Coronary bypass grafting after failed elective and failed emergent percutaneous angioplasty

J

THOR;\c CARDIOV;\SC SURG

1988;95:761-72

Coronary bypass grafting after failed elective and failed emergent percutaneous angioplasty Relative risks of emergent surgical intervention Emergency coronary artery bypass grafting after failed elective percutaneous transluminal coronary angioplasty can be performed with acceptable complication rates. Recently, however, a new class of patients with unsuccessful angioplasty has evolved with the use of thrombolytic therapy and emergent angioplasty as treatment for developing acute myocardial infarction. The efficacy of surgical intervention after failure of angioplasty in this setting has not been demonstrated. This report compares the results of coronary bypass done emergently after either failed elective or failed emergent angioplasty. Between March 1984 and September 1986; 1350 angioplasty procedures were performed at our institution, 393 for acute myocardial infarction. Of the 111 patients who came to operation, 42 had had unsuccessful elective angioplasty and 69 unsuccessful angioplasty done in the clinical setting of an evolving acute myocardial infarction detected by electrocardiographic criteria. Twenty-one of the 42 patients having unsuccessful electiveangioplasty (group nand 32 of the 69 with unsuccessful emergent angioplasty (group II) underwent emergency coronary artery bypass grafting. A retrospective nonparametric statistical comparison of the two groups was performed. Age, preoperative ejection fraction, distribution of vessels undergoing angioplasty, and number of vessels bypassed were not statistically different. AU group II patients received thrombolytic therapy, and a reperfusion catheter was used in over half the patients in each group. Three group I and six group II patients required a preoperative balloon pump, and half the patients in each group required postoperative inotropic support. One patient in group I (4.7 % ) and two patients in group II (6.2 % ) died (no significant difference). Only five patients in group I (23.8 % ) and 11 in group II (34.3 %) had enzymatic and electrocardiographic evidence of an acute myocardial infarction at discharge. Six patients in group II (15.6%) required reexploration for bleeding, versus none in group I (p = 0.04). Nonhemorrhagic complication rates, mean in-patient and acute care days, total hospital charges, and blood product utilization rates were not statistically different. These data indicate that emergency coronary artery bypass grafting can be performed when necessary in the setting of failed emergent percutaneous transluminal coronary angioplasty with results comparable to coronary bypass after failed elective angioplasty.

T. Bruce Ferguson, Jr., MD: Lawrence H. Muhlbaier, PhD,b Diane L. Salai, PA-C: and Andrew S. Wechsler, MD: Durham, N.«:

I

n recent years, technological advances in cardiology have significantly altered the treatment of certain forms of coronary artery disease. These advances have secondFrom the Departments of Surgery' and Community and Family Medicine," Duke university Medical Center. Durham. N.C. Read at the Thirteenth Annual Meeting of The Western Thoracic Surgical Association. Colorado Springs. Colo.• June 24-27. 1987. Address for reprints: T. Bruce Ferguson. Jr., MD. Box 2960. Duke University Medical Center. Durham. NC 27710.

arily impacted upon cardiovascular surgery by generating new subsets of patients with coronary disease requiring coronary artery bypass grafting. In 1977 Gruentzig, Senning, and Siegenthaler] introduced percutaneous transluminal coronary angioplasty (PTCA) as a nonsurgical means of myocardial revascularization. Currently, it has become the primary treatment modality for patients with single-vessel and some forms of two-vessel disease.i' Failure of these elective PTCA procedures created a subset of patients requiring emergent coronary artery bypass grafting. Murphy, Jones, 761

The Journal of

762

Ferguson et al.

Thoracic and Cardiovascular Surgery

Patients and methods

Group

IT

IN '321

Fig. I. Algorithm for the clinical course of the two groups analyzed in this study. All patients had unsuccessful PTCA and required surgical treatment on an emergent basis such that it was completed within 24 hours of the initial PTCA procedure.

and their colleagues'? initially described logistic and surgical techniques developed to care for these patients derived from the large PTCA experience at their institution. The experience at other centers has been similar."!' Controversy still exists as to the relative cost of each procedure, whether a designated operating room and surgical team should stand by for PTCA, and whether the time from onset of ischemia to reperfusion can be correlated with the rate and extent of postoperative myocardial infarction (MI).I2·13 Reported mortality rates are between 0% and 11% for emergent coronary bypass after failed elective PTCA, with perioperative MI rates of around 30%. Emergent coronary bypass for failed elective PTCA can thus be performed, when necessary, but with higher morbidity and mortality than is customary with elective bypass. 14. 15 In the past several years, newer interventional treatment schemes for patients with an evolvingclinical acute myocardial infarction (AMI) have generated an additional subset of patients requiring coronary bypass. Emergent PTCA performed in the setting of a clinical AMI has failed, and the patients require emergent operation." In contrast to elective PTCA failures, the efficacy and consequences of surgical intervention in this new population of patients undergoing unsuccessful emergent PTCA has not been widely reported. This retrospective analysis compares the results of coronary bypass performed emergently after failure of elective PTCA (group I) or emergent PTCA (group II).

During the period from March I, 1984, through Sept. 30, 1986, 1350 PTCAs were performed at our institution (Fig. I). Of these, 957 were scheduled, elective PTCAs. Group I (N = 21) was composed of those patients who had an unsuccessful scheduled, elective PTCA procedure and underwent emergent coronary bypass, defined as operation completed within 24 hours of the initial PTCA. Patients with unstable angina while receiving intravenous nitroglycerin were included in this group because they did not have evidence of acute ischemia or pain at the time of PTCA. Implementation of an emergency helicopter service beginning in March 1985 increased the number of patients transported to our institution with a diagnosis of clinical AMI, defined as chest pain of greater than 30 minutes' duration without response to conventional therapy. A large percentage of these patients were treated with thrombolytic therapy in transit. After triage they were taken immediately to the catheterization laboratory for emergency cineangiography followed by selective intracoronary thrombolysis or PTCA, or both, as indicated. Of the 1350 total patients, 393 underwent emergency PTCA for a clinical diagnosis of evolving AMI. Group II (N = 32) was composed of those patients who had unsuccessful emergent PTCA performed in the setting of a clinical AMI. Patients more than 6 hours into the ischemic process without successful intervention were not considered candidates for PTCA. Demographic and preoperative variables analyzed included age, sex, ejection fraction as determined at catheterization. clinical and electrocardiographic (ECG) evidence of AMI just before PTCA, thrombolytic therapy, and distribution of vessels subjected to PTCA. PTCA failure was a determination made by the intervening cardiologists. Dissection and reocclusion of the angioplasty vessel with the onset of acute chest pain, ECG changes, multiple arrhythmias, or cardiac decompensation were the most common causes of failure. After Dec. I, 1984, a reperfusion catheter was available for temporary perfusion of the myccardium distal to the angioplasty site." The intraaortic balloon pump (IABP) was inserted preoperatively for hemodynamic support when indicated, and these patients were classified as being in cardiogenic shock. Patients who could not be resuscitated with the reperfusion catheter, high-dose inotropic support, and the IABP and were at the point of death were not considered surgical candidates. Thrombolytic therapy was administered where indicated to group I patients at the time of PTCA failure and to group II patients as part of the interventional therapy for AMI. All patients in group II received intravenous or intracoronary thrombolytic therapy, or both. Criteria for PTCA were as follows: For the elective group I patients, PTCA was performed on the most significantly stenotic vessel in patients with one-, two-, and three-vessel disease. Multiple PTCAs in a single setting were not routinely performed during the time interval of this study. For the emergent group II patients, PTCA was routinely confined only to the infarct-related vessel, regardless of the extent of coronary disease. Patients with a clinical AMI and severe three-vessel disease, however, were not considered for PTCA and were referred directly for surgical treatment. Decreased ejection fraction on the ventriculogram and cardiogenic shock were not contraindications to PTCA in the emergent group.

Volume 95 Number 5

Coronary bypass after failed PTCA

May 1988

763

Table I. Emergent CABG after failed PTCA-perioperative data Group I: Elective PTCA

No. of patients Sex (M/F) Mean age (yr) 25th-75th percentile Preop. EF ('70) 25th-75th percentile AMI at admission Thrombolysis pre-PTCA AMI at PTCA Thrombolysis with PTCA Vessel treated by PTCA LAD CMA RCA LAD and RCA LAD and CMA Reperfusion catheter IABP Preop. Intraop. Cross-clamp time (min) 25th-75th percentile Bypass time (min) I 25th-75th percentile Redo CABG No. of vessels bypassed: 1 2 3 4 5 IMA grafts Single Double Postop. inotropic support No. of patients No. of days 25th-75th percentile

Group II: Emergent PTCA

p Value

21 17/4 55.9 ± 12.6 44-68 56.5 ± 8.3 54-60 0/21 (0%) 0/21 (0%) 0/21 (0%) 6/21 (28.6%)

32 24/8 55.9 ± 8.9 49.5-62.5 50.7 ± 11.5 41-59 32/32 (100%) 27/32 (84.4'l) 22/32 (68.9%) 32/32 (100%)

10/21 (47.6%) 1/21 (4.7%) 9/21 (42.9%) 1/21 (4.77<.) 0/21 (0%) 11/21 (52.4'ic)

19/32 (59.3'l) 1/32 (3.17<) 11/32 (34.4(l) 0/32 (0%) 1/32 (3.1%) 17/32 (53.1%)

3/21 (14.2%) 0/21 (0%) 29.7 ± 15.2 24-39 93.3 ± 52.9 69-114 4/21 (19.07<)

6/32 (18.7%) 2/32 (6.3'l) 25.0 ± 14.2 14-32 73.9 ± 41.8 44-98 2/32 (6.3'l)

5/21 9/21 6/21 1/21 0/21

(23.8'7c) (42.9'7c) (28.5%) (4.7%)

18/32 (56.3'7c) 8/32 (25.0'7c.) 3/32 (9.4%) 2/32 (6.3%) 1/32 (3.1%)

p = 0.048

10/21 (47.6'7c) 1/21 (4.7%)

15/32 (46.8'7c) 1/32 (3.1%)

p

= 0.948

10/21 (47.6%) 0.76 ± 0.94 0-1

18/32 (56.2%) 1.22 ± 1.54 0-2

p

= 0.285

p = 0.743 P = 0.829 p

= 0.048

p

= 0.586

p = 0.513

p = 0.439

p=0.103 p = 0.067

P = 0.425

CABG. Coronary artery bypass grafting: PTCA. percutaneous transluminal coronary angioplasty: EF. ejection fraction: AMI. acute myocardial infarction: LAD. left anterior descending; CMA. circumflex marginal artery: RCA. right coronary artery; lABP. intra-aortic balloon pump: [MA. internal mammary artery.

After determination that the PTCA had failed, the cardiac surgery and anesthesia teams on call were notified and the patient was transported to the first available operating room. Myocardial revascularization was performed as rapidly as possible. The internal mammary artery was harvested as a conduit when indicated. Systemic hypothermia (22° to 28° C), cold crystalloid cardioplegic solution, and topical hypothermic techniques were used in all but two patients. The area of myocardium at risk was grafted first when possible. Intraoperative variables analyzed included cross-clamp and cardiopulmonary bypass times, number of vessels bypassed, and number of internal mammary artery grafts. In the acute care unit, inotropic and IABP support were discontinued as the hemodynamic situation dictated. Coagulation parameters were used as a guide for specific therapeutic

intervention to treat postoperative bleeding. This therapy included packed red blood cells, fresh-frozen plasma, cryoprecipitate, platelets, aminocaproic acid, and protamine. Autotransfusion techniques were used when a large amount of postoperative hemorrhage was anticipated. ECG and cardiac isoenzyme (creatine kinase, CK-MB) data were obtained postoperatively on all patients. ECGs were interpreted by cardiologists not associated with the study. Development of new Q waves on serial postoperative ECGs was considered diagnostic of a peri operative AMI. CK-MB isoenzymes greater than 5% of total CK values or greater than 50 IU /L were considered diagnostic of infarction, obtained on three successive determinations. Lactic dehydrogenase isoenzyme measurements were not routinely performed. Lengths of stay in the hospital and acute care unit were

The Journal of

764

Thoracic and Cardiovascular Surgery

Ferguson et al.

GROUP I f--ECG---1

I-Enzyme-i

D

[[I

GROUP IT f--ECG--!

I- Enzyme-i

MI INDETERMINATE

~ NOMI

p' 0.088 (ECG) p' 0.700 (Enzyme)

I vs IT 2

3

2

3

Fig. 2. Perioperative ECG and enzymatic data for both groups, classified according to the three categories shown. There was no significant difference between the two groups. monitored, along with the total hospital charges incurred for each patient. Patient mortality, hemorrhagic and nonhernorrhagic complication, and blood-product utilization rates were determined. Mortality was defined as any hospital death. The study was performed retrospectively, with data acquisition from Duke Database for Cardiovascular Diseases and the medical record. Nonparametric univariate analyses were performed by use of the Wilcoxon rank sum test for continuous variables and x' test for discrete variables. The probability values displayed are those obtained for the univariate tests, without adjustment for multiple comparisons. The Bonferroni adjustment was used for multiple comparisons control. For multiple comparisons, a significance level of 0.05 divided by the number of multiple variables yielded a value necessary to consider the individual test significant.

Results As illustrated in Fig. 1, the failure rate for elective PTCA was 2.2% (21/957), and the failure rate for emergent PTCA was 8.1% (32/393). The 21 patients in group I and the 32 patients in group II underwent emergent coronary bypass as described earlier and formed the basis for comparison in this study. Demographic data for each of the two groups are listed in Table I. The male/female ratio and mean ages were similar in the two groups. Preoperative ejection fraction determined from the ventriculogram obtained at catheterization was lower in group II than group I, but this difference was not significantly different by univariate analysis. Both groups contained patients with one-, two-, and three-vessel disease. However, since patients who had a clinical AMI and severe three-vessel disease at catheterization were not considered for emer-

gent PTCA, there appeared to be a slightly higher percentage of cases of single-vessel disease in the group II patients than in the group I patients. PTCA-related data are also listed in Table I. With PTCA failure, six patients in group I received streptokinase or tissue plasminogen activator. Group II patients all had the diagnosis of a clinical AMI when they initially arrived for medical evaluation, based upon clinical evaluation and ECG data. Invasive therapy with intravenous streptokinase was begun on 27 of 32 patients during transport. Evaluation just before PTCA demonstrated that 100f32 patients had partial resolution of their symptoms and ECG findings, whereas the rest continued to have evidence of continuing ischemia. Five group II patients who did not receive intravenous streptokinase received intracoronary tissue plasminogen activator or streptokinase in the catheterization laboratory. Thus thrombolytic therapy was administered to all patients in group II before or in conjunction with PTCA. The distribution of vessels on which PTCA was attempted and failed was similar for the two groups (Table I). The left anterior descending coronary was most often attempted, followed in frequency by the right coronary and the left circumflex coronary arteries, respectively. In each group there was one double-vessel PTCA that failed. Temporary perfusion of myocardium distal to the PTCA site was established with a reperfusion catheter in over half the patients in each group. The catheters were kept patent with small doses of streptokinase (10,000 units every 5 to 10 minutes) and in all cases stabilized the clinical status of the patient. In one group II patient the guide wire and therefore the catheter could not be advanced beyond the dissection. There were no catheter-related complications in either group. Three patients in group I and six patients in group II had hemodynamic stability restored with the IABP placed preoperatively. Two additional patients in group II required balloon pump insertion in the operating room to be weaned from cardiopulmonary bypass. In group I, IABP support beginning in the operating room was not necessary. The distribution of vessels bypassed is shown in Table I, and the number ranged from one to five. The greatest percentage of patients in group I received two and three grafts whereas in group II over half the patients had single bypass grafting. This reflected a slightly greater percentage of patients with single-vessel disease in the emergent group. However, 9.4% of patients in this group had four or five bypasses, and the mean number of bypass grafts in each group as well as the distribution of the number of bypasses was not significantly different. Over half (52.4%) the

Volume 95 Number 5

Coronary bypass after failed PTCA

May 1988

patients in group I and 46.9% of patients in group II had a single internal mammary graft placed, whereas one patient in each group had bilateral mammary grafting. As might be expected from these data, the mean cross-clamp times and cardiopulmonary bypass times were not different between the two groups. Nineteen percent and 6.3% of group I and group II patients, respectively, were having reoperations. There was no difference in the percentage of patients requiring inotropic support or the duration of that support in the postoperative period (Table I). One patient in group I and two patients in group II with the IABP placed preoperatively did not require inotropic support in the postoperative period. There were three deaths in the total series. The one death in group I occurred in a 59-year-old man with a 54% ejection fraction who underwent PTCA of a stenotic saphenous vein graft from a previous bypass operation 13 years earlier. He had a reperfusion catheter and IABP placed in the catheterization laboratory and had three bypass grafts, but could not be weaned from cardiopulmonary bypass. Two patients died in group II. The first was a 55-year-old man who had an AMI. He was treated with intravenous streptokinase and underwent PTCA, during which the angioplasty balloon ruptured. A reperfusion catheter could not be passed beyond the obstruction. The patient was taken to the operating room and underwent single-vessel bypass grafting with saphenous vein. Eight defibrillation attempts were needed to restore stable cardiac rhythm after removal of the aortic cross-clamp, and intractable ventricular fibrillation developed 7 hours postoperatively. The second death occurred in a 58-year-old man who had an AMI and a 44% ejection fraction. He received streptokinase and single-vessel PTCA, which caused an intimal dissection. A reperfusion catheter was placed, and he underwent coronary bypass grafting with three saphenous vein grafts. He returned to the acute care unit receiving dopamine and epinephrine and died on the second postoperative day after an unexplained episode of hypotension followed by intractable supraventricular and ventricular arrhythmias. The mortality rates were not statistically different between the two groups. Perioperative MI data are illustrated in Figs. 2 and 3. The data were analyzed as follows. ECG and CK-MB results were independently categorized and labeled as (a) without evidence of perioperative MI (score I); (b) indeterminate, with nondiagnostic changes present on ECG or enzyme elevation not high enough to be diagnostic (score 2); and (c) clearly diagnostic of a perioperative MI (score 3). These results were not

75

765

GROUP I

50 25 >-

u

x: ., '" ~

~ ~

OMI

75

GROUP II

•

INDETERMINATE

~

NO MI

50 25

P-0.627 IvsII 2

3 ECG

4

5

+ ENZYME

6

Fig. 3. Composite ECG plus enzyme "scores" classified according to the categories shown. The composite scores for the indeterminate group ranged from a sum of 3 to a sum of 5. A sum of 2 indicated no evidence of either ECG or enzyme criteria for MI and a sum of 6 indicated that both ECG and enzyme criteria were indicative of perioperative MI.

significantly different between groups for either the ECG or enzyme data (Fig. 2). To select out patients who had evidence of a perioperative MI by both ECG and enzyme criteria, we combined the individual results into a composite "ECG plus enzyme score" for each patient. By these composite criteria, 23.8% of group I patients and 34.3% of group II patients had evidence of a perioperative MI (Table 11). Fig. 4 displays the perioperative MI data as a function of the clinical course of the two groups. Group I patients were without evidence of ischemia before PTCA. With failure of PTCA the 21 patients were stratified into three categories on the basis of the ECG plus enzyme data. In contrast, in all patients in group II the diagnosis of clinical AMI was made on admission. Before PTCA, but after some patients had received thrombolytic therapy, 10 of 32 had improved to the degree that their clinical AMI status was indeterminate. All patients then underwent unsuccessful PTCA and were stratified into the three categories. Of the 10 patients in the indeterminate clinical AMI status before PTCA, five had no evidence and five had indeterminate evidence of a perioperative MI. Conversely, all 22 patients in the positiveclinical AMI group before PTCA had either indeterminate or positive perioperative MI findings.

The Journal of

766

Thoracic and Cardiovascular

Ferguson et al.

Surgery

Table II. Emergent CABG after failed Pl'Cn-s-surgical results Group I: Elective PTCA Deaths Complications Hemorrhagic Nonhemorrhagic Postop. MI (ECG plus enzyme) Negative Indeterminate Positive Blood product utilization (total units) Blood/packed cells 25th-75th percentile Frozen plasma 25th-75th percentile Platelet packs 25th-75th percentile Cryoprecipitate 25th-75th percentile Hospital in-patient days 25th-75th percentile ICU days 25th- 75th percentile Total hospital charges ($) 25th-75th percentile

Group If: Emergent PTCA

1/21 (4.7'7r)

2/32 (6.2%)

p

0/21 (0.0%) 3/21 (14.3%)

6/32 (18.8%) 7/32 (21.9%)

p = 0.039 P = 0.722

5/21 (23.8%) 11/21 (52.4%) 5/21 (23.8%)

5/32 (15.6%) 16/32 (50.0%) 11/32 (34.3%)

p = 0.627

8.76 ± 4.97 4-11 5.43 ± 4.72 2-7 0.48 ± 0.60 0-1 0.48 ± 0.75 0-1 12.71 ± 7.33 9-13 1.90 ± 1.26 1-2 24.779.84 ± 8295.88 20290-27028

12.53 ± 8.70 7-16 8.70 ± 6.13 4-12 0.93 ± 1.39 0-1 1.17 ± 1.05 0-2 14.37 ± 10.24 9-15.5 4.44 ± 7.30 1-3.5 34.578.64 ± 24.692.57 21585-37232

p

=

=

0.819

0.146

p = 0.053 p = 0.362 p = 0.012 p = 0.964

p = 0.447 p

=

0.110

M I, Myocardial infarction; ECG. electrocardiogram: Enzyme: creatine kinase myocardial isoenzyme: ICC. intensive care unit.

Postoperative complications were categorized as hemorrhagic and nonhemorrhagic (Table III). Hemorrhagic complications were defined as those related to postoperative bleeding. Nonhemorrhagic complications consisted of other postoperative morbidity that contributed to a prolonged hospitalization. There were no hemorrhagic complications in group I. In contrast, six patients in group II (18.8%) were taken back to the operating room within 12 hours of operation for excessive mediastinal tube output or developing tamponade, or both. Four of the six had bleeding related to a coagulopathy, and two patients had a mechanical bleeding site that was sutured. This difference in hemorrhagic complication rates was significant (p = 0.039, Table II). In group I, 14.3% of patients had nonhemorrhagic complications. Two were pulmonary in origin and the third was cardiac tamponade, which developed on the fifth postoperative day shortly after the temporary atrial and ventricular pacing wires were removed. The patient was returned to the operating room for relief of the tamponade and recovered uneventfully. In group II, the nonhemorrhagic complication rate was 21.9%. The only IABP-related complication occurred in a woman in whom transient leg ischemia necessitated IABP removal. These rates of nonhemorrhagic complications between the two groups were not statistically different.

The mean number of units of packed red cells, plasma, platelets, and cryoprecipitate was greater in group II than in group I, which reflects the higher rate of bleeding complications in the emergent group (Table 11). These differences were not individually significantly different, however. Table II lists the mean number of hospital days, acute care unit days, and mean hospital charges to the patient or third party payer for both groups. Although the trend was for patients in group II to have longer intensive care and hospital stays, these differences were not great enough to be significant. Further, the hospital charges were not significantly different despite the rather large difference in the mean charges. Discussion Surgical revascularization of acutely ischemic myocardial tissue has been a controversial topic since the initial development of coronary bypass grafting by Johnson," Favaloro," and their associates in 1969. Very quickly, the techniques used for elective myocardial revascularization were applied to patients with transmural AMpo. 21 These initial attempts were made on patients in cardiogenic shock and yielded unacceptably poor results in terms of morbidity and mortality." With refinement of surgical techniques and the development of city-wide emergency medical teams, Mundth,"

Volume 95

Coronary bypass after failed PTCA 7 6 7

Number 5 May 1988

100r

I GROUP I (N·21)

50

~

~ 0 '" 100 ~ CT

Presentation GROUPll IN·321

PRE-PTCA

Postaperative

p·0627

;;e

INDETERMINATE

JI NO MI

50

o

Clinical AMI

Clinical AMI

ECG/Enzyme

MI

Fig. 4. Data from both groups correlating clinical status with perioperative MI results. The individual lines represent individual patients. See text for further details.

Berg," DeWood," and their associates in Spokane and Phillips and his colleagues":" in Des Moines have continued this effort, with currently reported mortality rates between 2.5% and 9%, depending on the extent of coronary disease. Critics of these studies have emphasized the absence of randomization, patient selection bias, and lack of actuarial survival data. 29. 3o Nevertheless, they demonstrate that, under certain conditions, myocardial revascularization in the setting of AMI can be performed with acceptable morbidity and mortality. Since 1980 the major focus in the management of AMI has been on limitation of infarct size and salvage of acutely ischemic myocardial tissue by active, interventional means." Catheterization of patients undergoing AMI confirmed that most acute ischemic events result from thrombogenic occlusion of a preexisting atherosclerotic lesion in a vessel supplying an area of myocardium at risk." Duration of ischemia, extent of collateral circulation, previous ischemic damage, and the myocardial oxygen supply/demand ratio are major determinants of regional and global myocardial viability.JJ·34 Of these, altering the duration of ischemia by invasive thrombolytic reperfusion techniques with streptokinase and tissue plasminogen activator had been demonstrated to be effective 60% to 80% of the time in the acute setting"" Restenosis, however, occurred in approximately 30% of those patients in whom reperfusion was achieved. 36. 37 PTCA, with or without antecedent or concomitant thrombolytic therapy, has been advocated as a more permanent nonsurgical method of

restoring flow to the ischemic area of myocardium. 38.39 Others have advocated PTCA as the optimal initial means of reperfusion." Interpretation of these data is complicated by the fact that very few prospective, randomized trials comparing these different types of interventional therapy exist." Factors still in question include the optimal timing of these procedures, the successful discrimination between myocardium which will benefit from reperfusion and that which will not because of the advanced nature of the ischemic injury, the long-term efficacy of these modes of therapy, and finally the role of surgical revascularization in this new, multimodality approach to the patient with AMI. In analogous fashion to failed elective PTCA, these interventional techniques have impacted on surgical treatment by generating a new subset of patients in whom emergent PTCA, performed in the setting of an evolving AMI, is unsuccessful. Unlike elective PTCA failure, however, operation on patients after unsuccessful emergent PTCA is associated with several potential complicating factors. These include a prolonged duration of ischemia before PTCA and therefore before surgical revascularization, the frequent presence of systemic thrombolysis, and a more unstable overall clinical situation by virtue of the fact that the entire sequence of events is a medical emergency. In addition, because of time constraints, performance of an optimal revascularization procedure including mammary grafting may be difficult." Intuitively, the risks of emergent surgical intervention

The Journal of

7 6 8 Ferguson et al.

Table m. Emergent CABG after failed PTCA-complications Hemorrhagic complications Group I: 0/21 (0%) Group II: 6/32 (18.8%) Reexploration for tamponade, chemical bleeding Reexploration for mechanical bleeding Reexploration for chemical bleeding Reexploration for chemical bleeding, Die, coagulopathy Reexploration for chemical bleeding, coagulopathy Reexploration for mechanical bleeding Nonhemorrhagic complications Group I: 3/21 (14.3%) Respiratory failure, reintubation Pneumonia, no reintubation Tamponade with pacing wire removal, postop. day 5 Group II: 7/32 (21.9%) Mediastinitis Respiratory failure, no reintubation ARDS caused by platelet transfusion reaction Respiratory failure, tracheostomy, ischemic leg from IABP Renal failure, no dialysis Bilateral pneumonia, no reintubation Acute condition within abdomen, gangrenous gallbladder on exploration DIe. Disseminated intravascular coagulopathy: ARDS, adult respiratory distress syndrome. Chemical bleeding = non-surgically correctable bleeding; mechanical bleeding = surgically correctable bleeding.

for failed emergent PTCA might be higher than those already established for surgical treatment after failed elective PTCA. The findings in this retrospective study, however, suggest that patients with unsuccessful emergent PTCA can be treated in much the same manner as patients with unsuccessful elective PTCA, with similar results. Comparison of these results with published mortality and morbidity statistics for elective coronary bypass is not valid, since in this study surgical input into the decision-making process did not begin until the PTCA had failed and the circumstances were no longer elective. It is of interest that the PTCA failure rate is different between the two groups in this study. The rate of 2.2% for elective PTCA is comparable to other reported figures.v" The failure rate for emergent PTCA in group II was four times as high, however. In addition to mechanical obstruction of the vessel, specific failure of thrombolytic therapy and the unstable clinical situation in general may account for this higher rate. It might be anticipated that the group II patients would be more difficult to manage intraoperatively than group I patients. All group II patients received thrombolytic therapy with attendant bleeding complications. The mean ejection fraction at angiography was slightly lower in this group with an evolving clinical AMI. A greater percentage of patients in group II were in

Thoracic and Cardiovascular Surgery

cardiogenic shock preoperatively, as defined by the requirement for IABP support. Despite these differences, however, the intraoperative course of both groups was essentially the same. Half the patients in each group received internal mammary artery grafts, and the aortic cross-clamp and cardiopulmonary bypass times were similar. The degree of inotropic support required to wean from bypass was not different. In this comparison, the distribution of vessels angioplastied and the distribution of number of vessels bypassed suggest that there was slightly more single-vessel disease in the emergent group than the elective group. Preoperative selectivity on the part of the cardiologist (excluding severe three-vessel disease with AMI from consideration for PTCA) and/ or intraoperative selectivity resulting from perceived time constraints could account for this. Nevertheless, these differences of distribution were not significantly different and comparison of the two groups was believed to be valid. The advantages of the reperfusion catheter in this setting have been previously demonstrated and are further confirmed in this study." The condition of more than 50% of patients in each group was stabilized with the catheter, because the area of jeopardized myocardium was reperfused much earlier than it otherwise would have been. Use of the catheter allowed increased flexibility in the manipulation of a busy elective operating schedule to accommodate the PTCA failure. Anatomic localization of the injured vessel distal to the dissection was made easier. The catheter was left in place during administration of the initial dose of cardiaplegic solution, which facilitated delivery directly to the area of myocardium that was ischemic as soon as the aorta was cross-clamped. It was then removed through the aortic clamp without difficulty. IMA grafts were placed in a higher percentage of patients in group I than has been previously reported for elective PTCA failures.v II. 14 Because of the superiority of this conduit," if a patient was in stable condition and the coronary anatomy suitable, the mammary artery was harvested either off or on cardiopulmonary bypass. The reperfusion catheter was essential for stabilization in certain patients during mammary harvesting. Similarly, even in the setting of a clinical AMI, with adequate stabilization the mammary artery was used in the group II patients without any measurable adverse effects. The development of a perioperative MI is a documented sequelae of elective PTCA failure and subsequent revascularization." II, 14 Interpretation of ECG and cardiac enzyme changes in this setting is difficult, however. To highlight those patients who either did or did not have a perioperative MI, we derived a composite

Volume 95 Number 5 May 1988

"ECG plus enzyme score" for each patient in each group (Fig. 3). A composite score of 2 indicated the absence of ECG or enzyme evidence of a perioperative MI, whereas a score of 6 was clearly diagnostic according to these criteria. The 23.8% MI rate for group I in this study is comparable to those reported in other series." In addition, 23.8% of patients in Group I had no evidence of an MI. The remaining patients had indeterminate changes, but Fig. 3 suggests that in a majority of these patients the changes were minimal. Analysis of the group II data is more complicated. It could be argued that all group II patients were destined for an MI in the absence of interventional therapy. Fig. 4 suggests that intervention even despite PTCA failure was partially successful in that 15.6% of patients had no evidence of an MI. Interestingly, the majority of these patients without evidence of a perioperative MI had some amelioration of their clinical status before PTCA. Presumably, reperfusion occurred earlier, either spontaneously or with thrombolytic therapy, in these patients than the other patients in group II. From this standpoint these patients might be expected to be at slightly less risk for development of a perioperative MI. Whether these patients represent a distinct subgroup of clinical AMI patients is unclear. The mortality rate in group I is comparable to those reported for emergent operation after elective PTCA.14,44 In our study, the mortality associated with emergent coronary bypass in the setting of failed emergent PTCA was not different from that for elective PTCA failure. By way of comparison, the in-hospital medical mortality for those patients with an AMI not undergoing PTCA was 11%.16 Not suprisingly, the only difference in postoperative complications between the groups related to hemorrhage. Six of the group II patients required reexploration, and in four of the six no mechanical bleeding site was found. Chemical coagulopathy was detected in all four patients and was thought to be related to thrombolytic therapy. Other studies from our institution have confirmed that streptokinase-induced coagulopathy is associated with additional postoperative morbidity." Finally, failure of PTCA in the setting of an AMI and subsequent emergent coronary bypass did not significantly prolong the number of in-patient days, intensive care days, or overall hospital charges. Further analyses of these three variables was limited by the small number of patients in each group. The increased occurrence of hemorrhagic complications in the group II patients probably contributed to the tendency of the three variables to be greater in this group, however. This study demonstrates that emergency coronary bypass can be performed after failed emergent PTCA

Coronary bypass after failed PTCA

769

done in the setting of an evolving MI with results similar to those for emergent bypass after failure of elective PTCA. As invasive techniques for AMI became more widespread, it is anticipated that this subset of patients admitted for emergency coronary bypass will increase. Improved surgical outcome will probably result from modifying four aspects in the management of these patients: (1) delay of PTCA if thrombolysis results in reperfusion until the thrombolytic effects have dissipated (> 12 hours); (2) improved management of systemic thrombolysis by early reversal and perhaps use of tissue plasminogen activator in lieu of streptokinase; (3) use of the reperfusion catheter whenever possible; (4) development of pharmacologic strategies to reduce adverse sequelae of reperfusion. When the surgeon's hand is forced by the circumstances, however, these data suggest that emergency coronary bypass in this setting can be performed with expectation of reasonable shortterm results. REFERENCES 1. Gruentzig A, Senning A, Siegenthaler W, Non-operative dilatation of coronary artery stenosis: percutaneous transluminal coronary angioplasty (PTCA). N Engl J Med 1979;30 I:61-8. 2. Gruentzig A. Percutaneous transluminal coronary angioplasty: six years' experience, Am Heart J 1984;I07: 818-9. 3. Block P. Percutaneous transluminal coronary angioplasty: role in the treatment of coronary artery disease. Circulation 1985;72(Pt 2):V 161-5. 4. Murphy DA, Craver JM, Jones EL, Gruentzig AR, King SB, Hatcher CR, Surgical revascularization following unsuccessful percutaneous transluminal coronary angioplasty. J THORAC CARDIOVASC SURG 1982;84:342-8. 5. Jones EL, Craver JM, Gruentzig AR, et al. Percutaneous transluminal coronary angioplasty: role of the surgeon. Ann Thorac Surg 1982;34:493-503. 6. Murphy DA, Craver JM, Jones EL, et al. Surgical management of acute myocardial ischemia following percutaneous transluminal coronary angioplasty: role of the intra-aortic balloon pump. J THORAC CARDIOVASC SURG 1984;87:332-9. 7. Jones EL, Murphy DA, Craver JM. Comparison of coronary artery bypass surgery and percutaneous transluminal coronary angioplasty including surgery for failed angioplasty. Am Heart J 1984;107:830-5. 8. Akins CW, Block Pc. Surgical intervention for failed percutaneous transluminal coronary angioplasty. Am J Cardiol 1984;53:I08C-II C. 9. Brahos GJ, Baker NH, Ewy G, et al. Aortocoronary bypass following unsuccessful PTCA: experience in 100 consecutive patients. Ann Thorac Surg 1985;40:7-10. 10. Killen DA, Hamaker WR, Reed W A. Coronary artery bypass following percutaneous transluminal coronary angioplasty. Ann Thorac Surg 1985;40:133-8.

The Journal of

770

Thoracic and Cardiovascular Surgery

Ferguson et al.

II. Pelletier LC, Pardini A, Renkin J, David PR, Hebert Y, Bourassa MG. Myocardial revascularization after failure of percutaneous transluminal coronary angioplasty. J THORAC CARDlOVASC SURG 1985;90:265-71. 12. Wilson JM, Dunn EJ, Wright CB, et al. The cost of simultaneous surgical standby for percutaneous transluminal coronary angioplasty. J THORAC CARDIOVASC SLRG 1986;91:362-70. 13. Block pe. Emergency surgery after percutaneous transluminal coronary angioplasty: he who calls the tune may have to pay the piper. Ann Thorac Surg 1985;40:1-2. 14. Golding LAR, Loop FD, Hollman JL, et al. Early results of emergency surgery after coronary angioplasty. Circulation 1986;74 (Pt 2):III-26-9. 15. Parsonnet V, Gielchinsky I, Hockberg M, Hussein SM, Fisch D, Rothfeld L. Emergency surgery after failed angioplasty [Abstract]. J Am Coil Cardiol 1987; 9:123A. 16. Stack RS, Hinohara T, Phillips HR, et at. Treatment of acute myocardial infarction with emergent coronary angioplasty [Abstract]. J Am Coil Cardiol 1987; 9:232A. 17. Hinohara T, Simpson JB, Phillips HR, et al: Transluminal catheter reperfusion: a new technique to reestablish blood flow after coronary occlusion during percutaneous transluminal coronary angioplasty. Am J Cardiol 1986; 57:684-6. 18. Johnson WD, Flemma RJ, Lepley D, Ellison EH. Extended treatment of severe coronary artery disease: a total surgical approach. Ann Surg 1969;170:460-7. 19. Favaloro RG, Effler DB, Groves LK. Severe segmental obstruction of the left main coronary artery and its divisions: surgical treatment by the saphenous vein graft technique. J THORAC CARDIOVASC SLRG 1970;43: I 0915. 20. Hill D, Kieth W J, Kelly JJ, et at. Emergency aortocoronary bypass for impending or extending myocardial infarction. Circulation 1971;43(Pt 2):1-105-12. 21. Dawson JT, Hall RJ, Hallman GL, et at. Mortality in patients undergoing coronary artery bypass after myocardial infarction. Am J Cardiol 1974;33:48-53. 22. Mundth ED, Buckley MJ, Leinbach RC, et al. Surgical intervention for the complications of acute myocardial ischemia. Ann Surg 1973;178:379-87. 23. Berg R, Selinger SL, Leonard JJ, et at. Immediate coronary artery bypass for acute evolving myocardial infarction. J THORAC CARDIOVASC SLRG 1981;81:493501. 24. DeWood MA, Heit J, Spores J, et at. Anterior transmural myocardial infarction: effects of surgical coronary reperfusion on global and regional left ventricular function. J Am Coil Cardiol 1983;1:1223-9. 25. DeWood MA, Spores J, Berg R, et at. Acute myocardial infarction: a decade of experience with surgical reperfusion in 701 patients. Circulation 1983;68(Pt 2):II8-14. 26. Phillips SJ, Kongtahworn C, Zeff RH, et at. Emergency coronary artery revascularization: a possible therapy for

27.

28.

29.

30.

31.

32.

33.

34. 35.

36.

37.

38.

39.

40.

41.

42.

acute myocardial infarction. Circulation I979;60(Pt 2):II241-50. Phillips 5J, Kongtahworn C, Skinner JR, et al Emergency coronary artery reperfusion: a choice therapy for evolving myocardial infarction. J THORAC CARDIOVASC SURG 1983;86:679-85. Phillips SJ, Zeff RH, Skinner JR, et at. Reperfusion protocol and results in 738 patients with evolving myocardial infarction. Ann Thorac Surg 1986;41: 119-28. Spencer Fe. Emergency coronary bypass for acute infarction: an unproven clinical experiment. Circulation 1983; 68:(Pt 2): I7-9. Silverman NA. The surgeon's role in the treatment of acute myocardial infarction. Surg Clin North Am 1985; 65:527-37. Gold HK, Cowley MJ, Palacios IF, et at. Combined intracoronary streptokinase infusion and coronary angioplasty during acute myocardial infarction. Am J Cardiol 1984;53: I 22C-5e. DeWood MA, Spores J, Notske R, et at. Prevalence of total occlusion during the early hours of transmural myocardial infarction. N Engl J Med 1980;303:897902. Goldman BS, Weisel RD. Surgical reperfusion of acute myocardial ischemia: a clinical review. J Cardiovasc Surg 1986; I: 167-99. Rentrop P, Blanke H, Karsh KR. Non-surgical reperfusion in acute myocardial infarction. Herz 1981;6:44-9. Krebber HJ, Schofer J, Mathey D, Montz R, Kalmar P, Rodewald G. Intracoronary thallium 201 scintigraphy as an immediate predictor of salvaged myocardium following intracoronary lysis. J THORAC CARDIOVASC 5LRG 1984;87:27-34. Gold HK, Leinbach RC, Palacios IF, et at. Coronary reocclusion after selective administration of streptokinase. Circulation 1983;68(Pt 2):1-50-4. Petrovich JA, Schneider JA, Taylor GJ, et at. Early and late results of operation after thrombolytic therapy for acute myocardial infarction. J THORAC CARDIOVASC SLRG 1986;92:853-8. Holmes DR, Smith HC, Vlietstra RE, et at. Percutaneous transluminal coronary angioplasty, alone or in combination with streptokinase therapy, during acute myocardial infarction. Mayo Clin Proc 1985;60:449-56. Erbel R, Pop T, Henrichs KJ, et at. Percutaneous transluminal coronary angioplasty after thrombolytic therapy: a prospective controlled randomized trial. J Am Coil Cardiol 1986;8:485-95. Hartzler GO, Rutherford BD, McConahay DR. Percutaneous transluminal coronary angiopIasty: application for acute myocardial infarction. Am J Cardiol 1984;53:117C21e. Jones EL. Reperfusion for evolving myocardial infarction. Medical illusion or therapeutic reality? Ann Thorac Surg 1986;41:117-8. Vanhaecke J, Flameng W, Sergeant P, et al. Emergency bypass surgery: late effects on size of infarction and

i

'Volume 95 Number 5 May 1988

43.

44.

45.

46.

47.

ventricular function. Circulation 1985;72(Pt 2):1117984. Cowley MJ, Dorros G, Kelsey SF, Van Raden M, Detre KM. Acute coronary events associated with percutaneous transluminal coronary angioplasty. Am J Cardiol 1984; 53:12C-6C. Roubin GS, Talley JD, Anderson HV, et al. Morbidity and mortality associated with emergency bypass graft surgery following elective coronary angioplasty [Abstract]. J Am Coli Cardiol 1987;9:124A. Ferguson TB, Hinohara T, Simpson J, Stack RS, Wechsler AS. Catheter reperfusion to allow optimal coronary bypass grafting following failed transluminal coronary angioplasty. Ann Thorac Surg 1986;42:399-405. Christian CB, Mack JW, Wetstein L. Current status of coronary artery bypass grafting for coronary artery atherosclerosis. Surg Clin North Am 1985;65:509-26. Lee KF, Mandell J, Rankin JS, Muhlbaier LH, Wechsler AS. Immediate versus delayed coronary grafting after streptokinase: postoperative blood loss and clinical results. J THoRAc CARDIOVASC SURG 1988;95:216-22.

Discussion (Papers by Morris (page 758( and Ferguson and associates [page 761/) Dr. Daniel J. Ullyot (San Francisco, Calif). I should like to focus on the delicate interface between surgeon and angioplasty physician. The literature on surgical standby for percutaneous transluminal coronary angioplasty (PTCA) suggests that 3% to 10% of patients undergo emergency surgical revascularization. The surgeon's role in the timing and decision to intervene surgically is largely passive. Although the hospital mortality after surgical revascularization under these circumstances is low, the myocardial infarction (MI) rate is distressingly high-30% by Q-wave criteria and perhaps 50% if enzyme data are included. Others have raised the question whether surgical intervention does anything to reverse MI under the circumstances of acute vessel occlusion after PTCA. Surgical standby may engender a false expectation of the safety of PTCA and lead to adventurism in the catheterization laboratory when the physician performing the PTCA knows that he can trigger a complex surgical undertaking automatically and on demand. I should like the authors to discuss what we think we are accomplishing by surgical intervention under these circumstances, other than covering an iatrogenic event by another intervention, and whether some patients might be better treated by nonoperative approaches. Dr. Douglas C. Morris (Atlanta, Ga.). One thing that has been stressed through the years is that, once the PTCA physician has reached 150 cases, he has gone through his learning curve. I think it does take 150 cases to get the primary success rate up. However, the learning curve extends beyond that point. Since surpassing 150 cases, I have learned how to deal with those patients whose condition is deteriorating. There are certain things that can be done to deal with this problem. I divide my patients into categories. One set of patients my surgeons do not see. The surgeons are on standby, but they

Discussion

771

have never seen the films. They expect that I can treat those patients, as do 1. There is another group that I think it is reasonable for me to attempt, but I want my surgeon to look at the films. If he challenges my plans, then we will discuss the correct approach in the case. If he does not challenge my plans, at least he knows what I am doing. He is a little more attuned to my scheduling of that particular high-risk patient. In another scenario, I can often determine about halfway through a case that the procedure is not going well. At that point, I call the operating room and ask about surgery and operating room availability. The surgeon may alter his plans. A few times we have made arrangements for the operation, and the PTCA has been successful. We would rather do that than have the patient get into trouble. At this point, we have had no deaths with our PTCAs at Crawford Long Hospital. I think also, as a PTCA physician, I have improved my ability to keep an artery temporarily open. I can often keep the guide wire across the lesion and continue to dilate intermittently until the surgery team is ready. Suggestions have been offered about inserting a perfusion catheter. Although the suggestion is good, the technique has some problems. Physicians may take too long trying to get the perfusion catheter into the artery while the patient's condition is deteriorating. Suggestions have also been made about putting the intra-aortic balloon pump (lABP) into such patients. That can be helpful, but again too much time can be taken trying to put a balloon in and thereby delaying the operation. There must be a very close working relationship between the PTCA physician and the surgeon. By working together, we can reduce the infarct rate with PTCA. Dr. Gordon Olinger (Milwaukee, Wis.). My question also relates to standby. I am not prepared to ask our PTCA physicians not to ask us for standby for elective cases. However, this new trend toward acute interventional cardiology and the idea that we should be standing by for every patient who is flown in by helicopter with an acute myocardial infarction (AMI) after either intravenous streptokinase or pending intracoronary streptokinase, with the notion that the patient might have a PTCA, places enormous demands on the operating room and on the surgeon. Furthermore, it is clear from the work that has been done in Michigan, and from other comments that have been made, that it is not appropriate for us to stand by for those cases. It may not even be appropriate for PTCA to be done in that time interval. I would be interested in what Dr. Morris has to say about this and what Dr. Ferguson's policy is at Duke. Obviously you reached these patients fairly soon. Were you standing by for them? Dr. Morris. Our policy is that the surgeons do not stand by for acute interventions with MIs. With the small group of patients who get to us within the first hour, we will let the surgeons know, because we believe these patients are more likely salvageable. If the patient arrives after a delay of 2 hours or more, and we make a full attempt at PTCA, it will be about 4 hours before coronary bypass can be attempted. At that point, the myocardium at jeopardy has been lost. All the surgeon will be able to do is to inflict a further insult on that myocardium. If we get the patient in very, very early-within the first hour, there is a greater chance of salvaging myocardium. However, as a routine, our surgeons are not on standby for those cases. I think it would be a waste of manpower. Dr. T. Bruce Ferguson (Durham, NiC]. Our policy at

The Journal of

772

Discussion

Duke is to not stand by for emergency PTCAs. There is a surgeon with an on-call beeper, and he is contacted when surgical intervention is deemed necessary by the cardiologist. If the surgeon agrees, the patient's condition is stabilized with a reperfusion catheter and IABP when necessary, and he is transported to the first available operating room. It is important to realize that we are dealing with a progressive sequence of events leading to clinical MI. The goal of interventional therapy is to alter the progression of this sequence of events. Our PTCA failure population are patients who have undergone triage at local hospitals, where almost all have received invasive intravenous thrombolytic therapy 30 to 45 minutes after arrival if their chest pain has not been resolved. They are transported to Duke University via helicopter and taken immediately to the interventional catheterization laboratory in the emergency room. Thus, in a large number of these patients, invasive therapy is begun early in the sequence of events that presumably would result in MI if allowed to continue. If PTCA is unsuccessful or inadvisable, a surgeon is consulted. Patients arriving 5, 6 and 7 hours after initial onset of chest pain are not considered surgical candidates. Perhaps this, to a degree, accounts for our slightly lower rate of MI in our PTCA-failure population. More importantly, our experience confirms an evolving management trend. Results from the TAMI trial demonstrate that early administration of tissue plasminogen activator plus emergent operation probably yields the best preservation of myocardial function, despite, as Dr. Morris indicated, a comparable rate of bleeding complications with tissue plasminogen activator or streptokinase. Thus the trend from a combined cardiology jcardiac surgery point of view is to reperfuse myocardium at risk as quickly as possible by whatever means are least damaging to the patient. Dr. Morris. I agree with that. I think we need to be very time conscious. Very early, that is when we want our surgeons. We do not think that sending the patients to later surgical treatment will help them. Dr. Quentin R. Stiles (Los Angeles, Calif). Dr. Ferguson, you commented on how the results may be improved in the future. You said that we should develop pharmacologic strategies to neutralize the adverse reperfusion sequelae. I wonder if you can give us a hint as to specifically what these undesirable sequelae are as you understand them, and what kind of pharmacologic steps should we be thinking about so that we can attempt to neutralize these? Dr. Ferguson. Dr. Stiles' pioneering work in the development of cardioplegia is widely known, and I thank him for his question. PTCA-failure patients requiring emergent operation provide a new patient population for which current techniques of cardioplegia that have so effectively reduced operative mortality for elective operations may need to be modified. This is initially done by use of the reperfusion catheter whenever possible. We have had very good success with this technique in providing patient stabilization before cardiopulmonary bypass, as well as providing a conduit for delivery of cardioplegic solution distal to the obstruction in the vessels supplying the area of myocardium at jeopardy. A biochemical answer to your question is more complicated. A large amount of data exists, both experimentally and clinically, related to agents that affect the potentially deleterious consequences of "reperfusion injury." Dr. Buckberg's group has done extensive work in this area, related to reperfusion after cardioplegic arrest of the heart. Others have

Thoracic and Cardiovascular Surgery

proposed free radical scavengers as agents to prevent reperfusion injury, although these data are less clear. Inclusion of high-energy phosphate precursors or strategies to facilitate resynthesis of high-energy phosphates, as Dr. Wechsler's laboratory has actively investigated, may all be applicable to this acute reperfusion circumstance. The goal of these interventions should be to prevent continuing cell death in myocardium that is potentially salvageable. The problem is that the actual duration of myocardial ischemia is going to be different in each one of these patients, and interventions applicable to one patient may not be applicable to another. This is an area of great promise, but considerable experimental work needs to be performed. Dr. Stiles. I suppose the area about which I am most confused right now is what to do about calcium ions in the immediate reperfusion state. Do you have any advice for me there? Dr. Ferguson. It has been hypothesized that calcium influx during ischemia is a major cause of reperfusion injury. Administration of calcium channel blocking agents at the time of ischemia has seemingly provided some benefit, whereas administration during the reperfusion phase has had only minimal effect. Whether these data could be applied to this particular clinical situa tion is unclear in my mind. If these data are correct, then calcium blockade should probably be administered at the time the patient is admitted with chest pain and interventional therapy is begun, rather than in the cardioplegic solution or reperfusion period. Again, this is an area that needs to be intensively investigated. Dr. David M. Follette (Santa Monica, Calif). Dr. Ferguson, we look at revascularization as a primary modality in AMI. Excellent data are available from Spokane and from Dr. Buckberg at UCLA, looking not only at acute revascularization but also avoidance of reperfusion injury. I am wondering if you have had enough experience to compare your results with primary coronary bypass for AMI with PTCA and streptolysis as the alternative form? PTCA and streptolysis seem from your data to have a higher mortality than the figures that have been reported from Spokane and some of the other centers that are doing primary operations in this instance. Dr. Ferguson. Our experience at Duke with primary surgical revascularization in AMI without prior cardiology intervention is actually very small, probably related to our somewhat unique geographic circumstance and active invasive cardiologists. Although our cardiology group as a whole is extremely aggressive, they have some guidelines as to whom they will submit to PTCA. Patients with severe three-vessel disease, left main disease, and, as Dr. Morris mentioned, patients requiring multiple PTCAs will be referred for emergent surgical intervention after thrombolytic therapy. Because of the intensive cardiology input up front, even without PTCA, I doubt that our experience with surgical intervention for AMI without prior invasive therapy will ever be significant. More important, as cardiologists become more convinced of the efficacy of invasive therapy for AMI, the number of patients admitted emergently for coronary bypass will increase. Those patients will have thrombolytic therapy on board and will have had unsuccessful PTCA. Strategies will need to be developed to ensure that these high-risk surgical patients are managed successfully.