Concentric hemorrhagic necrosis of the myocardium

Original Contributions CONCENTRIC HEMORRHAGIC NECROSIS OF T H E M Y O C A R D I U M A Morphological and Clinical Study Aw'um Gotlieb, M.D., F.R.C.P.(C.),* Serge Masse, M.D., F.R.C.P.(C.),'~ Jean Allard, M.D., F.R.C.S.(C.),+,.Anthony Dobell, M.D.,w and Shao-Nan Huang, M.D., F.R.C.P.(C.)I ] Abstract Forty-nine cases of concentric hemorrhagic necrosis o f the left ventricular myocardium were reviewed in a series of 97 autopsy studies in patients subjected to cardiopulmonary bypass for various types o f cardiac surgery. Tim gross and microscopic findings were analyzed in order to define the morphologic evolution of the lesion. The extent of involvement was graded by gross examination of serial transverse sections of the heart. Microscopically there were five major histologic changes, probably representing the sequential evolution of the lesion, i.e., contraction bands, subendocardial hemorrlmges, coagulative necrosis, healing by granulation tissue, and fibrosis. T h e location of the lesion coincided with tim vulnerable region of tim microcirculation. Owing to the implementation of new surgical teclmiques, the cases were subdMded into two groups, one covering the period from 1963 to 1970 and the other, 1971 to 1974. Concentric hemorrhagic necrosis was less frequent in the more recent group, but when it was present it was more pronounced in the individual heart. The lesion in the earlier group was milder but demonstrated a higher incidence of platelet nticrothrombi in the heart. In the recent cases concentric hemorrhagic necrosis tended to be more diffuse in aortocoronary bypass titan in valvular replacement surgery. We discuss one possible explanation for the development of this lesion, i.e., transient hypoxenfia occurring at the time of cardiopulmonary bypass, followed by reperfusion and accelerated necrosis with hemorrlmge.

*Teaching Fellow, Department of l'athology, McGill University Facultyof Medicine, Montreal, Quebec, Canada. tAssistant Professor, Deparnnent of Pathology,UniversityofSherbrooke,l'athologist,Sherbrooke University ttospital, Sherbrooke, Quebec, Canada. *Resident in Surgery,McGillUniversityFacultyof Medicineand RoyalVictoria 1lospital, Montreal, Quebec, Canada. w of Surgery, McGillUniversityFacultyof Medicine.Surgeon-in-Charge,Cardiovascular Thoracic Surgery, RoyalVictoria I lospital, Montreal,Quebec, Canada. ]]Professor of Pathology,McGillUniversityFacuhyof Medicine.Senior Pathologist,RoyalVictoria ttospital, Montreal,Quebec, Canada.

27

HUMAN PATHOLOGY-VOLUME 8, NUMBER 1 Jamtmy 1977 A distinct c l i n i c o p a t h o l o g i c e n t i t y known as concentric llemorrlmgic necrosis o f the left ventricular myocardium occurs as a serious and often fatal complication in some adult patients subjected to total cardioptdmonary bypass. ~ Clinically it is characterized by failure to maintain an adeqtmte cardiac output in the earl)" postoperative period. Often the cardiac output is inadequate to support life after termination of the cardiopuhnonary bypass. T h e lesion found at autopsy has distinct gross and microscopic features that differ from those of the usual ischemic infarct. Typically it does not follow the distribution o f a single coronary artery, being instead circumferential, either focal or diffuse, and involving the inner third to half of the myocardium. Microscopically, myofibrillar contraction bands and massive interstitial hemorrhage characterize the lesion. T h e entity has accounted for approximately one third of tbe deaths after cardiopulmonary bypass. ~ T h e mechanism of its development has not been completely elucidated. The present study is a review of our autopsy experience with concentric hemorrhagic necrosis. In this study it has been possible to reconstruct and document the evolution of the morphologic lesion as well as to speculate about some of the probable etiologic factors. We have been able to compare the morphologic and clinical findings in recent cases as opposed to patients operated upon several )'ears previously. Our findings are discussed with respect to transient anoxia and reperfusion. MATERIAL AND METHODS

28

Study of 97 autopsies in patients who had been subjected to total cardiopuhnonary bypass for various types of cardiac surgery was completed between January 1963 and December 1974. The heart specimens were available for systematic gross examination. The weight of the heart, the distribution and the extent of the lesions, the condition of the coronary arteries, and the flmctional and patlmlogic status of the vah'ular prosthesis and tile saphenous grafts were recorded. Serial transverse sections of both ventricles were

nmde extending from the apex to tile annulus, and photographs were taken. In addition, tissue from two cases encountered after 1969 was examined by postmortem angiography with Schlesinger's contrast medium. The clinical history and operative records were reviewed. The cause of death was determined by complete autopsy examination. Muhiple formalin fixed tissue blocks were taken for histologic examination, and sections were routinely stained with hematoxylin-phloxin and saffron, or hematoxylin and eosin. Masson trichrome and phosphotungstic acid-hematoxylin stains were also used in Solne

cases.

RESULTS

From J a n u a r y 1963 to December 1974 a total of 97 autopsies were performed on patients who had undergone cardiac surgery and had been on total cardio.pulmonary bypass. In order to emphasize certain changes occurring in the preceding four years, the autopsy cases were divided into two groups covering the periods 1963 to 1970 and 1971 to 1974. Group A (1963-1970) consisted of 60 cases encountered ill a series of 311 cardiac operations that required cardiopuhnonary bypass. Group B (1971-1974) consisted of 37 cases in a series of 447 cardiac operations. The mean age of group A patients was 46 years (range, 22 to 69 years) and of group B patients, 54 years (range, 23 to 72 years). Both groups showed a male to female ratio of 3 to 2. Forty-nine of 97 patients had concentric hemorrlmgic necrosis, group A having more patients with the lesion (37 of 60 or 62 per cent) than group B (12 of 37 or 32 per cent; Table 1). T h e operations performed are listed in Table I. In both groups valvular prosthesis surgery was the most common; however, in group B there were nine aortocoronary bypass procedures and two bypass grafts with aortic vah'e prosthesis, whereas in group A there were only two bypass grafts. T h e operation most frequently associated with concentric llemorrhagic necrosis in group A was aortic valve replacement. Thus, of 32 patients who died after aortic valve replacement alone or in

HEMORRHAGIC T A B L E 1.

N E C R O S I S O F T H E M Y O C A R D I U M - - G o r L I E B ET AL.

OPERATIVE PROCEDURES IN PATIENTS SUBJECTED TO TOTAL CARDIOPULMONARY BYPASS

Operative Procedures

I. Valve prosthesis Mitral Aortic Mitral and aortic Aortic and tricuspid Mitral and tricuspid Trivalvular Aortic prosthesis and saphenous vein bypass to left anterior descending coronary artery Aortic prosthesis and saphcnous vein bypass to right coronary and left anterior descending coronary artery 11. Saphenous vein graft To left coronary artery To right coronary artery To left anterior descending coronary artery and right iuternal mammary implam To right coronary and left anterior descending coronary artery Trivcssel bypass

Group A 1963-1970 ll'ilh CIIN Lesion Total

8 18 4 l 1 2

l 1

20(40%) 22(82%) 8(50%) 2(50%) 3(33%) 2(100%)

1000%) !*(100%)

Ill. Internal mammary to left anterior descending coronary artery IV. Congenital Correction for tetralogy Corrected transposition; atrial septal defect, ventricular septal defect repair; pulmonary valvotom)' V. Cardiac transplant

l

Group B 1971-1974 ll'ilh CttN l.esion Total

1 1 3 1 1

9(11%) 5(20%) 6(50%) 1(100%) 2(50%)

0

1(0)

0

1(0)

0

2(0)

l

1(100%)

2 I

5(40%) 1(100%)

o

i(o)

1(100%)

-

-

0

l (0)

-

-

l

1(100%)

37

60(62%)

12

37(32%)

*One had left internal mammary implant as well.

combination with a second valve replacement, 23 (72 per cent) showed the lesion. In group B, 13 patients died after operations that included aortic valve replacement, and only four (30 per cent) showed the lesion. The majority of the patients with the lesion died witifin the first 24 hours with the low output syndrome (Table 2). Tim causes of death beyond 10 days included pneumonia, cerebral complications, and sepsis. Bacterial endocarditis occurred as a cause of death only in patients surviving beyond one month. Patients not showing concentric hemorrhagic necrosis died of

transmural acute myocardial infarctions, puhnonary entboli, arrhythmias, pneumonia, lmmorrhage, tamponade, and valve prosthesis dysfunction. All the 97 patients who came to autopsy had left ventricuIar hypertrophy. In the case of tetralogy of Fallot, the right and the left ventricular myocardium was hyl)ertrophic and both areas showed marked concentric hemorrhagic necrosis. There were more patients in group A with marked llypertrophy than in group B. In group A 28 of the 60 patients (47 per cent) had hearts tlmt weighed over 700 gin., whereas only 9 of the 37 in group B had

29

HUMAN PATHOI.OGY--VOLUME 8, NUMBER 1 January 1977 T A B L E 2. C A u s e s OF DEATH IN 4 9 PATIENTS SHOW1NC; CONCENTRIC HEMORRIIAGIC NECROSIS OF TIlE ~|YOCARDIUM

Duration of Postoperative Survival Cause of Death

24 hours

1-3 days

4-10 days

11 days-I too.

Low o u t p u t s y n d r o m e Pneumoifia Tamponade Bacterial endocarditis Sepsis Cerebral complications Helnorrhage Arrhythlnia S u d d e n death

22 0 1 0 0 1 1 1 0

6 0 1 0 0 0 0 0 I

3 1 0 0 0 1 0 0 0

0 2 0 0 0 2 0 0 0

TOTAL

2-6

8

7~

~

hearts weighing o v e r 700 gm. Although each patient with concentric h e m o r r h a g i c necrosis showed left ventricular hypert r o p h y , the d e g r e e o f I w p e r t r o p h y was not m o r e striking t h a n in those patients d y i n g without tim lesion.

Gross Findings

30

T i m filll blown lesion o f concentric h e m o r r h a g i c necrosis was readily recognizable when tim heart was sectioned transversely. Tile earliest lesion ill tiffs series was seen in a patient who died t h r e e lmurs after c a r d i o p u h n o n a r y bypass. T i m i n n e r half o f tim m y o c a r d i u m was swollen anti bulging from the cut surface. It showed a concentric light red mottling. T i m macroenzymatic stain for succinic d e h y d r o g e n a s e showed concentric loss o f e n z y m e activity indicative o f earl}" ischemic injury. In patients who survived for 12 to 24 h o u r s the h e m o r r h a g i c lesions were grossly unmistakable (Fig. 1). In severe cases a u n i f o r m circumferential band o f h e n m r r h a g e involving tim inner half to two thirds o f the left ventricular myoc a r d i u m was seen f r o m the nfitral annulus to the apex o f the heart. In the least involved cases t h e r e was focal subendocardial h e m o r r h a g e , which was present in only one or two transverse sections and might have been overlooked if the left ventricle had not been cut in 1 cm. thick transverse sections. T h e demarcation f r o m the uninvolved area was well defined grossly. Foci o f uninvolved m y o c a r d i u m were often present in the papillary muscles and trabeculae carneae (Fig. 2).

More than 1 month 0 0 0 3 1 1 0 0 1

T i l e lesions were classified as marked, moderate, o r mild iuvolvement as seen oli gross inspection. A m a r k e d lesion was defined as o n e ill which there was complete circumferential distribntion and tile left ventricle was involved extensively from apex to annulus (Fig. 3). A m o d e r a t e lesion was o n e with either circumferential im'olvmnent in only one or two transverse cuts o r with several small focal snbendocardial lesions ill most of tim transverse slices. A mild lesion was one in which there were one o r two focal snbendocardial lesions in o n e or two transverse cuts but with no extension throngll several serial transverse slices. In g r o u p B there were seven cases (58 p e r cent) with m a r k e d lesions and five cases (42 p e r cent) with m o d e r a t e lesions; this g r o u p did not show any mild lesions. In g r o u p A, however, there were 12 cases (33 per cent) with m a r k e d lesions, 14 cases (39 per cent) with m o d e r a t e lesions, and 10 cases (28 per cent) with mild lesions. Tim e x t e n t o f these small mild lesions was c o n f i r m e d microscopically, althougll co,ttraction bands were occasionally seen ill adjacent grossly normal areas. In f o u r patients who survived for 14 to 18 days, tile hemorrhagic necrosis began to show resorption with f o r m a t i o n o f granulation tissue. Six patients survived for two m o n t h s to four }'ears. In t h r e e tile lesions were r e p r e s e n t e d by a semicircumferential thin band o f depressed gray-pink scar tissue in the inner half o f the myocardium, thus resembling subendocardial ischemic infarcts (Fig. 4). In only one o f these cases, the heart transplant, was there

HEMORRHAGIC

NECROSIS OF THE

M Y O C A R D I U M - - G o T L I E [ ~ E.T aL.

IIIJllllJlJllllJllllJliJllllJlllllllllllilillllllllliiiJliliilillllllllltliilll Figure 1. Transverse slice of heart with a marked gross lesion of concentric hemorrhagic necrosis involving the inner half of the left ventricular myocardium, including the papillary muscle and trabeculae carneae, in a patient dying 2-t hours after surgery.

LUETR,JC 11

1lllllllltillllilfltlllllilllllll[llltlIIIlIIltIlIllIlltlIIIIlilll

Figure 2. Transverse slice of heart with a moderate gross lesion of concentric hemorrhagic necrosis in which tile papillary muscles and trabeculae carneae are largely spared in a patient dying two days after surgery.

31

HUMAN

I)ATHOI~O{;Y-VOLUME

1 Jamm)y 1977

8, N U M B E R

Figure 3. Transverse slices o f the heart with a marked lesion of concentric hemorrhagic necrosis extending through all levels of the left ventricular myocarditm~ in a patient dying two days after surgery.

I

~.::~mI~= ~"

, ~ " ~ '

)

. , ' ~ " " " ' . :.

~ . . . . : , ?~ . .....

"

~ ,'5.~. )~':.::"-

"~---~'.~ ."~ ; , . ' - ~ ' ~ r

3~ 4.".-- - .~ ,~"~,~, ~_'7-~.::-~ ~.. - " "

~,:~.

" ~'r

"-'

--

.

=.

". . . .

-;'~

-.'~

.,~'~'~..~ I

~:;

a ~ I )

' '(

_

Figure 4. ttistologic section of left vemricular myocardium showing fibrosis of tile inner half of tile myocardium, including the trabeculae carneae and papillary muscle, in a patient dying two months after surgery. (Hematoxylin and eosin stain. X40.)

32

HEMORRHAGIC

NECROSIS OF THE

M Y O C A R D I U M - - G O T L I E B ET AL.

Figure 5. Gross and histologic sections of left ventricular myocardium showing focal, somewhat translucent loose fibrous tisst,e (arrows) at the boundary of inner third and outer two-thirds of the left ventricular wall sparing the trabeculae carneae and papillary muscle in a patient dying two months after surgery. (Hematoxylin and eosin stain, x40.)

significant coronary atherosclerosis. O f the other three cases, two showed a narrow band of scar sharply defined fiom the adjacent myocardium at the middle third of the left ventricular wall, undoubtedly representing a healed lesion (Fig. 5), and the other showed only focal subendocardial fibrosis with no significant coronary artery atherosclerosis.

Microscopic Fin dings There are five types of light microscopic clmnges described in tile literature, namely, contraction bands, subendocardial hemorrhage, co agulative necrosis, calcification, and subendocardial fibrosis. ~ All these lesions were found in varying degrees and various combinations in this series. Contraction bands and interstitial edema were most prevalent in the patients who died within the first 24 hours (Fig. 6). Interstitial llemorrhage and coagulative necrosis accompanying a mild degree of leucocytic infiltration and fibrinoid necrosis o f the blood vessels in the diseased myocardium were characteristic and constituted the hallmarks of the full blown lesion in patients who survived for one to ten days. In earlier lesions the interstitial hemorrhage was associated with contrac-

tion bands occurring in tile same area (Fig. 7). In more advanced lesions tile Ilemorrhage was most often present within necrotic myocardinnn Tile more advanced lesions showed varying degrees of repair with resorption of necrotic muscle, infiltration with macroplmges, proliferation of fibroblasts, anti formation of granulation tissue (Fig. 8). By two to three months a typical band of scar tissue was left as the residual stignm in the subendocardial region of the myocardium. These areas were often clmracterized by hemosiderin laden macrophages. Patients who showed healed lesions did not have any lfistory of the dranmtic postoperative low output syndrome. Fibrin platelet thromboemboli and occasional foreign body emboli were seen in the intramyocardial coronary arteries. Platelet thrombi were found in the areas of hemorrhabdc necrosis in 20 of the 25 cases in group A and in four of the eight patients in group B dying within the first three days. The number of tbrombi identified in each of the four cases in g r o u p B was markedl)' less than those in group A. Except for a patient with disseminated intravascnlar coagulation, platelet thrombi were rarely seen in patients who survived beyond four days. However, in most cases the n n m b e r o f

33

H U M A N I ' A T H O L O G Y - - V O 1 . U M E 8, N U M B E R I Jamm U 1977

Figure 6. tlistologic section of subendocardial region showing the presence of irregular transverse contraction bands in tile myocardial fibers in a patient dying '24 hours after surgery. (llmnatoxylin ~.ltlC[ COSillstain. x250.) thrombi s c a n t y to Two shortly

34

s e e m e d to b e d i s p r o p o r t i o n a t e l y a c c o u n t f o r t h e e x t e n s i v e lesions. patients from group A who died after the surgical procedure

s h o w e d n u m e r o u s p l a t e l e t a g g r e g a t e s in s m a l l a r t e r i e s a n d c a p i l l a r i e s . I t was e s t i m a t e d t h a t in a cross s e c t i o n a l a r e a o f 2.5 cm. t h e r e w e r e at least 200 foci o f

Figure 7. ttismlogic section of subendocardial region showing association of corm'action bands and interstitial henlorrhage in a patient dying two days after surgery. (! Icmatoxylin and eosin stain. ::<250.)

HEMORRHAGIC

NECROSIS OF THE

M Y O G A R D I U M - - G o T L X E B ET ALo

Figure 8. ttistologic scction-of subendocardial region showing mononuclear.cell infiltration and proliferation of granulation tissue in a patient dying 14 days after surgery. (ltematoxylin and cosin stain, x250.)

platelet aggregates obstructing the microvasculature. In these two patients, in whoin platelet thrombi were most massive in the heart, a few platelet thrombi were f o u n d in the p u h n o n a r y arteries, kidneys, and brain. However, there was no evidence o f infarction in these organs. In one patient who died 10 days following s u r g e r y with diffuse intravascular coagulation and P s e u d o m o n a s pnetnnonia, the intramyocardial .microvasculature showed fresh platelet thrombi as well as platelet thrombi that had u n d e r g o n e organization and endothelialization, indicative o f a process occurring several days prior to death. T h e hearts o f patients who had u n d e r gone saphenous vein bypass s u r g e r y were markedly different f r o m those in most o f the cases o f valvular s u r g e r y in that the coronary arteries showed severe atherosclerosis with significant stenosis generally involving all three vessels. T h e myocardium showed gross and lnicroscopic evidence o f dense scarring characteristic of r e m o t e infarcts, as well as m a n y small recent microinfarcts with proliferation o f fibroblasts, infiltration o f macrophages, and early fibrosis. In three cases t h e r e was

significant snbendocardial fibrosis in the left ventricular myocardium. Six o f 11 patients showed recent lesions with comraction bands, interstitial lleinorrhage, and coagulative necrosis o f myofibers. Five died within f o u r days and tile sixth patient died nine days following surgery. In all cases the lesions were recognized grossly as being either m a r k e d l y or inoderately henlorrhagic; however, the h e m o r r h a g e was m o r e focal arid m o r e irregularly continent than that in the valve surgery patients. Microscopically the contraction bands, h e m o r r h a g e , and necrosis were often present focally in all layers, althougll there was a definite tendency for subenocardial distribution to occur. In the cases with subendocardial fibrosis, the lesions were often adjacent to the fibrotic area, where h e m o r r h a g e s were occasionally seen. Platelet t h r o m b i were sometimes f o u n d in vessels in the necrotic myocardium. DISCUSSION

Tile clinicopathologic entity o f concentric h e m o r r h a g i c necrosis is well docu-

35

HUMAN I'ATItOLOGY--VOLUME 8, NUMBER 1 Janualy 1977

36

mented in our large series of autopsies. We have been able to follow the morphologic evolution of the lesion, confirming the findings of Henson et al? We have shown tlmt the disorder is less fi'equent now than in the past in cardiopulmonary bypass patients coming to autopsy. This may be because patients are being operated upon earlier in the course of the disease and because some patients with advanced myocardial damage are being advised not to undergo surgical treatment. In addition, the importance of normal coronary artery perfusion pressure is now recognized as being desirable in all patients and essential in those with coronary arteriosclerosis, since in tiffs latter group flow may be totally pressure dependent. I f the perfusion pressure is suboptimal, ventricular fibrillation during cardiopulmonary bypass may direct blood away from the subendocardial zone and hypertroplfied myocardium is particularly vulnerable. 3 Although the lesions in the patients undergoing aortocoronary bypass surgery were clearly subendocardial in distribution, there were similar focal lesions seen througlaout all layers of the myocardinm. This probably reflects the increased vulnerability of the entire myocardium in patients with extensive coronary artery disease. T h e etiology of concentric hemorrlmgic necrosis cannot be ascertained from the morphologic data presented. The current theory is that hypoxia plays a major role.4 T h e endocardial myocardium has a vulnerable blood supply and thus can be more easily made hypoxic. Thus a possible explanation for the subendocardial location of the lesion is that this area is rendered hypoxic during the surgical procedure. Studying the conditions present during cardiopulmonary bypass, Buckberg 4 has shown experimentally tlmt in the hypertrophied heart undergoing ventricular fibrillation, the subendocardial perfusion is decreased out of proportion to the oxygen demands of that region, leading to anoxia. In two cases that we studied by postmortem angiography we found, as was described by Taber, 5 that there was an absence of microvascular filling in the inner half of the myocardinm where areas of gross hemorrlmgic necrosis were

present. Whether this lack of filling is a cause of tile hypoxia o1"a resuh of it cannot be ascertained. The only morphologic basis for the lack of perfusion was the presence of platelet thrombi in the small arterioles o f the involved subendocardial region. In reviewing our cases chronologically, however, we noted a definite decline in the presence of platelet thrombi when comparing group B to group A. Henson ~ also noted fewer microthrombi than was reported earlier by Morales.6 There is a striking resemblance between the microscopic features of concentric hemorrlmgic necrosis and those of lesions produced by transient anoxia and reperfusion. Experiments in dogs with reperfusion after coronary artery occlusion have shown lesions characterized by myofibrillar fragmentation, hemorI'hage, and necrosis. In addition, microt b r o m b i were seen in the areas o f infarction5 These changes were not evident after three hours of occlusion, but 9began to appear after reperfusion. It has also been shown that experimental occlusion followed by reperfnsion increases the rate of myocardial necrosis, s It is possible that there may be two critical events occurring in the development of the morphologic lesion of concentric hemorrhagic necrosis. The first may be a relative decrease in perfilsion in the subendocardial region. The etiology for this is not known and is not apparent from our stud),. The second stage is one of accelerated necrosis due to reperfusion characterized by marked interstitial llemorrhage and necrosis. Tiffs would imply injury to the microvasculature. In a study not included in this report, we compared the ultrastructure of the right and left subendocardial microvasculature in postmortem material in two cases and found varying degrees of endothelial injury characterized by endothelial denudation exposing the subendothelial collagenous fibrils or basement membrane. Leucocytes and platelets were seen adhering to the vessel wall. These clmnges were definitely more prominent in the left ventricular vessels than in the right. We feel that ventricular wall biopsy will be necessary to further investigate the nature of the injury in the nficrovasculature. There is another lesion referred to as

HEMORRHAGIC

N E C R O S I S O F T H E M Y O C A R D I U M - - G O T L I E t ~ ET AL.

"stone heart" in which the heart goes into a contractile state during operation and cardiopuhnonary bypass cannot be discontinued because of inadequate cardiac output? These hearts show severe myocardial hypertrophy and significant interstitial fibrosis as well as contraction bands. We have not encountered such lesions in o u r series o f patients.

3.

4.

5.

ACKNOSVLEDGMENT

T h e authors wish to acknowledge the technical help o f Mrs. Hassmig Minassian and Mr. David More.

REFERENCES 1. Henson, D. E., Najafi, M. D., Callaghan, R., Coogan, P., Julian, O. C., and Eisenstein, R.: Myocardial lesions following open heart surgery. Arch. Path., 88:423, 1969. 2. Najafi, H., Henson, D., Dye, W. S., Javid, H., ttunter, J. A., Callaghan, R., Eisenstein, R., and

6. 7.

8.

9.

Julian, O. C.: Left ventricnlar hemorrhagic necrosis. Ann. Thorac. Surg., 7:550, 1969. Becker, R. M., Shizgal, H. M., and Dobell, A. R. C.: Distribution of coronary blood flow during cardio-pulmonary bypass in pigs. Ann. Thorac. Surg., 16:228, 1973. Itottentrott, C. E,, To~vers, B., Knrkji, II. J., Maloney, J. Y., and Buckberg, G.: The hazard of ventricular fibrillation in hypertrophied ventricles during cardio-pulmonary bypass. J. Thorac. Cardiovas. Surg., 66:742, 1973. Taber, R. E., Morales, A. R., and Fine, G.: Myocardial necrosis and the postoperative low cardiac output syndrome. Ann. Thorac. Surg., 4:12, 1967. Morales, A. R., Fine, G., and Taber, R. E.: Cardiac st, rgery and myocardial necrosis. Arch. Path., 83:71, 1967. I~'mg, T. W., Corday, E., Gold, If., Meerbaum, S., Rubins, S., Costantini, C., Hirose, S., Osher, J., and Rosen, Y.: Consequences of reperfusion after coronary occlusion. Effects on hemodynamic and regional myocardial metabolic ftmction. Am. J. Cardiol., 33:69, 1974. Herdson, P. B., Sommers, 1t. M., and Jennings, R. B.: A comparative study of the fine structure of normal and ischemic dog myocardium with special reference to earl)" changes following temporary occlusion of a coronary artery. Am. J. Path., 46:367, 1965. Cooley,D. A., Reul, G. J., and Wukasch, D. C.: Ischemic contracture of the heart: "'stone heart." Am. J. Cardiol., 29:575, 1972.

Department of Pathology Pathology Institnte McGill University 3775 University Street Montreal, Quebec H3A 2B4 Canada (Dr. Hnang)

37