Thrombus length discrepancy on dual-phase CT can predict clinical outcome in acute ischemic stroke

Abstract

OBJECTIVES: The thrombus length may be overestimated on early arterial computed tomography angiography (CTA) depending on the collateral status. We evaluated the value of a grading system based on the thrombus length discrepancy on dual-phase CT in outcome prediction.

METHODS: Forty-eight acute ischemic stroke patients with M1 occlusion were included. Dual-phase CT protocol encompassed non-contrast enhanced CT, CTA with a bolus tracking technique, and delayed contrast enhanced CT (CECT) performed 40s after contrast injection. The thrombus length discrepancy between CTA and CECT was graded by using a three-point scale: G0 = no difference; G1 = no difference in thrombus length, but in attenuation distal to thrombus; G2 = difference in thrombus length. Univariate and multivariate analyses were performed to define independent predictors of poor clinical outcome at 3 months.

RESULTS: The thrombus discrepancy grade showed significant linear relationships with both the collateral status (P?=?0.008) and the presence of antegrade flow on DSA (P?=?0.010) with good interobserver agreement (κ?=?0.868). In a multivariate model, the presence of thrombus length discrepancy (G2) was an independent predictor of poor clinical outcome [odds ratio?=?11.474 (1.350-97.547); P =0.025].

CONCLUSIONS: The presence of thrombus length discrepancy on dual-phase CT may be a useful predictor of unfavourable clinical outcome in acute M1 occlusion patients.EY POINTS:

? Early arterial phase CTA may underestimate thrombus length. ? Thrombus length discrepancy grade reflects collateral status or presence of antegrade flow. ? Outcome prediction may be better with thrombus length grade than collateral score.