Abstract 94: Patient Selection using CT Perfusion (CTP) Imaging Improves Effectiveness of Stroke Thrombolysis in the 0-4.5 Hour Time Window

Background: The role of CT perfusion (CTP) in thrombolysis decisions remains controversial and there are no studies that compare outcomes of thrombolysis in patients with or without mismatch on CT perfusion imaging. Methods: We analysed registry data between Jan 2009 and December 2010 for patients thrombolysed within 0-4.5 hours of stroke onset in whom CTP studies were performed prior to thrombolysis. The centre followed thrombolysis guidelines but patients >80 years were included. CTP was not obligatory in the treatment protocol and failure to demonstrate a mismatch was not a contraindication to thrombolysis. We retrospectively analysed data for estimated CTP mismatch of ≥ 100% according to pre-defined criteria and compared outcomes of thrombolysed patients showing perfusion mismatch with those showing no mismatch. Findings: The sample included 160 patients aged between 32-95 years of whom 63 had no mismatch and 97 had a significant mismatch. The two groups were comparable for mean age (73 v 70 years, p=0.18), sex (49% v 54% male, p=0.75), premorbid Rankin Score (mRS) 0-2 (81% v 92%, p=0.77), vascular risk factors profile, mean baseline BP (148/87 v 148/79 mm Hg, p=0.92), mean blood glucose (6.6 v 6.6 mmols/L, p=0.98) and mean National Institute of Health Stroke Scale (NIHSS) score (14.0 v 12.6,p=0.12). Patients who had mismatch prior to thrombolysis showed lower mean 24 hour NIHSS score (7.6 v 11.8, p=0.002) and greater mean 24 hour improvement in NIHSS score (5.1 v 2.0, p=0.010). A higher proportion of patients with mismatch achieved mRS 0-1 and mRS 0-2 at 3 months (36% v 18%, p= 0.012 and 51% v 32%, p=0.015 respectively) but there were no differences in symptomatic sICH rates (1.1% v 0%). Mortality (29% v 18%) and any intracranial haemorrhages (19% v 13%) were lower in mismatch patients but did not achieve significance. Regression analyses showed that PCT mismatch prior to thrombolysis was an independent predictor of both early improvement and functional outcomes at 3 months. Conclusion: Stroke patients who have perfusion mismatch on CTP imaging prior to thrombolysis within the 4.5 hour time window show better early and 3 month outcomes compared with those in whom mismatch cannot be demonstrated. Patient selection using multimodal CT may improve the effectiveness of thrombolysis.