Statistics from Altmetric.com
Stroke is a well known complication of coronary catheterisation but there are only few reports on the incidence of stroke after coronary angiography (CoAng) and angioplasty (PTCA),1 2and the unique characteristics of this complication are not well documented. We therefore examined stroke rate after cardiac procedures in our institution, and whether certain stroke patterns are more common than others.
We retrospectively examined medical records of all patients that were coded in the computerised hospital database with discharge diagnoses of both stroke and CoAng or PTCA between the years 1987 and 1994. Only patients that had a stroke within 24 hours after the cardiac procedure were included. All patients were evaluated by a neurologist. The neurological deficit was classified as mild (for example, mild hemiparesis or dysphasia) or severe (for example, severe hemiparesis or aphasia). Patients with non-localising neurological symptoms such as vertigo were not included in the study. Brain CT was performed in all patients, as soon as the haemodynamic condition allowed it, for identification of haemorrhagic events. We further classified ischaemic events as embolic in patients with hyperacute presentation of stroke and maximal neurological deficit at onset. We then compared the stroke group with a control group for the frequency of cerebrovascular disease risk factors, for the indication for the cardiac procedure, and for the degree of coronary disease that was subsequently found. The control group consisted of 48 randomly selected patients who underwent CoAng or PTCA without developing stroke. All catheterised patients routinely received antiplatelet therapy before, and heparin during the cardiac procedure. The extent of coronary disease was graded as mild (single vessel disease), moderate (double vessel disease), or severe (triple vessel disease). A χ2 test and Fisher’s exact test were used for statistical evaluation.
A total of 7151 patients underwent cardiac procedures (3460 had PTCAs and 3691 had CoAngs) during the study period. Seventeen patients had strokes that fulfilled the entry criteria, accounting for an incidence of 0.23%. Sixteen patients had an ischaemic stroke and one had a haemorrhage. The great majority of the ischaemic strokes were severe (13/16, 81.5%) and occurred during the cardiac procedure itself (12/16, 75%). All of the strokes evolved hyperacutely, with maximal deficit at onset, including those that appeared several hours after the catheterisation. Brain CT obtained in the acute phase, showed acute infarcts in the distribution of the middle cerebral artery in five patients and was interpreted as normal in all other patients. The severity, distribution, and hyperacute presentation strongly suggest that the neurological events were embolic in nature.
Stroke locations were distributed unevenly between the main vascular territories. Thirteen patients (81.5%) had a left hemispheric stroke, whereas only one patient (6%) had a minor right hemispheric event (p<0.0001, χ2), and two patients (12.5%) had a vertebrobasilar event (p=0.0002, χ2).
There were three deaths and one patient was left with severe disability. All three deaths occurred during the initial hospital stay and were related to the acute stroke.
Table 1 shows that the group of stroke patients did not differ from the control group for risk factors for cerebrovascular disease and severity of coronary disease. However, the indication for the cardiac procedure in most stroke patients was urgent (75% with unstable angina pectoris or acute myocardial infarction), whereas it was non-urgent in the majority of controls (63% with stable angina pectoris or congestive heart failure). This difference was significant (p=0.01, χ2 ).
The stroke rate of 0.23% after cardiac procedures found in our study is similar to other reports.1 2 Identification of patients who are prone to have a stroke as well as mechanisms for its development might enable us to lower the risk of this serious complication. Brown and Topol1 reported that stroke patients had more risk factors for cerebrovascular disease than controls. Alternatively, our study suggests that performance of cardiac procedures on an urgent basis due to acute activity of the heart disease may increase the risk of cerebrovascular complication. This may be explained by the presence of a generalised hypercoaguable state at the time of the procedure.
The clinical features of most stroke patients in our series are typical for an embolic event. Our finding that these emboli involve preferentially the left hemisphere, may provide a clue as to their source and mechanism of dislodgement. We cannot rule out completely a selection bias in this retrospective study, as minor right hemispheric events may escape diagnosis more often than left hemispheric ones. However, such a bias does not hold for major neurological events. In addition, all patients were subject to close monitoring in an intensive care setting and a consulting neurologist was routinely called in all cases of suspected neurological deficit, including minor ones, thus minimising a selection bias. Therefore, we think that the relative abundance of major left hemispheric strokes is genuine.
Uneven distribution of stroke location does not support a cardiac origin of emboli. Alternatively, brain emboli could arise in these patients from the aortic arch or carotid arteries. Such emboli could preferentially reach the left hemisphere due to the vascular anatomy of the aortic arch region. We suggest that manipulation of the catheter tip at the distal bend of the aortic arch, which is closer to the origin of the left common carotid artery, may cause dislodgment of atheromatous material preferentially into that artery. It was previously suggested that scraping the ascending aorta while searching for the ostia of coronary arteries was the cause of stroke.3 However, Tunick et al 4 showed that most protruding aortic atheromas are located distal to the innominate artery, and are therefore a cause for left hemispheric strokes in most patients. This is in good agreement with our proposed mechanism for embolism in CoAng/PTCA patients. Indeed, one of our more recent patients underwent transoesophageal echocardiography after the stroke, and an atheromatous plaque was found in the aortic arch near the origin of the left common carotid artery.
In two studies in which cardiac catheterisations were performed by inserting the catheter at a right antecubital5 or a brachial6 site, an increased incidence of vertebrobasilar strokes was found. In these cases the catheter tip may have released atheromatous material while advancing in the subclavian artery, preferentially into the vertebral system, sparing the carotid arteries. In our study, the femoral artery was used for catheter insertion in all stroke patients except for one. The one stroke patient in whom the catheter was inserted via the radial artery had a vertebrobasilar event.
In conclusion, we suggest that cardiac catheterisation still carries a measurable risk for stroke, especially in patients with acute coronary disease. We also suggest that stroke location depends on the route of catheterisation, with left hemispheric strokes being more common when the femoral artery is used for access. This finding calls for special care to be taken to avoid excessive catheter manipulation near the origin of the left common carotid artery.
This work was supported by the Sol Irwin Juni endowment fund, Jerusalem, Israel.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.