Effects of vasoactive drugs on flows through left internal mammary artery and saphenous vein grafts in man

J

THORAC CARDIOVASC SURG

1991;102:730-5

Effects of vasoactive drugs on flows through left internal mammary artery and saphenous vein grafts • In man Vasoactive agents are commonly used in the postcardiopulmonary bypass period to elevate the mean arterial pressure of myocardial revascularization patients. Concern exists that administration of vasoactive agents in this setting may affect flow through saphenous vein and internal mammary artery grafts. Twenty-eight patients were randomly assigned to receive one of the six two-drug combinations of phenylephrine, norepinephrine, and epinephrine. After termination of cardiopulmonary bypass baseline, hemodynamic measurements and electromagnetic flow probe measurements of saphenous vein and internal mammary artery graft flow were made. The first agent was then infused to elevate mean arterial pressure 20 mm Hg. After 5 minutes of stability, hemodynamic and graft flow measurements were repeated. The infusion was terminated, 5 minutes of stability were obtained, and baseline measurements were repeated. The second agent was then infused, and measurements were repeated after a 5-minutestabilization period. Phenylephrine induced a nonsignificant increase in saphenous vein graft flow (68 ± 31 versus 81 ± 49 ml/min) and a significant decrease in internal mammary artery graft flow (40 ± 16 versus 32 ± 12 ml/min). Norepinephrine induced a significant increase in saphenous vein graft flow (SO ± 39 versus 97 ± 39 ml/min) and no significant change in internal mammary artery graft flow (44 ± 20 versus 45 ± 20 ml/min~ Epinephrine induced a significant increase in both saphenous vein (82 ± 38 versus 96 ± 40 ml/min) and internal mammary artery (38 ± 12 versus 55 ± 24 ml/min) graft flows. We conclude that administration of vasoactive agents in the postcardiopulmonary bypass period may significantly affect saphenous vein and internal mammary artery graft flows.

James A. DiNardo, MD,a Arthur Bert, MD,b Michael J. Schwartz, MD, PhD,c Robert G. Johnson, MD,d Robert L. Thurer, MD,d and Ronald M. Weintraub, MD,d

Tucson, Ariz., Providence, R.I., and Boston, Mass.

Currently both saphenous vein (SV) and internal mammary artery (IMA) are used as conduits in patients undergoing coronary revascularization. The long-term From the CharlesA. Dana Research Instituteof Beth Israel Hospital, the Departments of Anaesthesia? and Surgery," Beth IsraelHospital, Harvard Medical School, Boston, Mass., the Department of Anesthesiology, Rhode Island Hospital,R.I.,band the Department of Anesthesiology, University of Arizona Health Sciences Center, Tucson,Ariz." Received for publication Feb. 5, 1990. Accepted for publication July 9, 1990. Addressfor reprints: James A. DiNardo, MD, ClinicalAssistantProfessor of Anesthesiology, Departmentof Anesthesiology, University of Arizona Health Sciences Center, 150I North CampbellAve., Tucson,AZ 85724.

12/1/24527

730

patency of IMA grafts is superior to that of SV grafts, 1, 2 and, despite the fact that IMA grafts tend to have immediate flows one half to one third that supplied by an SV to the same coronary bed.' the flow capacity of IMA grafts can increase in response to long-term demand both immediatelyv' and over time. 6, 7 These characteristics, along with improved long-term patient survival, have made the IMA the coronary artery bypass graft of choice.' There has been recent concern that the response of engrafted IMA to exogenously administered vasoactive agents may be different from that seen in engrafted SV. 9 In particular, there has been concern that an enhanced response of engrafted IMA to vasopressors may compromise graft flow in the immediate postcardiopulmonary bypass (CPB) period." 10 To date there has been no controlled trial of the effect of commonly administered

Volume 102 Number 5 November 1991

Effectsof vasoactive drugs on venograft flows 7 3 I

200

100

180

90

160

80 70

140 c

E

120

E

100

<,

~

0

u:::

80 60 40 20

c

I ?~ •

'E

60

E

50

<,

;:

•

0 u::: 40

30 20

I

:

Control

Phenylephrine

p=0.15

10 0

0

I

I

c-o.ooa

•

•

Control

Phenylephrine

Fig. 1. Plot of SV graft flows during control and phenylephrine infusion periods for all patients. Mean ± standard deviation for both periods and p value are shown

Fig. 2. Plot of left IMA graft flows during control and phenylephrine infusion periods for all patients. Mean ± standard deviation and p value are shown.

vasoactive agentsonflows inthe engraftedIMAs and SVs in humansin the post-CPBperiod. The aim of this study wasto comparethe effecton IMA and SV graft flows in man of three vasoactive agents (phenylephrine, norepinephrine, and epinephrine) commonly used to elevate mean arterial pressure (MAP) in the post-CPB period.

obtained, control hemodynamic and graft flow measurements were made. The first vasoactive drug was then administered as an infusion to elevate MAP 20 mm Hg. When necessary atrial or atrioventricular sequential pacing was used to maintain heart rate at the control rate. After a 5-minute period of stability at the increased MAP with a steady-state infusion of the vasoactive agent, hemodynamic and graft flow measurements were repeated. The vasoactive agent was then discontinued. After 5 minutes of stability a second set of control hemodynamic and graft flow measurements were made. The second vasoactive drug was then administered as an infusion to elevate MAP 20 mm Hg. When necessary, atrial or atrioventricular sequential pacing was used to maintain heart rate at the control rate. After a 5-minute period of stability at the increased MAP with a steady-state infusion of the vasoactive agent, hemodynamic and graft flow measurements were made. Graft flow measurements. An appropriately sized squarewave electromagnetic flow probe (Cliniflow II, model FM 701D; Carolina Medical Electronics, Inc., Winston-Salem, N.C.) was used to make all left IMA and SV graft flow measurements. This device provides automatic probe calibration and utilizes input of the patient's hematocrit value to increase flow measurement accuracy. A portion ofthe left IMA was dissected free of overlying connective tissue by the surgeon to facilitate accurate placement of the probe. During the steady state, graft flows were seen to vary by only I to 2 ml/rnin. The accuracy of the flow probes was documented in a pilot study. A l-minute collection of the free flow effluent of 10 free left IMAs and SVs in a medicine cup was compared with electromagnetic flow probe measurements. The flow probes were consistently found to be accurate to within I to 2 ml/rnin. Statistical analysis. Each patient served as his own control for each of the two vasoactive agent infusions. The significance of hemodynamic and graft flowchanges was tested by Student's t test for paired differences. Differences between groups was tested by analysis of variance. Correlation was performed with Pearson's correlation coefficient (r). Probability values less than 0.05 were considered statistically significant.

Methods Study population. The protocol was approved by the investigational committee for human study at the Beth Israel Hospital in Boston. Informed consent was obtained from study subjects. Criteria for admission into the study group were (I) elective coronary artery bypass operation, (2) ejection fraction >40%, (3) a left IMA used as a single graft to the left anterior descending coronary artery, (4) an SV used as a nonsequential coronary artery bypass graft, and (5) hemodynamic stability after termination of CPB without the use of any vasoactive agents. Twenty-eight patients were studied, 26 men and 2 women. The average age was 61.7 years. Study protocol. After admission into the study each patient was randomly assigned to receive two of three possible vasoactive agents (pheynlephrine, norepinephrine, and epinephrine) in random order. Sixteen patients received a phenylephrine infusion, 21 a norepinephrine infusion, and 19 an epinephrine infusion. All patients received their morning doses of antianginal medications. Patients were anesthetized with 100% oxygen, lorazepam, pancuronium, or vecuronium and fentanyl (500 to 100 /-Lg/kg) or sufentanil (l0 to 20 /-Lg/kg). No inhalation anesthetic agents were used during the study protocol. All patients had a 20-guage Teflon catheter inserted into a radial artery and had a thermodilution pulmonary artery catheter inserted by way of the right internal jugular vein. The study commenced after discontinuation of CPB and protamine administration. At this point efforts were made to keep surgical stimulation minimal. All measurements were made with the chest open. Once hemodynamic stability was

732

The .lournal of Thoracic and Cardiovascular

DiNardo et al.

Surgery

200

100

180

90

160

80

140

70

c:

c:

·E 120 <, E 100 ~

0

u:::

·E 60 <,

E

50

~

0

80

u::: 40

60

30

40

20

20

p=0.01

0

10 0

Control

Norepinephrine

Control

Norepinephrine

Fig. 3. Plot of SV graft flowsduring control and norepinephrine infusion periods for all patients. Mean ± standard deviation for both periods and p value are shown.

Fig. 4. Plot of left IMA graft flowsduring control and norepinephrine infusion periods for all patients. Mean ± standard deviation for both periods and p value are shown.

Results For all patients the second set of controlhemodynamics and controlgraft flows was not significantly different from the first set. Furthermore, randomization produced similar control conditions in each of the three study groups (phenylephrine, norepinephrine, and epinephrine). The number of patients receiving ,a-adrenergic blockers and calciumchannelblockers in each groupwas similar. There were no differences in graft flows based on the order in which the two vasoactive agents were administered. In addition, there was no correlation between the doseof a vasoactive agent administered and graft flows; that is, patients who received high doses of a vasoactive agent to elevateMAP had graft flow responses similarto thosepatientswhoreceived lowerdoses of the samevasoactive agent. Finally, graft flow responses to all three vasoactive agents were similar in patients with both low (60 to 75 mm Hg) and high (76 to 90 mm Hg) baseline MAP. Dose requirements. The mean doseof agent required to elevatethe MAP 20 mm Hg was as follows: phenylephrine 76 ± 31 J.Lg/min or 0.87 ± 0.37 J.Lg/kg/min; norepinephrine 3.9 ± 2.8 J.Lg/min or 0.049 ± 0.036 J.Lgfkgjmin; and epinephrine 2.5 ± 1.3 J.Lg/min or 0.032 ± 0.Ql5 J.Lgfkg/min. Graft flows. Phenylephrine infusion induced an increase in flow through SV grafts. SV graft flow (mean ± standard deviation) was68 ± 31 ml/rnin during the control period and 81 ± 49 mljmin during the phenylephrine infusion period (p = 0.15). The SV graft flows during control and infusion periods for all 16 patients are shown in Fig. 1. Phenylephrine infusion

induced a decrease in flow through left IMA grafts. Left IMA graft flow (mean ± standard deviation) was 40 ± 16 mljmin during the controlperiod and 32 ± 12 ml/rnin during the phenylephrine infusion period (p = 0.008).The left IMA graft flows duringcontrol and infusion periods for all 16 patients are shown in Fig. 2. Norepinephrine infusion induced an increase in flow through SV grafts. SV graft flow (mean ± standard deviation) was 80 ± 39 mljmin during the controlperiod and 97 ± 49 ml/rnin during the norepinephrine infusionperiod (p =0.01). The SV graft flows duringcontrol and infusion periods for all 21 patientsare shown in Fig. 3. Norepinephrine infusion induced no change in flow through left IMA grafts. Left IMA graft flow (mean ± standard deviation) was 44 ± 20 ml/rnin during the controland 45 ± 20 mljmin during the norepinephrine infusion period (p = 0.8). The left IMA graft flows during control and infusion periods for all 21 patients are shown in Fig. 4. Epinephrine infusion induced an increase in flow through SV grafts. SV graft flow (mean ± standard deviation) was 82 ± 38 ml/min during the controlperiod and 96 ± 40 mljmin during the epinephrine infusion period (p = 0.03).The SV graft flows duringcontrol and infusion periods for all 19 patients are shown in Fig. 5. Epinephrine infusion inducedan increase in flow through left IMA grafts. Left IMA graft flow (mean ± standard deviation) was 38 ± 12 ml/rnin during the control period and 55 ± 24 ml/rnin during the epinephrine infusion period (p = 0.001).The left IMA graft flows duringcontrol and infusion periods for all 19 patients are shown in Fig. 6. Hemodynamics. Phenylephrine infusion increased

Volume 102

Effects of vasoactive drugs on venograft flows 7 3 3

Number 5 November 1991

MAP from 75 ± 9 to 94 ± 11 mm Hg (p = 0.0001). Heart rate during phenylephrine infusion remained unchanged comparedwithcontrol(79 ± 7versus77 ± 9 beats/min; p > 0.1). Phenylephrine infusion induced a reduction in cardiac index from 3 ± 0.8 to 2.6 ± 0.5 L/min/m 2 (p = 0.0003). There were no changes in pulmonary capillarywedgepressure(9.5 ± 3 versus10 ± 3 mm Hg) or central venous pressure (8.3 ± 3 versus 8.5 ± 3 mm Hg) during phenylephrine infusion. Phenylephrine infusion wasassociatedwith an increasein left ventricular stroke work index from 33.3 ± 7.2 to 38.8 ± 8.2 gm-m/beat/m' (p = 0.0002). Norepinephrine infusion increasedMAP from 75 ± 9 to 97 ± 10 mm Hg (p = 0.0001). Heart rate during norepinephrine infusion remained unchanged from control (82 = 6 versus82 ± 6 beats/min;p > 0.1). Norepinephrine infusion induced no change in cardiac index (2.9 ± 0.8versus2.7 ± 0.8 L/min/m2;p = 0.54).There were no changes in pulmonary capillary wedgepressure (9.1 ± 2 versus9.7 ± 3 mm Hg) or central venouspressure (8.5 ± 3 versus 8.9 ± 3 mm Hg) during norepinephrine infusion. Norepinephrineinfusion was associated with an increase in left ventricular stroke work index from 29.4 ± 9 to 40.3 ± 13.2 gm-m/beat/rrr' (p = 0.00(1). Epinephrine infusion increased MAP from 74 ± 8 to 92 ± 12 mm Hg (p = 0.0001). Heart rate during epinephrine infusion remained unchanged from control (80 ± 8 versus 80 ± 10 beats/min; p > 0.1). Epinephrine infusion induced no change in cardiac index (2.6 ± 0.5versus2.9 ± 0.8 L/min/m2;p = 0.07).There wasnochangein central venous pressure(8.4 ± 3 versus 8.4 ± 3mm Hg) during epinephrineinfusion. There was a small increase in pulmonary capillary wedge pressure during epinephrine infusion (8.3 ± 2 versus9.8 ± 2 mm Hg;p = 0.01). Epinephrineinfusion was associatedwith an increase in left ventricular stroke work index from 32.3 ± 10 to 40.4 ± 14.3 gm-m/beat/m? (p =0.0001). Discussion It iscommonly necessary to use vasoactive drugs in the post-CPB periodto stabilizethe conditionof patients who have undergonecoronaryartery revascularization. Vasoactive drugs, in addition to altering variables that determine myocardial oxygenconsumption, may have effects onthecaliberof newlyengrafted left IMA and SV grafts. Obviously a reduction in conduit caliber and a potential reduction in myocardialbloodflow are undesirablein the presence of increased myocardial oxygen consumption. Therefore we chose to study the effectsof clinically relevant doses of three vasoactive drugs on IMA and SV graft flows in the post-CPB period. Phenylephrine infusion produced a nonsignificant

200 180 160

1==

140 c

120 "E <,

E 100 ;;:

0

u:::

80 60 40 20 0

p=0.03 Control

Epinephrine

Fig. 5. Plot of SV graft flows during control and epinephrine infusion periods for all patients. Mean ± standard deviation for both periods and p value are shown.

increasein SV graft flow and a significantreductionin left IMA graft flow. Previous studies in man have demonstrated an increasein SV and IMA graft flow after bolus injection of phenylephrine. I I. 12 These increased flows paralleledincreasesin myocardialoxygenconsumptionas reflected in the rate-pressure product (heart rate X MAP). In a canine model where cardiac output, heart rate, and blood pressure were kept constant, phenylephrine infusion (2 ~g/kg/min) produced significant reductions in both IMA and SV graft flows." It was suggested that theseflow reductions weredue to constrictionoflarge epicardial vessels, of collateral vessels, or of the grafts themselves." In fact, isolated segments of both human'< 14 and canine'> SV and IMA constrict when exposed to phenylephrine. The difference betweenSV and IMA graft flow observed under similar hemodynamic conditions during phenylephrine infusion in our study suggests differences in the responsiveness of human SV and IMA to a pure a-adrenergic agent. However, the constriction responses of isolated human SV and IMA segments to phenylephrine have been shown to be similar.!' Because SV is more distensible than IMA 16 it is conceivable that elevations in MAP are more likely to offsetflow reductionsinducedby phenylephrine vasoconstriction in SV grafts compared with IMA grafts. In addition, the internal diameter of an IMA is generally much smallerthan that of an SV. With an SV, inflow can be thought of as very large relative to coronary artery outflow. Therefore identical percent reductions in graft cross-sectional area are much more likelyto be flow limiting in the smaller caliber IMA than in SV. An increase in myocardial oxygen consumption accompanied phenylephrine infusion as reflected in the

The Journal of Thoracic and Cardiovascular

734 DiNardo et al.

Surgery

100 90 80

70 c::

'E <, E ;:

60 50

0 u::: 40

30 20

I

10

p=0.001

0 Control

Epinephrine

Fig. 6. Plot of left IMA graft flows during control and epinephrine infusion periods for all patients. Mean ± standard deviation for both periods and p value are shown. increased left ventricular stroke work index. Although no evidence of myocardial ischemia was observed during phenylephrine infusion, it is conceivable that a reduction in IMA graft flow in the face of increased myocardial oxygen consumption could be detrimental to myocardial oxygen balance. Norepinephrine has been demonstrated to cause constriction of isolated segments of both humanl- 14 and canine" SV and IMA. However, greater constriction is produced by norepinephrine in isolated human SV segments than in isolated human IMA segments.':' This may be explained by the fact that while both postsynaptic a I-adrenergic and a2-receptors inducing vasoconstriction exist in SVs, only postsynaptic ai-adrenergic receptors are present in IMAs. 13 In a canine model where cardiac output, heart rate, and blood pressure were kept constant, norepinephrine infusion (0.1 J,Lgjkgjmin) produced significant increases in both IMA and SV graft flows.? Because dpjdt was not constant, it is possible that these graft flowincreases were due to an increase in myocardial oxygen consumption." In contrast to phenylephrine, norepinephrine caused no reduction in IMA graft flowin this study. This may be due to a weak vasoconstricted effect of norepinephrine on IMA. In addition, at similar heart rates and blood pressure, norepinephrine would be expected to increse myocardial oxygen consumption more than phenylephrine because of ~ ,-enhanced contractility. It is possiblethat the additional myocardial oxygen consumption requirements induced by norepinephrine infusion reduced the resistance in the native coronary circulation such that IMA graft flow remained constant despite direct vasoconstriction of the IMA.

Previous studies in man have demonstrated an increase in SV and IMA graft flow after bolus injection of epinephrine. I I , 12, 17 Dogs having undergone grafting of the IMA to the left anterior descending coronary artery, with subsequent ligation of this artery proximal to the anastomosis, demonstrate increases in IMA graft flow after bolus injection of epinephrine that appear to parallel increases in systolic blood pressure.!? When hypovolemia exists, epinephrine administration produces less dramatic increases in systolic blood pressure and decreases in IMA graft flow.18 In a canine model where cardiac output, heart rate, and blood pressure were kept constant, epinephrine infusion (0.05 J,Lgjkgjmin) produced significant increases in IMA flow and significant reductions in SV graft flows." To date the effect of epinephrine on isolated segments of SV and IMA has not been investigated. In this study epinephrine infusion increased flow through both SV and IMA graft. These flow increases may be due to increases in myocardial oxygen consumption, increases in MAP, lack of a significant graft and native coronary artery vasoconstrictive effect, or a combination of these factors. After engrafting of an IMA it is common clinical practice to infuse nitroglycerin in the post-CPB period. This practice is based on sound experimental evidence that nitroglycerin causes relaxation ofisolated caninel- 19 and human!" IMA segments. In addition, in a canine model where cardiac output, heart rate, and blood pressure were kept constant, nitroglycerin administration (1 J,Lgjkgjmin) produced significant increases in IMA graft flows." 19 Although nitroglycerin would be expected to attenuate or abolish the vasoconstriction induced by phenylephrine, norepinephrine, and epinephrine, it is impossible to draw any conclusions from this study about IMA graft flows when nitroglycerin and vasopressors are administered concomitantly. Furthermore it should be emphasized that the ability of nitroglycerin to attenuate or abolish vasoconstriction induced by phenylephrine, norepinephrine, and epinephrine is not due to a-adrenergic receptor blockade but to direct relaxation of vascular smooth muscle. In this protocol SV graft flows were studied in nonsequential grafts to the circumflex artery, or to large septal or diagonal branches. In addition, the SV graft studies revascularized areas of myocardium that had a normal or hypokinetic contraction pattern by ventriculography at the time of cardiac catheterization. Therefore it would be expected that these areas would have a large amount of recruitable, salvageable myocardium compared with akinetic and dyskinetic segments.P This makes regional myocardial oxygen consumption a more important variable in the determination of SV graft flow than it would

Volume 102 Number 5 November 1991

bein grafts to the right coronary circulation or to regions with little salvageable myocardium. In summary, administration of phenylephrine, norepinephrine, and epinephrine to elevate MAP has significant effects on human SV and IMA graft flows in the post-CPB period. In this study phenylephrine, compared with epinephrine and norepinephrine, adversely affects IMA graft flow. Administration of exogenous vasoactive agents probably affects graft flow through the interaction of several mechanisms. Administration of vasoactive agents may (1) directly change native coronary vascular resistance, (2) increase myocardial oxygen consumption and subsequently alter native coronary resistance, (3) directly change graft resistance, and (4) change graft perfusion pressure by altering systemic blood pressure. Although this study does not clearly delineate the site or sites of action of these vasoactive agents, it provides insight into the effect of these agents on graft flows in a common clini -;al setting.

REFERENCES 1. Lytle BW, Loop FD, Cosgrove DM, RatliffNB, Easley K, Taylor ·PL. Long-term (5 to 12 years) serial studies of internal mammary artery and saphenous vein coronary grafts. J THORAC CARDIOVASC SURG 1985;89:248-58. 2. Singh RN, Sosa JA, Green GE. Long-term fate of the internal mammary and saphenous vein grafts. J THORAC CARDIOVASC SURG 1983;86:359-63. 3. Flemma RJ, Singh HM, Tector AJ, Lepley D, Frazier BL. Comparative hemodynamic properties of vein and mammary artery in coronary bypass operations. Ann Thorac Surg 1975;20:619-27. 4. Lee CN, Orszulak TA, Schaff HV, Kaye MP. Flow capacity ofthe canine internal mammary artery. J THORAC CARDIOVASC SURG 1986;91:405-10. 5. Schmidt DH, Blau F, Hellman C, Grzelak L, Johnson D. Isoproterenol-inducedflowresponsesin mammary and vein bypassgrafts. J THORAC CARDIOVASC SURG 1980;80:31926. 6. Singh RN, Sosa JA. Internal mammary artery: a "live" conduit for coronary bypass. J THORAC CARDIOVASC SURG 1984;87:936-8. 7. Sing RN, Beg RA, Kay EB. Physiologicaladaptability: the secret of success of the internal mammary artery grafts. Ann Thorac Surg 1986;41:247-50.

Effects ofvasoactive drugs on venograft flows 7 3 5

8. Loop FD, Lytle BW, Cosgrove DM. New arteries for old. Circulation 1989;79(pt 2):140-5. 9. Jett GK, Arcidi JM, Dorsey LMA, Hatcher CR, Guyton RA. Vasoactive drug effect on blood flow in internal mammary artery and saphenous vein grafts. J THORAC CARDIOVASC SURG 1987;94:2-11. 10. Beavis RE, Mullany CJ, Cronin KD, et al. An experimental in vivostudy of the canine internal mammary artery and its response to vasoactive drugs. J THORAC CARDIOVASC SURG 1988;95:1059-66. 11. McCormick JR, Kaneko M, Baue AE, Geha AS. Blood flow and vasoactive drug effects in internal mammary and vein bypass grafts. Circulation 1975;52:173-80. 12. Geha AS, Krone RJ, McCormick JR, Baue AE. Selection of coronary bypass: anatomic, physiological, and angiographic considerations ofvein and mammary artery grafts. J THORAC CARDIOVASC SURG 1975;70:414-31. 13. Weinstein JS, Grossman W, Weintraub RM, Thurer RL, Johnson RG, Morgan KG. Differences in alpha-adrenergic responsiveness between human internal mammary arteries and saphenous veins. Circulation 1989;79:1264-70. 14. He G-W, Rosenfeldt FL, Buxton BF, Angus JA. Reactivity of human isolated internal mammary artery to constrictor and dilator agents: implications for treatment of internal mammary artery spasm. Circulation 1989;80(pt2): 1141-50. 15. Hei G-W, Angus JA, Rosenfeldt FL. Reactivity of the canine isolated internal mammary artery, saphenous vein, and coronary artery to constrictor and dilator substances: relevance to coronary bypass graft surgery. J Cardiovasc PharmacoI1988;12:12-22. 16. Dobrin P, Canfield T, Moran J, Sullivan H, Pifarre R. Coronary artery bypass: the physiological basis for differences in flow with internal mammary and saphenous vein grafts. J THORAC CARDIOVASC SURG 1977;74:445-54. 17. Grondin CM, Lesperance J, Bourassa MG, Campeau L. Coronary artery grafting with saphenous vein or internal mammary artery: comparison of late results in two consecutive series of patients. Ann Thorac Surg 1975;20:605-18. 18. von Segesser LK, Lehmann K, Turina M. Deleterious effects of shock in internal mammary artery anastomoses. Ann Thorac Surg 1989;47:575-9. 19. Jett GK, Arcidi JM, Hatcher CR, Abel PW, Guyton RA. Vasodilator drug effects on internal mammary artery and saphenous veingrafts. J Am Coli Cardiol1988;11 :1317-24. 20. Helfant RH, Bodenheimer MM, Banka VS. Asynergy in coronary heart disease: evolving clinical and pathophysiologic concepts. Ann Intern Med 1977;87:475-82.