- Email: [email protected]
Measurement of regional myocardial blood flow
J
THoRAc CARDIOVASC SURG
1988;96:775-81
Measurement of regional myocardial blood flow Application of the electrolytic hydrogen clearance method in man To assess myocardial perfusion intraoperatively and to evaluate the adequacy of coronary bypass grafting, we measured regional myocardial blood flow by the electrolytic hydrogen clearance method in 49 patients. Group I comprised 10 patients with nonischemic heart disease and group II, 39 patients with ischemic heart disease undergoing coronary bypass grafting. Group II was subdivided according to the percent stenosis of the coronary arteries supplying the ventricular regions: group lla, <75 % stenosis; group lIb, ~75% stenosis. Mean myocardial blood flows were 154 ± 7, 145 ± 5, and 98 ± 9 m1/min/loo gm in groups I, Ila, and lIb, respectively (p < 0.01, group lIb versus groups I and Ila), Mean blood flows were 161 ± 19, 159 ± 12, 78 ± 12, and 59 ± 15 m1/min/loo gm in areas of the left anterior descending coronary artery with <50 %, 75 %, 90 %, and 99 % stenosis in group Il, In areas with a totally occluded left anterior descending coronary artery with collaterals, mean flow was 90 ± 15 m1/min/loo gm. The mean myocardial blood flows were 40 ± 7 and 100 ± 14 m1/min/loo gm in areas with anterior Q wave and non-Q wave infarction, respectively (p < 0.01). After cardiopulmonary bypass, the mean flow increased from 99 ± 11 to 150 ± 7 m1/min/loo gm in the grafted areas in group lIb (p < 0.01~ but it did not change in group I or Ila, The electrolytic hydrogen clearance method provided quantitative evaluation of myocardial perfusion and recovery from hypoperfusion by coronary bypass grafting. This method was especially useful in patients undergoing mammary artery grafting.
Michio Kawasuji, MD, Fumio Kawajiri, MD, and Takashi Iwa, MD, Kanazawa, Japan
Severe atherosclerotic obstruction in the coronary arteries can reduce regional myocardial blood flow (MBF). The hemodynamic objective of coronary bypass operations is to restore adequate blood flow to the jeopardized myocardial segments. Myocardial perfusion is estimated by electrocardiography, myocardial scintigraphy, coronary angiography, and left ventriculography before and after operation. Measurements of regional MBF by means of inert gas such as xenon 133 have been reported."! However, surgeons have few tools to evaluate myocardial perfusion at the time of operation. Methods presently used to evaluate results of coronary bypass, such as electromagnetic flow, do not measure blood flow per unit mass of myocardium. Some methods From the Department of Surgery (I), Kanazawa University School of Medicine, Kanazawa, Japan. Note: The review process for the paper was conducted by Dr. Dwight C. McGoon.
Received for publication March 18, 1987.
reported to measure regional MBF during operation are complicated and have limitations.': 5 The hydrogen clearance method has been used to measure tissue blood flow/" Stosseck, Lubbers, and Cortin" reported a hydrogen clearance method for which electrochemically generated hydrogen gas is used. Koshu and associates" modified the method and reported excellent correspondence between blood flow values obtained by their method and the hydrogen inhalation method. Experimental results indicated that regional MBF could be measured by the electrolytic hydrogen clearance method. I I We l 2 have already reported the first clinical application of this method to measure MBF. The present study was performed (I) to quantitate myocardial perfusion intraoperatively and to correlate MBF with the degree of coronary artery disease and (2) to evaluate the adequacy of myocardial tissue perfusion in the grafted areas by measuring the increase in regional MBF.
Accepted for publication April 12, 1988.
Patients and methods
Address for reprints: Michio Kawasuji, MD, Department of Surgery (I), Kanazawa University School of Medicine, 13-1 Takararnachi, Kanazawa 920. Japan.
This study was conducted on 49 patients undergoing elective cardiac operations with cardiopulmonary bypass (CPB) at Kanazawa University Medical Center between September
775
The Journal of
7 76
Thoracic and Cardiovascular Surgery
Kawasuji, Kawajiri, Iwa
Reference electrode
T 1/2 =O.67mln
. . fIO['~m"~'"'C __"" ~:
Fig. 1. A, Schematic presentation of methods for measuring intraoperative MBF. B, Time course of polarographic current caused by hydrogen oxidation and its logarithm. Hydrogen concentration increased by electrolysis of tissue water during the first 50 seconds and then decreased because of washout by blood flow. MBF was calculated from the half period of the change in polarographic current. 1985 and May 1987. They were divided into 10 patients with nonischemic heart disease (group I) and 39 patients with ischemic heart disease (group II). The patients in group I were between 31 and 62 years old. They had normal coronary arteries. Two of the patients had an atrial septal defect, four had mitral stenosis, and four had the Wolff-Parkinson-White syndrome. Operative procedures were closure of atrial septal defect in two patients, open mitral commissurotomy in two, mitral valve replacement in two, and division of the atrioventricular accessory conduction pathway in four. The patients in group II were between 6 and 69 years old. Thirteen of the patients had single-vessel coronary disease, 12 had doublevessel disease, 10 had triple-vessel disease, and four had a left main trunk lesion. The operative procedure was single coronary bypass grafting in 13 patients, double coronary bypass grafting in 20 patients, and triple coronary bypass grafting in six patients. An internal mammary artery was used as a graft to the left anterior descending coronary artery (LAD) in 12 of the patients. Segments of the great saphenous vein were used as grafts to other coronary arteries. Coronary cineangiograms were obtained after administration of nitroglycerin (0.4 mg sublingually). The grading of coronary artery stenosis was expressed according to the reporting system for coronary artery disease by the American Heart Association." The diagnosis of myocardial infarction was established by chest pain, elevation of total creatine kinase MB level, and electrocardiographic change. Myocardial infarction was divided into two groups: Q wave and non-Q wave infarction." The diagnosis of anterior Q wave infarction was established by development of a Q wave of 0.04 second or more in any precordial lead. The diagnosis of anterior non-Q wave infarction was established by electrocardiographic criteria as persistent new T wave or ST segment depression, or both, and absence of new Q wave. All patients were anesthetized with nitrous oxide and fentanyl, intubated, and the lungs ventilated with a ventilator. The studies were performed during the operation, with a
midsternal incision for cardiac exposure. Radial artery pressure, central venous pressure, and the electrocardiogram were monitored continuously on a Hewlett-Packard polygraph recorder (Hewlett-Packard Company, Andover, Mass.). MBF was first measured before CPB. Myocardial preservation through a single period of aortic clamping was achieved by administration of a cold (4° C) crystalloid potassium cardiaplegic solution. MBF was measured at cardiac standstill just after infusion of a cardioplegic solution. Intracardiac repair or coronary bypass grafting was then performed. MBF was again measured after weaning from CPB and stabilization of hemodynamics. Regional MBF was measured with an electrolytic hydrogen clearance tissue blood flowmeter RBF-2 and a data analyzer BDA-I-2 (Biomedical Science Inc., Kanazawa, Japan). The electrode was a Teflon-coated bipolar electrode with a diameter of 0.5 mm and contained a platinum-iridium wire of 130 JIm (electrode A) and a platinum wire of 300 JIm (electrode B). This needle electrode was inserted into the anterior wallsof the left and right ventricles (Fig. I). A reference electrode was connected to the subcutaneous tissue. A direct current of 50 to 100 JIA was passed for 50 seconds between electrode B and the reference electrode for electrolysis of tissue water. Hydrogen gas was generated as follows: 2H 30+ + 2e-
--+
Hzt + 2HzO
Simultaneously, the apparatus monitored the changes in hydrogen concentration by recording the polarographic current caused by hydrogen oxidation at electrode A as follows: Hz --+ 2H+ + 2eThe half period of the change in hydrogen concentration was calculated. The true regional blood flow (F) was calculated from the equation '0: F = 69.3j(T A F B = 69.3jTB
-
FB) (mljminjloo gm)
Volume 96 Number 5 November 1988
Regional myocardial blood flow
NS
Ol o Q
.... c
~ 200
E ~
1. 2
.Q u.
J
"tl
0 0
iii
.
III "tl III
•• •
100
u 0
••
P
•
..•
• • I Q •
-..-.. • •
• I
• •
...T I
--2 •
J
>-
• II •
~
--Q Cl
• 0'----=----=--=----,----=-------' Group I Group IIa Group IIb Fig. 2. MBF before CPB for all patients. The mean value of MBF was lower in group lIb than groups I and lIb (p < 0.01). where T A is the half period of the hydrogen washout curve with blood flow, Ta is the half period of the hydrogen washout curve without blood flow, and Fa is the false blood flow by hydrogen gas diffusion. Tawas measured at the beginning of cardioplegic arrest, that is, during cessation of coronary circulation. Cumulative data were expressed as mean ± standard error of the mean. Statistical comparisons were made by Student's t test to detect significant (p < 0.05) differences between the measured variables. The paired t test was used to compare prebypass to postbypass changes within the same patient, and the unpaired t test was used to compare the differences between the groups.
Results MBF before CPB. Fig. 2 shows the MBF data for all patients before CPB. The mean MBF for group I was 154 ± 7 ml/min/l00 gm. The mean MBF in the anterior walls of the right and left ventricles was 154 ± 9 and 152 ± 11 ml/min/l00 gm, respectively (no significant difference [NS]). The mean MBF for group I, which showed a narrow range of deviation, was considered to be the normal value for MBF by this method. The areas supplied by coronary arteries with <75% diameter stenosis and s; 75% diameter stenosis constituted groups IIa and Ilb, respectively. The mean MBF for group IIa, 145 ± 5 ml/min/100 gm, was not significantly different from the normal value. The mean MBF for group Ilb, 98 ± 9 ml/min/l00 gm (range 5 to 215), was significantly lower than that for group I or IIa (p < 0.01). The hemodynamic data before CPB showed no significant difference between groups I and II (Table I).
....c
E200 ....
!
~
.2 u,
"tl 0 0
III
I I
P
• a
•
•
_100 .!!!
0
•
I
• •
:~
U
0
>:!:
<50
8
I
•
.
"tl III
collateral $
• collateral e
•
·1 ·2 T
777
• •• • 90
0 0
if
..
:f 0 0
8
75 99 100% LAD percent stenosis
Fig. 3. Relationship of MBF to preoperative angiographic assessment of coronary stenosis. The region of the LAD was selected for this study. Mean value of MBF was low in the areas of LAD with ~ 90% stenosis.
Fig. 3 shows the relationship of MBF to the angiographic assessment of the percent coronary stenosis. The anterior wall of the left ventricle was selected for this study. The mean MBFs in the areas supplied by an LAD with <50%, 75%, 90%, and 99% stenosis were 161 ± 19, 159 ± 12, 78 ± 12, and 59 ± 15 nil/min/ 100 gm, respectively. The mean MBF in the areas supplied not by totally occluded LAD but by collateral vessels from the right coronary artery was 90 ± 15 ml/min/l00 gm. Fig. 4 shows the relationship of MBF to the presence of myocardial infarction in group lIb. The anterior wall of the left ventricle supplied by the LAD was again selected for this study. The mean MBF in areas with Q wave infarction, 40 ± 7 ml/min/100 gm, was significantly lower than that in areas with non-Q wave infarction, 100 ± 14 ml/min/100 gm, or that in noninfarcted areas, 120 ± 11 nil/min/ 100 gm. Severity of LAD stenosis resulting in Q wave infarction was 90% in two patients, 99% in three patients, and 100% with collaterals in three patients. Severity of LAD stenosis resulting in non-Q wave infarction was 90% in three patients, 99% in one patient, and 100% with collaterals in four patients. MBF after CPB. MBF after CPB was measured in 57 areas; 10 areas in group I, 16 areas in group IIa, and 31 areas in group lIb. Table I shows the cumulative MBF and hemodynamic data for groups I and II. Before CPB there was no difference in MBF between
The Journal of Thoracic and Cardiovascular Surgery
7 7 8 Kawasuji, Kawajiri, Iwa
Table I. MBF and hemodynamic data before and after CPB Group II Group 110
Group I MBF (mi/min/IOO gm) Before CPB
NS
153 ± 12
After CPB Heart rate (beats/min) Before CPB
NS After CPB MAoP (mm Hg) Before CPB
NS After CPB CVP (mm Hg) Before CPB
NS After CPB Hematocrit (%) Before CPB After CPB
[151 ± 5
t
~NS~
~NS~
Group lib
147 ± II J
~*~
149 ± 10
~NS~
NS
[91.7 ± 5.8
~NS~
80.2 ± 3.8J
94.4 ± 3.9
~NS~
95.0 ± 2.8
C5.8 ± 3.6
~NS~
79.2 ± 2.0J
77.3 ± 2.1
~NS~
75.5 ± 1.9
[9.3 ± 1.2
~NS~
7.4 ± 0.6J
8.8 ± 0.9
~NS~
9.2 ± 0.8
[39.9 ± 1.5
~NS~
38.0± 0.7J
32.1 ± 1.5
~NS~
32.4 ± 0.8
99 ± 1]
t
150 ± 7
t
t
t
MBF. Myocardial blood flow; CPB. cardiopulmonary bypass; NS. not significant; MAoP, mean aortic pressure; CVP. central venous pressure. •p
tp
< 0.05. < 0.01.
groups I and IIa, whereas the mean MBF was significantly lower in group lIb than in group I or IIa (p < 0.01). The mean MBF increased from 99 ± II to 150 ± 7 rnl/min/l00 gm after revascularization in Group lIb (p < 0.05) and approximated a normal value. In group I the hematocrit value was lower after CPB, but there was no significant change in hemodynamics after CPB. In Group II the heart rate and central venous pressure were significantly higher and mean systemic blood pressure and hematocrit value were significantly lower after CPB than before CPB. Fig. 4 shows the relationship of MBF change by revascularization to the severity of anterior myocardial infarction. The mean MBF increased from 39 ± 9 to 127 ± 15 rnl/min/l00 gm in areas with Q wave infarction (p < 0.01). It increased from 95 ± 18 to 143 ± 18 rnl/min/l00 gm in areas with non-Q wave infarction (NS) and from 120 ± 21 to 161 ± 7 mlj min/IOO gm in noninfarcted areas (p < 0.01). The mean MBF after revascularization was significantly higher in noninfarcted areas than in areas with Q wave infarction (p < 0.01). The mean MBF was compared between the areas of the LAD grafted by saphenous vein grafts and internal mammary artery grafts. Before revascularization the mean MBF was 103 ± 12 mljmin/l00 gm (n = 19) in the areas of saphenous vein grafts and 93 ± 16 ml/
min/l00 gm (n = 12) in the areas of mammary artery grafts (NS). After revascularization the mean MBF was 150 ± 9 rnl/min/l00 gm in the areas of saphenous vein grafts and 150 ± 12 rnl/min/l00 gm in the areas of mammary artery grafts. Discussion Measurement of regional MBF provides a direct and quantitative method of evaluating myocardial perfusion. Many approaches, including radioactive isotope uptake methods," radioactive microsphere methods,": 17 and inactive gas clearance methods.v-" have been applied to measure regional MBF. However, intraoperative measurements of MBF have a number of limitations. Hydrogen clearance methods have been used to measure tissue blood flow. Hyman' introduced a quantitative polarographic technique of measuring tissue hydrogen concentration, and Aukland, Bower, and Berliner" modified this technique. Blood flow rates in animal kidneys and hearts were measured by the hydrogen clearance method and found to correlate well with those obtained by venous outflow collection. Kinoshita and associates" measured MBF by the hydrogen clearance method and reported a good correspondence between MBF values obtained by hydrogen clearance and the radioactive microsphere method. Stosseck, Lubbers, and Cottin? measured tissue blood flow by using electrochemically
Volume 96 Number 5 November 1988
Regional myocardial blood flow
--
P
01
P
0
$2
P <0.01
....C
NS
P
NS
P
~
j
0
u,
"t' 0 0
iii
.
100
NS
Ii
NS
II
~ 200
E
.Il!
779
f .•
tf
"t'
Cll U
..
1q
0
>:E
pre post QOMI
pre
post
Non-Q OMI
pre
post OMI(-)
Fig. 4. Relationship of MBF to severity of myocardial infarction and MBF change after coronary bypass grafting. The anterior wall of the left ventricle was selected for this study. Mean values of prebypass and postbypass MBF were significantly lower in the areas of Q wave infarction than the notinfarcted areas. OMI; Old myocardial infarction.
generated hydrogen gas. More recently, Koshu and associates'? modified the equation proposed by Stosseck's group and indicated that the values calculated by their equation correlated well with those obtained by the hydrogen inhalation method. The electrolytic hydrogen clearance blood flowmeter generates hydrogen gas electrochemically and records its clearance by monitoring the change in polarographic current caused by hydrogen oxidation. As a preliminary experiment, one of us" used the electrolytic hydrogen clearance method to measure the MBF in canine hearts and confirmed that this method was applicable for measuring MBF. We'thave reported that this method is safe and can be adapted for use in patients. In the clinical setting, circulatory arrest was not obtained. Therefore, T B was measured at cardioplegic arrest during CPB. Values of F B obtained in this manner correlated with those obtained in the experiments. The electrolytic hydrogen clearance method has many advantages: Measurements can be done without the use of hydrogen gas. Less time is required than for the inhalation method. Regional tissue blood flowcan be measured repeatedly in a given patient. Tissue blood flow is determined as an absolute value (milliliters per minute per 100 gm). There are no harmful effects such as electrical hazard or mechanical damage by a needle electrode. The present study showed that measurement
of MBF by the electrolytic hydrogen clearance method is useful for assessing myocardial perfusion in humans. In the present study, group II was divided into groups IIa and lIb according to the degree of coronary artery stenosis <75% and ;;;;75%, because 75% stenosis is one of the clinical criteria for bypass grafting. The mean value of MBF was normal in group IIa but was significantly decreased in group lIb. Smith and associates' recorded intraoperative Xenon 133 clearance curves and demonstrated that MBF was reduced distal to the coronary lesion with >80% reduction in diameter. The present study showed reduction in MBF in the areas of LAD with ;;;; 90% stenosis. It is noteworthy that the mean MBF in the areas supplied only by collateral vessels was as much as the mean MBF in the areas of LAD with 90% stenosis. This observation is in accord with studies by Schwarz" and Sesto and Schwarz," who examined the effect of collaterals on myocardial function and found that collaterals effectively produced partial revascularization and functionally converted total occlusion to approximately a 90% stenosis. The present study showed that the mean MBF was significantly lower in areas with Q wave infarction than in areas with non-Q wave infarction. This study demonstrated a low, narrow range of MBF in areas with Q wave infarction and a relatively high and wide range of
7 80
The Journal of Thoracic and Cardiovascular Surgery
Kawasuji, Kawajiri, Iwa
MBFs in areas with non-Q wave infarction. This result and postbypass MBF data support that a Q wave infarction is a more homogenous infarction that involves less surviving myocardium than a non-Q wave infarc-
tion." Measurement of MBF before and after revascularization makes it possible to assess intraoperatively the adequacy of coronary bypass graft for regional tissue perfusion. Electromagnetic flow measures only bypass graft flow irrespective of its distribution.' Highly variable graft flow might be expected, because the magnitude of the distal runoff varies and the fraction of distal flow in a grafted coronary artery depends on the graft/coronary artery cross-sectional area ratio and severity of proximal stenosis in the recipient artery." On the other hand, the hydrogen clearance method measures actual nutrient blood flow per unit mass of myocardium. The results of the present study confirm that coronary bypass grafts restore MBF in hypoperfused regions. What will decide the MBF after revascularization? In the regions where the prebypass MBF was normal, negligible increases in MBF were observed. Postbypass MBF depends upon the severity of infarction. The mean MBF after revascularization was significantly lower in areas with Q wave infarction than in areas with non-Q wave infarction and in noninfarcted areas. The function of coronary bypass grafts might affect the MBF value after revascularization. This study showed that the adequacy of mammary artery grafts in restoring MBF was the same as that of saphenous vein grafts as a whole. Assessing graft function by measuring MBF is especially useful in the cases of mammary artery grafts." The mammary artery is dissected as a pedicle containing two accompanying veins, endothoracic fascia, and adjacent tissue. Thus it is difficult to measure the flow of mammary artery grafts by an electromagnetic flowmeter. One patient who received a mammary artery graft to the LAD showed a low postbypass MBF. His postoperative graft angiogram showed stenosis at the anastomosis. Robicsek" reported a simple test to determine the efficiency of mammary artery grafts. It involved measuring the temperature increase in the grafted area and was not a quantitative method of evaluating graft function. Potential problems accompanying intraoperative measurements should be considered. Many factors other than coronary stenosis, including heart rate, myocardial contractility, wall tension, and hematocrit value, can markedly alter coronary blood flow." Surgical trauma, anesthesia, and drugs might have influenced MBF. The heart rate and aortic pressure were not markedly
different from the values obtained in the unanesthetized state. No' patient had anemia, hypoxia, or hypotension. The factors of surgical trauma and anesthesia should have influenced the ischemic and nonischemic patients similarly. After CPB, hemodilution was seen in the ischemic and nonischemic patients in the same degree. Inotropic agents were administered in most patients with ischemic and nonischemic heart disease after CPB. Although nitroglycerin was administered in all patients with ischemic heart disease, the hemodynamics before and after CPB showed no significant difference between the ischemic and nonischemic patients. Therefore, we thought that revascularization was the main factor in the change of MBF. In conclusion, the present study showed that measurement of MBF by the electrolytic hydrogen clearance method was useful to evaluate myocardial perfusion during coronary bypass operations. It provided a direct and quantitative method of evaluating myocardial perfusion and could be used to determine the efficiency of coronary bypass grafts in restoring MBF at the time of operation. REFERENCES 1. Cannon PJ, Dell RB, Dwyer EM Jr. Measurement of regional myocardial perfusion in man with 133xenon and a scintillation camera. J Clin Invest 1972;51:964-77. 2. Korbuly DE, Formanek A, Gypser G, et al. Regional myocardial blood flow measurements before and after coronary bypass surgery: a preliminary report. Circulation 1975;52:38-45. 3. Cannon PJ, Radioisotope studies of the regional myocardial circulation. Circulation 1975;51 :955-63. 4. Greene DG, Klocke F, Schiemert GL, Bunneli IL, Wittenberg SM, Lajos T. Evaluation of venous bypass grafts from aorta to coronary artery by inert gas desaturation and direct flowmeter techniques. J Clin Invest 1972;51 :191-6. 5. Smith SC Jr, Gorlin R, Herman MY, Taylor WJ, Collins JJ Jr. Myocardial blood flow in man: effects of coronary collateral circulation and coronary artery bypass surgery. J Clin Invest 1972;51 :2556-65. 6. Kety SS. The theory and applications of the exchange of inert gas at the lungs and tissue. Pharmacol Rev 1951;3:141. 7. Hyman ES. Linear system for quantitating hydrogen at a platinum electrode. Circ Res 1961;9:1093-7. 8. AukIand K, Bower BF, Berliner RW. Measurement of local blood flow with hydrogen gas. Circ Res 1964;14:164-87. 9. Stosseck K, Lubbers DW, Cottin N. Determination of local blood flow (microflow) by electrochemically generated hydrogen: construction and application of the measuring probe. Pflugers Arch 1974;348:225-38.
Volume 96 Number 5
Regional myocardial blood flow
November 1988
10. Koshu K, Kamiyama K, Oka N, Endo S, Takaku A, Saito T. Measurement of regional blood flow using hydrogen gas generated by electrolysis. Stroke 1982;13:483-7. II. Kawajiri F. Experimental and clinical studies of myocardial ischemia by measurement of blood flow. J Juzen Med Soc 1987;96:674-85. 12. Kawasuji M, Kawajiri F, Aoyama T, Sakakibara N, Takemura H, Iwa T. Intraoperative determination of the efficiency of mammary artery graft by hydrogen clearance method. Jpn J Thorac Surg 1986;39:624-6. 13. A reporting system on patients evaluated for coronary artery disease. Report of the Ad Hoc Committee for grading of coronary artery disease, Council on Cardiovascular Surgery, American Heart Association. Circulation 1975;51:5. 14. Maisel AS, Ahnve S, Gilpin E, et al. Prognosis after extension of myocardial infarct: the role of Q wave or non-Q wave infarction. Circulation 1985;71:211-7. 15. Mueller TM, Marcus ML, Ehrhardt JC, Chaudhuri T, Abbound FM. Limitations of thallium-201 myocardial perfusion scintigrams. Circulation 1976;54:640-6. 16. Selwyn AP, Shea MJ, Foale R, et al. Regional myocardial and organ blood flow after myocardial infarction: application of the microsphere principle in man. Circulation 1986;73:433-43. 17. Cohen MY, Yipintsoi T. Restoration of cardiac function
18.
19.
20.
21.
22.
23.
781
and myocardial flow by collateral development in dogs. Am J Physiol 1981;240:H811-9. Kinoshita M, Takayama Y, Miyazaki M, et al. Measurement of myocardial blood flow by means of hydrogen clearance curve: the comparison with radioactive microsphere method. J Jpn Coli Angiol 1981;21:183-8. Schwarz F. Correlation between the degree of coronary artery obstruction and myocardial dysfunction. In: Shaper W, ed. The pathophysiology of myocardial perfusion. Amsterdam: North-Holland Biomedical Press, 1979;30544. Sesto M, Schwarz F. Regional myocardial function at rest and after rapid ventricular pacing in patients after myocardial revascularization by coronary bypass graft or by collateral vessels. Am J Cardiol 1979;43:920-8. Obertson 18, Smith JC, Robel SB, Spencer MP, Mansfield PB, Sauvage LR. Origin of down-stream flow in nonobstructed coronary stenosis. Arch Surg 1973;107:764-70. Robicsek F. A simple test to determine the efficiency of mammary artery grafts during operation. Ann Thorac Surg 1985;39:388. Marcus ML. Metabolic regulation of coronary blood flow. In: Marcus ML, ed. The coronary circulation in health and diseases. New York: McGraw-Hill Book Company, 1983;69-73.
THoRAc CARDIOVASC SURG
1988;96:775-81
Measurement of regional myocardial blood flow Application of the electrolytic hydrogen clearance method in man To assess myocardial perfusion intraoperatively and to evaluate the adequacy of coronary bypass grafting, we measured regional myocardial blood flow by the electrolytic hydrogen clearance method in 49 patients. Group I comprised 10 patients with nonischemic heart disease and group II, 39 patients with ischemic heart disease undergoing coronary bypass grafting. Group II was subdivided according to the percent stenosis of the coronary arteries supplying the ventricular regions: group lla, <75 % stenosis; group lIb, ~75% stenosis. Mean myocardial blood flows were 154 ± 7, 145 ± 5, and 98 ± 9 m1/min/loo gm in groups I, Ila, and lIb, respectively (p < 0.01, group lIb versus groups I and Ila), Mean blood flows were 161 ± 19, 159 ± 12, 78 ± 12, and 59 ± 15 m1/min/loo gm in areas of the left anterior descending coronary artery with <50 %, 75 %, 90 %, and 99 % stenosis in group Il, In areas with a totally occluded left anterior descending coronary artery with collaterals, mean flow was 90 ± 15 m1/min/loo gm. The mean myocardial blood flows were 40 ± 7 and 100 ± 14 m1/min/loo gm in areas with anterior Q wave and non-Q wave infarction, respectively (p < 0.01). After cardiopulmonary bypass, the mean flow increased from 99 ± 11 to 150 ± 7 m1/min/loo gm in the grafted areas in group lIb (p < 0.01~ but it did not change in group I or Ila, The electrolytic hydrogen clearance method provided quantitative evaluation of myocardial perfusion and recovery from hypoperfusion by coronary bypass grafting. This method was especially useful in patients undergoing mammary artery grafting.
Michio Kawasuji, MD, Fumio Kawajiri, MD, and Takashi Iwa, MD, Kanazawa, Japan
Severe atherosclerotic obstruction in the coronary arteries can reduce regional myocardial blood flow (MBF). The hemodynamic objective of coronary bypass operations is to restore adequate blood flow to the jeopardized myocardial segments. Myocardial perfusion is estimated by electrocardiography, myocardial scintigraphy, coronary angiography, and left ventriculography before and after operation. Measurements of regional MBF by means of inert gas such as xenon 133 have been reported."! However, surgeons have few tools to evaluate myocardial perfusion at the time of operation. Methods presently used to evaluate results of coronary bypass, such as electromagnetic flow, do not measure blood flow per unit mass of myocardium. Some methods From the Department of Surgery (I), Kanazawa University School of Medicine, Kanazawa, Japan. Note: The review process for the paper was conducted by Dr. Dwight C. McGoon.
Received for publication March 18, 1987.
reported to measure regional MBF during operation are complicated and have limitations.': 5 The hydrogen clearance method has been used to measure tissue blood flow/" Stosseck, Lubbers, and Cortin" reported a hydrogen clearance method for which electrochemically generated hydrogen gas is used. Koshu and associates" modified the method and reported excellent correspondence between blood flow values obtained by their method and the hydrogen inhalation method. Experimental results indicated that regional MBF could be measured by the electrolytic hydrogen clearance method. I I We l 2 have already reported the first clinical application of this method to measure MBF. The present study was performed (I) to quantitate myocardial perfusion intraoperatively and to correlate MBF with the degree of coronary artery disease and (2) to evaluate the adequacy of myocardial tissue perfusion in the grafted areas by measuring the increase in regional MBF.
Accepted for publication April 12, 1988.
Patients and methods
Address for reprints: Michio Kawasuji, MD, Department of Surgery (I), Kanazawa University School of Medicine, 13-1 Takararnachi, Kanazawa 920. Japan.
This study was conducted on 49 patients undergoing elective cardiac operations with cardiopulmonary bypass (CPB) at Kanazawa University Medical Center between September
775
The Journal of
7 76
Thoracic and Cardiovascular Surgery
Kawasuji, Kawajiri, Iwa
Reference electrode
T 1/2 =O.67mln
. . fIO['~m"~'"'C __"" ~:
Fig. 1. A, Schematic presentation of methods for measuring intraoperative MBF. B, Time course of polarographic current caused by hydrogen oxidation and its logarithm. Hydrogen concentration increased by electrolysis of tissue water during the first 50 seconds and then decreased because of washout by blood flow. MBF was calculated from the half period of the change in polarographic current. 1985 and May 1987. They were divided into 10 patients with nonischemic heart disease (group I) and 39 patients with ischemic heart disease (group II). The patients in group I were between 31 and 62 years old. They had normal coronary arteries. Two of the patients had an atrial septal defect, four had mitral stenosis, and four had the Wolff-Parkinson-White syndrome. Operative procedures were closure of atrial septal defect in two patients, open mitral commissurotomy in two, mitral valve replacement in two, and division of the atrioventricular accessory conduction pathway in four. The patients in group II were between 6 and 69 years old. Thirteen of the patients had single-vessel coronary disease, 12 had doublevessel disease, 10 had triple-vessel disease, and four had a left main trunk lesion. The operative procedure was single coronary bypass grafting in 13 patients, double coronary bypass grafting in 20 patients, and triple coronary bypass grafting in six patients. An internal mammary artery was used as a graft to the left anterior descending coronary artery (LAD) in 12 of the patients. Segments of the great saphenous vein were used as grafts to other coronary arteries. Coronary cineangiograms were obtained after administration of nitroglycerin (0.4 mg sublingually). The grading of coronary artery stenosis was expressed according to the reporting system for coronary artery disease by the American Heart Association." The diagnosis of myocardial infarction was established by chest pain, elevation of total creatine kinase MB level, and electrocardiographic change. Myocardial infarction was divided into two groups: Q wave and non-Q wave infarction." The diagnosis of anterior Q wave infarction was established by development of a Q wave of 0.04 second or more in any precordial lead. The diagnosis of anterior non-Q wave infarction was established by electrocardiographic criteria as persistent new T wave or ST segment depression, or both, and absence of new Q wave. All patients were anesthetized with nitrous oxide and fentanyl, intubated, and the lungs ventilated with a ventilator. The studies were performed during the operation, with a
midsternal incision for cardiac exposure. Radial artery pressure, central venous pressure, and the electrocardiogram were monitored continuously on a Hewlett-Packard polygraph recorder (Hewlett-Packard Company, Andover, Mass.). MBF was first measured before CPB. Myocardial preservation through a single period of aortic clamping was achieved by administration of a cold (4° C) crystalloid potassium cardiaplegic solution. MBF was measured at cardiac standstill just after infusion of a cardioplegic solution. Intracardiac repair or coronary bypass grafting was then performed. MBF was again measured after weaning from CPB and stabilization of hemodynamics. Regional MBF was measured with an electrolytic hydrogen clearance tissue blood flowmeter RBF-2 and a data analyzer BDA-I-2 (Biomedical Science Inc., Kanazawa, Japan). The electrode was a Teflon-coated bipolar electrode with a diameter of 0.5 mm and contained a platinum-iridium wire of 130 JIm (electrode A) and a platinum wire of 300 JIm (electrode B). This needle electrode was inserted into the anterior wallsof the left and right ventricles (Fig. I). A reference electrode was connected to the subcutaneous tissue. A direct current of 50 to 100 JIA was passed for 50 seconds between electrode B and the reference electrode for electrolysis of tissue water. Hydrogen gas was generated as follows: 2H 30+ + 2e-
--+
Hzt + 2HzO
Simultaneously, the apparatus monitored the changes in hydrogen concentration by recording the polarographic current caused by hydrogen oxidation at electrode A as follows: Hz --+ 2H+ + 2eThe half period of the change in hydrogen concentration was calculated. The true regional blood flow (F) was calculated from the equation '0: F = 69.3j(T A F B = 69.3jTB
-
FB) (mljminjloo gm)
Volume 96 Number 5 November 1988
Regional myocardial blood flow
NS
Ol o Q
.... c
~ 200
E ~
1. 2
.Q u.
J
"tl
0 0
iii
.
III "tl III
•• •
100
u 0
••
P
•
..•
• • I Q •
-..-.. • •
• I
• •
...T I
--2 •
J
>-
• II •
~
--Q Cl
• 0'----=----=--=----,----=-------' Group I Group IIa Group IIb Fig. 2. MBF before CPB for all patients. The mean value of MBF was lower in group lIb than groups I and lIb (p < 0.01). where T A is the half period of the hydrogen washout curve with blood flow, Ta is the half period of the hydrogen washout curve without blood flow, and Fa is the false blood flow by hydrogen gas diffusion. Tawas measured at the beginning of cardioplegic arrest, that is, during cessation of coronary circulation. Cumulative data were expressed as mean ± standard error of the mean. Statistical comparisons were made by Student's t test to detect significant (p < 0.05) differences between the measured variables. The paired t test was used to compare prebypass to postbypass changes within the same patient, and the unpaired t test was used to compare the differences between the groups.
Results MBF before CPB. Fig. 2 shows the MBF data for all patients before CPB. The mean MBF for group I was 154 ± 7 ml/min/l00 gm. The mean MBF in the anterior walls of the right and left ventricles was 154 ± 9 and 152 ± 11 ml/min/l00 gm, respectively (no significant difference [NS]). The mean MBF for group I, which showed a narrow range of deviation, was considered to be the normal value for MBF by this method. The areas supplied by coronary arteries with <75% diameter stenosis and s; 75% diameter stenosis constituted groups IIa and Ilb, respectively. The mean MBF for group IIa, 145 ± 5 ml/min/100 gm, was not significantly different from the normal value. The mean MBF for group Ilb, 98 ± 9 ml/min/l00 gm (range 5 to 215), was significantly lower than that for group I or IIa (p < 0.01). The hemodynamic data before CPB showed no significant difference between groups I and II (Table I).
....c
E200 ....
!
~
.2 u,
"tl 0 0
III
I I
P
• a
•
•
_100 .!!!
0
•
I
• •
:~
U
0
>:!:
<50
8
I
•
.
"tl III
collateral $
• collateral e
•
·1 ·2 T
777
• •• • 90
0 0
if
..
:f 0 0
8
75 99 100% LAD percent stenosis
Fig. 3. Relationship of MBF to preoperative angiographic assessment of coronary stenosis. The region of the LAD was selected for this study. Mean value of MBF was low in the areas of LAD with ~ 90% stenosis.
Fig. 3 shows the relationship of MBF to the angiographic assessment of the percent coronary stenosis. The anterior wall of the left ventricle was selected for this study. The mean MBFs in the areas supplied by an LAD with <50%, 75%, 90%, and 99% stenosis were 161 ± 19, 159 ± 12, 78 ± 12, and 59 ± 15 nil/min/ 100 gm, respectively. The mean MBF in the areas supplied not by totally occluded LAD but by collateral vessels from the right coronary artery was 90 ± 15 ml/min/l00 gm. Fig. 4 shows the relationship of MBF to the presence of myocardial infarction in group lIb. The anterior wall of the left ventricle supplied by the LAD was again selected for this study. The mean MBF in areas with Q wave infarction, 40 ± 7 ml/min/100 gm, was significantly lower than that in areas with non-Q wave infarction, 100 ± 14 ml/min/100 gm, or that in noninfarcted areas, 120 ± 11 nil/min/ 100 gm. Severity of LAD stenosis resulting in Q wave infarction was 90% in two patients, 99% in three patients, and 100% with collaterals in three patients. Severity of LAD stenosis resulting in non-Q wave infarction was 90% in three patients, 99% in one patient, and 100% with collaterals in four patients. MBF after CPB. MBF after CPB was measured in 57 areas; 10 areas in group I, 16 areas in group IIa, and 31 areas in group lIb. Table I shows the cumulative MBF and hemodynamic data for groups I and II. Before CPB there was no difference in MBF between
The Journal of Thoracic and Cardiovascular Surgery
7 7 8 Kawasuji, Kawajiri, Iwa
Table I. MBF and hemodynamic data before and after CPB Group II Group 110
Group I MBF (mi/min/IOO gm) Before CPB
NS
153 ± 12
After CPB Heart rate (beats/min) Before CPB
NS After CPB MAoP (mm Hg) Before CPB
NS After CPB CVP (mm Hg) Before CPB
NS After CPB Hematocrit (%) Before CPB After CPB
[151 ± 5
t
~NS~
~NS~
Group lib
147 ± II J
~*~
149 ± 10
~NS~
NS
[91.7 ± 5.8
~NS~
80.2 ± 3.8J
94.4 ± 3.9
~NS~
95.0 ± 2.8
C5.8 ± 3.6
~NS~
79.2 ± 2.0J
77.3 ± 2.1
~NS~
75.5 ± 1.9
[9.3 ± 1.2
~NS~
7.4 ± 0.6J
8.8 ± 0.9
~NS~
9.2 ± 0.8
[39.9 ± 1.5
~NS~
38.0± 0.7J
32.1 ± 1.5
~NS~
32.4 ± 0.8
99 ± 1]
t
150 ± 7
t
t
t
MBF. Myocardial blood flow; CPB. cardiopulmonary bypass; NS. not significant; MAoP, mean aortic pressure; CVP. central venous pressure. •p
tp
< 0.05. < 0.01.
groups I and IIa, whereas the mean MBF was significantly lower in group lIb than in group I or IIa (p < 0.01). The mean MBF increased from 99 ± II to 150 ± 7 rnl/min/l00 gm after revascularization in Group lIb (p < 0.05) and approximated a normal value. In group I the hematocrit value was lower after CPB, but there was no significant change in hemodynamics after CPB. In Group II the heart rate and central venous pressure were significantly higher and mean systemic blood pressure and hematocrit value were significantly lower after CPB than before CPB. Fig. 4 shows the relationship of MBF change by revascularization to the severity of anterior myocardial infarction. The mean MBF increased from 39 ± 9 to 127 ± 15 rnl/min/l00 gm in areas with Q wave infarction (p < 0.01). It increased from 95 ± 18 to 143 ± 18 rnl/min/l00 gm in areas with non-Q wave infarction (NS) and from 120 ± 21 to 161 ± 7 mlj min/IOO gm in noninfarcted areas (p < 0.01). The mean MBF after revascularization was significantly higher in noninfarcted areas than in areas with Q wave infarction (p < 0.01). The mean MBF was compared between the areas of the LAD grafted by saphenous vein grafts and internal mammary artery grafts. Before revascularization the mean MBF was 103 ± 12 mljmin/l00 gm (n = 19) in the areas of saphenous vein grafts and 93 ± 16 ml/
min/l00 gm (n = 12) in the areas of mammary artery grafts (NS). After revascularization the mean MBF was 150 ± 9 rnl/min/l00 gm in the areas of saphenous vein grafts and 150 ± 12 rnl/min/l00 gm in the areas of mammary artery grafts. Discussion Measurement of regional MBF provides a direct and quantitative method of evaluating myocardial perfusion. Many approaches, including radioactive isotope uptake methods," radioactive microsphere methods,": 17 and inactive gas clearance methods.v-" have been applied to measure regional MBF. However, intraoperative measurements of MBF have a number of limitations. Hydrogen clearance methods have been used to measure tissue blood flow. Hyman' introduced a quantitative polarographic technique of measuring tissue hydrogen concentration, and Aukland, Bower, and Berliner" modified this technique. Blood flow rates in animal kidneys and hearts were measured by the hydrogen clearance method and found to correlate well with those obtained by venous outflow collection. Kinoshita and associates" measured MBF by the hydrogen clearance method and reported a good correspondence between MBF values obtained by hydrogen clearance and the radioactive microsphere method. Stosseck, Lubbers, and Cottin? measured tissue blood flow by using electrochemically
Volume 96 Number 5 November 1988
Regional myocardial blood flow
--
P
01
P
0
$2
P <0.01
....C
NS
P
NS
P
~
j
0
u,
"t' 0 0
iii
.
100
NS
Ii
NS
II
~ 200
E
.Il!
779
f .•
tf
"t'
Cll U
..
1q
0
>:E
pre post QOMI
pre
post
Non-Q OMI
pre
post OMI(-)
Fig. 4. Relationship of MBF to severity of myocardial infarction and MBF change after coronary bypass grafting. The anterior wall of the left ventricle was selected for this study. Mean values of prebypass and postbypass MBF were significantly lower in the areas of Q wave infarction than the notinfarcted areas. OMI; Old myocardial infarction.
generated hydrogen gas. More recently, Koshu and associates'? modified the equation proposed by Stosseck's group and indicated that the values calculated by their equation correlated well with those obtained by the hydrogen inhalation method. The electrolytic hydrogen clearance blood flowmeter generates hydrogen gas electrochemically and records its clearance by monitoring the change in polarographic current caused by hydrogen oxidation. As a preliminary experiment, one of us" used the electrolytic hydrogen clearance method to measure the MBF in canine hearts and confirmed that this method was applicable for measuring MBF. We'thave reported that this method is safe and can be adapted for use in patients. In the clinical setting, circulatory arrest was not obtained. Therefore, T B was measured at cardioplegic arrest during CPB. Values of F B obtained in this manner correlated with those obtained in the experiments. The electrolytic hydrogen clearance method has many advantages: Measurements can be done without the use of hydrogen gas. Less time is required than for the inhalation method. Regional tissue blood flowcan be measured repeatedly in a given patient. Tissue blood flow is determined as an absolute value (milliliters per minute per 100 gm). There are no harmful effects such as electrical hazard or mechanical damage by a needle electrode. The present study showed that measurement
of MBF by the electrolytic hydrogen clearance method is useful for assessing myocardial perfusion in humans. In the present study, group II was divided into groups IIa and lIb according to the degree of coronary artery stenosis <75% and ;;;;75%, because 75% stenosis is one of the clinical criteria for bypass grafting. The mean value of MBF was normal in group IIa but was significantly decreased in group lIb. Smith and associates' recorded intraoperative Xenon 133 clearance curves and demonstrated that MBF was reduced distal to the coronary lesion with >80% reduction in diameter. The present study showed reduction in MBF in the areas of LAD with ;;;; 90% stenosis. It is noteworthy that the mean MBF in the areas supplied only by collateral vessels was as much as the mean MBF in the areas of LAD with 90% stenosis. This observation is in accord with studies by Schwarz" and Sesto and Schwarz," who examined the effect of collaterals on myocardial function and found that collaterals effectively produced partial revascularization and functionally converted total occlusion to approximately a 90% stenosis. The present study showed that the mean MBF was significantly lower in areas with Q wave infarction than in areas with non-Q wave infarction. This study demonstrated a low, narrow range of MBF in areas with Q wave infarction and a relatively high and wide range of
7 80
The Journal of Thoracic and Cardiovascular Surgery
Kawasuji, Kawajiri, Iwa
MBFs in areas with non-Q wave infarction. This result and postbypass MBF data support that a Q wave infarction is a more homogenous infarction that involves less surviving myocardium than a non-Q wave infarc-
tion." Measurement of MBF before and after revascularization makes it possible to assess intraoperatively the adequacy of coronary bypass graft for regional tissue perfusion. Electromagnetic flow measures only bypass graft flow irrespective of its distribution.' Highly variable graft flow might be expected, because the magnitude of the distal runoff varies and the fraction of distal flow in a grafted coronary artery depends on the graft/coronary artery cross-sectional area ratio and severity of proximal stenosis in the recipient artery." On the other hand, the hydrogen clearance method measures actual nutrient blood flow per unit mass of myocardium. The results of the present study confirm that coronary bypass grafts restore MBF in hypoperfused regions. What will decide the MBF after revascularization? In the regions where the prebypass MBF was normal, negligible increases in MBF were observed. Postbypass MBF depends upon the severity of infarction. The mean MBF after revascularization was significantly lower in areas with Q wave infarction than in areas with non-Q wave infarction and in noninfarcted areas. The function of coronary bypass grafts might affect the MBF value after revascularization. This study showed that the adequacy of mammary artery grafts in restoring MBF was the same as that of saphenous vein grafts as a whole. Assessing graft function by measuring MBF is especially useful in the cases of mammary artery grafts." The mammary artery is dissected as a pedicle containing two accompanying veins, endothoracic fascia, and adjacent tissue. Thus it is difficult to measure the flow of mammary artery grafts by an electromagnetic flowmeter. One patient who received a mammary artery graft to the LAD showed a low postbypass MBF. His postoperative graft angiogram showed stenosis at the anastomosis. Robicsek" reported a simple test to determine the efficiency of mammary artery grafts. It involved measuring the temperature increase in the grafted area and was not a quantitative method of evaluating graft function. Potential problems accompanying intraoperative measurements should be considered. Many factors other than coronary stenosis, including heart rate, myocardial contractility, wall tension, and hematocrit value, can markedly alter coronary blood flow." Surgical trauma, anesthesia, and drugs might have influenced MBF. The heart rate and aortic pressure were not markedly
different from the values obtained in the unanesthetized state. No' patient had anemia, hypoxia, or hypotension. The factors of surgical trauma and anesthesia should have influenced the ischemic and nonischemic patients similarly. After CPB, hemodilution was seen in the ischemic and nonischemic patients in the same degree. Inotropic agents were administered in most patients with ischemic and nonischemic heart disease after CPB. Although nitroglycerin was administered in all patients with ischemic heart disease, the hemodynamics before and after CPB showed no significant difference between the ischemic and nonischemic patients. Therefore, we thought that revascularization was the main factor in the change of MBF. In conclusion, the present study showed that measurement of MBF by the electrolytic hydrogen clearance method was useful to evaluate myocardial perfusion during coronary bypass operations. It provided a direct and quantitative method of evaluating myocardial perfusion and could be used to determine the efficiency of coronary bypass grafts in restoring MBF at the time of operation. REFERENCES 1. Cannon PJ, Dell RB, Dwyer EM Jr. Measurement of regional myocardial perfusion in man with 133xenon and a scintillation camera. J Clin Invest 1972;51:964-77. 2. Korbuly DE, Formanek A, Gypser G, et al. Regional myocardial blood flow measurements before and after coronary bypass surgery: a preliminary report. Circulation 1975;52:38-45. 3. Cannon PJ, Radioisotope studies of the regional myocardial circulation. Circulation 1975;51 :955-63. 4. Greene DG, Klocke F, Schiemert GL, Bunneli IL, Wittenberg SM, Lajos T. Evaluation of venous bypass grafts from aorta to coronary artery by inert gas desaturation and direct flowmeter techniques. J Clin Invest 1972;51 :191-6. 5. Smith SC Jr, Gorlin R, Herman MY, Taylor WJ, Collins JJ Jr. Myocardial blood flow in man: effects of coronary collateral circulation and coronary artery bypass surgery. J Clin Invest 1972;51 :2556-65. 6. Kety SS. The theory and applications of the exchange of inert gas at the lungs and tissue. Pharmacol Rev 1951;3:141. 7. Hyman ES. Linear system for quantitating hydrogen at a platinum electrode. Circ Res 1961;9:1093-7. 8. AukIand K, Bower BF, Berliner RW. Measurement of local blood flow with hydrogen gas. Circ Res 1964;14:164-87. 9. Stosseck K, Lubbers DW, Cottin N. Determination of local blood flow (microflow) by electrochemically generated hydrogen: construction and application of the measuring probe. Pflugers Arch 1974;348:225-38.
Volume 96 Number 5
Regional myocardial blood flow
November 1988
10. Koshu K, Kamiyama K, Oka N, Endo S, Takaku A, Saito T. Measurement of regional blood flow using hydrogen gas generated by electrolysis. Stroke 1982;13:483-7. II. Kawajiri F. Experimental and clinical studies of myocardial ischemia by measurement of blood flow. J Juzen Med Soc 1987;96:674-85. 12. Kawasuji M, Kawajiri F, Aoyama T, Sakakibara N, Takemura H, Iwa T. Intraoperative determination of the efficiency of mammary artery graft by hydrogen clearance method. Jpn J Thorac Surg 1986;39:624-6. 13. A reporting system on patients evaluated for coronary artery disease. Report of the Ad Hoc Committee for grading of coronary artery disease, Council on Cardiovascular Surgery, American Heart Association. Circulation 1975;51:5. 14. Maisel AS, Ahnve S, Gilpin E, et al. Prognosis after extension of myocardial infarct: the role of Q wave or non-Q wave infarction. Circulation 1985;71:211-7. 15. Mueller TM, Marcus ML, Ehrhardt JC, Chaudhuri T, Abbound FM. Limitations of thallium-201 myocardial perfusion scintigrams. Circulation 1976;54:640-6. 16. Selwyn AP, Shea MJ, Foale R, et al. Regional myocardial and organ blood flow after myocardial infarction: application of the microsphere principle in man. Circulation 1986;73:433-43. 17. Cohen MY, Yipintsoi T. Restoration of cardiac function
18.
19.
20.
21.
22.
23.
781
and myocardial flow by collateral development in dogs. Am J Physiol 1981;240:H811-9. Kinoshita M, Takayama Y, Miyazaki M, et al. Measurement of myocardial blood flow by means of hydrogen clearance curve: the comparison with radioactive microsphere method. J Jpn Coli Angiol 1981;21:183-8. Schwarz F. Correlation between the degree of coronary artery obstruction and myocardial dysfunction. In: Shaper W, ed. The pathophysiology of myocardial perfusion. Amsterdam: North-Holland Biomedical Press, 1979;30544. Sesto M, Schwarz F. Regional myocardial function at rest and after rapid ventricular pacing in patients after myocardial revascularization by coronary bypass graft or by collateral vessels. Am J Cardiol 1979;43:920-8. Obertson 18, Smith JC, Robel SB, Spencer MP, Mansfield PB, Sauvage LR. Origin of down-stream flow in nonobstructed coronary stenosis. Arch Surg 1973;107:764-70. Robicsek F. A simple test to determine the efficiency of mammary artery grafts during operation. Ann Thorac Surg 1985;39:388. Marcus ML. Metabolic regulation of coronary blood flow. In: Marcus ML, ed. The coronary circulation in health and diseases. New York: McGraw-Hill Book Company, 1983;69-73.