Coronary Artery Bypass Without CardioPulmonarv Bvpass Albert J. Pfister, MD, M. Salah Zaki, MD, Jorge M. Garcia, MD, Luis A. Mispireta, MD, Paul J. Corso, MD, Anjum G. Qazi, MD, Steven W. Boyce, MD, Thomas R. Coughlin, Jr, MD, and Patricia Gurny, MD Washington Hospital Center, Washington, DC, and the Washington Adventist Hospital, Takoma Park, Maryland
The purpose of this article is twofold: to describe our technique for performing coronary artery bypass grafting without cardiopulmonary bypass (off pump) and to demonstrate that this operation is safe, in terms of mortality and certain indices of morbidity. Very little has been published in regard to off-bypass operations. From 1985 through 1990, 220 patients underwent operation off bypass; 220 on-pump controls were retrospectively matched for number of grafts, left ventricular function, and date of operation. Groups were compared in terms of mortality and ten indicators of morbidity. The same analysis was performed for ten subgroups. We found no statistically significant difference between groups in mortality
(off pump, 1.4% [3/2201; on pump, 2.4% [5/2201), which held across all subgroups. Patients undergoing operation off pump required blood far less often (not transfused: off pump, 72.7%[160/2201; on pump, 54.6%[116/2201;p = 0.005 by Fisher’s exact test), and the low output state occurred statistically less frequently off pump (offpump, 5.5% [12/2201; on-pump, 12.7% [28/2201; p = 0.01 by Fisher’s exact test). Further research should be directed to which subgroups can be operated on to advantage off pump and which, if any, groups of patients should be confined to on-bypass operations.
D
bypass machine itself, hypothermia, or myocardial preservation techniques that were employed. Federico Benetti [5] reported on 30 operations done off bypass between 1980 and 1983, in which the patients were selected on the basis of arterial caliber and site, without regard to clinical state or left ventricular function. It was recently reported [6] that Doctor Benetti has extended his experience in off-bypass procedures, prompted in part by limited resources and funds available in Argentina. In this report we will describe our technique for CABG without CPB. By considering a variety of postoperative risks we will compare our results with the outcomes in a matched control group of patients operated on with CPB. We will show that this technique is safe and useful despite the theoretical loss of cardioprotection, and that certain subsets of patients may have reduced morbidity and mortality when having operation off pump.
uring the nearly three decades of coronary artery bypass grafting (CABG) experience in the United States there has been little interest in operating without cardiopulmonary bypass (CPB). Coronary artery bypass grafting with CPB has had enormous success, and the relative ease of working on the arrested heart undoubtedly makes most cardiac surgeons favor CPB in most instances. Cardiopulmonary bypass has been associated with several adverse effects. Included among these are mechanical trauma to blood components, activation of complement, impaired hemostasis, and diminished oxygen delivery. In certain cases CABG may be safer without CPB. Coronary artery bypass grafting without CPB was successfully performed on dogs in the early 1970s [l].Ankeney [2] reported to The Society of Thoracic Surgeons in 1972 a study of 143 patients who underwent CABG without CPB. Three years later Trapp and Bisarya [3] described their off-bypass technique used on 63 patients. In 1984, Akins and associates [4] studied the interventricular septal motion in 22 postoperative patients having had bypass grafts, 11of whom were operated on without CPB. Although the 11 patients who underwent bypass grafting without CPB did not have the abnormal septal motion observed in the cohort having bypass grafting with CPB, it is unclear whether this observation was related to the Presented at the Twentyeighth Annual Meeting of The Society of Thoracic Surgeons, Orlando, FL, Feb 3-5, 1992. Address reprint requests to Dr Pfister, 1706 New Hampshire Ave, NW, Washington, DC 20009.
0 1992 by The Society of Thoracic Surgeons
(Ann Thorac Surg 1992;54:1085-92)
Material and Methods Patients Patients whose operation was performed off bypass were selected according to certain criteria. Only those patients requiring grafts in the right coronary artery or left anterior descending coronary artery systems were eligible. Circumflex grafts are performed by torquing the heart for exposure, and our group has elected to do those on bypass. Among those eligible, our group preferentially of patients off pump: operates on the (1) patients who need only grafts in the left anterior descending and right coronary artery systems; (2) patients with heavily calcified aortas, for whom the large cannulas 0003-4975/92/$5.00
1086
PFISTER ET AL CABG WITHOUT CPB
pose an added risk of dissection, rupture, or embolization; (3) patients who have had previous cardiac operations; (4) patients with prior cerebrovascular accidents; (5) patients on dialysis or with very impaired renal function; and (6) Jehovah's Witnesses. Our patients were referred from urban, suburban, and rural areas in the mid-Atlantic states and were operated on in two hospitals, a 900-bed inner-city private hospital and a 300-bed suburban community hospital. Our experimental group is composed of 220 patients who had CABG alone during the period 1985 to 1990 (with only 9 of the operations having been done before 1987) without being put on CPB. The average age of this group is 61 years, with a range of 36 to 88 years. Each patient had one or two grafts in some combination of right coronary artery, left anterior descending, and diagonal, and no patient had any other operation performed within the same anesthesia time as the CABG procedure. The total number of new grafts done in this group is 291. Between 1985 and 1990 a total of 8,336 patients had CABG done by our group. Of these, 1,733 patients had one or two grafts performed on CPB. From this subgroup, we selected a control group of 220 patients with 291 new grafts who underwent CABG alone, on bypass, between the years 1985 and 1990. The control group was matched for year of operation, number of new grafts, and left ventricular function. In this group of patients undergoing operation on bypass, the mean pump time was 32 minutes (range, 13 to 253 minutes) and the mean aortic cross-clamp time was 17 minutes (range, 0 to 66 minutes). The average age of the control group is 59 years with a range of 32 to 86 years. We considered ten variables affecting postoperative morbidity and mortality: hypertension, diabetes, ventricular function, renal disease, history of stroke, age, sex, reoperation, number of grafts, and graft site. We selected 11 dependent variables that we believed were related to mortality and short-term morbidity: in-hospital mortality, perioperative myocardial infarction, low output state, mechanical support, postoperative confusion, postoperative stroke, blood transfusions, superficial chest wound infections, mediastinitis, days in the intensive care unit, and days in the hospital.
Anesthetic and Surgical Technique Anesthetic management of patients having CABG done off bypass does not differ from that employed in those patients having operation on pump. It is important to have an inotrope and an antiarrhythmic in readiness because the technique involves a certain amount of cardiac manipulation. The anesthesiologist must maintain the patient in an adequate volume status before the start of the distal anastomosis. Close monitoring of the electrocardiogram for ischemia is essential. In the surgical procedure itself, it is helpful to tack up the cut edge of the pericardium to the chest wall to provide a firm surface on which to prop the heart as the operative field is immobilized. If it is necessary to support the heart, bolsters of laparotomy pads soaked in warm saline solution are most appropriate.
Ann Thorac Surg 1992;54:108592
There are several guidelines specific to the main sites of distal anastomoses. Right coronary lesions are approached by providing gentle traction on the acute margin of the heart with sutures of 2-0 silk. These sutures are then held to the drapes with hemostats to provide appropriate exposure of the right coronary artery. The artery is encircled proximal and distal to the planned arteriotomy with fine silicone rubber (Quest Medical, Inc, Dallas, TX) loops or fine silk sutures, which also serve as tourniquets to interrupt blood flow. The left anterior descending coronary artery system is approached by supporting the heart on several warm, moist laparotomy pads. Proximal and distal control is usually best obtained with small, spring-loaded Bogart clamps. Alternatively, simple pressure with the blunt end of a forceps can control bleeding when the arteriotomy is made. In general the area of the planned arteriotomy can be further immobilized with traction sutures of pledgeted suture material placed to the right and left of the planned incision. Ten thousand units of heparin is administered and allowed time to circulate before the arteriotomy is made. Gentle pressure maintained by the first assistant at sites proximal and distal to the artery aids in immobilizing it, and the anastomosis is then performed in the standard fashion. It is quite helpful to have a second assistant use a fine suction tip to keep the operative field dry. The proximal and distal clamps on the artery are released prior to snugging up the anastomosis. After completion of the procedure, the heparin is totally reversed with protamine. Should the heart demonstrate ischemic signs at any time during this procedure, the surgeon must be prepared to abandon the technique, fully anticoagulate, and place the patient on CPB.
Statistical Analysis We used Fisher's exact test to examine the effects of the two techniques on those variables with categorical data. For ratio data we employed the nonparametric MannWhitney unpaired test to determine statistical significance.
Results We eliminated renal failure and history of stroke from this analysis because there were very few cases in either group. Additionally, postoperative cerebrovascular accidents and confusion were very infrequent events, occurring equally in on-bypass and off-bypass patients. Any documented history of high blood pressure would cause a patient to be categorized as hypertensive for this study. Similarly, history of insulin use or treatment by oral hypoglycemics or diabetic diet would cause the patient to be classified as diabetic. Ventricular impairment was defined as "mild" if the ejection fraction was 0.35 to 0.45, "moderate" if 0.20 to 0.40, and "severe" if less than 0.20. The variable "redo" indicated the presence of any prior open cardiothoracic operation-nearly always previous CABG. Patients having operation off pump are not at increased risk of mortality. Because the incidence of mortality is
PFISTER ET AL CABG WITHOUT CPB
Ann Thorac Surg 1992:54:1085-92
Table 1. Mortality Off Pump On Pump
Grouv Total (Off 220/0n: 220) Hypertension (Off 115/0n: 111) Diabetic (Off 42/0n: 48) Redo (Off 47/0n: 19) Elderly (Off: 24/0n: 19) Female (Off: 57/0n: 66) Normal LV (Off 149/0n: 149) Imp LV (Off 71/0n: 71) Mildly Imp LV (Off 40/0n: 40) Mod-Sev Imp LV (Off 3110n: 31) One graft (Off 149/0n: 149) Two grafts (Off 71/0n: 71) Right graft (Off 30/0n: 31) Left graft (Off 190/0n: 189)
LV = left ventricle; Imp = impaired; NS = not significant. severely;
(%)
(%)
1.4 0 2.4 2.1 4.2 0 0.7 2.8 5.0 0 2.0 0 3.3 1.1
2.4 2.7 4.2 10.5 10.6 1.5 1.4 4.2 0 9.7 2.0 2.8 0 2.6
Mod-Sev
=
p Value NS NS NS
NS NS NS NS NS NS NS
NS NS NS NS
moderately or
low, it was not possible with a sample of this size to determine if, in fact, patients having operation off bypass had a statistically significant advantage in terms of mortality, as a visual inspection of the data might suggest (Table 1). The data in Table 2 convincingly demonstrate that patients operated on without CPB are more likely to avoid blood transfusions entirely. Only a few subgroups failed to attain statistically significant differences between onpump and off-pump groups, but all of those groups did exhibit a numerical advantage for off-pump cases in this study. Postoperative low output syndrome was defined as the condition in which inotropic drug support was required for more than 24 hours for the purpose of generating an adequate cardiac output. The incidence of low output and
Table 2. Percentage of Patients Not Transfused" Group Total Hypertension Diabetic Redo Elderly Female Normal LV Imp LV Mildly Imp LV Mod-Sev Imp IV One graft Two grafts Right graft Left graft
P
Off Pump
On Pump
72.7 74.8 69.0 70.2 54.2 63.2 75.4 66.2 67.5 64.6 72.5 73.2 60.0 74.4
54.6 52.2 39.6 47.4 36.8 43.9 58.4 46.5 50.0 41.9 61.0 40.8 48.4
The purpose of this article is twofold: to describe our technique for performing coronary artery bypass grafting without cardiopulmonary bypass (off pump) and to demonstrate that this operation is safe, in terms of mortality and certain indices of morbidity. Very little has been published in regard to off-bypass operations. From 1985 through 1990, 220 patients underwent operation off bypass; 220 on-pump controls were retrospectively matched for number of grafts, left ventricular function, and date of operation. Groups were compared in terms of mortality and ten indicators of morbidity. The same analysis was performed for ten subgroups. We found no statistically significant difference between groups in mortality
(off pump, 1.4% [3/2201; on pump, 2.4% [5/2201), which held across all subgroups. Patients undergoing operation off pump required blood far less often (not transfused: off pump, 72.7%[160/2201; on pump, 54.6%[116/2201;p = 0.005 by Fisher’s exact test), and the low output state occurred statistically less frequently off pump (offpump, 5.5% [12/2201; on-pump, 12.7% [28/2201; p = 0.01 by Fisher’s exact test). Further research should be directed to which subgroups can be operated on to advantage off pump and which, if any, groups of patients should be confined to on-bypass operations.
D
bypass machine itself, hypothermia, or myocardial preservation techniques that were employed. Federico Benetti [5] reported on 30 operations done off bypass between 1980 and 1983, in which the patients were selected on the basis of arterial caliber and site, without regard to clinical state or left ventricular function. It was recently reported [6] that Doctor Benetti has extended his experience in off-bypass procedures, prompted in part by limited resources and funds available in Argentina. In this report we will describe our technique for CABG without CPB. By considering a variety of postoperative risks we will compare our results with the outcomes in a matched control group of patients operated on with CPB. We will show that this technique is safe and useful despite the theoretical loss of cardioprotection, and that certain subsets of patients may have reduced morbidity and mortality when having operation off pump.
uring the nearly three decades of coronary artery bypass grafting (CABG) experience in the United States there has been little interest in operating without cardiopulmonary bypass (CPB). Coronary artery bypass grafting with CPB has had enormous success, and the relative ease of working on the arrested heart undoubtedly makes most cardiac surgeons favor CPB in most instances. Cardiopulmonary bypass has been associated with several adverse effects. Included among these are mechanical trauma to blood components, activation of complement, impaired hemostasis, and diminished oxygen delivery. In certain cases CABG may be safer without CPB. Coronary artery bypass grafting without CPB was successfully performed on dogs in the early 1970s [l].Ankeney [2] reported to The Society of Thoracic Surgeons in 1972 a study of 143 patients who underwent CABG without CPB. Three years later Trapp and Bisarya [3] described their off-bypass technique used on 63 patients. In 1984, Akins and associates [4] studied the interventricular septal motion in 22 postoperative patients having had bypass grafts, 11of whom were operated on without CPB. Although the 11 patients who underwent bypass grafting without CPB did not have the abnormal septal motion observed in the cohort having bypass grafting with CPB, it is unclear whether this observation was related to the Presented at the Twentyeighth Annual Meeting of The Society of Thoracic Surgeons, Orlando, FL, Feb 3-5, 1992. Address reprint requests to Dr Pfister, 1706 New Hampshire Ave, NW, Washington, DC 20009.
0 1992 by The Society of Thoracic Surgeons
(Ann Thorac Surg 1992;54:1085-92)
Material and Methods Patients Patients whose operation was performed off bypass were selected according to certain criteria. Only those patients requiring grafts in the right coronary artery or left anterior descending coronary artery systems were eligible. Circumflex grafts are performed by torquing the heart for exposure, and our group has elected to do those on bypass. Among those eligible, our group preferentially of patients off pump: operates on the (1) patients who need only grafts in the left anterior descending and right coronary artery systems; (2) patients with heavily calcified aortas, for whom the large cannulas 0003-4975/92/$5.00
1086
PFISTER ET AL CABG WITHOUT CPB
pose an added risk of dissection, rupture, or embolization; (3) patients who have had previous cardiac operations; (4) patients with prior cerebrovascular accidents; (5) patients on dialysis or with very impaired renal function; and (6) Jehovah's Witnesses. Our patients were referred from urban, suburban, and rural areas in the mid-Atlantic states and were operated on in two hospitals, a 900-bed inner-city private hospital and a 300-bed suburban community hospital. Our experimental group is composed of 220 patients who had CABG alone during the period 1985 to 1990 (with only 9 of the operations having been done before 1987) without being put on CPB. The average age of this group is 61 years, with a range of 36 to 88 years. Each patient had one or two grafts in some combination of right coronary artery, left anterior descending, and diagonal, and no patient had any other operation performed within the same anesthesia time as the CABG procedure. The total number of new grafts done in this group is 291. Between 1985 and 1990 a total of 8,336 patients had CABG done by our group. Of these, 1,733 patients had one or two grafts performed on CPB. From this subgroup, we selected a control group of 220 patients with 291 new grafts who underwent CABG alone, on bypass, between the years 1985 and 1990. The control group was matched for year of operation, number of new grafts, and left ventricular function. In this group of patients undergoing operation on bypass, the mean pump time was 32 minutes (range, 13 to 253 minutes) and the mean aortic cross-clamp time was 17 minutes (range, 0 to 66 minutes). The average age of the control group is 59 years with a range of 32 to 86 years. We considered ten variables affecting postoperative morbidity and mortality: hypertension, diabetes, ventricular function, renal disease, history of stroke, age, sex, reoperation, number of grafts, and graft site. We selected 11 dependent variables that we believed were related to mortality and short-term morbidity: in-hospital mortality, perioperative myocardial infarction, low output state, mechanical support, postoperative confusion, postoperative stroke, blood transfusions, superficial chest wound infections, mediastinitis, days in the intensive care unit, and days in the hospital.
Anesthetic and Surgical Technique Anesthetic management of patients having CABG done off bypass does not differ from that employed in those patients having operation on pump. It is important to have an inotrope and an antiarrhythmic in readiness because the technique involves a certain amount of cardiac manipulation. The anesthesiologist must maintain the patient in an adequate volume status before the start of the distal anastomosis. Close monitoring of the electrocardiogram for ischemia is essential. In the surgical procedure itself, it is helpful to tack up the cut edge of the pericardium to the chest wall to provide a firm surface on which to prop the heart as the operative field is immobilized. If it is necessary to support the heart, bolsters of laparotomy pads soaked in warm saline solution are most appropriate.
Ann Thorac Surg 1992;54:108592
There are several guidelines specific to the main sites of distal anastomoses. Right coronary lesions are approached by providing gentle traction on the acute margin of the heart with sutures of 2-0 silk. These sutures are then held to the drapes with hemostats to provide appropriate exposure of the right coronary artery. The artery is encircled proximal and distal to the planned arteriotomy with fine silicone rubber (Quest Medical, Inc, Dallas, TX) loops or fine silk sutures, which also serve as tourniquets to interrupt blood flow. The left anterior descending coronary artery system is approached by supporting the heart on several warm, moist laparotomy pads. Proximal and distal control is usually best obtained with small, spring-loaded Bogart clamps. Alternatively, simple pressure with the blunt end of a forceps can control bleeding when the arteriotomy is made. In general the area of the planned arteriotomy can be further immobilized with traction sutures of pledgeted suture material placed to the right and left of the planned incision. Ten thousand units of heparin is administered and allowed time to circulate before the arteriotomy is made. Gentle pressure maintained by the first assistant at sites proximal and distal to the artery aids in immobilizing it, and the anastomosis is then performed in the standard fashion. It is quite helpful to have a second assistant use a fine suction tip to keep the operative field dry. The proximal and distal clamps on the artery are released prior to snugging up the anastomosis. After completion of the procedure, the heparin is totally reversed with protamine. Should the heart demonstrate ischemic signs at any time during this procedure, the surgeon must be prepared to abandon the technique, fully anticoagulate, and place the patient on CPB.
Statistical Analysis We used Fisher's exact test to examine the effects of the two techniques on those variables with categorical data. For ratio data we employed the nonparametric MannWhitney unpaired test to determine statistical significance.
Results We eliminated renal failure and history of stroke from this analysis because there were very few cases in either group. Additionally, postoperative cerebrovascular accidents and confusion were very infrequent events, occurring equally in on-bypass and off-bypass patients. Any documented history of high blood pressure would cause a patient to be categorized as hypertensive for this study. Similarly, history of insulin use or treatment by oral hypoglycemics or diabetic diet would cause the patient to be classified as diabetic. Ventricular impairment was defined as "mild" if the ejection fraction was 0.35 to 0.45, "moderate" if 0.20 to 0.40, and "severe" if less than 0.20. The variable "redo" indicated the presence of any prior open cardiothoracic operation-nearly always previous CABG. Patients having operation off pump are not at increased risk of mortality. Because the incidence of mortality is
PFISTER ET AL CABG WITHOUT CPB
Ann Thorac Surg 1992:54:1085-92
Table 1. Mortality Off Pump On Pump
Grouv Total (Off 220/0n: 220) Hypertension (Off 115/0n: 111) Diabetic (Off 42/0n: 48) Redo (Off 47/0n: 19) Elderly (Off: 24/0n: 19) Female (Off: 57/0n: 66) Normal LV (Off 149/0n: 149) Imp LV (Off 71/0n: 71) Mildly Imp LV (Off 40/0n: 40) Mod-Sev Imp LV (Off 3110n: 31) One graft (Off 149/0n: 149) Two grafts (Off 71/0n: 71) Right graft (Off 30/0n: 31) Left graft (Off 190/0n: 189)
LV = left ventricle; Imp = impaired; NS = not significant. severely;
(%)
(%)
1.4 0 2.4 2.1 4.2 0 0.7 2.8 5.0 0 2.0 0 3.3 1.1
2.4 2.7 4.2 10.5 10.6 1.5 1.4 4.2 0 9.7 2.0 2.8 0 2.6
Mod-Sev
=
p Value NS NS NS
NS NS NS NS NS NS NS
NS NS NS NS
moderately or
low, it was not possible with a sample of this size to determine if, in fact, patients having operation off bypass had a statistically significant advantage in terms of mortality, as a visual inspection of the data might suggest (Table 1). The data in Table 2 convincingly demonstrate that patients operated on without CPB are more likely to avoid blood transfusions entirely. Only a few subgroups failed to attain statistically significant differences between onpump and off-pump groups, but all of those groups did exhibit a numerical advantage for off-pump cases in this study. Postoperative low output syndrome was defined as the condition in which inotropic drug support was required for more than 24 hours for the purpose of generating an adequate cardiac output. The incidence of low output and
Table 2. Percentage of Patients Not Transfused" Group Total Hypertension Diabetic Redo Elderly Female Normal LV Imp LV Mildly Imp LV Mod-Sev Imp IV One graft Two grafts Right graft Left graft
P
Off Pump
On Pump
72.7 74.8 69.0 70.2 54.2 63.2 75.4 66.2 67.5 64.6 72.5 73.2 60.0 74.4
54.6 52.2 39.6 47.4 36.8 43.9 58.4 46.5 50.0 41.9 61.0 40.8 48.4
