- Email: [email protected]
Hemorrhage from myocardial revascularization
J
THORAC CARDIOVASC SURG
82:768-772, 1981
Hemorrhage from myocardial revascularization Four groups of 12 dogs each had ligation of the left anterior descending coronary artery (LAD) with subsequent release of the ligature and confirmed reperfusion. After 24 hours of reperfusion, the hearts were removed. sliced at I em intervals along the transverse axis. and stained with triphenyltetrazolium chloride. Measurements of the infarct size as a percentage of the left ventricular area (I) and of hemorrhage as a percentage of infarct size (H) were made. The duration of ligation was 3 hours in Group I. 6 hours in Group II. 18 hours in Group Ill, and 30 hours in Group IV. No significant difference in infarct size was found among the groups. Percent hemorrhage was 25.2% of infarct area in Group I. 28.3% in Group II. 18.1 % in Group Ill. and 0.7% in Group IV. If reperfusion hemorrhage into an acute myocardial infarct is deleterious. these data suggest that danger to be decreased at 18 hours and absent at 30 hours after acute coronary occlusion.
Thomas J. Vander Salm, M.D.,* Linda A. Pape, M.D.,** Janet Price, B.A.,*** and Marianne Burke, B.A.,**** Worcester. Mass.
With the clinical success of coronary artery bypass grafts for elective myocardial revascularization, extensions of its use for other than chronic angina have been attempted. One of these attempts has been to truncate the course of a myocardial infarction by emergency revascularization in the hope that the size of the necrosis will be diminished. The report by DeWood and associates 1 of 187 patients treated with elective early revascularization following a myocardial infarction suggests that good results require re-establishment of coronary blood flow within 6 hours of the infarct. Bolooki and colleagues- demonstrated angiographic improvement in left ventricular function when the coronary grafts were completed within 4 hours of the infarction. However, logistic difficulties militate against the possibility of such early revascularization in most patients. Disquieting clinical" and laboratory'- 4 studies suggest adverse effects from early postinfarction rev ascularization, especially if not completed within the early grace period. From the University of Massachusetts Medical Center, Worcester, Mass. Received for publication March 10, 1981. Accepted for publication April I, 1981. Address for reprints: Thomas J. Vander Salm, M.D., Department of Surgery, University of Massachusetts Medical Center, Worcester, Mass. 01605. *Associate Professor of Surgery. **Assistant Professor of Medicine. ***Department of Cardiology. ****University of Massachusetts Medical School, Worcester, Mass.
768
Montoya and associates" showed that reperfusion hemorrhage occurred in dogs after 3 hours of coronary occlusion and that reperfusion after 5 hours of occlusion was associated with even greater hemorrhage. They also showed the same phenomenon in man after postmyocardial infarct revascularization. However, they did not show how long after coronary occlusion this revascularization hemorrhage would no longer occur. In patients with intractable postinfarction angina, early revascularization may be necessary. For them, the question of the duration of jeopardy from revascularization hemorrhage assumes major importance. This study was designed to define that period in dogs.
Methods Adult mongrel dogs were anesthetized with sodium nembutal, 35 mg/kg. Through a left thoracotomy, the left anterior descending coronary artery (LAD) was isolated near the first diagonal branch. The LAD was occluded immediately proximal or distal to the first diagonal branch with an umbilical tape over a narrow gauge plastic tubing, the site chosen to create a moderately large infarct. Lidocaine, 50 mg, was given intravenously just before the ligation to reduce the tendency toward ventricular fibrillation. After intervals of from 3 to 30 hours, the ligature was removed, the vessel externally inspected for confirmation of restored distal flow, and the thoracotomy closed. After 24 hours of LAD reperfusion, the dog was killed with a concentrated potassium chloride solution injected intravenously. The heart was immediately removed and patency of the LAD confirmed by gross appearance after serial
0022-5223/811110768+05$00.50/0 © 1981 The C. V. Mosby Co.
Volume 82
Hemorrhage from myocardial revascularization
Number 5
769
November, 1981
o Norma l ~
Infarct
• Hemorrhage
Fig. 1. lllustrative sections from one heart with respective normal, infarcted, and hemorrhagic areas.
Table I. Infarct area (as percentage of ventricular area) and hemorrhagic area (as percentag e of infarct area) in the four groups Group
Occlusion time (hr )
Percent infarct
I
3 6 18 30
24.8 27.4 31.1 31.2
II III
IV
9.15 11.81 9.07 13.53
Percent hemorrhage
SD
25.2 28.3 18.1* 0.7 *t
17.91 30.43 22.15 1.12
'Group III and IV together different from groups [ and II at 0.01 level by Duncan test. tDifferen l from groups I. II. and III at the 0 .01 level by Duncan test.
sectioning. The heart was then serially sliced at 1 em intervals along the transverse axis (parallel to the atrioventricular groove) and stained with triphenytetrazolium chloride for 20 minutes at 37° C. The hemorrhage was grossly visible, usually appearing blue-black. The infarcted area was unstained and the normal myocardium was stained dark red. All slices below the ligation were placed under a glass slab. Ventricular infarct and hemorrhagic areas were traced onto tracing paper . Relative areas were measured by weighing cutouts of the tracings . Infarcted area as a percent of total left ventricular area and hemorrhage area as a percent of infarcted area were calculated for each group . The interventricular septum was assigned as left ventricle for these calculations. The dogs were divided into four groups of about 12 surviving dogs. The LAD occlusion was maintained for 3 hours in Group I, 6 hours in Group II , 18 hours in Group III, and 30 hours in Group IV . In Groups I and
II, anesthesia was maintained for the duration of the LAD occlusion; in Groups III and I V the chest was closed and the dogs allowed to reawaken before being reanesthetized for removal of the occlusion. All dogs were allowed to reawaken during the 24 hours of reperfusion . Means and standard deviations within each group were calculated. Differences between groups were analyzed by a Duncan procedure . Linear regression analysis was used to assess correlation between percent infarct and percent hemorrhage.
Results The hemorrhage always occurred within the areas of clearly defined infarction and tended to be located in the central subendocardial area of the infarcted tissue (Fig. 1). In Group I (3 hour ligation) the mean infarct area (I) was 24.8 % of the left ventricular area below the level of ligation , and the hemorrhagic area (H) was
770
The Journal of Thoracic and Cardiovascular Surgery
Vander Salm et al,
50 , - - - - - - - -
•
40
30 I
""
..
••
•
•• ••
.. •.
..
20
.
•
•
•
..
.. H = 21.2 + 0.28 I r = 0.657
.. 10
•
•
..
•
•
• •
..
Legend
.. - 3hrs. • - 6hrs. • -18 hrs. 0
20
40
%1
60
80
100
Fig. 2. Linear regression of hemorrhagic areas (as a percent of infarct areas) on infarct size (as a percentage of ventricular areas). %H, Percent hemorrhage. %/. Percent infarction. 25.2% of the infarcted area; in Group II (6 hour ligation), I = 27.4% and H = 28.3%; in Group III (18 hour ligation), I = 31.1% and H = 18.1%; and in Group IV (30 hour ligation), I = 31.2% and H = 0.7% (Table I). Analysis for statistical differences between the groups by a Duncan test showed no difference at the 0.05 level in the size of infarct. For percent hemorrhage, however, Group IV (30 hour ligation) was less than the other three groups at a 0.01 level, and taken together, Groups III (18 hour ligation) and IV (30 hour ligation) were less than Groups I (3 hour ligation) and II (6 hour ligation), again at the 0.01 level by the Duncan test. In Groups I, II, and III, amount of hemorrhage as a percent of infarct was predicted by the regression equation, H = 21.2 + 0.28 I, and had a correlation coefficient of 0.657. Correlation coefficients for the individual groups were as follows: Group I (3 hours) = 0.563, Group II (6 hours) = 0.754, and Group III (18 hours) = 0.834 (Fig. 2).
Discussion Early revascularization following a myocardial infarction may salvage jeopardized myocardium and conceivably could reduce the sequelae of a large myocardial infarction. Several studies support this thesis." 5-8 All these studies show that restoration of coronary perfusion within 4 to 6 hours of occlusion reduces the size of the infarct that occurs with permanent ligation. Clinical evidence also supports this thesis. Bolooki and Vargas? revascularized 13 patients within 4 hours of the onset of a myocardial infarction. All patients survived and all showed postoperative reduction of left ventricular dyskinesia and a corresponding improvement in left ventricular function. DeWood and associates! have enthusiastically revascularized patients with uncomplicated acute myocardial infarctions. Among their patients revascularized within 6 hours of the start of symptoms, the hospital mortality was 2%; in a similar group, treated medically, the hospital mortality was 11.5%. Both laboratory and clinical evidence suggest that
Volume 82 Number 5 November, 1981
the grace period for postinfarct revascularization is short and that delays in performing revascularization will produce results as bad as or worse than those with no .revascularization. Bolooki and co-workers" found that reperfusion of the occluded canine LAD at 2 hours resulted in a smaller infarct size than did a control ligation (no reperfusion), but that reperfusion after 3 hours did not decrease infarct size. Mathur, Guinn, and Burris? measured infarct size after LAD ligation in dogs. After permanent ligation, 92% of animals had an infarct greater than 15% of left ventricular mass, whereas after 2 hour and 5 hour temporary ligations with reperfusion, 50% and 100%, respectively, had greater than 15% infarction. Banka, Chadda, and Helfant'" evaluated left ventricular function in dogs by strain gauge implantation. Reperfusion after 30 to 45 minutes of ligation resulted in a return to normal contraction. Reperfusion after a 2 hour ligation, however, always accentuated the contraction abnormalities seen with ligation alone. The large clinical series reported by DeWood's group! also suggests a grace period. Although revascularization at a mean of 5.3 hours after the start of infarct symptoms did improve the mortality when compared with medical treatment, revascularization at a mean of 9.3 hours later did not. Since none of the group revascularized early had preoperative elevation of creatine kinase, the severity of the myocardial infarction is difficult to assess. Bolooki and Vargas" revascularized 25 patients after acute myocardial infarction. In the 13 operated upon less than 4 hours after acute ischemic symptoms began, there were no deaths, and they had improved left ventricular function on postoperative angiograms. However, of the 12 operated upon more than 4 hours after the symptoms began, two died. All 10 survivors of late revascularization demonstrated increased postoperative left ventricular dyskinesia with decreased left ventricular function. Hemorrhage into a reperfused infarct has been described as a deleterious consequence of revascularization delayed beyond the early grace period. Capone and Most!' occluded the LAD of pigs, and in those reperfused after 15 minutes or occluded for 3 hours without reperfusion, minimal hemorrhage was seen. However, in a group reperfused after 1 hour, intramyocardial hemorrhage was severe. Althaus and associates." found no difference in infarct size in pigs when the LAD was ligated permanently or for 3 hours with reperfusion. However, the reperfused hearts all had hemorrhage into the infarct. Bresnahan and colleagues'? found a correlation between infarct size and hemorrhage. After 5 hours of canine LAD occlusion followed by reperfusion, half
Hemorrhage from myocardial revascularization
77 1
the dogs had 42% less infarct than predicted (from quantitative measurement of early creatine phosphokinase [CPK] release), whereas half had 100% more infarct than predicted. Those dogs with larger infarcts had ten times as much hemorrhage as did the dogs with small infarcts. Bresnahan's group!" concluded that hemorrhage itself may cause larger than predicted infarction, but they based their conclusions on estimates of infarct size from serum CPK levels. As Vatner and co-workers!' have shown, release of coronary occlusion increases serum CPK and thereby leads to an overestimate of infarct size. We too found correlation between infarct size and hemorrhage size. The most likely explanation for this finding, however, is not that hemorrhage causes increased infarction but that larger infarcts predispose to larger areas of reperfusion hemorrhage because of increasing damage to blood vessels. Lie and associates" studied 44 patients who died after coronary revascularization. When an infarct had occurred within 1 to 7 days preoperatively or up to 12 hours postoperatively, eight of 14 patients (57%) had a grossly visible hemorrhagic infarct. Infarcts occurring 1 to 14 days postoperatively showed hemorrhage in ~e of 13 patients (38%) and those occurring 15 to 90 days postoperatively showed hemorrhage in one of 17 patients (6%). Montoya and associates" noted hemorrhage into infarcts in three patients who were revascularized during an acute myocardial infarction and subsequently died. Because of this discouraging result, they studied dogs after LAD ligation. In those ligated permanently, no hemorrhage was seen in the infarct. In those reperfused after 3 hours of ligation, the infarct size was reduced but 83% had some hemorrhage. In those reperfused after 5 hours of ligation, the infarcts were larger and all were markedly hemorrhagic. All these studies seem to point, in both animals and man, to a period early after the onset of myocardial ischemia/infarction when revascularization is safe and to a subsequent period when revascularization may be deleterious-perhaps because of the hemorrhage into the infarct. This hemorrhage, chronologically, correlates with the time of increased vascular permeability seen after LAD ligation in dogs.'" Most centers have not adopted the practice of DeWood and his surgical colleagues 1 ofimmediate surgical revascularization as the primary treatment of uncomplicated myocardial infarctions. The evidence of complications, in both man and animal, has been such that most surgeons prefer not to revascularize an acute myocardial infarction. The increasing evidence for successful treatment of acute myocardial infarction by thrombus lysis with streptokinase again raises the numerous ques-
772
The Journal of Thoracic and Cardiovascular Surgery
Vander Salm et al.
tions about the consequences and timing of postinfarction reperfusion. Unfortunately, although many studies define the beginning of the potentially harmful period as several hours from the onset of ischemia, the end of the danger period is not known. The need for an answer to this question is forced upon us by patients who have continuing angina after an infarct. What is the earliest time following the very early grace period when coronary artery bypass grafts will not cause intrainfarct hemorrhage? We used intrainfarct hemorrhage as a marker of possible damage from revascularization. In dogs, a sequence of LAD occlusion and then reperfusion caused marked hemorrhage at 3 and 6 hours of occlusion. A small decrement was seen after 18 hours of occlusion, and hemorrhage was nearly absent after 30 hours of occlusion. The results in our first two groups are similar to those of Montoya and associates. 3 The reduced hemorrhage in our 18 hour group and in the 24 hour group of White and associates, 16 and the near absence of hemorrhage in our 34 occlusion group, define the time limitation of the reperfusion hemorrhage. These results may not be translatable to clinical experience. First, neither the onset nor offset of the period of danger from revascularization hemorrhage will necessarily be the same in man as it is in dogs. Second, the presence or absence of reperfusion hemorrhage may not predict a clinically unsafe or safe period. To date, the significance of reperfusion hemorrhage is not known. Whether it is a cause of acute increases in wall stiffness or delayed healing remains uncertain and whether it is a cause of or merely a marker for a poor prognosis remains to be determined. Reperfusion bleeding intd'the infarct occurred at reperfusion as early as 3 hours but diminished at 18 hours and was nearly absent at 30 hours. In man, if a similar time scale applies, and if coronary revascularization is necessary early after myocardial infarction but after the first several hours of infarction, it might be wise to delay the operation for at least 1 day from the onset of symptoms so as to prevent intrainfarct hemorrhage. Obviously, refractory postinfarct angina demands a careful assessment of the risk to infarcted myocardium from early revascularization and the risk to ischemic myocardium from extension of the infarct. We wish to thank Carol Mainville, Ph.D., for her assistance with the statistical evaluation. REFERENCES DeWood MA, Spores J, Notske RN, Lang HT, Shields JP, Simpson CS, Rudy LW, Grunwald R: Medical and
2
3
4
5
6
7
8
9 10
II 12
13
14
15
16
surgical management of myocardial infarction. Am J Cardiol 44:1356, 1979 . Bolooki H, Kotler MD, Lottenberg L, Dresnick S, et al: Myocardial revascularization after acute infarction. Am J Cardiol 36:395, 1975 Montoya A, Mulet J, Pifarre R, Brynjolfsson G, Moran JM, Sullivan HJ, Gunnar RM: Hemorrhagic infarct following myocardial revascularization. J THORAC CARDIOVASC SURG 75:206, 1978 Lie JT, Lawrie GM, Morris GC, Winters WL: Hemorrhagic myocardial infarction associated with aortocoronary bypass revascularization. Am Heart J 96:295, 1978 Reimer KA, Lowe JE, Rasmussen MM, Jennings RB: The wavefront phenomenon of ischemic cell death. I. Myocardial infarct size vs duration of coronary occlusion in dogs. Circulation 56:786, 1977 Ginks WR, Sybers HD, Maroko PR, Covell JW, Sobel BE, Ross J: Coronary artery reperfusion. II. Reduction of myocardial infarct size at I week after the coronary occ1usion. J Clin Invest 51 :2717, 1972 Mathur VS, Guinn GA, Burris WH: Maximal revascularization (reperfusion) in intact conscious dogs after 2 to 5 hours of coronary occlusion. Am J Cardiol 36:252, 1975 Smith GT, Soeter JR, Haston HH, McNamara 11: Coronary reperfusion in primates. Serial electrocardiographic and histologic assessment. J Clin Invest 54: 1420, 1974 Bolooki H, Vargas A: Myocardial revascularization after acute myocardial infarction. Arch Surg 111: 1216, 1976 Banka VS, Chadda KD, Helfant RH: Limitations of myocardial revascularization in restoration of regional contraction abnormalities produced by coronary occlusion. Am J Cardiol 34:164, 1974 Capone RJ, Most AS: Myocardial hemorrhage after coronary reperfusion in pigs. Am J Cardiol 41:259, 1978 Althaus U, Janett J, Scholl E, Riedwyl H: Effects of myocardial revascularization following acute coronary occlusion in pigs. Eur J Clin Invest 6:7, 1976 Bresnahan GF, Roberts R, Shell WE, Ross J, Sobel BE: Deleterious effects due to hemorrhage after myocardial reperfusion. Am J Cardiol 33:82, 1974 Vatner SF, Baig H, Manders, WT, Maroko PR: Effects of coronary artery reperfusion on myocardial infarct size calculated from creatine kinase. J Clin Invest 61: 1048, 1978 West PN, Connors JP, Clark RE, et al: Compromised microvascular integrity in ischemic myocardium. Lab Invest 38:677, 1978 White FC, Sanders M, Bloor CM: Regional redistribution of myocardial blood flow after coronary occlusion and reperfusion in the conscious dog. Am J Cardiol 42:234, 1978
THORAC CARDIOVASC SURG
82:768-772, 1981
Hemorrhage from myocardial revascularization Four groups of 12 dogs each had ligation of the left anterior descending coronary artery (LAD) with subsequent release of the ligature and confirmed reperfusion. After 24 hours of reperfusion, the hearts were removed. sliced at I em intervals along the transverse axis. and stained with triphenyltetrazolium chloride. Measurements of the infarct size as a percentage of the left ventricular area (I) and of hemorrhage as a percentage of infarct size (H) were made. The duration of ligation was 3 hours in Group I. 6 hours in Group II. 18 hours in Group Ill, and 30 hours in Group IV. No significant difference in infarct size was found among the groups. Percent hemorrhage was 25.2% of infarct area in Group I. 28.3% in Group II. 18.1 % in Group Ill. and 0.7% in Group IV. If reperfusion hemorrhage into an acute myocardial infarct is deleterious. these data suggest that danger to be decreased at 18 hours and absent at 30 hours after acute coronary occlusion.
Thomas J. Vander Salm, M.D.,* Linda A. Pape, M.D.,** Janet Price, B.A.,*** and Marianne Burke, B.A.,**** Worcester. Mass.
With the clinical success of coronary artery bypass grafts for elective myocardial revascularization, extensions of its use for other than chronic angina have been attempted. One of these attempts has been to truncate the course of a myocardial infarction by emergency revascularization in the hope that the size of the necrosis will be diminished. The report by DeWood and associates 1 of 187 patients treated with elective early revascularization following a myocardial infarction suggests that good results require re-establishment of coronary blood flow within 6 hours of the infarct. Bolooki and colleagues- demonstrated angiographic improvement in left ventricular function when the coronary grafts were completed within 4 hours of the infarction. However, logistic difficulties militate against the possibility of such early revascularization in most patients. Disquieting clinical" and laboratory'- 4 studies suggest adverse effects from early postinfarction rev ascularization, especially if not completed within the early grace period. From the University of Massachusetts Medical Center, Worcester, Mass. Received for publication March 10, 1981. Accepted for publication April I, 1981. Address for reprints: Thomas J. Vander Salm, M.D., Department of Surgery, University of Massachusetts Medical Center, Worcester, Mass. 01605. *Associate Professor of Surgery. **Assistant Professor of Medicine. ***Department of Cardiology. ****University of Massachusetts Medical School, Worcester, Mass.
768
Montoya and associates" showed that reperfusion hemorrhage occurred in dogs after 3 hours of coronary occlusion and that reperfusion after 5 hours of occlusion was associated with even greater hemorrhage. They also showed the same phenomenon in man after postmyocardial infarct revascularization. However, they did not show how long after coronary occlusion this revascularization hemorrhage would no longer occur. In patients with intractable postinfarction angina, early revascularization may be necessary. For them, the question of the duration of jeopardy from revascularization hemorrhage assumes major importance. This study was designed to define that period in dogs.
Methods Adult mongrel dogs were anesthetized with sodium nembutal, 35 mg/kg. Through a left thoracotomy, the left anterior descending coronary artery (LAD) was isolated near the first diagonal branch. The LAD was occluded immediately proximal or distal to the first diagonal branch with an umbilical tape over a narrow gauge plastic tubing, the site chosen to create a moderately large infarct. Lidocaine, 50 mg, was given intravenously just before the ligation to reduce the tendency toward ventricular fibrillation. After intervals of from 3 to 30 hours, the ligature was removed, the vessel externally inspected for confirmation of restored distal flow, and the thoracotomy closed. After 24 hours of LAD reperfusion, the dog was killed with a concentrated potassium chloride solution injected intravenously. The heart was immediately removed and patency of the LAD confirmed by gross appearance after serial
0022-5223/811110768+05$00.50/0 © 1981 The C. V. Mosby Co.
Volume 82
Hemorrhage from myocardial revascularization
Number 5
769
November, 1981
o Norma l ~
Infarct
• Hemorrhage
Fig. 1. lllustrative sections from one heart with respective normal, infarcted, and hemorrhagic areas.
Table I. Infarct area (as percentage of ventricular area) and hemorrhagic area (as percentag e of infarct area) in the four groups Group
Occlusion time (hr )
Percent infarct
I
3 6 18 30
24.8 27.4 31.1 31.2
II III
IV
9.15 11.81 9.07 13.53
Percent hemorrhage
SD
25.2 28.3 18.1* 0.7 *t
17.91 30.43 22.15 1.12
'Group III and IV together different from groups [ and II at 0.01 level by Duncan test. tDifferen l from groups I. II. and III at the 0 .01 level by Duncan test.
sectioning. The heart was then serially sliced at 1 em intervals along the transverse axis (parallel to the atrioventricular groove) and stained with triphenytetrazolium chloride for 20 minutes at 37° C. The hemorrhage was grossly visible, usually appearing blue-black. The infarcted area was unstained and the normal myocardium was stained dark red. All slices below the ligation were placed under a glass slab. Ventricular infarct and hemorrhagic areas were traced onto tracing paper . Relative areas were measured by weighing cutouts of the tracings . Infarcted area as a percent of total left ventricular area and hemorrhage area as a percent of infarcted area were calculated for each group . The interventricular septum was assigned as left ventricle for these calculations. The dogs were divided into four groups of about 12 surviving dogs. The LAD occlusion was maintained for 3 hours in Group I, 6 hours in Group II , 18 hours in Group III, and 30 hours in Group IV . In Groups I and
II, anesthesia was maintained for the duration of the LAD occlusion; in Groups III and I V the chest was closed and the dogs allowed to reawaken before being reanesthetized for removal of the occlusion. All dogs were allowed to reawaken during the 24 hours of reperfusion . Means and standard deviations within each group were calculated. Differences between groups were analyzed by a Duncan procedure . Linear regression analysis was used to assess correlation between percent infarct and percent hemorrhage.
Results The hemorrhage always occurred within the areas of clearly defined infarction and tended to be located in the central subendocardial area of the infarcted tissue (Fig. 1). In Group I (3 hour ligation) the mean infarct area (I) was 24.8 % of the left ventricular area below the level of ligation , and the hemorrhagic area (H) was
770
The Journal of Thoracic and Cardiovascular Surgery
Vander Salm et al,
50 , - - - - - - - -
•
40
30 I
""
..
••
•
•• ••
.. •.
..
20
.
•
•
•
..
.. H = 21.2 + 0.28 I r = 0.657
.. 10
•
•
..
•
•
• •
..
Legend
.. - 3hrs. • - 6hrs. • -18 hrs. 0
20
40
%1
60
80
100
Fig. 2. Linear regression of hemorrhagic areas (as a percent of infarct areas) on infarct size (as a percentage of ventricular areas). %H, Percent hemorrhage. %/. Percent infarction. 25.2% of the infarcted area; in Group II (6 hour ligation), I = 27.4% and H = 28.3%; in Group III (18 hour ligation), I = 31.1% and H = 18.1%; and in Group IV (30 hour ligation), I = 31.2% and H = 0.7% (Table I). Analysis for statistical differences between the groups by a Duncan test showed no difference at the 0.05 level in the size of infarct. For percent hemorrhage, however, Group IV (30 hour ligation) was less than the other three groups at a 0.01 level, and taken together, Groups III (18 hour ligation) and IV (30 hour ligation) were less than Groups I (3 hour ligation) and II (6 hour ligation), again at the 0.01 level by the Duncan test. In Groups I, II, and III, amount of hemorrhage as a percent of infarct was predicted by the regression equation, H = 21.2 + 0.28 I, and had a correlation coefficient of 0.657. Correlation coefficients for the individual groups were as follows: Group I (3 hours) = 0.563, Group II (6 hours) = 0.754, and Group III (18 hours) = 0.834 (Fig. 2).
Discussion Early revascularization following a myocardial infarction may salvage jeopardized myocardium and conceivably could reduce the sequelae of a large myocardial infarction. Several studies support this thesis." 5-8 All these studies show that restoration of coronary perfusion within 4 to 6 hours of occlusion reduces the size of the infarct that occurs with permanent ligation. Clinical evidence also supports this thesis. Bolooki and Vargas? revascularized 13 patients within 4 hours of the onset of a myocardial infarction. All patients survived and all showed postoperative reduction of left ventricular dyskinesia and a corresponding improvement in left ventricular function. DeWood and associates! have enthusiastically revascularized patients with uncomplicated acute myocardial infarctions. Among their patients revascularized within 6 hours of the start of symptoms, the hospital mortality was 2%; in a similar group, treated medically, the hospital mortality was 11.5%. Both laboratory and clinical evidence suggest that
Volume 82 Number 5 November, 1981
the grace period for postinfarct revascularization is short and that delays in performing revascularization will produce results as bad as or worse than those with no .revascularization. Bolooki and co-workers" found that reperfusion of the occluded canine LAD at 2 hours resulted in a smaller infarct size than did a control ligation (no reperfusion), but that reperfusion after 3 hours did not decrease infarct size. Mathur, Guinn, and Burris? measured infarct size after LAD ligation in dogs. After permanent ligation, 92% of animals had an infarct greater than 15% of left ventricular mass, whereas after 2 hour and 5 hour temporary ligations with reperfusion, 50% and 100%, respectively, had greater than 15% infarction. Banka, Chadda, and Helfant'" evaluated left ventricular function in dogs by strain gauge implantation. Reperfusion after 30 to 45 minutes of ligation resulted in a return to normal contraction. Reperfusion after a 2 hour ligation, however, always accentuated the contraction abnormalities seen with ligation alone. The large clinical series reported by DeWood's group! also suggests a grace period. Although revascularization at a mean of 5.3 hours after the start of infarct symptoms did improve the mortality when compared with medical treatment, revascularization at a mean of 9.3 hours later did not. Since none of the group revascularized early had preoperative elevation of creatine kinase, the severity of the myocardial infarction is difficult to assess. Bolooki and Vargas" revascularized 25 patients after acute myocardial infarction. In the 13 operated upon less than 4 hours after acute ischemic symptoms began, there were no deaths, and they had improved left ventricular function on postoperative angiograms. However, of the 12 operated upon more than 4 hours after the symptoms began, two died. All 10 survivors of late revascularization demonstrated increased postoperative left ventricular dyskinesia with decreased left ventricular function. Hemorrhage into a reperfused infarct has been described as a deleterious consequence of revascularization delayed beyond the early grace period. Capone and Most!' occluded the LAD of pigs, and in those reperfused after 15 minutes or occluded for 3 hours without reperfusion, minimal hemorrhage was seen. However, in a group reperfused after 1 hour, intramyocardial hemorrhage was severe. Althaus and associates." found no difference in infarct size in pigs when the LAD was ligated permanently or for 3 hours with reperfusion. However, the reperfused hearts all had hemorrhage into the infarct. Bresnahan and colleagues'? found a correlation between infarct size and hemorrhage. After 5 hours of canine LAD occlusion followed by reperfusion, half
Hemorrhage from myocardial revascularization
77 1
the dogs had 42% less infarct than predicted (from quantitative measurement of early creatine phosphokinase [CPK] release), whereas half had 100% more infarct than predicted. Those dogs with larger infarcts had ten times as much hemorrhage as did the dogs with small infarcts. Bresnahan's group!" concluded that hemorrhage itself may cause larger than predicted infarction, but they based their conclusions on estimates of infarct size from serum CPK levels. As Vatner and co-workers!' have shown, release of coronary occlusion increases serum CPK and thereby leads to an overestimate of infarct size. We too found correlation between infarct size and hemorrhage size. The most likely explanation for this finding, however, is not that hemorrhage causes increased infarction but that larger infarcts predispose to larger areas of reperfusion hemorrhage because of increasing damage to blood vessels. Lie and associates" studied 44 patients who died after coronary revascularization. When an infarct had occurred within 1 to 7 days preoperatively or up to 12 hours postoperatively, eight of 14 patients (57%) had a grossly visible hemorrhagic infarct. Infarcts occurring 1 to 14 days postoperatively showed hemorrhage in ~e of 13 patients (38%) and those occurring 15 to 90 days postoperatively showed hemorrhage in one of 17 patients (6%). Montoya and associates" noted hemorrhage into infarcts in three patients who were revascularized during an acute myocardial infarction and subsequently died. Because of this discouraging result, they studied dogs after LAD ligation. In those ligated permanently, no hemorrhage was seen in the infarct. In those reperfused after 3 hours of ligation, the infarct size was reduced but 83% had some hemorrhage. In those reperfused after 5 hours of ligation, the infarcts were larger and all were markedly hemorrhagic. All these studies seem to point, in both animals and man, to a period early after the onset of myocardial ischemia/infarction when revascularization is safe and to a subsequent period when revascularization may be deleterious-perhaps because of the hemorrhage into the infarct. This hemorrhage, chronologically, correlates with the time of increased vascular permeability seen after LAD ligation in dogs.'" Most centers have not adopted the practice of DeWood and his surgical colleagues 1 ofimmediate surgical revascularization as the primary treatment of uncomplicated myocardial infarctions. The evidence of complications, in both man and animal, has been such that most surgeons prefer not to revascularize an acute myocardial infarction. The increasing evidence for successful treatment of acute myocardial infarction by thrombus lysis with streptokinase again raises the numerous ques-
772
The Journal of Thoracic and Cardiovascular Surgery
Vander Salm et al.
tions about the consequences and timing of postinfarction reperfusion. Unfortunately, although many studies define the beginning of the potentially harmful period as several hours from the onset of ischemia, the end of the danger period is not known. The need for an answer to this question is forced upon us by patients who have continuing angina after an infarct. What is the earliest time following the very early grace period when coronary artery bypass grafts will not cause intrainfarct hemorrhage? We used intrainfarct hemorrhage as a marker of possible damage from revascularization. In dogs, a sequence of LAD occlusion and then reperfusion caused marked hemorrhage at 3 and 6 hours of occlusion. A small decrement was seen after 18 hours of occlusion, and hemorrhage was nearly absent after 30 hours of occlusion. The results in our first two groups are similar to those of Montoya and associates. 3 The reduced hemorrhage in our 18 hour group and in the 24 hour group of White and associates, 16 and the near absence of hemorrhage in our 34 occlusion group, define the time limitation of the reperfusion hemorrhage. These results may not be translatable to clinical experience. First, neither the onset nor offset of the period of danger from revascularization hemorrhage will necessarily be the same in man as it is in dogs. Second, the presence or absence of reperfusion hemorrhage may not predict a clinically unsafe or safe period. To date, the significance of reperfusion hemorrhage is not known. Whether it is a cause of acute increases in wall stiffness or delayed healing remains uncertain and whether it is a cause of or merely a marker for a poor prognosis remains to be determined. Reperfusion bleeding intd'the infarct occurred at reperfusion as early as 3 hours but diminished at 18 hours and was nearly absent at 30 hours. In man, if a similar time scale applies, and if coronary revascularization is necessary early after myocardial infarction but after the first several hours of infarction, it might be wise to delay the operation for at least 1 day from the onset of symptoms so as to prevent intrainfarct hemorrhage. Obviously, refractory postinfarct angina demands a careful assessment of the risk to infarcted myocardium from early revascularization and the risk to ischemic myocardium from extension of the infarct. We wish to thank Carol Mainville, Ph.D., for her assistance with the statistical evaluation. REFERENCES DeWood MA, Spores J, Notske RN, Lang HT, Shields JP, Simpson CS, Rudy LW, Grunwald R: Medical and
2
3
4
5
6
7
8
9 10
II 12
13
14
15
16
surgical management of myocardial infarction. Am J Cardiol 44:1356, 1979 . Bolooki H, Kotler MD, Lottenberg L, Dresnick S, et al: Myocardial revascularization after acute infarction. Am J Cardiol 36:395, 1975 Montoya A, Mulet J, Pifarre R, Brynjolfsson G, Moran JM, Sullivan HJ, Gunnar RM: Hemorrhagic infarct following myocardial revascularization. J THORAC CARDIOVASC SURG 75:206, 1978 Lie JT, Lawrie GM, Morris GC, Winters WL: Hemorrhagic myocardial infarction associated with aortocoronary bypass revascularization. Am Heart J 96:295, 1978 Reimer KA, Lowe JE, Rasmussen MM, Jennings RB: The wavefront phenomenon of ischemic cell death. I. Myocardial infarct size vs duration of coronary occlusion in dogs. Circulation 56:786, 1977 Ginks WR, Sybers HD, Maroko PR, Covell JW, Sobel BE, Ross J: Coronary artery reperfusion. II. Reduction of myocardial infarct size at I week after the coronary occ1usion. J Clin Invest 51 :2717, 1972 Mathur VS, Guinn GA, Burris WH: Maximal revascularization (reperfusion) in intact conscious dogs after 2 to 5 hours of coronary occlusion. Am J Cardiol 36:252, 1975 Smith GT, Soeter JR, Haston HH, McNamara 11: Coronary reperfusion in primates. Serial electrocardiographic and histologic assessment. J Clin Invest 54: 1420, 1974 Bolooki H, Vargas A: Myocardial revascularization after acute myocardial infarction. Arch Surg 111: 1216, 1976 Banka VS, Chadda KD, Helfant RH: Limitations of myocardial revascularization in restoration of regional contraction abnormalities produced by coronary occlusion. Am J Cardiol 34:164, 1974 Capone RJ, Most AS: Myocardial hemorrhage after coronary reperfusion in pigs. Am J Cardiol 41:259, 1978 Althaus U, Janett J, Scholl E, Riedwyl H: Effects of myocardial revascularization following acute coronary occlusion in pigs. Eur J Clin Invest 6:7, 1976 Bresnahan GF, Roberts R, Shell WE, Ross J, Sobel BE: Deleterious effects due to hemorrhage after myocardial reperfusion. Am J Cardiol 33:82, 1974 Vatner SF, Baig H, Manders, WT, Maroko PR: Effects of coronary artery reperfusion on myocardial infarct size calculated from creatine kinase. J Clin Invest 61: 1048, 1978 West PN, Connors JP, Clark RE, et al: Compromised microvascular integrity in ischemic myocardium. Lab Invest 38:677, 1978 White FC, Sanders M, Bloor CM: Regional redistribution of myocardial blood flow after coronary occlusion and reperfusion in the conscious dog. Am J Cardiol 42:234, 1978