Electrocardiographic abnormalities in patients presenting with strokes.

The development of changes in the electrocardiogram during the acute phase of cerebrovascular lesions is a well-recognized phenomenon. Burch, Meyers, and Abildskov (1954) defined these changes as typically consisting of inversion of the T waves, which were of considerable amplitude and width, in standard and chest leads, prolongation of the QTc interval and the presence of U waves. They encountered these abnormalities in cases of subarachnoid haemorrhage, intracerebral haemorrhage, and cerebrovascular accidents of non-specified type. Subsequently, the changes were also reported in cases of head injury (Hersch, 1961) and meningitis and space-occupying lesions (Hersch, 1964). Shuster (1960) examined the heart at necropsy in three patients with a cerebrovascular accident who had shown these ECG changes during life and found it to be normal. Koskelo, Punsar, and Sipila (1964), on the other hand, found subendocardial haemorrhages in their three cases of subarachnoid haemorrhage and suggested these were the cause of the ECG changes. However, there have been many case reports, of which Russell (1966) has supplied a thorough recent review, in which the heart was normal, hence subendocardial haemorrhage cannot be invoked as the cause of the ECG changes in all cases. Cropp and Manning (1960) observed the development of the ECG changes during manipulation of the orbital surface of the frontal lobe of the brain during the approach to an anterior cerebral aneurysm and suggested that they arose from area 13, which subserves autonomic function. The present study was undertaken to test this hypothesis and to elucidate the clinical problem which arises when ECG changes are encountered in a patient with a stroke.