Comparison of T2*GRE and DSC-PWI for hemorrhage detection in acute ischemic stroke patients: Pooled analysis of the EPITHET, DEFUSE 2, and SENSE 3 stroke studies

Shalini A Amukotuwa; Nancy J Fischbein; Gregory W Albers; Stephen Davis; Geoffrey A Donnan; Jalal B Andre; Roland Bammer

doi:10.1177/1747493019858781

Comparison of T2*GRE and DSC-PWI for hemorrhage detection in acute ischemic stroke patients: Pooled analysis of the EPITHET, DEFUSE 2, and SENSE 3 stroke studies

Int J Stroke. 2020 Feb;15(2):216-225. doi: 10.1177/1747493019858781. Epub 2019 Jul 10.

Authors

Shalini A Amukotuwa^{1

2}, Nancy J Fischbein³, Gregory W Albers⁴, Stephen Davis⁵, Geoffrey A Donnan⁶, Jalal B Andre⁷, Roland Bammer^{3

6}

Affiliations

¹ Department of Radiology, University of Melbourne, Melbourne, Victoria, Australia.
² Monash Imaging, Monash Health, Clayton, Victoria, Australia.
³ Department of Radiology, Stanford University, Stanford, CA, USA.
⁴ Stanford Stroke Center, Stanford University School of Medicine, Stanford, CA, USA.
⁵ Department of Neurology, University of Melbourne, Parkville, Victoria, Australia.
⁶ Florey Department of Neuroscience and Mental Health, University of Melbourne, Parkville, Victoria, Australia.
⁷ Department of Radiology and Imaging Services, University of Washington Medical Center, Seattle, WA, USA.

PMID: 31291850
DOI: 10.1177/1747493019858781

Abstract

Aims: The objective of this study was to compare the diagnostic performance of the baseline pre-contrast images of dynamic susceptibility contrast perfusion-weighted imaging (DSC-PWI) with conventional T2*gradient recalled echo (GRE) imaging for detection of hemorrhage in acute ischemic stroke patients.

Material and methods: T2*GRE and DSC-PWI from 393 magnetic resonance imaging scans from 221 patients enrolled in three prospective stroke studies were independently evaluated by two readers blinded to clinical and other imaging data. Agreement between T2*GRE and DSC-PWI for the presence of hemorrhage, and acute hemorrhagic transformation, was assessed using the kappa statistic. Inter-reader agreement was also assessed using the kappa statistic.

Results: Agreement between the baseline images of DSC-PWI and T2*GRE regarding the presence of hemorrhage was almost perfect (k_{reader 1} : 0.90, 95% confidence interval 0.86-0.95 and k_{reader 2} : 0.91, 95% confidence interval 0.87-0.96). Agreement between the sequences was still higher for detection of acute hemorrhagic transformation (k_{reader 1} : 0.94, 95% confidence interval 0.91-0.98 and k_{reader 2} : 0.95, 95% confidence interval 0.92-0.98). Inter-reader agreement for detection of hemorrhage was also almost perfect for both T2*GRE (k: 0.95, 95% confidence interval 0.91-0.98) and DSC-PWI (k: 0.96, 95% confidence interval 0.93-0.99). Acute hemorrhagic transformation detected on T2*GRE was missed on DSC-PWI by one or both readers in 5/393 (1.3%) scans.

Conclusion: The almost perfect statistical agreement between DSC-PWI and conventional T2*GRE suggests that DSC-PWI is sufficient for hemorrhage screening prior to thrombolysis in stroke patients. T2*GRE can therefore be omitted when DSC-PWI is included, thereby shortening the acute ischemic stroke magnetic resonance imaging protocol and expediting treatment. Trial registration: ClinicalTrials.gov Identifier: NCT02586415.

Keywords: Brain bleed; brain microbleeds; cerebral hemorrhage; cerebral infarction; magnetic resonance imaging; stroke.

Publication types

Comparative Study
Research Support, N.I.H., Extramural
Research Support, Non-U.S. Gov't

MeSH terms

Adult
Aged
Aged, 80 and over
Brain / diagnostic imaging*
Cerebral Hemorrhage / complications
Cerebral Hemorrhage / diagnostic imaging*
Female
Humans
Ischemic Stroke / complications
Ischemic Stroke / diagnostic imaging*
Magnetic Resonance Imaging / methods*
Male
Middle Aged
Perfusion Imaging / methods*

Associated data

ClinicalTrials.gov/NCT02586415

Abstract

Publication types

MeSH terms

Associated data

Grants and funding