Delayed-phase CT angiography is superior to arterial-phase CT angiography at localizing occlusion sites in acute stroke patients eligible for intra-arterial reperfusion therapy.

We compared the accuracy for localizing arterial occlusion sites between delayed-phase computed tomography angiography (CTA) and arterial-phase CTA in acute stroke patients with large vessel occlusion. Institutional Review Board approval was obtained, and informed consent was waived. For patients treated with intra-arterial thrombolysis within a 6 hour window between January 2009 and November 2011, we retrospectively assessed the arterial occlusion sites of pre-treatment, delayed-phase and arterial-phase CTA and compared these with digital subtraction angiography (DSA) findings. The positive predictive value for detection of vessel occlusion for both CTA methods was derived using DSA as a reference standard. The outcomes were compared using the McNemar test. Inter-observer disagreement within each modality was assessed using the Kendall W test. Among 73 patients treated with intra-arterial thrombolysis, 24 (32.9%) underwent both arterial-phase and delayed-phase CTA, and 66 (90.4%) underwent arterial-phase CTA only. For 24 patients undergoing both arterial-phase and delayed-phase CTA, the delayed-phase CTA detected concordant occlusion sites with DSA in 21 patients, yielding a positive predictive value of 87.5%, whereas the arterial-phase CTA detected this in 14 patients, for a positive predictive value of 58.3% (p=0.013). When discordant with DSA, arterial-phase or delayed-phase CTA indicated occlusions in more proximal sites than DSA. Delayed-phase CTA more precisely localized the occlusion site than arterial-phase CTA in acute stroke patients with large vessel occlusion. Thus, delayed-phase CTA findings could guide neurointerventionists in selecting intra-arterial thrombolysis modalities.