PubMed1999
To clarify the relationship between long-term prognosis of patients with stroke and their MRI findings, 103 patients with initial cerebral thrombosis, who survived more than three months after the ictus, were studied for five years. The mean age of 98 patients (T group), who were followed up completely, was 73.1 years-old and 65 were men. The age-matched controls consisted of two groups: 65 subjects, who had hypertension and/or diabetes without a history of stroke (R group), and 85 subjects, who had any hypertension, diabetes and stroke (N group). MRI findings were divided into six categories: 1) types of causative lesion, 2) grades of periventricular hyperintensity (none, rims/caps, patchy, diffuse PVH), 3) number of spotty lesions, 4) presence of silent infarction. 5) ventricular dilatation, and 6) extents of brain atrophy. Types of causative lesion were subdivided into 3 subtypes; infarction of the perforating artery territory (P type), infarction of the cortical artery territory (C type), and brainstem infarction (B type). The presence of vascular risks and dementia, and the extent of activity of daily living (ADL) were assessed. The P, C, and B types were identified by MRI in 46, 36, and 16 of the T group, respectively. Motor impairment, dementia, and an ADL status of complete dependence at discharge were also seen in 84, 44, and 22, respectively. In the T group, 33 patients died during five years, which resulted in a cumulative mortality rate of 33.7% and an annual mortality rate of 8.2%. Based on log-rank analysis, the survival rate of the T group revealed was significantly lower than those of the R and N groups. The recurrent rate in the T group (annual stroke recurrence rate was 4.0%) was higher than in the R and N groups, but stroke recurrence was not the cause of death and two thirds of deaths were due to aspiration pneumonia and/or asphyxia. Cox hazard regression analysis for death due to respiratory diseases showed that the hazard ratios of infarction, patchy PVH, and more than 4 spotty lesions were 8.87 (p < .001), 0.31 (p = .058), and 0.44 (p = .098), respectively. Compared to the survival group, rates of complete dependence in ADL, dementia, and brain atrophy were significantly higher in the death group with low incidences of the P type and patchy PVH, which indicated small vessel disease. These findings suggested that in patients with cerebral thrombosis, even in the chronic phase, care should be taken to prevent pneumonia and/or asphyxia due to bulbar palsy. Furthermore, no MRI findings were distinct predictors of long-term prognosis, although infarction based on the small vessel disease had rather good outcome in terms of respiratory disease.