Characterization of the Cardiac Effects of Acute Subarachnoid Hemorrhage in Dogs
Background and Purpose We know that significant cardiac involvement can occur in patients with acute intracranial hemorrhage, particularly in those with subarachnoid hemorrhage. These patients may present with electrocardiographic abnormalities that were previously thought to be benign. However, many die of cardiovascular sequelae, which suggests more serious cardiac problems. To characterize the cardiac, rhythmic, and myocardial disturbances that occur 2 to 4 hours after subarachnoid hemorrhage, we conducted an experimental study using autologous blood (7.9±0.3 mL) injected into the right frontal lobe and subarachnoid space in canines.
Methods Nine adult mongrel dogs were anesthetized with isoflurane and their rectal temperatures maintained at 37°C. Electrocardiogram, heart rate, mean arterial pressure, mean pulmonary artery pressure, and intracranial pressure were continuously measured. Transesophageal echocardiography was performed to assess myocardial wall motion changes and aortic and pulmonary flow velocities before, immediately after, and 2 and 4 hours after intracranial hemorrhage. Blood samples were collected and analyzed for catecholamines and cardiac enzymes, and cardiac output was measured. Animals were killed at 2 to 4 hours after subarachnoid hemorrhage, and a piece of the myocardium was freeze-clamped for analysis of tissue catecholamines. Light and electron microscopy were used for histopathologic assessment.
Results Subarachnoid hemorrhage produced significant increases in intracranial pressure, cardiac output, and aortic and pulmonary flow velocities. Also, significant changes in creatine kinase and catecholamines were observed. Electrocardiographic recordings showed changes of tachycardia, ST-segment depression, inverted T wave, and premature ventricular contractions in four animals at 1 to 5 minutes after injection, and echocardiographic changes were evident in all animals at 20 to 240 minutes. Microscopic examination of the heart showed evidence of myocardial changes in one animal with the use of light microscopy and in nine with the use of electron microscopy.
Conclusions This study demonstrates the high incidence of cardiac involvement, specifically wall motion abnormalities, that occur after subarachnoid hemorrhage and suggests the importance of continuous cardiac monitoring, particularly echocardiographic measurements, in those patients.
More than 500 000 people suffer strokes annually in the United States. One fifth of these cases are due to acute ICH, and of those 10% to 20% will ultimately die from the effects of this bleeding. Recent research is examining the role that the heart plays in these deaths.1
For many years, clinicians have known that ECG abnormalities are seen in 50% to 70% of patients with ICH.2 3 4 These changes, including T-wave inversion, ST depression, prolongation of the QT interval, and prominent U waves, have been reported to occur in the first 2 weeks and may take up to 6 weeks to resolve.5 While these findings are abnormal, they were first thought to be transient and inconsequential rather than a harbinger of cardiac disease. However, there are additional abnormalities, including wall motion disturbances, increases in plasma CK, and postmortem findings, that suggest distinct histological abnormalities occurring in the myocardium secondary to brain injury.1 6 7 Patients with SAH are particularly prone to subsequent cardiac effects. Pollick et al1 reported a study of 13 patients with no prior cardiac disease who sustained SAH and were studied with two-dimensional echocardiography. They found that patients with left ventricular wall motion abnormalities tended to have a higher neurological grade, greater increase in initial CK and myocardial isoenzyme levels, higher initial heart rate, and more frequent inverted ECG T waves. Three of 4 patients with abnormal echocardiograms died, whereas all 9 patients without wall motion abnormalities survived.
To date, there has been no systematic study of the etiology, functional significance, and timing of these latter abnormalities; nonetheless, a certain number of patients with acute ICH and in particular SAH do succumb and may have been greatly adversely affected by these changes. Based on this observed relationship, we undertook a study to determine whether the onset and timing of these abnormalities might have an independent value in predicting poor outcome.
Materials and Methods
This experimental study was conducted according to the guidelines and standards of the US Public Health Services for the use and care of laboratory animals and was approved by the Institutional Animal Care and Use Committee of Allegheny-Singer Research Institute.
Nine adult male mongrel dogs (weight, 20 to 25 kg) were sedated with 20 mg/kg thiopental IV and placed under general anesthesia (isoflurane) introduced through an endotracheal tube at an initial concentration of 1% in 100% oxygen. A catheter was inserted into a cephalic vein, an arterial catheter was placed into the left femoral artery, and a Swan-Ganz catheter was placed into the pulmonary artery by a right femoral approach. Rectal temperature was measured and maintained at 37°C by a heated blanket. ICP was recorded by means of an intraparenchymal transducer probe, placed in the right frontal lobe by twist drill craniotomy (Camino Laboratories). Standard ECG electrodes were placed on the body and limbs, and ECG was continuously recorded.
In 7 of 9 animals, 8 mL of autologous blood was withdrawn from a cephalic vein and injected into the right frontal lobe and subarachnoid space through a separate cranial hole fitted with a standard three-way device. The mean±SD volume injected for the 9 animals was 7.9±0.3 mL. The 8-mL value was chosen on the basis of data from experiment 1, in which we determined that this amount produces an ICP level above 25 mm Hg. In experiment 2, only 7 mL was injected because the animal’s ICP increased unexpectedly beyond the desired level.
Transesophageal echocardiography was performed before, immediately after, and at 1, 2, and 4 hours after SAH to assess wall motion changes and aortic and pulmonary flow velocities. At the same time, cardiac output was determined by thermal dilution, and a blood sample was collected. Baseline and post-SAH plasma levels of catecholamines and creatinine phosphokinase (total, isoenzymes, and isoforms) were measured. The following parameters were measured: systemic arterial blood pressure; pulmonary artery systolic, diastolic, and wedge blood pressures; ECG; and ICP. Data were collected and analyzed with SPSS software with the use of repeated measures analysis tests.
The animals were observed for up to 4 hours after hemorrhage. They were killed by intravenous lethal injection of sodium thiopental once myocardial wall motion abnormalities were detected. A piece of the myocardium was freeze-clamped for tissue catecholamine analysis. The brain, heart, and adrenals were harvested and preserved for histopathologic study.
Subarachnoid blood accumulation was found in all animals by gross inspection on autopsy. Brain sectioning showed blood present in the parenchyma of the brain in 2 animals and in the third ventricle in another, and 1 animal had a subdural hematoma.
Immediately after SAH was induced, the ICP increased in all animals and remained elevated until the animals were killed 2 to 4 hours after SAH (Table 1⇓). In addition, SAH produced significant increases in mean arterial blood pressure, mean pulmonary artery pressure, heart rate, cardiac output, and aortic and pulmonary flow velocities (Table 1⇓). The ECG showed changes including tachycardia, ST-segment depression, inverted T wave, and premature ventricular contractions in 4 animals at 1 to 5 minutes after injection (Fig 1⇓), and echocardiographic changes were evident at 20 to 240 minutes in all animals (Table 2⇓). Also, there were significant increases in total CK enzymes and in the cardiac isoenzymes of CK (CK-MB) (Table 2⇓). CK-MB isoforms MB2/MB1 were greater than 1.6 (normal, ≈1.0 in humans, unknown in animals) in 4 animals 2 hours after SAH compared with 0.98 when measured at baseline. When serum was analyzed for circulating catecholamines, norepinephrine was found to be higher at 5 minutes, and epinephrine was significantly (P<.05) higher at 2 hours after SAH (Table 2⇓). Cardiac tissue catecholamines were not elevated and did not correlate with the circulating catecholamines.
Light Microscopic Examination
Six animals were originally subjected to standard histological examinations. Little myocardial damage was seen. There was only one case in which lesions were detected that were consistent with myofiber vacuolar and contraction band changes. In this specimen, several microscopic lesions were present, but no overt myofiber necrosis per se was detected, probably because of the short time (2 to 4 hours) between the induction of SAH and euthanasia. The microscopic changes appeared to be consistent with early evolving lesions. Some fiber appeared slightly ragged or irregular, and there were prominent eosinophilic transverse bands, which represent the earliest stages of myofiber degeneration. We opted to continue only electron microscopic examination, since light microscopy was found to be inefficient in detecting early lesions.
Electron Microscopic Examination
In contrast to light microscopy, changes on electron microscopy were seen in all 9 animals. Individual myocardial cell lesions with a distinct swollen and loose appearance were observed. There were many affected areas showing more unequivocal disruption of the cell architecture and a more serious disappearance of cross striations (Fig 2⇓). Vacuolization areas that can be identified with early edema formation were also present, as was infiltration by fibroblasts and macrophages. Left ventricular endocardial microinfarcts in 3 dogs and right atrial infarction in 1 dog could be seen.
Our study describes the acute occurrence of systemic and cardiac changes and shows that cerebral influence on cardiac activity is constant, ie, that these cerebral pathophysiological states are expressed in the heart in a very short period (2 to 4 hours). In this study, we found ECG changes in 4 of 9 animals.
It was first thought that serum electrolytes are a cause of these cardiac abnormalities; however, this opinion remains uncorroborated.8 9 It is now believed that such abnormalities are caused either by hypothalamic stimulation or disinhibition or by alteration of the dynamic feedback and reverberating circuitry within the brain stem.10 Our finding of a rapid onset of these changes, with perturbations of the central nervous system, strongly suggests that these changes are due to neural rather than humoral effects.11 Oppenheimer et al12 13 also suggested a neural mechanism for the brain-heart connection by showing lethal arrhythmias when stimulating the insular cortex and a strong reciprocal connectivity with the limbic system.14 In addition, it is now believed that these ECG abnormalities can be present in several intracranial processes including cerebral infarction, glioma, head trauma, intraparenchymal hemorrhage, meningitis, psychological stress situations, seizures, and migraines.4 15 16 17 Surgical manipulation of the basal forebrain region has also been reported to cause similar but reversible intraoperative ECG changes.1
Although hemodynamic changes were recognized as early as 1900 by Cushing,18 neither they nor the early cardiovascular responses to SAH have been widely identified or appreciated. We show that there is an immediate response of increased heart rate, cardiac output, and aortic and pulmonary flow velocities. Arterial blood pressure and pulmonary artery pressure are significantly higher after 2 hours of SAH. Autonomic control is constantly exerted over arterial blood pressure, peripheral vascular resistance, and cardiac rate, rhythm, metabolism, contractility, and output. The integration of visceral and somatic afferent impulses with appropriate neural output is the constant responsibility of the central nervous system in its role of cardiovascular regulation. Although the role of the parasympathetic nervous system in cardiac regulation and in pathological conditions is perhaps not completely understood, it is nonetheless believed that the sympathetic nervous system exerts the major influence on cardiac function. It has been postulated that ECG changes result from autonomic nervous system abnormalities arising from area 13 in the orbitofrontal cortex.15
Many theories are proposed to explain cerebral-cardiac pathophysiological interactions.12 Catecholamines have been previously reported to be elevated in patients with SAH. We confirm in our model an elevation of epinephrine immediately after SAH, with a decline when measured again after 2 hours, and a persistent increase in norepinephrine levels. There is considerable experimental evidence pointing to a relationship between increased norepinephrine levels and rate of repolarization, which puts the heart at risk for ventricular arrhythmias.
The types of histological lesions found in the case of light microscopy have been described by others. Previous studies have demonstrated focal myocardial necrosis, ie, contraction band necrosis in conjunction with experimentally produced SAH in animal models.7 Also, there are recent reports of animal models with catecholamine infusion and patients with pheochromocytomas, which produce identical myocardial lesions.6 Other cardiac lesions described include both focal and subendocardial cardiac myofibrillar degeneration.19 20 It has been shown that the neurogenic cardiac lesion will occur even in adrenalectomized animals, which argues against a hormonal mechanism. These lesions can be blocked by measures that interfere with the sympathetic levels of the autonomic nervous system, such as ganglionic blockade and antiadrenergic drugs. Because we terminated our experiments at 2 to 4 hours after SAH, there may not have been enough time for a distinct histopathology to appear with light microscopy.
Pathological analysis initially was believed to be unrevealing in patients with ICH and cardiac disturbance. However, a unifying hypothesis developed when it was subsequently discovered on autopsy that microscopic subendocardial hemorrhages existed.17 20 21 It was believed that in some patients these lesions may be reversible and resolve before death.21 A more recent finding, now considered to be characteristic, is manifested by myofibrillar degeneration, focal myocytolysis, histiocytic infiltration, and diffusely necrotic heart zones. These changes have been summarized by the term “myofibrillar necrosis” and are commonly seen at the ultrastructural level.17 21 22 23 With electron microscopy, we also were able to see more lesions than those observed using standard light microscopy techniques.
Yuki et al22 reported that patients with SAH who had ECG abnormalities consistent with acute myocardial dysfunction and left ventriculographic findings were free of symptoms if bleeding aneurysms could be clipped. They believed that coronary vasospasm after SAH may lead to a wide spectrum of clinical cardiac abnormalities, from transient coronary ischemia, to myocardial dysfunction or “stunning” without necrosis, to a state of irreversible myocardial infarction.
While many have thought that the cardiac irregularities seen in these patients were benign, self-limited, irreversible, or nontreatable,16 our results indicate that multisegmental wall motion dyskinesia may be important as an etiological factor, which may be manifest through suboptimal cardiac tissue perfusion. The CK-MB enzyme consists of two subfractions, CK-MB1 and CK-MB2, known as CK-MB isoforms. CK-MB2 is believed to be the tissue-specific enzyme that is released into the serum when myocardial necrosis occurs and is subsequently converted to CK-MB1. Therefore, it is possible that this ratio of CK-MB2 to CK-MB1 may be an early serological marker for myocardial necrosis in patients with ECG changes associated with SAH. In our study we observed an increasing trend in 4 animals, which might suggest that myocardial cell necrosis occurred. We believe that CK-MB isoforms may be useful for the early evaluation of ECG changes and subsequent development of segmental left ventricular dysfunction during the course of an intracerebral bleed. Clearly, a longer study would be beneficial to evaluate the use of CK-MB isoforms in detecting cellular damage.
Our preliminary findings, as noted in this report, represent observations in which we attempted to analyze cardiac function beyond mere rhythm disturbances or isoenzyme expression. Two-dimensional echocardiography provides an analysis of heart function that would appear to be the final denominator for neural expression. We suggest that an improved outcome may be obtained by careful cardiac observations in cases of ICH and by treating those with multisegmented wall motion abnormalities in an effort to improve cardiac contractility, output, and ultimate performance. We have developed an experimental model to characterize the immediate and often ignored response to acute ICH in an attempt to resolve the exact pathophysiological mechanism of these cardiac events. The cause, ultimate clinical importance, and treatment of the brain-heart connection in pathological states remain to be elucidated.
Selected Abbreviations and Acronyms
This study was supported by grant 93-024-1P from Allegheny-Singer Research Institute, Pittsburgh, Pa. The authors would like to thank Leslie Arelt, Melissa Krukenberg, Maureen Miller, and Steve Cherilla for their excellent technical support and Nancy Lynch for the editorial services provided.
- Received October 23, 1995.
- Revision received December 28, 1995.
- Accepted January 16, 1996.
- Copyright © 1996 by American Heart Association
Andreoli A, di Pasquale G, Pinelli G, Grazi P, Tognetti F, Testa C. Subarachnoid hemorrhage: frequency and severity of cardiac arrhythmias: a survey of 70 cases studied in the acute phase. Stroke. 1987;18:558-564.
Brouwers PJAM, Wijdicks EFM, Hasan D, Verneulan M, Wever EFD, Frericks H, van Gijn J. Serial electrocardiographic recording in aneurysmal subarachnoid hemorrhage. Stroke. 1989;20:1162-1167.
Hersch C. Electrocardiographic changes in subarachnoid hemorrhage, meningitis, and intracranial space-occupying lesions. Br Heart J. 1964;26:785-793.
Harries AD. Subarachnoid hemorrhage and the electrocardiogram: a review. Postgrad Med J. 1981;57:294-296.
Doshi R, Neil-Dwyer G. Hypothalamic and myocardial lesions after subarachnoid hemorrhage. J Neurol Neurosurg Psychiatry. 1977;40:821-826.
Feutz V, Gormseu J. Electrocardiographic patterns in patients with cerebrovascular accidents. Circulation. 1962;25:22-28.
Davies KR, Gelb AW, Manninen PH, Boughner DR, Bisnaire D. Cardiac function in aneurysmal subarachnoid hemorrhage: a study of electrocardiographic and echocardiographic abnormalities. Br J Anaesth.. 1991;67:58-63.
Oppenheimer SM, Cechetto DF, Hachinski VC. Cerebrogenic cardiac arrhythmias: cerebral electrocardiographic influences and their role in sudden death. Arch Neurol. 1990;47:513-519.
Yasui Y, Breder CD, Saper CB, Cechetto DF. Autonomic responses and efferent pathways from the insular cortex in the rat. J Comp Neurol. 1990;303:355-374.
Cropp GJ, Manning GW. Electrocardiographic changes simulating myocardial ischemia and infarction associated with spontaneous intracranial hemorrhage. Circulation. 1960;22:25-38.
Weidler DJ. Myocardial damage and cardiac arrhythmias after intracranial hemorrhage: a critical review. Stroke. 1974;5:759-764.