MRI vs CT for Baseline Imaging Evaluation in Acute Large Artery Ischemic Stroke: A Subanalysis of the SWIFT-DIRECT Trial

Abstract

Background and objectives: Whether MRI or CT is preferable for the evaluation of patients with suspected stroke remains a matter of debate, given that the imaging modality acquired at baseline may be a relevant determinant of workflow delays and outcomes with it, in patients with stroke undergoing acute reperfusion therapies.

Methods: In this post hoc analysis of the SWIFT-DIRECT trial that investigated noninferiority of thrombectomy alone vs IV thrombolysis (IVT) + thrombectomy in patients with an acute ischemic anterior circulation large vessel occlusive stroke eligible to receive IVT within 4.5 hours after last seen well, we tested for a potential interaction between baseline imaging modality (MRI/MR-angiography [MRA] vs CT/CT-angiography [CTA]) and the effect of acute treatment (thrombectomy vs IVT + thrombectomy) on clinical and safety outcomes and procedural metrics (primary analysis). Moreover, we examined the association between baseline imaging modality and these outcomes using regression models adjusted for age, sex, baseline NIH Stroke Scale (NIHSS), occlusion location, and Alberta Stroke Program Early CT Score (ASPECTS) (secondary analysis). Endpoints included workflow times, the modified Rankin scale (mRS) score at 90 days, the rate of successful reperfusion, the odds for early neurologic deterioration within 24 hours, and the risk of symptomatic intracranial hemorrhage. The imaging modality acquired was chosen at the discretion of the treating physicians and commonly reflects center-specific standard procedures.

Results: Four hundred five of 408 patients enrolled in the SWIFT-DIRECT trial were included in this substudy. Two hundred (49.4%) patients underwent MRI/MRA, and 205 (50.6%) underwent CT/CTA. Patients with MRI/MRA had lower NIHSS scores (16 [interquartile range (IQR) 12-20] vs 18 [IQR 14-20], p = 0.012) and lower ASPECTS (8 [IQR 6-9] vs 8 [IQR 7-9], p = 0.021) compared with those with CT/CTA. In terms of the primary analysis, we found no evidence for an interaction between baseline imaging modality and the effect of IVT + thrombectomy vs thrombectomy alone. Regarding the secondary analysis, MRI/MRA acquisition was associated with workflow delays of approximately 20 minutes, higher odds of functional independence at 90 days (adjusted odds ratio [aOR] 1.65, 95% CI 1.07-2.56), and similar mortality rates (aOR 0.73, 95% CI 0.36-1.47) compared with CT/CTA.

Discussion: This post hoc analysis does not suggest treatment effect heterogeneity of IVT + thrombectomy vs thrombectomy alone in large artery stroke patients with different imaging modalities. There was no evidence that functional outcome at 90 days was less favorable following MRI/MRA at baseline compared with CT/CTA, despite significant workflow delays.

Trial registration information: ClinicalTrials.gov Identifier: NCT03192332.

Figure 1. Adjusted Clinical and Procedural Outcome Distribution and Workflow Metrics by Baseline Imaging Modality (CT vs MRI)

Descriptive primary and secondary outcomes by baseline imaging (MRI vs CT). (A) mRS at the 90-day visit, (B) door-to-needle time for patients receiving IVT, (C) door-to-puncture time, (D) door-to-reperfusion if cs-eTICI 2b/3, (E) postinterventional reperfusion (cs-eTICI 2b/3), (F) ICH, separated in hemorrhagic infarction, aICH, sICH, and unkICH. aICH = asymptomatic ICH; eTICI = expanded treatment in cerebral infarction score; HI = hemorrhagic infarction; ICH = intracranial hemorrhage; mRS = modified Rankin scale; sICH = symptomatic ICH; unkICH = ICH with unknown symptomatology.