Although diabetes mellitus (DM) has been found to be a strong risk factor for ischemic stroke and is correlated with the worst of outcomes, it has yet to be determined whether or not the stroke subtypes and distribution of ischemic lesions differ in patients with DM as compared to other ischemic stroke patients. We studied 215 consecutive patients with first-time episodes of ischemic stroke at Kyoto Second Red Cross Hospital. Patients with cardiogenic embolism were excluded. MRI examinations were performed on all patients and 158 of the patients underwent cerebral angiography. The patients were classified into 4 groups according to their symptomatic lesions : 118 patients with lacunar infarcts, 62 with non-lacunar infarcts in the supratentorial region, 20 with brainstem infarcts, and 15 with TIAs. The distribution of ischemic lesions, based on the MRIs, was subsequently evaluated for all patients. The lesions included lacunar infarcts (3-15 mm in diameter), cortical and subcortical non-lacunar infarcts, brainstem infarcts, and cerebrallar infarcts. The lacunar infarcts were further divided into 2 subgroups : those located in the territory of the lenticulostriate and subcortical perforating arteries, and those in the thalamoperforating and thalamogeniculate arteries. The brainstem infarcts were also divided into 2 subgroups : the branch atheromatous type (infarcts extending to the surface of the pontine basis), and the lipohyalinosis type (infarcts located within the pons, but not extending to the pontine surface). The relationship between the distribution of ischemic lesions and risk factors, including hypertension (HT), DM, hyperlipidemia (HL), smoking (SM), and heart disease, was investigated. The complication rate of DM was significantly higher in patients with pontine infarcts of the branch atheromatous type than in those without pontine lesions. In contrast, the complication rates of HT, HL, SM, and heart disease were not statistically different between the two groups. The numbers of lacunae were counted, and classified according to the following scheme : 0 (absent), single lacuna (one lacuna), and multiple lacunae (more than two lacunae). Although multiple lacunae, in both the lenticulostriate artery and thalamoperforant artery territories, were found to be significantly correlated with HT, there was no correlation with DM, HL, or SM. Non-lacunar supratentorial infarcts, and cerebellar infarcts showed no correlation with HT, DM, HL, or SM. The stenotic lesions in the extracranial and intracranial arteries identified by cerebral angiography were classified into 4 groups : Grade 0, <30% stenosis ; Grade 1, 30-60% ; Grade 2, 60-99% ; and Grade 3, occlusion. Although higher-grade stenotic lesions tended to be correlated with aging, such lesions were not significantly correlated with other risk factors. DM was also correlated with pontine infarcts of the branch atheromatous type. However, DM showed no correlation with brainstem infarcts of the lipohyalinosis type, thalamic lacunae, or cerebellar infarcts. It was thus not possible to conclude that all infarcts within the territory of vertebrobasilar arteries were associated with DM. The mechanism whereby DM is associated with such lesions requires further clarification.
View full abstract