Br Heart J 1988;59:196-200

Coronary thrombosis and platelet/fibrin microemboliin death associated with acute myocardial infarction

RICHARD J FRINK, PIERCE A ROONEY Jr, JAMES O TROWBRIDGE,JAMES P ROSE*

From the *Department of Cardiovascular Research and the Department of Pathology, Mercy Hospital ofSacramento, California, and the Division of Forensic Pathology, Sacramento County Coroner's Office

SUMMARY The frequency and clinical significance of platelet/fibrin microemboli in the micro-circulation were investigated in 24 patients whose deaths (before and during hospital admission)were associated with acute myocardial infarction. An acute coronary thrombus was present in allthe hearts. In nine hearts an acute thrombus was found in more than one major epicardial coro-

nary artery. A total of 35 acute thrombi were found in the 24 hearts. Platelet/fibrin microemboliwere found in 19 (79%) hearts. Eighteen patients died in hospital. The hearts of 16 of these cases

showed microemboli; 16 had important arrhythmias or various forms of heart block; 13 showedacute pathological changes in the conduction system. Fourteen of the deaths in hospital were

primarily the result of cardiogenic shock and four were primarily caused by arrhythmia. Six of thedeaths that occurred before admission to hospital were regarded as being arrhythmic in origin.Three of these showed microemboli and the other three had acute pathological changes in theconduction system. Microemboli were found in two (24%) of 12 control hearts.Coronary thrombosis was found in most deaths caused by acute myocardial infarction and

platelet/fibrin microemboli were present in the majority of such hearts. These may arise from thecoronary thrombus in the larger upstream vessel supplying the microcirculation.

Myocardial infarction is caused by or is closely asso-ciated with acute coronary thrombosis. Recent stud-ies show that the frequency of occlusion in acutemyocardial infarction decreases during the first 24hours of a transmural infarction.1 These data sug-gest that lysis of thrombi and even of occlusions mayoccur in the course of acute infarction and may con-vert an occlusion into a non-occlusive thrombus.The exposed surface of a non-occlusive thrombusis often covered with friable, loosely attachedplatelet/fibrin aggregates.2 Some of these fragmentsundoubtedly pass distally to the microcirculation.3 4The clinical significance of these microemboli hasnot been fully determined, but they have beenreported in association with lethal rhythm dis-turbances, particularly ventricular arrhythmias,

Requests for reprints to Dr Richard J Frink, Heart Research Foun-dation of Sacramento, 3900 J Street, Sacramento, California 95819,USA.

Accepted for publication 13 August 1987

myocardial infarction, and sudden cardiac death.5 - 7This report will show that many patients who die ofacute myocardial infarction not only have acutecoronary thrombosis, but also have platelet/fibrinmicroemboli in the microcirculation, commonly dis-tal to an acute coronary thrombus. In addition, mostof these patients have important arrhythmias, andthey often have acute pathological changes in theconduction system.

Patients and methods

The hearts of 24 patients (17 men and seven women,aged 32-93 (mean 59)) who died of a microscopicallyconfirmed acute myocardial infarction were subjec-ted to a detailed pathological study of the coronaryarteries, conduction system, and myocardium. Sixpatients died suddenly out of hospital and theirhearts were obtained from the Sacramento Countycoroner's office. No electrocardiographic data wereavailable for these six patients, and their deaths areassumed to have been caused by arrhythmia. None

196

on 17 February 2019 by guest. P

rotected by copyright.http://heart.bm

j.com/

Br H

eart J: first published as 10.1136/hrt.59.2.196 on 1 February 1988. D

ownloaded from

Coronary thrombosis and platelet/fibrin microemboli in death associated with acute myocardial infarction 197of them had evidence of myocardial rupture or con-gestive changes in the lungs. Eighteen patients diedafter admission to the hospital and for all of them wehad monitor strips or electrocardiograms showingthe presence or absence of arrhythmias. Thirteenpatients had been given anticoagulants, heparin, orwarfarin.

Hearts were selected on the basis of (a) micro-scopic evidence of acute myocardial infarction; (b)satisfactory injection of the coronary arteries with abarium gelatin mass; (c) absence of previous heartoperation, coronary angioplasty, or streptokinase.No patients meeting these criteria were excluded.The cause of death in 14 hospital patients was car-diogenic shock; a lethal arrhythmia, confirmed elec-trocardiographically had caused the deaths of theother four hospital patients. The infarctions wereclassified as transmural (full thickness of the ventric-ular wall) or non-transmural.We used a postmortem, angiographic, photo-

graphic technique with a coloured barium gelatinmass to study the hearts.2 Briefly, hearts wereobtained fresh and uncut at the necropsy table, thecoronary arteries were cannulated, and injected witha coloured barium gelatin mass. All hearts wereinjected within 24 hours of death.

After fixation in formalin, the entire length of allmajor epicardial coronary arteries was dissectedfrom the heart and decalcified. X rays were taken ofthe dissected coronary arteries before and afterdecalcification. Figure 1 shows a typical bariumfilled coronary artery. Arteries were then cut every2-3 mm, embedded in paraffin and sectioned at6 gm. These coronary artery sections were stainedwith haematoxylin eosin and also with Martius scar-let blue stain to detect fresh fibrin. Ten or moreblocks of myocardium were taken from standardisedsites and any suspected areas of infarction. Blocks oftissue containing the sinus and atrioventricularnodes, His bundle, and proximal bundle brancheswere removed and embedded for histological study.All sections of myocardium, atrioventricular node,and sinus node were stained with haematoxylineosin and Martius scarlet blue. All microscopicalsections were thoroughly searched for platelet/fibrinmicroemboli. The colouring material used in theinjection mass identified the myocardial distributionof a given coronary artery. Thus fragments found inthe distribution of a given coronary artery wereassumed to have originated in or to have passeddown through this artery.An acute coronary thrombus in the main epicar-

dial arteries was defined as an attached, organisedmass of fibrin and platelets that stained appropri-ately with the Martius scarlet blue stain.Platelet/fibrin microemboli were identified micro-

Fig 1 x ray taken ofan epicardial right coronary arteryafter decalcification. Asterisk indicates area close to coronaryostium and arrows show the site offilling defects proximaland distal to the obstruction, which was confirmed to be anacute thrombus.

scopically as organised groups of platelets and fibrin,with clearly identifiable fibrin strands, present in asmall myocardial arteriole. Platelet aggregates with-out fibrin were not counted in this study. Post-mortem fibrin was distinguished as a looser, morefilamentous structure, with no organisation of thefibrin.

CONTROL PATIENTSThe hearts from 12 men (aged 19-8 1, mean 36) wereexamined to determine the frequency of micro-emboli in patients whose deaths were caused by non-cardiac disease, accidents, or violence (n = 10) andalso those whose deaths were cardiogenic but non-coronary in origin (n = 2; myocarditis, hypertrophicsubaortic stenosis). Of the 10 patients dying of non-cardiac disease, accidents, or violence, six heartswere injected in a manner identical to that describedabove and four were processed and examined with-out the use of injection material. The two cardio-genic control hearts were injected with the barium

on 17 February 2019 by guest. P

rotected by copyright.http://heart.bm

j.com/

Br H

eart J: first published as 10.1136/hrt.59.2.196 on 1 February 1988. D

ownloaded from

198

gelatin mass. All controi hearts were preparedwithin 24 hours of death. Three patients, all ofwhom had died of non-cardiogenic causes, diedwithin six days of admission to the hospital. Theremainder were all out-of-hospital deaths.

Results

The table shows the pathological findings in all 24cases, subdivided according to transmural or non-

transmural infarction. The term "global infarction"indicates that acute infarction was found throughoutthe myocardium. An acute coronary thrombus was

found in all the hearts. These thrombi were firmlyattached to the arterial wall and had irregular sur-

faces with many loosely attached fragments. In ninehearts an acute coronary thrombus was found inmore than one epicardial coronary artery. In the 24patients, 35 acute coronary thrombi were found inseparate major epicardial coronary arteries. Theinfarction was distal to an acute thrombus in eachheart, with the exception of those with globalinfarction. Microemboli were present in 19 (79%) ofall hearts, and most were found in the distribution ofan artery that contained an acute thrombus. Figure 2shows the appearance of a typical microembolus.

Frink, Rooney, Trowbridge, Rose

Microemboli were also found in three patients withleft ventricular mural thrombi. In two of thesepatients, microemboli were found in the distributionof an artery that did not contain an acute coronary

thrombus.Although all hearts contained a coronary throm-

bus, emboli were more common in patients withtransmural infarction than in patients with non-

transmural infarction. The difference, however, was

not significant (p > 010). Five of the six patients

with non-transmural infarction died outsidehospital: three had microemboli, and four hadpathological changes in the conduction system.

Eighteen deaths occurred after admission to thehospital. All died within 17 days of admission; eightwithin the first 48 hours. There were significantarrhythmias, over and above single atrial or ventric-ular unifocal extrasystoles in 16 patients; five ofthese also developed various forms of new onset

heart block. The forms of heart block included 2:1block, right bundle branch block, left bundle branchblock, atrioventricular dissociation, and junctionalrhythm. There was ventricular tachycardia or ven-

tricular fibrillation (other than as an agonal rhythm)in 12 (60%) patients. Two patients did not show any

type of arrhythmia. The conduction system was

Table Site of thrombus in relation to location of infarction and distal microemboli in patients with transmural andnon-transmural infarction

Site of Site of AnticoagulantsPatient Site of acute thrombosis infarct location microemboli given

Transmural infarction1 LAD Ant None No2 RC Inf Inf/Lat Yes3 LAD Ant Ant Yes4 LAD, RC Inf Ant/Inf/Avn Yes5 LAD Ant Ant/Inf Yes6 LAD Ant Ant No7 LAD Global Ant/Inf/Lat No8 LAD, Circ, RC Global Ant/Inf No9 LAD, Circ, RC Global Ant/Lat No10 LAD Ant Ant/Lat Yes11 LAD Ant None Yes12 RC Inf Inf/Sep/Lat Yes13 LAD, RC Inf Ant/Inf/Sep Yes14 LAD Ant Ant/Sep/Inf Yes15 LAD, RC Global Ant/Inf Yes16 Circ, RC Inf Inf Yes17 LAD RC Ant/Lat Ant/Inf/HB Yes18 LAD Global Ant/Lat YesSubtotal 15 3 9 16

Non-transmural infarction19 LAD RC Inf None No20 LAD Ant Septal No21 LAD Ant/Inf None No22 LAD RC Inf None No23 RC Inf Inf No24 LAD Septal Inf/Lat NoSubtotal 5 3 3Total 20 3 12 19 (79"O)

LAD, left anterior descending coronary artery; RC, right coronary artery; Circ, circumflex coronary artery; Ant, anterior; Lat, lateral;Global, infarction found diffusely throughout the myocardium; Inf, inferior; Sep, septal; Avn, atrioventricular node; HB, His bundle.

on 17 February 2019 by guest. P

rotected by copyright.http://heart.bm

j.com/

Br H

eart J: first published as 10.1136/hrt.59.2.196 on 1 February 1988. D

ownloaded from

Coronary thrombosis and plateletlfibrin microemboli in death associated with acute myocardial infarction 199

not affect the detectability of the microemboli. Theconduction system was examined in all controlhearts and no acute changes were found.

Discussion

Fig 2 Photomicrograph showing small intramyocardialarteriole containing a microembolus (black arrow).M, myocardium; P, platelets. Solid black arrows showfibrinstrands forming islands ofplatelets.

abnormal in 13 (72%) patients. The crux of theheart was supplied by the right coronary artery in 23hearts.Of 13 patients on anticoagulant treatment, 11

were considered to be fully anticoagulated, and ofthese, 10 showed microemboli.

CONTROL STUDIESCoronary thrombosis was not found in any of thecontrol hearts and there was no evidence of coronarydisease in any of them. Two patients showed a sin-gle, but well defined, platelet/fibrin microembolus inthe myocardial microcirculation. These two patientshad been admitted to hospital for three and six daysbefore death, both had severe intracranial hae-morrhage, and both hearts were injected with thebarium gelatin mass. No microemboli were found inany of the other control patients, whether or not theyhad been injected with the barium gelatin mass.

This indicates that the use of the injection-Mass does

PLATELET/FIBRIN MICROEMBOLI IN THEMICROCIRCULATIONWe found platelet/fibrin microemboli in mostpatients who died of acute myocardial infarction,especially of transmural infarction. We found thatthe more comprehensive the search for microemboli,the more we found. There are three possible sourcesof the platelet/fibrin microemboli and these are notnecessarily mutually exclusive. First, the formationof microemboli may be an agonal event or may occursoon after death. Platelet aggregation during acuteillness has been reported8; for this reason, pureplatelet aggregations were not counted in this study.The fragments we considered to have been formedbefore death were tightly knit, compact, well-formed fragments of fibrin, with or without platelets(fig 2), unlike the loose, unorganised appearance ofpostmortem fibrin.A second possibility is that microemboli may

occur as a result of increased platelet reactivity,which is sometimes seen in patients with acute myo-cardial infarction.9 The condition is, as yet, poorlydefined, and there are no pathological studies forcomparison. Except in the two cases of intracranialhaemorrhage, however, microemboli were onlyfound in association with a thrombus. Thus it isunlikely that they represent a primary event.The third possibility is that the microemboli may

originate from an acute coronary thrombus locatedin the upstream coronary artery. Most microemboliin this study were distal to a coronary thrombus,which could have served as a source for these frag-ments. In addition, those patients with ventricularmural thrombi had microemboli in the micro-circulation in the distribution of an artery that didnot contain a thrombus. Furthermore, in a previousstudy of coronary thrombosis and sudden death,2platelet/fibrin microemboli were found in mostcases. The thrombi in this and the current studywere characterised by their irregular surface andloosely attached fragments. Clearly, the thrombus isin a dynamic state of change undergoing episodicformation and dissolution as a result of the flow ofblood, platelet disaggregation, and fibrinolysis.1 4 10The body of evidence thus supports the notion thatplatelet/fibrin microemboli originate from a prox-imal thrombus.Microemboli were found in the microcirculation

irrespective of the length of hospital stay. But nei-ther-the time of embolisation nor the state of the

on 17 February 2019 by guest. P

rotected by copyright.http://heart.bm

j.com/

Br H

eart J: first published as 10.1136/hrt.59.2.196 on 1 February 1988. D

ownloaded from

200 Frink, Rooney, Trowbridge, Rosefibrinolytic system at the time of death were known.None the less, as all patients died as a result of amyocardial infarction and the majority of thepatients had microemboli, it is possible thatplatelet/fibrin microemboli contributed to the hae-modynamic and/or electical instability in thesepatients. The present data show that suchembolisation is a possibility irrespective of the age ofthe infarction.

Anticoagulation seemed to have little effect onthe presence or absence of microemboli. Theeffectiveness, therefore, of anticoagulants to retardthe thrombotic process and reduce the frequency ofmicroemboli is questionable.

CLINICAL MANIFESTATIONS OF MICROEMBOLIAn important question is the relation betweenmicroemboli and death associated with acute myo-cardial infarction. Numerous investigators have pro-vided evidence that microemboli may be involved infocal fibrosis,1' subendocardial injury,'2 unstableangina pectoris,'3 and sudden cardiac death.267 Toour knowledge, however, no correlation betweenmicroemboli and death associated with acute myo-cardial infarction has yet been made. The presentfinding of a high frequency of microemboli indicatesthat the two might be related. The above evidencealso suggests that microemboli do have pathogeneticpotential and that they are apparently capable ofcausing various clinical symptoms.The relation between microemboli and arrhyth-

mias is not clear. Important arrhythmias or recentonset heart block, assuming that all out-of-hospitaldeaths were arrhythmic in origin, were present in92% of patients in this study. Microemboli werefound in 79%. Most (61%) arrhythmias were ven-tricular in origin. Conduction system abnormalitywas also present in 61% of cases. The occurrence ofarrhythmias, therefore, correlated reasonably wellwith the presence of microemboli, as well as withsome types of conduction system abnormalities. Thepathological findings, however, could not be cor-related with specific arrhythmias. In several casesmicroemboli were found in or immediately adjacentto the conduction system. This evidence does notprove that microemboli caused the arrhythmias orheart block, but neither does it rule out that possi-bility. At the very least microemboli should beconsidered a possible mechanism for producingtransient, severe, reversible ischaemia of focal areasof the myocardium and/or the conduction system.

In summary, platelet/fibrin microemboli in themicrocirculation are common in patients dying ofacute myocardial infarction and they seem to bederived from a proximal thrombus. Arrhythmiasand acute pathological changes in the conduction

system are also common in these cases; whether ornot the arrhythmias are related p platelet/fibrinmicroemboli is unknown and requirfs further study.

This study was supported in part by the HeartResearch Foundation of Sacramento and MercyHospital of Sacramento. We thank Ken Shiba forphotographic assistance and Dr Louis Ostrach forreviewing the article.

References

1 Dewood MA, Spores J, Notske R, et al. Prevalence oftotal coronary occlusion during the early hours oftransmural myocardial infarction. N Engl J Med1980;303:897-902.

2 Frink RJ, Trowbridge JO, Rooney PA. Nonobstructivecoronary thrombosis in sudden cardiac death. Am JCardiol 1978;42:48-51.

3 Mustard JF. Platelets and thrombosis in acute myo-cardial infarction. Hosp Pract 1972;7:1 15-28.

4 Jorgensen L. The role of platelet embolism from crum-bling thrombi and of platelet aggregates arising inflowing blood. In: Sherry S, Brinkhous KM, GentonE, Stengle JM, eds. Thrombosis. Washington, DC:National Academy of Sciences, 1969:506-33.

5 Jorgensen L, Rowsell HC, Hovig T, Glynn MF, Mus-tard JF. Adenosine diphosphate-induced plateletaggregation and myocardial infarction in swine. LabInvest 1967;17:616-44.

6 El-Maraghi N, Genton E. The relevance of platelet andfibrin thromboembolism of the coronary micro-circulation, with special reference to sudden cardiacdeath. Circulation 1980;62:936-44.

7 Davies MJ, Thomas AC, Knapman PA, HangartnerJR. Intramyocardial platelet aggregation in patientswith unstable angina suffering sudden ischemiccardiac death. Circulation 1986;73:418-27.

8 Hampton JR, Mitchell JRA. Abnormalities in plateletbehaviour in acute illnesses. Br Med J 1966;i:1078-80.

9 Mehta P, Mehta J. Platelet function studies in coronaryartery disease. V. Evidence of enhanced plateletmicrothrombus formation activity in acute myo-cardial infarction. Am J Cardiol 1979;43:757-60.

10 Kinlough-Rathbone RL, Packham MA, Mustard JF.Vessel injury, platelet adherence, and platelet sur-vival. Arteriosclerosis 1983;3:529-46.

11 Mills NL, Ochsner JL. Distal thromboembolism andproximal coronary arteriosclerotic lesions. Surgery1972;72: 1030-6.

12 Moore S, Belbeck LW, Evans G, Pineau S. Effects ofcomplete or partial occlusion of a coronary artery.Lab Invest 1981;44:151-7.

13 Falk E. Unstable angina with fatal outcome: dynamiccoronary thrombosis leading to infarction and/orsudden death. Autopsy evidence of recurrent muralthrombosis with peripheral embolization culmin-ating in total vascular occlusion. Circulation 1985;71:699-708.

on 17 February 2019 by guest. P

rotected by copyright.http://heart.bm

j.com/

Br H

eart J: first published as 10.1136/hrt.59.2.196 on 1 February 1988. D

ownloaded from