Background and Purpose— As a reliable scoring system to detect the risk of symptomatic intracerebral hemorrhage after thrombectomy for ischemic stroke is not yet available, we developed a nomogram for predicting symptomatic intracerebral hemorrhage in patients with large vessel occlusion in the anterior circulation who received bridging of thrombectomy with intravenous thrombolysis (training set), and to validate the model by using a cohort of patients treated with direct thrombectomy (test set). Methods— We conducted a cohort study on prospectively collected data from 3714 patients enrolled in the IER (Italian Registry of Endovascular Stroke Treatment in Acute Stroke). Symptomatic intracerebral hemorrhage was defined as any type of intracerebral hemorrhage with increase of ≥4 National Institutes of Health Stroke Scale score points from baseline ≤24 hours or death. Based on multivariate logistic models, the nomogram was generated. We assessed the discriminative performance by using the area under the receiver operating characteristic curve. Results— National Institutes of Health Stroke Scale score, onset-to-end procedure time, age, unsuccessful recanalization, and Careggi collateral score composed the IER-SICH nomogram. After removing Careggi collateral score from the first model, a second model including Alberta Stroke Program Early CT Score was developed. The area under the receiver operating characteristic curve of the IER-SICH nomogram was 0.778 in the training set (n=492) and 0.709 in the test set (n=399). The area under the receiver operating characteristic curve of the second model was 0.733 in the training set (n=988) and 0.685 in the test set (n=779). Conclusions— The IER-SICH nomogram is the first model developed and validated for predicting symptomatic intracerebral hemorrhage after thrombectomy. It may provide indications on early identification of patients for more or less postprocedural intensive management.

No abstract

Background: Unenhanced computed tomography (CT) is the imaging technique used in acute stroke. In some cases it is unable to detect damage even 24–48 h after symptom onset. The aim of our work was to evaluate the diagnostic value of diffusion-weighted imaging (DWI) in the post-acute phase of cerebral ischemia in patients in whom CT did not yield a definitive diagnosis. Methods: We retrospectively evaluated DWI findings in 214 patients, out of a series of 1,680 patients admitted to our hospital following the acute onset of focal neurological symptoms, in whom non-contrast CT, performed within 30 h of symptom onset, was normal (123), incongruous, i.e. a marked hypodensity indicative of an old infarct or a slight hypodensity not consistent with the clinical findings (66), or leukoaraiotic, i.e. diffuse chronic hypodensities in the periventricular white matter (25). Results: DWI showed signs of recent brain ischemia in 125/214 (58%) patients: 64/123 (52%) with a normal CT, 41/66 (62%) with an incongruous CT, and 20/25 (80%) with leukoaraiosis (p = 0.027). Multiple lesions were detected in 16/125 (16%) patients, while single lesions were ≤2 cm in 83/109 (76%) cases. DWI showed signs of ischemia in 22/91 (24%) patients with TIA and in 103/123 (84%) patients with stroke (p = 0.0001). Conclusion: In a quite high proportion of patients with recent symptoms of transient ischemic attack/ischemic stroke and a repeat non-diagnostic CT, DWI may help in the diagnosis of ischemic stroke and in shedding light on the underlying pathogenic mechanism.

No abstract