Online newspaper of professor Yasser Metwally

The author: Professor Yasser Metwally

http://yassermetwally.com

THE CARDIAC CONSEQUENCES OF STROKE

Just over 50 years ago, Aschenbrenner 1 noted that neurologic lesions can be associated with electrocardiographic (ECG) changes in young patients with presumed normal myocardia and cardiac vasculature. A similar period has elapsed since the first report of the association between acute stroke and ECG changes. 3 Despite this and the numerous individual reports that followed, these observations have either been largely ignored or treated as a curiosity of little significance. Only recently has the physiologic basis of these findings become the subject of intense study, further extending the validity of the clinical observations into the fields of central autonomic control and sudden death.

This article concentrates on the nature of the ECG changes and their probable cause following stroke, their implications for arrhythmogenesis and cardiac dysfunction, and their putative role in clinical outcome and morbidity.

STROKE AND THE HEART

• Electrocardiographic Effects

The first account in the western literature of an association between acute stroke and ECG changes appeared in 1947, 3 Four cases were described, with at least two patients suffering a subarachnoid hemorrhage; precise neurologic details were lacking. ECG patterns suggestive of acute myocardial ischemia predominantly involving the left ventricular endocardium were identified. Subsequently, Burch, 2 studied the phenomenon in a more systematic fashion.

Seventeen abnormal ECGs were separated from recordings obtained from stroke patients admitted to the Charity Hospital of Louisiana throughout 1950. The predominant diagnosis was intracerebral or subarachnoid hemorrhage (14 of 17) based on spinal tap abnormalities. The recordings were obtained between 24 hours and 7 days after the stroke. A triad of changes was illustrated, comprising prolongation of the QT interval, T waves of increased amplitude and duration, and abnormal U waves, especially in the septal leads. Although the study sample was small, the changes were most frequently observed after subarachnoid hemorrhage, followed by intracerebral hemorrhage, and then by ischemic stroke. Subsequent uncontrolled studies indicated that acute hemorrhage, whether intracerebral or subarachnoid, is associated with ECG changes of the kind described by Burch, 2 in 61% to 71% of cases; such changes are observed in 5% to 17% of ischemic strokes. 2,11, 14, 45

Ischemic heart and cerebrovascular diseases frequently coexist in the same patient; moreover, they share similar risk factors. Any ECG changes following stroke could therefore be caused by exacerbation of coincident coronary artery disease. Dimant and Grob, 11 compared recordings from 100 consecutive acute stroke patients (ischemic stroke, intracerebral and subarachnoid hemorrhage) taken within 3 days of admission, with those from age/sex matched patients admitted for carcinoma of the colon. ST depression and prolongation of the QT interval were seven times more frequent in patients after acute cerebrovascular events than in the control patients. T wave inversion and ventricular premature beats were likewise four times more common in the acute stroke group. Although this would suggest an association between stroke and ECG changes, ischemic heart disease was three times more prevalent in the stroke group than among the controls, however.

Lavy and colleagues, 25 specifically addressed the problem of associated cardiac disease as a cause of the ECG changes following stroke. The ECGs of acute ischemic stroke or intracerebral hemorrhage patients with no history of heart disease and a normal recent recording, or patients whose recordings were unremarkable on admission, were analyzed for changes occurring after their cerebrovascular events. Of 52 consecutive patients, 25 had no evidence of previous cardiac disease. An ischemic stroke was suffered by 18 of the 25 patients, and 8 (44%) showed either a recent onset ischemia-pattern ECG or a cardiac arrhythmia (the precise cardiac details were scant). Of the seven patients with hemorrhagic stroke and no evidence of previous cardiac disease, five demonstrated new onset cardiac arrhythmias.

In a further attempt to control for the effects of concomitant coronary artery disease, Goldstein, 15 compared the ECGs of 53 acute stroke patients (subarachnoid and intracerebral hemorrhage, and ischemic stroke) taken within 24 hours of admission, with tracings taken an average of 4 months earlier. A control group comprised 63 age sex matched patients admitted for reasons other than stroke or a cardiac cause whose previous tracings were also available. Abnormal prolongation of the QT interval not seen in previous recordings was observed in 32% of the stroke group and in 2% of the controls. New T wave inversion was apparent in 15% of the stroke group and abnormal U waves in 13%; neither appeared as a new feature in the admission ECGs of the control group. These differences were highly significant. All three ECG abnormalities occurred together in 8% of the stroke group and in 1% of the control group (in this latter case associated with hypokalemia). It seems unlikely that the new ECG findings were entirely due to the coincidence of acute cerebrovascular or cardiac events or that the strokes in all these patients were due to a cardioembolic cause.

Direct evidence confirms that ECG changes may occur even in the presence of normal coronary arteries and in the absence of acute ischemic changes. In Cropp and Manning’s, 8 series of 29 patients who had developed ECG changes following subarachnoid hemorrhage, 8 patients died, 5 of whom were autopsied. None of these patients showed evidence of coronary artery disease or myocardial ischemia. Similar findings were noted by Shuster, 41 and Tobias et al, 48. In Goldstein’s study, 8 of the 37 patients who died were autopsied. All of these patients had elevated creatine phosphokinase (CPK) levels and died from an intracerebral hemorrhage. No evidence of an acute ischemic cardiac event was found. In Fentz and Gormsen’ s, 14 series of acute cerebral infarctions confirmed by autopsy, none of the patients who died and who had ECG abnormalities showed evidence of coronary occlusion. Connor 7 also reported a patient who died following stroke without autopsy evidence of coronary stenosis.

The ECG effects induced by acute stroke are often evanescent, resolving with little residuum over a period of days to months. 14, 10, 41 This argues against myocardial infarction as a cause of the changes that would be expected to produce a persistent ECG deficit in the majority of cases.

• Cardiac Arrhythmias

The ECG effects described indicate changes in ventricular depolarization. This may be associated with a propensity to the development of cardiac arrhythmias, especially of a ventricular nature. Lavy et al, 25 demonstrated a 39% incidence of new arrhythmias (including ventricular arrhythmias) in the admission ECGs of acute ischemic stroke and intracerebral hemorrhage patients not known to have previous heart disease. In a similar study, Goldstein, 15 found a 25% incidence of new arrhythmias after acute stroke admissions of all types, compared to 3% in the control group. Of these cardiac arrhythmias, the most common was atrial fibrillation that occurred with a frequency of 9%. Postadmission recent-onset atrial fibrillation was not found in the control group. Of course, it is conceivable that strokes in this group of patients may have been caused by embolization from the heart attendant on the development of atrial fibrillation. There was no difference in the incidence of new ventricular arrhythmias between the stroke and control groups. In a study of intracerebral hemorrhage to which group patients were admitted if there was no history of heart disease and in whom a recent ECG taken before the neurologic event was unremarkable, the incidence of ventricular arrhythmias was 10%, 50 This correlated with the demonstration of a temporoparietal location of the hemorrhage on computed tomography (CT) scan. The postevent cardiac rhythm status was assessed using single ECGs taken at an unspecified time. In a study by Dimant and Grob,11 in which the ECG taken within 3 days of the patients’ admission was compared with ECGs taken during a similar time period in patients with carcinoma of the colon, atrial fibrillation was the most common arrhythmia (21%), and ventricular arrhythmias occurred at a frequency of 13%. For the control group the figures were 2% for atrial fibrillation and 3% for ventricular arrhythmias. No allowance was made for preexisting heart disease.

In all of these studies the patients were not investigated by continuous cardiac recording. Consequently, this is likely to lead to an under-representation of ventricular arrhythmias, which are often of short duration. Moreover, cardiac embolization from recent-onset atrial fibrillation may account for the apparent frequency of this arrhythmia among stroke cases. These considerations may explain the discrepancy between the nature of the arrhythmias anticipated from the ventricular repolarization changes noted in poststroke ECGs and their observed frequency after stroke.

In two studies, 32,38 of cardiac arrhythmias using Holter data obtained from patients with cerebral infarction, intracerebral hemorrhage, or transient ischemic attacks, the incidence of ventricular arrhythmias was higher than previously reported. In the study by Rem et al, 38 60% of patients demonstrated such a rhythm disturbance. In Norris et al, 32 investigations the figure was 24%. Neither of these studies, however, controlled for preexisting cardiac disease.

In studies of cardiac rhythm disturbances following subarachnoid hemorrhage, ventricular arrhythmias are exceptionally common and correlate with prolongation of the QT interval, 12 The overall incidence of cardiac arrhythmias reaches 98% in this condition, with multifocal ventricular premature beats occurring in 54% of patients, couplets in 40%, and unsustained ventricular tachycardia in 29%. 46 Torsades de pointes, a highly malignant form of ventricular arrhythmia that is difficult to treat, occurred in 4% of such patients. 12 No association with a history of cardiac disease has been noted. The discrepancy between these studies and those just discussed may be due in part to the method of arrhythmia assessment. In subarachnoid hemorrhage patients, arrhythmia frequency was assessed by continuous cardiac monitoring and therefore may have been more likely to detect ventricular rhythm disturbances. Moreover, atrial fibrillation may not be frequently encountered, as embolization from a cardiac source plays no part in causing this condition. On the other hand, the effects of a subarachnoid hemorrhage are seldom unifocal, with vasospasm, raised intracranial pressure, and ventricular and intracerebral hemorrhages often found to produce multiple effects in the neuraxis. This plethora of intracerebral events might be expected to influence cardiac dynamics in a different fashion from a unifocal lesion and to generate a different variety of arrhythmias.

STROKE-INDUCED ELECTROCARDIOGRAPHIC CHANGES AND PROGNOSIS

If stroke-induced ECG effects lead to an increase in the incidence of cardiac arrhythmias, then the prognosis should be adversely affected by their presence. In Goldstein’s study, 15 a mortality of 80% was seen in those patients demonstrating malignant ventricular arrhythmias (tachycardia, fibrillation; asystole), compared to 23% of stroke patients not showing these changes. No other ECG parameter reflected mortality that did not also correlate with a history of ischemic heart disease. Lavy et al, 25 found a mortality of 69% in patients with ischemic or intracerebral hemorrhagic stroke showing recent-onset ECG changes (ST segment or T wave changes, or arrhythmias) on admission. This compared with a mortality of 0% in patients without ECG changes.

In subarachnoid hemorrhage, the presence of QT interval prolongation, of ST wave depression, or of abnormal U waves is associated with the angiographic demonstration of vasospasm and a poor prognosis. 9 There appears to be no correlation with a history of cardiac disease or with the patients’ neurologic state.

Following admission to the hospital, approximately 6% of stroke patients die suddenly and unexpectedly over the ensuing month. 42 This does not appear to correlate with neurologic condition. It is conceivable that these patients may succumb to a fatal cardiac arrhythmia, although this has not been demonstrated with direct monitoring. The evidence so far presented, however, indicates that those stroke patients developing ECG changes may be at risk and possibly should be monitored until these changes regress.

CAUSES OF ELECTROCARDIOGRAPHIC CHANGES FOLLOWING STROKE

The similarity of stroke-induced ECG changes to those induced by hypokalemia, 40 suggests low potassium levels as a possible cause. No association has been found between the occurrence of ECG changes after subarachnoid hemorrhage and plasma potassium levels. 9, 11, 45 Cruickshank et al, 10 however, demonstrated a correlation between ECG changes after subarachnoid hemorrhage and total exchangeable body potassium. These measurements were made at least 1 week after the stroke, during which time other factors such as inadequacy of nutrition might intervene to affect potassium stores. The majority of the potassium values were in the low-normal range and not outside of it. Thus, the significance of this study is questionable.

Considerable evidence has accrued that stroke is associated with myocardial damage, although, as already discussed, this does not appear to be ischemic in nature. Increases in certain plasma enzymes are indicative of cardiac damage. Dimant and Grob, 11 found elevated CPK levels in 29% of patients admitted following acute stroke; none of the control group had raised plasma enzyme levels. The presence of increased levels correlated with an increased incidence of T wave inversion, ST depression, and conduction defects. Mortality was higher in the stroke patients with elevated CPK levels compared with those whose levels were within the normal range. In Goldstein’s study, 15 no correlation was established between ECG abnormalities and raised plasma CPK levels. Mortality was likewise increased in those patients with the higher enzyme levels, however. Neither of these studies specifically measured the cardiac isoenzyme component of the plasma-total CPK levels, so that other factors, such as skeletal muscle damage, may have contributed to the observed in creases; this may account for the discrepancies.

Norris, 33 systematically investigated changes in the cardiac isoenzyme CPK- MB following stroke. Although 44% of all the 230 strokes investigated were found to have raised total CPK levels, the specific cardiac isoenzyme (CPK- MB) was elevated in only 11%. Despite the total plasma CPK being raised in 66% of control patients, no increase in the specific cardiac isoenzyme occurred except in those with obvious causes such as myocardial infarction. A significant correlation was demonstrated between elevation of CPK-MB and the presence of ECG changes and cardiac arrhythmias.

Unlike the situation following myocardial infarction, CPK-MB levels rise slowly after stroke and reach a peak approximately 4 days after the event. 33 This is further evidence against the ischemic nature of the associated cardiac injury.

If myocardial damage indicated by rising CPK-MB levels is not ischemic in origin (based on the time period of enzyme changes and autopsy studies) is there any other pathology that could account for these changes? In 1933 Neuburger, 30 described scattered subendocardial hemorrhages in patients dying during epileptic seizures. Subsequently, these hemorrhages were observed in the hearts of patients dying within a few days of their stroke. 7, 24, 44 In addition, the hemorrhages involved the conducting system. Greenhoot and Reichenbach, 16 noted that the cardiac pathology (termed myocytolysis) following stroke encompassed a wider range of changes than subendocardial hemorrhage. These changes included scattered foci of swollen myocytes surrounded by infiltrating monocytes, interstitial hemorrhages, and myofibrillary degeneration. Moreover, the lesions were centered around intracardiac nerves rather than blood vessels, suggesting that humoral or ischemic factors were not likely the cause.

Changes in the distribution and nature of formazan granules after treatment with nitroblue tetrazolium have been observed in the hearts of patients dying after acute cerebrovascular events. It is believed these precede histologic evidence of myocytolysis. Kolin and Norris, 23 showed such changes in 89% of patients dying following subarachnoid hemorrhage, in 71% of those dying after an intracerebral hemorrhage, and in 52% of ischemic strokes. Similar abnormalities were seen in only 26% of hearts from control patients dying of different causes (septic shock, thyroid crisis). Interestingly, these changes are not demonstrable in the hearts of patients who survive for longer than 2 weeks after their stroke. This partly parallels the time course of the ECG changes seen in this condition.

Myocytolysis has been described in the hearts of patients dying following catecholamine treatment for shock or as a result of pheochromocytoma. 17 Similar changes have been induced in dogs on stimulation of the left stellate ganglion. The involvement of catecholamines in the cause of this pathologic change also results from the finding of identical lesions in the hearts of animals infused with either epinephrine or norepinephrine. 22, 31, 47 Moreover, identical ECG changes to those described earlier may be induced in humans and animals treated with norepinephrine or epinephrine. This has been attributed to a direct effect on cardiac repolarization, independent of reflex changes induced by blood pressure effects resulting from catecholamine infusion. 22 26 Both the ECG changes and the myocardial damage can be abolished following human subarachnoid hemorrhage by treatment with phentolamine and propranolol. 29 It is therefore suggested that in certain stroke cases there is an increase in sympathoadrenal tone with resultant myocardial damage as evidenced by the ECG and CPK changes previously described.

Direct measurements of plasma catecholamine levels following stroke have indicated significant elevations after cerebral infarction or hemorrhage compared to control groups. 28, 37, 49 In the case of cerebral infarction, plasma levels of norepinephrine, epinephrine, and dopamine were unrelated to age, blood pressure, severity of infarction, or level of stress within the stroke group (assessed by plasma cortisol levels). 28 A significant correlation was established between heart rate and plasma norepinephrine levels in both the control and the stroke groups. The location of the stroke, whether cortical or brain stem, did not significantly influence plasma norepinephrine levels.,28 Plasma norepinephrine levels were elevated to a greater extent than plasma epinephrine levels in this study. This, together with the inability to demonstrate a relationship between plasma cortisol and plasma norepinephrine levels in the stroke group, suggests that some of this catecholamine is derived from an extraadrenal source, most likely neural. Consequently, sympathetic neural activity may be increased after stroke, an assumption for which there is evidence from animal studies (see the following discussion)

ELECTROCARDIOGRAPHIC CHANGES AND STROKE LOCATION

There is little available evidence correlating specific stroke location with ECG changes and arrhythmias. Norris et al, 32 demonstrated a statistically significant increase in the occurrence of ventricular and atrial premature beats and atrial fibrillation following hemispheric, as opposed to brain stem, stroke. In a study of intracerebral hemorrhage, left frontal hematomas were associated with prolongation of the QT interval and T wave abnormalities. 50 Sinus bradycardia and premature ventricular beats accompanied nontraumatic temporoparietal hematomas, whereas sinus tachycardia was more common after thalamic or basal ganglia hemorrhages. Brain stem hemorrhages were more likely to be accompanied by atrial fibrillation or premature atrial beats. In this study, patients with preexisting cardiac diseases, including arrhythmias, were excluded.

EXPERIMENTAL ELUCIDATION OF CARDIAC EFFECTS OF STROKE

Recently, the insular cortex, which in humans lies beneath the frontoparietal and superior temporal opercula, has been identified as a site intimately concerned with the elaboration of autonomic function and its integration with behavior. 4 The posterior left insular cortex of the rat contains a site of cardiac chronotropic representation. 34 Phasic microstimulation linked to the R wave of the ECG of rostral sites within this region results in tachycardia, whereas stimulation of caudal sites generates bradycardia. Prolonged phasic stimulation within the insular cortex results in increasing atrioventricular and interventricular heart block, prolongation of the QT interval, ST segment depression, and the ultimate death of the animal in asystole. 36 This correlates with elevation of plasma norepinephrine but not epinephrine (implying a neural source, 27) and the demonstration of cardiac myocytolysis.

Experimental stroke in the cat induced by the application of a ligature around the left middle cerebral artery has been associated with an increase in plasma norepinephrine, epinephrine, and dopamine levels compared to sham-operated controls. 43 Moreover, these changes were seen only when the insular cortex was involved within the infarcted area. Those animals whose strokes spared the left insular cortex demonstrated plasma catecholamine levels comparable to those of sham-operated animals. In the rat stroke model, infarction of cortical territory supplied by the middle cerebral artery, including the insular cortex, was also associated with an increase in circulating plasma catecholamines, as well as with evidence of cardiac myocytolysis. 6 Sham-operated animals showed no such changes in circulating catecholamine levels or evidence of myocytolysis.

Recent studies have indicated increases in somatosympathetic reflexes following stroke in the cat but not in the rat. 18 In addition, increases in renal sympathetic nerve activity have been demonstrated in the rat stroke model. 19

The available experimental evidence indicates that infarction of a large area of cortex including the left insula results in elevation of plasma catecholamines and an increase in sympathetic nerve activity. Whether these infarcted areas include the posteriorly located insular zone containing cardiac control sites is unclear. If so, it would appear that the overall cardiac effect of removal of this area is a disinhibition of sympathetic function. This is evidenced by the association of cardiac myocytolysis (a marker of increased sympathetic activity) with strokes, including the insular cortex in the rat.

If the left insular cortex contains a site of cardiac representation, might there be a difference in the occurrence of cardiac pathology with right- and left-sided lesions?

EFFECTS OF AGE AND STROKE LATERALIZATION

Although most strokes occur in the elderly, and baroreceptor sensitivity decreases while catecholamine sensitivity increases with advancing age, 13, 17, 39 most experimental strokes are carried out in young animals. We studied six groups of Wistar rats, three groups (young, adult, and senescent) undergoing middle cerebral artery occlusion (MCAO), and three corresponding sham-operated control groups. The senescent MCAO group had a 61% mortality rate, and the animals that died showed the greatest increase in sympathetic nerve activity and a significant increase in the QT interval, 5 which we have shown experimentally to be a prelude to fatal cardiac arrhythmias. 35, 36

The side of the experimental stroke also makes a difference. Right-sided MCAO in the rat results in significantly greater increases in plasma norepinephrine and QT interval prolongation than does left-sided MCAO. 19

CONCLUSION

Evidence accumulated over the past 50 years overwhelmingly indicates that acute stroke can damage even a normal heart. In the presence of coronary artery disease, these cerebrocardiac effects are likely to be synergistic. The frequency of ventricular arrhythmias has probably been under-represented in ischemic stroke patients. They are likely to contribute to the sudden, unexpected mortality seen following that condition, as well as to stroke extension due to arrhythmia-induced hypotension.

The insular cortex is involved in the control of heart rate and rhythm. Its inclusion within an area of cerebral infarction in animal models of stroke reproduces some of the ECG and cardiac pathologic changes seen in this condition.

It is suggested that patients with acute stroke should receive continuous cardiac monitoring for 3 days following their event. Those with ECG evidence of ventricular repolarization changes should be monitored until these have resolved. Such patients, as well as those with insular involvement or those with co-existing ischemic heart disease, may be at particular risk for sudden death or stroke extension due to cardiac arrhythmia.

---

References

1. Aschenbrenner R, Bodechtel G: Ueber EKG veranderungen bei hirntumorkranken. Klinische Wochenschrift 17:298-302, 1938

2. Burch GE, Meyers R, Abildskov JA: A new electrocardiographic pattern observed in cerebrovascular accidents. Circulation 9:719-723, 1954

3. Byer E, Ashman R, Toth LA: Electrocardiogram with large upright T waves and long Q-T intervals. Am Heart j 33:796-799, 1947

4. Cechetto DF, Saper CB: Role of the cerebral cortex in autonomic function. In Loewy AD, Spyer KM (eds): The Autonomic Nervous System: Central Regulation of Autonomic Functions. Oxford, Oxford University Press, 1990, pp 208-223

5. Cechetto DF, Wilson JX, Smith KE, et al: Effect of age on cardiac and autonomic function in the rat stroke model. Stroke 21:174, 1990

6. Cechetto DF, Wilson JX, Smith KE, et al: Autonomic and myocardial changes in middle cerebral artery stroke models in the rat. Brain Research 502:296-305, 1989

7. Connor R: Heart damage associated with intracranial lesions. Brit Med J 3:29-31, 1968

8. Cropp Gj, Manning GW: Electrocardiographic changes simulating myocardial ischemia and infarction associated with spontaneous intracranial hemorrhage. Circulation 22:25-38, 1960

9. Cruickshank JM, Neil-Dwyer G, Brice J: Electrocardiographic changes and their prognostic significance in subarachnoid hemorrhage. J Neurol Neurosurg Psych 37:755-759, 1974

10. Cruickshank JM, Neil-Dwyer G, Stott A: Possible role of catecholamines, corticosteroids, and potassium in the production of electrocardiographic abnormalities associated with subarachnoid hemorrhage. Brit Heart J 36:697-706, 1974

11. Dimant J, Grob D: Electrocardiographic changes and myocardial damage in patients with acute cerebrovascular accidents. Stroke 8:448-455, 1977

12. DiPasquale G, Pinelli G, Andreoli A, et al: Holter detection of cardiac arrhythmias in intracranial subarachnoid hemorrhage. Am j Cardiol 59:596-600, 1987

13. Duke PC, Wade JG, Hickey RF: The effects of age on baroreceptor reflex function in man. Can Anesth Soc J 23:111-124, 1976

14. Fentz V, Gormsen J: Electrocardiographic patterns in patients with cerebrovascular accidents. Circulation 25:22-28, 1962

15. Goldstein DS: The electrocardiogram in stroke: Relationship to pathophysiological type and comparison with prior tracings. Stroke 10:253-259, 1979

16. Greenhoot JH, Reichenbach DD: Cardiac injury and subarachnoid hemorrhage: A clinical, pathologic, and physiologic correlation. J Neurosurg 30:521-531, 1969

17. Gribbin B, Pickering TG, Sleight P, et al: Effect of age and high blood pressure on baroreflex sensitivity in man. Circulation Research 29:424-431, 1971

18. Hachinski VC, Smith KE, Ciriello J, et al: The insula and altered sympathetic activity after experimental stroke. In Stober T (ed): Central Nervous System Control of the Heart. Boston, Martinus Nijhoff, 1986, pp 181-186

19. Hachinski VC, Wilson JX, Tichenoff Lj, et al: Asymmetry of sympathetic consequences of experimental stroke. Stroke 22:139, 1991

20. Hoffbrand BI, Morgan BDG: Functional significance of electrocardiographic changes associated with subarachnoid haemorrhage. Lancet 1:844-845, 1965

21. KJouda M, Brynjolfsson G: Cardiotoxic effects of electrical stimulation of the stellate ganglia. Ann N Y Acad Sci 156:271-280, 1969

22. Kolin A, Kvasnicka J: Pseudoinfarction pattern of the QRS complex in experimental cardiac hypoxia induced by noradrenalin. Cardiologia 43:362-370, 1963

23. Kolin A, Norris JW: Myocardial damage from acute cerebral lesions. Stroke 15:990- 993, 1984 24. Koskelo P, Punsar S, Sipila W: Subendocardial hemorrhage and ECG changes in intracranial bleeding. Brit Med j 1:1479-1480, 1964

25. Lavy S, Yaar I, Melamed E, et al: The effect of acute stroke on cardiac functions as observed in an intensive stroke care unit. Stroke 5:775-780, 1974

26. Lepeschkin E, Marchet H, Schroeder G, et al: Effect of epinephrine and norepinephrine on the electrocardiogram of 100 normal subjects. Am j Cardiol 5:594-603, 1960

27. Micalizzi ER, Pals DT: Evaluation of plasma norepinephrine as an index of sympathetic neuron function in the conscious unrestrained rat. Life Sciences 24:2071-2076, 1979

28. Myers M, Norris JW, Hachinski VC: Plasma norepinephrine in stroke. Stroke 12:200- 204, 1981

29. Neil-Dwyer G, Walter P, Cruickshank JM, et al: Effect of propranolol and phentolamine on myocardial necrosis after subarachnoid haemorrhage. BMJ 1:990-992, 1978

30. Neuberger K: Ueber die Herzmuskelveraenderungen bei Epileptikern und ihre Bezichungen zur Angina pectoris. Frankfurter Zeitschrift Pathologie 46:14-42, 1933

31. Niles NR, Zavin JD, Morikado RN: Histochemical study of effects of hypoxia and isoproterenol on rat myocardium. Am j Cardiol 22:381-388, 1968

32. Norris,JW, Froggatt GM, Hachinski VC: Cardiac arrhythmias in acute stroke. Stroke 4:392-396, 1978

33. Norris JW, Hachinski VC, Myers M, et al: Serum cardiac enzymes in stroke. Stroke 10:548-553, 1979

34. Oppenheimer SM, Cechetto DF: Cardiac chronotropic organization of the rat insular cortex. Brain Research 533:66-72, 1990

35. Oppenheimer SM, Hachinski VC, Wilson JX, et al: The insula and cardiac arrhythmias: Implications for stroke. Stroke 21:174-175, 1990

36. Oppenheimer SM, Wilson JX, Guiraudon C, et al: Insular cortex stimulation produces lethal cardiac arrhythmias: A mechanism of sudden death? Brain Research 550:115- 121, 1991

37. Peerless Sj, Griffiths JC: Plasma catecholamines following subarachnoid hemorrhage. Ann Roy Coll Phys Surg (Canada) 5:8-49, 1972

38. Rem JA, Hachinski VC, Boughner DR, et al: Value of cardiac monitoring and echocardiography in TIA and stroke patients. Stroke 16:950-956, 1985

39. Rowe JW, Troen BR: Sympathetic nervous system and aging in man. Endocrine Reviews 1:167-169, 1980

40. Shamroth L: An Introduction to Electrocardiography, ed 5. London, Blackwell Scientific Publications, 1974, p 82

41. Shuster S: The electrocardiogram in subarachnoid hemorrhage. Br Heart J 22:316- 320, 1960

42. Silver F, Norris JW, Lewis A, et al: Early mortality following stroke: A prospective review. Stroke 15:492-496, 1984

43. Smith KE, Hachinski VC, Gibson Cj, et al: Changes in plasma catecholamine levels after insula damage in experimental stroke. Brain Research 375:182-185, 1986

44. Smith RP, Tomlinson BE: Subendocardial hemorrhages associated with intracranial lesions. j Path Bact 68:327-334, 1954

45. Stern S, Lavy S, Carmon A, et al: Electrocardiographic patterns in hemorrhagic stroke. j NeuroI Sci 8:61-67, 1968

46. Stober T, Anstatt Th, Sen S, et al: Cardiac arrhythmias in subarachnoid hemorrhage. Acta Neurochir 93:37-44, 1988

47. Szakas J, Cannon A: L-norepinephrine myocarditis. Am j Clin Path 30:425-434, 1958

48. Tobias SL, Bookatz Bj, Diamond TH: Myocardial damage and electrocardiographic changes in acute cerebrovascular hemorrhage: A report of three cases and review. Heart Lung 16:521-526, 1987

49. Tomomatsu T, Ueba Y, Matsumoto T: ECG observations and urinary excretion of catecholamines in cerebrovascular accident. Jpn Circ j 28:905-912, 1964

50. Yamour Bj, Sridharan MR, Rice, et al: Electrocardiographic changes in cerebrovascular hemorrhage. Am Heart j 99:294-300, 1980