Silent New DWI Lesions within the First Week after Stroke

Abstract

Background: The rate of cerebral infarct recurrence is determined by clinical examination. Routine neurological examination is less sensitive than cerebral imaging in detecting new cerebral lesions. We aimed to determine the rate of new diffusion-weighted imaging (DWI) lesions at 2 time points after stroke and to identify factors associated with them. Methods: Patients who were hospitalized with acute ischemic stroke underwent DWI at 3 time points (within 24, 48 and 144 h after stroke onset, respectively). Scans were made anonymous and reviewed in a random order. Lesions on DWI were delineated manually by blinded investigators. Then, coregistered DWI templates were analyzed for new ischemic lesions on the corresponding follow-up DWI. New lesions had to be separate and lesion growth was not considered. Univariate and multivariable logistic regression analyses were performed to define predictors of new DWI lesions. Results: A total of 159 patients were enrolled in the study. Clinical stroke recurrence was detected in 2.5% of patients. A new cerebral lesion was detected in 5.7% of patients between first and second imaging (first interval) and 23.3% between second and third imaging (second interval). In univariate analyses, thrombolysis and multiple lesion pattern were associated with new lesions within the first interval. Ipsilateral carotid stenosis, multiple lesion pattern, vessel recanalization, atrial fibrillation, older age and higher NIHSS were associated with new lesions within the second interval. In multivariable analysis, ipsilateral carotid stenosis, recanalization and multiple lesion pattern remained independently associated with any new lesions. Conclusions: New DWI lesions occur more often than routine neurological examination suggests. Thrombolysis was associated with very early new DWI lesions within the first interval, ipsilateral carotid stenosis and spontaneous recanalization with new DWI lesions within the second interval.