|Paper title||Stroke and stroke mimics in the acute setting: role of multimodal CT|
|Form of presentation||Poster|
Stroke can be mimicked by non-vascular conditions. Due to emphasis on rapid admission and treatment of acute ischemic stroke patients, it is crucial to identify these conditions to avoid unnecessary treatments and potential complications. We reviewed the prevalence of stroke mimic conditions at our Center and investigated the performance of the multimodal-computed tomography (CT) protocol (unenhanced and contrast enhanced brain CT, CT-angiography and CT-perfusion) to discriminate stroke mimics from ischemic stroke.
We retrospectively searched from our Radiology database and selected all imaging studies performed for clinical suspicion of acute ischemic stroke between January 2015 and December 2016. Patients with multimodal-CT protocol at admission and at least one follow-up imaging study (brain CT or MR) were included. We compared the results of the multimodal-CT protocol with the final clinical diagnosis at discharge.
424 patients were included. In 94 cases a stroke mimic condition was eventually diagnosed, with a prevalence of stroke mimics of 23% including 30 seizures, 11 migraines with aura attack, 10 functional disorders, 8 infections, 7 brain tumors, 5 peripheral vertigo, 4 metabolic causes, 3 syncope, 2 transient global amnesia, 1 hypoglycemia, 1 subdural hematoma, 1 cervical epidural hematoma, 1 dural fistula and 10 non-vascular undetermined causes. Multimodal-CT sensitivity, specificity and accuracy were 22%/99%/82% overall and 33%/99%/89% when excluding a priori CT-undetectable stroke mimic conditions (functional disorders, syncope and peripheral vertigo). Multimodal CT revealed 12/30 peri-ictal changes, 7/7 brain tumors, 1/1 dural fistula and 1/1 subdural hematoma. CT-perfusion played a key role in the detection of seizure related stroke mimic conditions, where brain CT and CT-angiography are tipically silent.
5/12 seizures and 6/7 tumors showed a hyperperfusion pattern (increased cerebral blood volume and flow +/- reduced mean transit time) while 7/12 seizures and 1/7 tumors showed a hypoperfusion pattern (increased mean transit time +/- reduced cerebral blood volume and flow).
Multimodal-CT demonstrated low sensitivity but rather high specificity in the diagnosis of stroke mimics in the acute setting. Multimodal CT ability to identify stroke mimics may represent an asset in the diagnostic management of patients with suspected stroke.