Abstract 158: The Malignant CTP Imaging Profile Predicts Worse Functional Outcomes

    loading  Checking for direct PDF access through Ovid


Introduction: The Malignant MRI profile, defined as a large lesion on DWI or PWI (Tmax>10s), has been proposed as a marker of poor outcome despite reperfusion. It is not known if a corresponding malignant CT perfusion (CTP) profile can be used to identify stroke patients with a poor prognosis despite timely reperfusion.

Hypothesis: The Malignant CTP profile identifies stroke patients with poor clinical outcomes despite endovascular reperfusion.

Methods: The NIH-funded CTP to predict Response to recanalization in Ischemic Stroke Project (CRISP) prospectively enrolled acute ischemic stroke patients undergoing intra-arterial thrombectomy. CTP was obtained prior to the procedure and followed by a post-procedural MRI within 36 hours. The CTP Malignant profile was pre-specified as an infarct core (rCBF <30%) ≥70mL or a lesion with severe hypoperfusion (Tmax>10s) ≥100mL. Poor functional outcome was defined as a mRS 5-6 at 90 days. We evaluated performance of the pre-specified Malignant CTP profile for predicting poor functional outcome. We then performed an ROC analysis to optimize the ischemic core and Tmax>10s volumes for predicting poor outcome with high specificity (≥90%).

Results: Among 190 patients who underwent endovascular therapy, and had technically adequate CTP and 90-day outcome data, 51 (27%) had the Malignant CTP profile (45 on Tmax10 criteria alone, 6 on both infarct core and Tmax10 criteria). The Malignant CTP profile was associated with an increased rate of poor outcome (26% vs 14%; OR = 2.2; 95% CI 0.98-4.8; p=0.08). In patients with reperfusion (n=170), the percent of poor outcomes was significantly higher among patients with the Malignant CTP profile (27% vs 12%, p=0.02; OR = 3.1; 95% CI 1.3-7.4, adjusted for age). Based on ROC analysis, a CTP infarct core >50 mL or a Tmax>10s lesion >150 mL predicted poor outcome with high specificity (90%), but corresponding sensitivity was low (28%) and PPV was modest (36%, or 9/25).

Conclusion: Although presence of the Malignant CTP profile doubled the likelihood of poor outcome, only 1 out of 3 patients with this profile who had endovascular reperfusion experienced a poor outcome. This suggests that a subset of patients with the Malignant CTP profile may benefit from endovascular therapy.

Related Topics

    loading  Loading Related Articles