Abstract 3071: Automated CT Perfusion Processing For the Evaluation of Transient Ischemic Attack

Abstract

Introduction: The diagnosis of transient ischemic attack (TIA) can be difficult. We evaluated the yield of automated CT perfusion (CTP) in addition to non contrast head CT and CT angiography for the detection of acute ischemic lesions among TIA patients and compare it with multimodal MRI [diffusion weighted imaging (DWI) and perfusion weighted imaging (PWI)] results. Methods: Consecutive patients with a final diagnosis of transient neurological symptoms secondary to ischemia who underwent acute CT, CTP and CTA were enrolled. A subset of the patients underwent multimodal MRI. Presence of symptomatic stenosis was assessed on CTA by the treating team at the time of patient evaluation. The presence and location of acute ischemic lesions was assessed on: CT, CTP, DWI, and PWI. The rater was blinded to the clinical presentation. The presence of a MR or CT perfusion lesion was assessed using TMax. TMax maps for both CT and MR were automatically generated by RApid processing of PerfusIon and Diffusion (RAPID) software. Results: Thirty-three patients were enrolled: median age was 66 years old (IQR 58-82); median ABCD2 score was 4 (IQR 3-5); median delay from symptom onset to CTP was 4.6 hours (IQR 2-9.6). MRI was performed in 23 (70%) patients after a median delay of 20.4 hours (IQR 8.3-30.6) after symptom onset and 5.5 hours (IQR 3.4-20.8) after CTP. No patient experienced recurrence between CTP and MRI. Non-contrast head CT did not demonstrate any acute ischemic lesions. CTA found 3 symptomatic vessel lesions. CTP revealed a focal ischemic lesion in 11 patients (33%). The lesion location concurred with the presumed symptom side in all but one patient, and with the 3 symptomatic vessel lesions found on CTA. DWI was performed in 23 patients, and was positive in 7 (30%). Three of these patients had a negative early CTP. CTP was positive in 3 patients with subsequent normal DWI. MR perfusion was performed in 17 patients and found an acute ischemic lesion in 8 (47%) of them. Four of these MR PWI positive patients also had a positive CTP. In 3 patients with a negative MR PWI, CTP detected an acute ischemic lesion. At least 1 of the 3 modalities (CTP, DWI, PWI) was positive among 10/17 (59%) of patients. Conclusion: The results of this exploratory study suggest that automatically processed CTP increases the yield of head CT and CTA for the evaluation of TIA patients. In some cases CTP found evidence of transient focal ischemia that was not detected by MRI, in others CTP was negative and DWI or PWI was positive. These findings suggest that CTP and MRI may be complementary techniques to confirm the ischemic nature of transient neurological symptoms.