angina as required up to 1 hour before catheterization. All patients were premeditated with diazepam, 10 mg orally, 1 hour before the procedure.

At catheterization a thermodilution coronary sinus blood flow catheter adapted with a pacing electrode (Wilton-Web- ster Laboratories) was introduced through left basilic vein or right femoral vein, and positioned in the coronary sinus with the proximal thermistor approximately 1 cm distal to the os- tium of the coronary sinus. Its position was confirmed by the injection of a small amount of contrast medium (less than 0.5 ml). A pigtail catheter was then placed in the left ventricle, and a left Judkins coronary catheter positioned in the left main coronary ostium, confirmed by a minimal amount of contrast. Patients were excluded from the study if the left main coronary artery was less than 2 cm long to avoid inade- quate mixing. Systemic blood pressure, left ventricular end- diastolic pressure and thermodilution coronary sinus flows were recorded, and blood samples for aortic and coronary sinus lactate determinations were obtained at rest and after each intervention.

Coronary sinus pacing was begun at 10 beats/min above the baseline heart rate, and the rate was increased by 10 beats/min every minute unit1 moderate angina was produced (at a se- verity of about 5 on a scale of 0 to 10). One millicurie of tech- netium-9_9m-labeled human albumin microspheres was then introduced at a steady rate over approximately 10 seconds into the left main coronary artery and flushed with 5 ml of normal saline solution. In 10 patients 40 pg of nitroglycerin (NTG) was injected into the left main coronary artery and in the other 10 patients a similar volume of saline solution was injected instead of nitroglycerin. Within 15 seconds, 0.3 mCi of in- dium-111-labeled human albumin microspheres was injected in the same manner as the technetium-99m. Hemodynamic variables, coronary sinus blood flow measurements and samples for lactate were obtained before and immediately

FIGURE 1. Circumferential count profile of a to- mographic slice, representing the average counts per pixel for each 18’ segment. The assignment of myocardial perfusion territories is illustrated here with segments 1 to 3 and 18 to 20 classified as right coronary artery, segments 4 to 9 as cir- cumflex and segments 10 to 17 as left anterior descending territory (including diagonal branches).

E f 100

8 I 2 90 1 60

after intracoronary nitroglycerin administration during steady-state pacing. Pacing was then terminated and mea- surements were repeated after the resolution of angina.

Coronary arteriography in multiple projections and left ventricular angiography were then performed by standard techniques. After catheterization was completed the patient was transferred to the nuclear medicine laboratory for single photon emission tomography. The patient was positioned under the gamma camera (General Electric Model 4OOT), which was fitted with a parallel-hole, general-purpose coili- mator interfaced to a Medical Data Systems computer. Each image was acquired for 30 seconds at 6’ intervals for a total of 180°. Dual isotope simultaneous acquisition was performed using a 20% window centered at 140 keV for technetium-99m and a 15% window centered at I72 keV for the lower peak of indium-111 to ensure perfect registration of myocardial re- gions before and after NTG. The images of each isotope were acquired onto separate 64 X 64 matrices with an average count of 30,000 counts/image for technetium-99m and 9,000 counts/image for indium-111.

The pre- and post-NTG flow distribution, as represented by technetium and indium images respectively, were recon- structed using Medical Data Systems A2 software and real- igned into transverse circular slices perpendicular to the iong axis of the left ventricle. Four to 5 equidistant slices from apex to base in which the left ventricular myocardium could be visualized as a perfect doughnut were retained for analysis, and identical reconstruction and slice selection were done for each technetium and indium image pair. Image pairs situated immediately at the apex or basal valvular plane of the left ventricle were excluded from the analysis. The left ventricular portion of each slice was then separated circumferentially into 18” segments and the average counts per pixel of each segment were plotted as a profile, and with their vascular territories assigned as illustrated in Figure 1. This profile was con-

z 70

i 60-

& 50-

; 40-

2 30-

E P

20-

z! lo-

I I I 1 1 0 60 120 180 240 300 360

CIRCUMFERENTIAL PROFILE ANGLE

A I f LEFT f

A RIGHT

, DIAGONAL , ANTERIOR I

I I DESCENDING IC~~~ENRAyRYI

I VASCULAR TERRITORIES

TOMOGRAPHIC SLICE 160

240

36010 DEGREES

1272 INTRACORONARY N~TFKIGLYCER~N AND CORONARY FLOW

TABLE I Coronary Anatomy and Changes in Coronary Sinus Blood Flow and Distribution After lntracoronary Nitroglycerin

Changes in No. of Stenosis- Coronary Myocardial Tomo Coronary Degree and Sinus Flow Segments with

PI Slices Territory Location b-4 Increased Flow

1

2

5

6

7

8

9

10

5

5

5

LAD LAD Circ RCA LAD Circ RCA Normal

Ki RCA LAD Circ RCA

k% Circ RCA LAD Circ RCA LAD Circ RCA LAD LAD Circ RCA

>90% Normal Normal

~,“:~I O-50% >90%

Normal 50-75 % >90% >90% 75-90 % 100% >90% 50-75 % Normal 75-90 % 75-90 % 50-75 % 100% 75-90 % 75-90 Y/o

;;I:$ Normal ’ Normal Normal

5 kE

>90% 75-90%

RCA ~~ Normal

Mid Prox

Prox Mid

Prox Prox Prox Prox Prox Prox Mid Prox Prox Prox Prox Mid Prox Prox

Mid Prox

Prox Prox

15 (9%)

50 (28%)

11 (10%)

17 (9%)

15 (10%)

20 (21%)

14 (23%)

8 (8%)

46 (21%)

9/24 (38%) 12/16 (75%) 17/30 (57%)

19/32 (59%) 17124 (71%)

39/70 (55%) 13/32 (41%) 15124 (64%)

13/40 (33%) 25130 (83%)

6/40 (15%) 13/18 (72%)

5/12 (42%)

35140 (88%) 17/30 (57%)

15/32 (47%) 19124 (79%)

9/16 (56%) 6/16 (37%)

14124 (58%)

14140 (35%) 20/30 (67%)

Territories distal to a stenosis are considered affected by that stenosis. Right coronary artery segments were not analyzed.

Circ = circumflex coronary artery; LAD = left anterior descending coronary artery; Prox = proximal; RCA = right coronary artery; Tomo = tomographic.

strutted for each corresponding slice before and after NTG treatment.

As the microsphere method used in this study reflected only distribution and not absolute flow, and a different amount of activity was injected for the technetium and indium micro- spheres to optimize imaging, the count profiles derived from the reconstructed tomograms are not directly comparable. To determine the relative changes in flow distribution for each segment after NTG compared with that before NTG, the

PATIENT J.R.

2wK%

% FLOW CHANGE

-100% 0 180 360

MYOCARDIAL SEGMENTS

FIGURE 2. Profile of the percent flow change in each 18’ segment, calculated by subtracting the pre- from the postnitroglycerin profile, then dividing by the prenitroglycerin profile.

counts in the post-NTG images were normalized for total activity and relative changes in total coronary sinus flow ac- cording to the formula:

N eorr = N x Total myocardial counts pre x CSBF Post Total myocardial counts post CSBF Pre

where N = post-NTG profile of the raw data, N,, = post- NTG profile corrected for flow and counts, CSBF pre = cor- onary sinus blood flow before NTG, and CSBF post = coronary sinus blood after NTG.

The original pre-NTG profile was then subtracted from the flow-corrected post-NTG profile, and the absolute difference in counts was obtained. To compare data between patients this difference curve was divided by the pre-NTG profile, thus providing a profile of the percentage change in flow related to the original pre-NTG values (Fig. 2). This curve represented relative segmental changes in flow after intracoronary NTG.

The coronary arteriograms in multiple views were inter- preted by 2 independent observers without knowledge of the clinical data, and the severity and location of the stenotic le- sions were determined. The location of affected segments were then determined by following the vascular distribution ter- ritories outlined in Figure 1 adjusted according to the location and size of compromised distal perfusion regions. To char- acterize the degree of compromise in perfusion in each to- mographic segment, the upstream coronary anatomy for in- dividual patients was used to classify the segments into mild, moderate, severe or noncompromised perfusion categories. Tomographic segments supplied by vessels with less than 50%

May 1,1985 THE AMERICAN JOURNAL OF CARDIOLOGY Volume 55 1273

TABLE II Hemodynamic, Metabolic and Coronary Flow Response to Pacing and lntracoronary Nitroglycerin

(beazmin) DAP LVEDP CSF

(mdl) MVL

(mm Hg) (mm Hgl (ml/min) (mg/min)

Baseline 70f 9 72f 14 12f9 132 f 21 1.8 f 1.6 2.1 f 1.6 Pacing-Angina llOf9’ 83 f 13’ 11 f8 182 f31 l -0.7 f 1.1 -0.5 f 1.0 IC NTG 110f 12’ 83f 11’ lOf9 221 f 34.t 1.0 f 1.4t 0.7 f 1.3

l p <O.Ol compared with baseline. t p <0.05 pre- vs post-IC NTG. Values are mean f standard deviation. CSF = coronary sinus flow; DAP = diastolic pressure; HR = heart rate; IC = intracoronary; LE = lactate extraction; LVEDP = left ventricular

end-diastolic pressure; MVL = myocardial lactate flux; NTG = nitroglycerin.

luminal narrowing were classified as noncompromised. Seg- ments perfused by vessels with 50 to 75% narrowings were classified as mildly compromised, 76 to 90% narrowings as moderately compromised and greater than 90% as severely compromised, with presumed corresponding reductions in- coronary flow reserve.

The 10 patients in the control group were analyzed to es- tablish the intrinsic variability of the technique and to de- lineate the distribution of labeled particles in the absence of any intervention. By pooling the data from these patients we established a mean and standard deviation for each 18” myocardial segment. A range of 2 standard deviations on ei- ther side of the mean was taken as the 95% confidence limits, outside of which would indicate a significant redistribution of flow.

The different curves from the patients who received NTG were then examined segment by segment for the left coronary territories. If the difference in flow distribution exceeded the 2 standard deviation range established by the control group, ‘a significant increase in flow was considered to have taken place in that 18” segment. By this means, the number and percent of segments showing increased flow in each category were determined. To quantitate the magnitude of change in flow, the total counts before and after NTG administration for all segments within each category were summed and the mean change in flow for each perfusion category was determined.

Ethical considerations: The experimental protocol was approved by the Committee on Human Experimentation in February 1982. Each patient gave informed consent before the study. The total radiation dose to the myocardium (target organ) from 1 mCi of technetium-99m and 0.3 mCi of in-

-1 ’ ‘P<.Ol **p<.o5

I

ggg PACING

0 -- PRE.P~CING PACE TO ANGINA POST NTG OFF PACING

INTERVENTION

FIGURE 3. Coronary sinus blood flow (CSBF) at rest and during pacing, before and after intracoronary nitroglycerin. NTG = nitroglycerin. *Between prepacing and pacing to angina. l ‘Between pra and post- NTG during pacing.

dium-111 which each patient received was 1.6 rads. The total body radiation dose was 0.07 rads.

Statistics: Comparison of multiple interventions on the same measurements was done using analysis of variance and Ducan’s multiple range testing for subgroups. Comparison of proportions was done with chi-square test. A p value <0.05 was considered significant.

Results

Patient characteristics: Sixteen men and 4 women were studied. The average age was 51 f 8 years in the control group and 55 f 9 years in the study group. The distribution and severity of the coronary lesions in the nitroglycerin group are listed in Table I. One patient (no. 3) had no significant obstructive disease and no evidence of coronary spasm by ergonovine provocation on repeat catheterization. All patients were paced to an average heart rate of 110 f 9 beats/min from a baseline heart rate of 70 f 9 beats/min to reach a steady state of moderate angina (on a scale 5 to 10). All patients who received intracoronary NTG continued to have angina during pacing. None of the patients had any complica- tion from the procedure.

Hemodynamic and lactate changes: The hemo- dynamic and metabolic changes for the patients who received NTG are summarized in Table II, and the coronary sinus flow and lactate data are plotted in Figures 3 and 4. The arterial blood pressure at rest and after NTG showed no significant changes, indicating that this dose of intracoronary NTG did not have any

LACTATE

EXTRACTION (Wdl)

2 “P<.OS

t

V m PACING .l

PRE.PACING PACE TO POST NTG OFF PACING ANGINA

FfGURE 4. Net coronary lactate flux at rest and during pacing, before and after intraccronary nitroglycerin (NTG). l ‘pm- vs post-nitroglycerin during pacing.

1274 INTRACORONARY NITROGLYCERIN AND CORONARY FLOW

TABLE III Percent Redistribution Profile for Each of the 18’ Myocardlal Segments in the Control Patients

Sement Mean SD -2 SD +2 SD

1

3 :+ 314

z 2:

! 8.:

: 1:7

:: f :Y

;: -1:3 -;j

:: 2::

:; i::

:i E 3:3 20

SD = standard deviation.

-9.9 16.5

t::

1::

-10.0 -9.4 17.1 16.2

-10.2 -14.1 16.3 19.4

t :l

S:!

-16.0 -15.2 17.7 16.0

t:;

-15.3 -14.0 16.7 19.9

K

-10.2 -11.2 16.4 13.6

-11.1 -12.9 1i.I

;::

t:;

-10.1 -10.7 11:7 16.2

;:B

-9.4 -9.7 16.9 16.4

t :: -9.1 -6.2 16.6 15.2 -6.7 15.2

significant systemic effect. Left ventricular end-dia- stolic pressure also did not change significantly after nitroglycerin administration. Coronary sinus flow in- creased after pacing and increased further after nitro- glycerin. There was net lactate extraction in these pa- tients at rest; however, during pacing there was net lactate production. After NTG administration there was a significant trend back to extraction but it did not re- turn to prepacing levels.

Control data: The mean and standard deviation for each myocardial segment in the control group are tab- ulated in Table III. Figure 5 shows the two standard deviation control range as a profile.

Effect of nitroglycerin on redistribution of flow: Number of segments with significant redistribution: The number of segments that demonstrated a signifi- cant increase in flow for each patient and the total for, each perfusion category are detailed in Tables I and IV, respectively. Many segments showed a significant in- crease in flow after NTG. However, the moderately compromised regions allowed the greatest number of segments with an increase in coronary flow (79%),

CONTROL RANGE

MYOCARDIAL SEGMENTS

FIGURE 5. Two standard deviation (S.D.) control range of the percent change in coronary flow in the absence of pharmacologic interven- tion.

whereas the most severely compromised regions showed the smallest increase (25%).

Quantity of redistribution of flow after nitroglycerin: All the categories showed a net increase in perfusion. The mean increase in coronary flow in normal segments was 23%. The mildly and moderately compromised zones demonstrated a greater increase in coronary flow change (33% and 44%, respectively, p <O.OOl compared with control). The most severely compromised catego- ries showed only a modest increase of 15%, significantly less than the normal zones (p <O.OOl compared with control).

Discussion

Technique: Dual-labeled microspheres: The dual- labeled microsphere technique has been used previously to study changes in regional coronary flow in patientslo The advantages of this technique are that it has direct quantitative potential without the need of background subtraction. The high target to background ratio also increases the statistical reliability of the data after to- mographic reconstruction. It also allows the direct comparison of regional blood flow before and after an intervention.

One potential source of error of the intracoronary technique is preferential streaming of particles down the left coronary artery. This is an important consid- eration if absolute flows were being measured. However, in this study it was not the absolute but the relative change before and after an intervention in which we were interested. If streaming does occur, it should be consistent between the injections of particles before and after the intervention. Hence the importance of pre- cisely adhering to a consistent injection protocol for the microspheres must be emphasized. It is also possible that streaming could be different between low flow zones and normal flow zones. This is accounted for by having a range of normal and low flow zones in the control data. To further diminish the risk of a chance error in the measurements we adopted the rigorous limits of f2 standard deviations around the mean change in distribution of flow that occurred when saline solution alone was injected in the left main coronary artery.

Coronary sinus flow and single-photon emission computed tomography: A previous studylo of intra- coronary and intravenous nitroglycerin on coronary flow distribution had 2 important limitations: there were no absolute measurements of coronary blood flow, and planar imaging was used, which has the inherent limi- tations of superimposition of structures. In this study absolute measurements of coronary blood flow were obtained using coronary sinus thermodilution catheters and superimposition of structures was obviated using single-photon emission computed tomography.

Mechanisms of action of intracoronary nitro- glycerin: Early studies by Gorlin et al3 showed that NTG did not increase coronary blood flow in patients with coronary artery disease. Yet Ganz and Marcus4 reported that NTG did increase coronary blood flow in about half of patients with pacing-induced angina. They postulated that NTG increased coronary blood flow

May I.1985 THE AMERICAN JOURNAL OF CARDIOLOGY Volume 55 1275

TABLE IV Relation Between Severity of Stenosis and Number of Segments Responding to Nitroglycerin

lschemic Category Significant Increkse

in Flow After NTG Total No.

of Segments p Value l

Normal lschemic

Mild stenosis Moderate stenosis Severe stenosis

142(63%) 226

1;: <El 176 <o:oo 1

l Compared with normal regions. NS = not significant; NTG = nitroglycerin.

mainly in the relatively nonischemic regions, where the arteriolar dilatation was not already maximal, even though they did not measure regional coronary flow. Conversely, Mehta and Pepine6 found that NTG did increase the great cardiac vein flow in patients with left anterior descending branch stenosis during pacing- induced angina, but they used sublingual rather than intracoronary NTG, thus confounding the direct with the systemic effect of the drug.

As expected, in normal territories of perfusion, in- tracoronary NTG increased regional coronary blood flow, by 23%. In the mildly and moderately ischemic regions, significantly more segments of myocardium showed an increase in coronary blood flow, and the overall increases in coronary flow in these perfusion categories of territories were 33% and 44%, respectively. In the most ischemic regions, only 52% of segments showed an increase in coronary blood flow, and the overall increase in coronary blood flow was limited to 15%, less than the normal zone values. Thus, NTG ap- parently increased coronary blood flow in all regions of the myocardium, but preferential redistribution was toward the mildly and moderately ischemic territories, and there was least increase in the most ischemic territories.

These observations fit the hypothesis that during coronary ischemia, due to local accumulation of meta- bolic factors, the conductance or resistance vessels to the ischemic regions dilate to varying degrees, de- pending on the severity of ischemia. The intrinsic action of NTG in mildly and moderately flow-compromised zones with some residual coronary reserve still intact is to further dilate these vessels that are not yet fully di- lated. Added to this effect may be the opening of small collateral vessels that are not apparent on routine con- trast angiography. In the severely ischemic regions the vessels behind the stenoses may be already fully dilated, with no further flow reserve, and nitroglycerin failed to exert much further effect to increase regional blood flow. This would conform to the observation that NTG did not increase flow enough to relieve pacing-induced an- gina, but did improve lactate metabolism. During pac- ing-induced angina, there was net lactate production and the increased coronary blood flow was the result of increased cardiac output and increased demand. However after NTG, there was less lactate production, likely due to partial relief of ischemia in the mildly and moderately compromised zones. The severely com- promised zones very likely continued with a similar

degree of ischemia due to little net change in coronary blood flow.

Our data differ from those of Brown et al,7 who used a high-resolution angiographic technique to determine that the stenotic vessels can vasodilate with NTG and that coronary vascular resistance decreased the most (38%) in severe stenosis. However, they did not measure total coronary blood flow, and assumed physiologic changes followed directly the anatomic changes. Our method monitors the actual net physiologic changes independent of anatomic changes, and should be a more accurate measurement of the variable under investi- gation. However, we have not attempted to determine if the net change is predominantly related to conduc- tance or resistance vessel dilatation.

Potential limitations: Coronary sinus flow mea- surements: One limitation of the study is that the total coronary blood flow, rather than the selective great cardiac vein flow, was measured. This was done to achieve consistent results and a readily reproducible technique. This would necessarily include flow from both the right and left coronary artery territories. We believe that the drawback was minor compared with the advantages, as the coronary sinus captures most of the left coronary artery venous drainage. Because the in- tervention was administered only in the left coronary artery, any changes resulting from the intervention would be reflected in the coronary sinus blood flow. No intervention was made on right coronary flow and, thus, no change in right coronary contribution to the coronary sinus flow would be expected to confound the results.

Single-photon emission computed tomography: The problems of tissue attenuation and of partial vol- ume effects were not specifically addressed in this study. However, as we were interested only in relative changes and not in absolute quantitation of coronary blood flow, the effects of attenuation and partial volume response were minimized by selecting two labels with close energy peaks. The influence of the patient’s cardiac and res- piratory motions were also minimized by acquiring both tracer images simultaneously, thus ensuring perfect registration of the pre- and post-NTG images. However due to the limits of resolution of the system, endocar- dial/epicardial flow redistribution was not resolved.

Effect of collaterals: The effect of NTG on collateral flow was not addressed by this study. As our imaging technique studied only the left coronary artery, it would fail to account for flow from collaterals originating from the right coronary artery. Hence, patients with angio-

1276 INTRACORONARY NITROGLYCERIN AND CORONARY FLOW

graphic collaterals from the right to the left coronary artery (left anterior descending or circumflex) were excluded from this study. There may still exist in our study patients with significant collaterals from the right coronary artery or within left coronary territory itself not demonstrated by angiography.ll We would expect such collaterals to be relatively small and as such not to have contributed large quantities of unmeasured cor- onary flow. However, the limited increase in flow in the most severely ischemic territory could have been sup- plemented with some collateral flow. In addition, ni- troglycerin administered systemically, rather than by the direct intracoronary route, could still enhance flow to the most stenosed regions through collaterals from other coronary territories. This method, however, pro- vides an excellent potential of assessing collateral flow from the left coronary to right coronary territory, since the right coronary territory is not labeled with activity unless such collaterals exist.

Acknowledgment: We thank Louise Williams for con- siderable work in manuscript preparation. We also thank Bette Shurvell for assistance in manuscript preparation.

References

1. Willlam W, Amsferdam EA, Mascn DT. Hemcdynamic effects of nitro- glycerin in acute myocardial infarction. Decrease in ventricular preload at the expense of cardiac output. Circulation 1975;51:421.

2. Fam W, McGregor M. Effect of coronary vasodilator drugs on retrograde !lo; in areas of chronic myocardial ischemia. Cir Res 1964;15:355-

3. Gcrfk R, Brachfeld N, MacLecd C, Sopp P. Effect of nitroglycerin on the coronary circulation in patients with coronary artery disease cr increased left ventricular work. Circulation 1972;46:680-889.

4. Ganz W, Marcus HS. Failure of intracoronary nftroglycerin to alleviate

6. i&l cing-induced angina. Circulation 1972;48:880-889.

efn L, Frfesfm GC, Lkhtlen Pf?, Roes RS. The effect of nifmglycarln on the systemic and coronary circulation in man and dogs. Circulation 1968;33:107-116.

6. Meffta J, Peplns CJ. Effect of sublingual nitroglycerin on r 7

lonal flow in patients with and without coronary disease. Circulation 197 ;5&803-7.

7. BrcwnRG,BcfscnE,PetsfscnRB,PlsfceCD,Dcd9sHT.Them&anisrns for nitroglycerin action: Stenosis vasod5ation as a major component of the d response. Circulation 1981;84:1089-1100.

6. k%egor M. Pathogenesis of angina pectoris and role of nitrates in relief of myocardlal ischemia. Am J Med 1983;74(68):21-27.

9. Gculd KL, Hamflton GW, Llpsccmb K, Rltchle JL, Kennady JW. Method for assessing stress-induced regional malperfusion during coron

T arte-

riography. Experimental validation and clinical application. Am J ardiol 1974;34:557-585.

10. McEwan MP, Berman ND, Mcrch JE, Fe@fn D, M&a 7

hlfn PR. Effect of intravenous and intracoronary nitroglycerin on left ventr cufar wall motion and perfusion in patients with coronary artery disease. Am J Cardiol 1981;47:102-108.

11. Goldstein RE, Silnscn EB, Scherer JL, Senln9er RP, Grehl TM, Epstein SC: Intraoperative coronary collateral function in patients with coronary occlusive disease-nitroglycerin responsiveness and angiographic cor- relations. Circulation 1974;49:298-308.