Objectives: To identify predictors and possible pathogenetic mechanisms of early neurological deterioration in patients with acute ischemic strokes and to evaluate their impact on clinical outcome.
Design: Case series.
Setting: University hospital's stroke unit.
Patients: A continuous series of 152 patients with first-ever ischemic hemispheric strokes were hospitalized within 5 hours of onset, evaluated with the Canadian Neurological Scale, and underwent a computed tomographic (CT) scan. The initial subset of 80 patients also underwent angiography. A repeated CT scan or autopsy was performed within 5 to 9 days of a patient's stroke. Progressing neurological deficit was defined as a decrease of one point or more in the global neurological scale score during hospitalization, when compared with that at entry.
Results: The conditions of 39 patients (26%) deteriorated during the initial 4 days; 20 patients (51%) had an impaired level of consciousness, and 19 patients (49%) had impaired limb strength and/or speech. They had been hospitalized earlier and had higher serum glucose levels at admission; the baseline CT scans of these patients showed an early focal hypodensity and initial mass effect more frequently. On the repeated CT scan (144 patients) or at autopsy (eight patients), patients with a progressing course more frequently had large infarcts, severe mass effect, and hemorrhagic infarction. We found no differences with regard to demographic data, medical history, and treatments that were given; only subcutaneous heparin calcium was more frequently administered to patients with a progressing course. Twenty-two (27%) of the 80 patients who underwent angiography had a progressing course, of whom 20 (91%) had an intracranial and/or extracranial arterial occlusion, with collateral blood supply in seven patients (35%). Logistic regression analysis showed that the independent predictors of progression were the serum glucose levels at admission and the early focal hypodensity with cortical and corticosubcortical locations, with the positive predictive values of the latter being 34% (95% confidence interval [CI], 26% to 42%) and 57% (95% CI, 47% to 67%), respectively. Among patients who underwent angiography, logistic regression analysis showed a significant correlation between carotid siphon occlusion and a progressing course. The 30-day case-fatality ratio and disability (Barthel index, < 60) were higher in patients with a progressing course (36% and 54% vs 12% and 35%, respectively).
Conclusions: Early stroke deterioration is still an event that is difficult to predict; it is largely determined by cerebral edema following an arterial occlusion, as indicated by an early focal hypodensity and initial mass effect on the baseline CT scan. Since early deterioration anticipates a bad outcome in 90% of patients, it might be used as an early surrogate end point in therapeutic trials.