Systematic Review of Perfusion Imaging With Computed Tomography and Magnetic Resonance in Acute Ischemic Stroke: Heterogeneity of Acquisition and Postprocessing Parameters
A Translational Medicine Research Collaboration Multicentre Acute Stroke Imaging Study
Background and Purpose—Heterogeneity of acquisition and postprocessing parameters for magnetic resonance– and computed tomography–based perfusion imaging in acute stroke may limit comparisons between studies, but the current degree of heterogeneity in the literature has not been precisely defined.
Methods—We examined articles published before August 30, 2009 that reported perfusion thresholds, average lesion perfusion values, or correlations of perfusion deficit volumes from acute stroke patients <24 hours postictus. We compared acquisition parameters from published studies with guidance from the Acute Stroke Imaging Research Roadmap1. In addition, we assessed the consistency of postprocessing parameters.
Results—Twenty computed tomography perfusion and 49 perfusion-weighted imaging studies were included from 7152 articles. Although certain parameters were reported frequently, consistently, and in line with the Roadmap proposals, we found substantial heterogeneity in other parameters, and there was considerable variation and underreporting of postprocessing methodology.
Conclusions—There is substantial scope to increase homogeneity in future studies, eg, through reporting standards.
Computed tomography perfusion (CTp) and magnetic resonance (MR) –derived perfusion-weighted imaging (PWI) hold promise for patient selection for reperfusion therapies in ischemic stroke by defining tissue viability.2 However, clinical trials that use perfusion imaging to support an extended time window for thrombolysis have, to date, been inconclusive.3 Although this may be because of a number of methodological issues related to perfusion thresholds,4,5 relative and absolute tissue compartment volumes,6,7 and definitions of tissue at risk,5 it may also reflect differences in acquisition and postprocessing of perfusion data. Indeed, the Acute Stroke Research Imaging Roadmap consensus statement1 has recently encouraged adherence to common acquisition protocols. In this study, we assessed the heterogeneity of current perfusion-based stroke studies to evaluate current practice and variation within the literature.
The search strategy and data extraction, which incorporated manuscripts published up to August 2009, have been described in detail elsewhere.5 Manuscript selection criteria for this study were:
Adult (age ≥18 years) stroke patients <24 hours postictus, distinguishable from other patients described in publications by the same research group
Report of perfusion characteristics from studies using either first-pass CTp or MR-derived PWI for the following:
Threshold values for tissue compartments
Mean perfusion values in different tissue compartments
Correlation of deficit volumes on perfusion imaging with lesions on other imaging modalities
Subjects with hemorrhagic stroke, venous infarction, or chronic occlusive cerebrovascular disease.
Studies of perfusion techniques other than first-pass bolus tracking CTp or MR-PWI, e.g., arterial spin labeling, CTp from triphasic helical technique, or other steady-state techniques.
Studies of technical development/optimization of imaging parameters for CT or PWI techniques.
Studies using both duplicate data and analyses from other larger included studies.
A review of 7153 articles yielded 49 MR-PWI articles and 20 CTp articles (Supplemental Table; http://stroke.ahajournals.org).5 The manuscripts of all but 3 studies were received by the publishing journal before publication of the Roadmap.
Although several parameters were consistent with the Roadmap, acquisition parameters were heterogeneous, sometimes with a large range noted (Table 2). For example, in CT studies, there were; number of slices ranging from 1 to 4); rate of injection ranging from, 2 to 20 mL/s; electric parameters of 9 different combinations of peak kilovoltage and milliamperes; and for volume of iodine in contrast, 8 combinations in 10 papers, ranging from 10.5 to 18.5g. For MR studies, there were: 24 combinations of echo time/repetition time, number of phases of acquisition ranging from 20 to 60 phases, and number of slices ranging from 7 to 40s. Postprocessing parameters were also heterogeneous (Table 2).
In CT and magnetic resonance imaging perfusion studies, many acquisition and postprocessing parameters are frequently unreported or heterogeneous. The impact of the observed heterogeneity is likely to be complex, and when combined with variable definitions of penumbra,5 may explain, at least in part, the failure to replicate promising initial results using imaging selection and end points when CTp and PWI are undertaken across multiple centers.3 The effect of underreporting is less clear, but the Roadmap proposals now give future authors the opportunity to state, at least in a generic manner, that such acquisition parameters have been performed to an agreed standard, even if some acquisition parameters are deliberately omitted. Improved homogeneity of acquisition parameters may aid clinical trial conduct as well as translation to clinical practice.
Limitations of this study include focused coverage of the literature and the use of a consensus statement (the Roadmap) as a reference standard. In addition, practice may have changed since the analysis.
In conclusion, although word space is limited in journals, many details could easily be provided in an online version. Perhaps a “standard for reporting perfusion imaging” (STRPI) could be implemented following consensus on the EQUATOR Network and by journal editors.
Sources of Funding
This study was supported by the Translational Medicine Research Collaboration (NS-EU-082). J.W. and K.M. were supported by the Scottish Funding Council through the Scottish Imaging Network, A Platform for Scientific Excellence (SINAPSE) Collaboration (http://www.sinapse.ac.uk). R.T. was supported by a grant from The Stroke Association (Registered charity SC037789). K.D. was supported by the Patrick Berthoud Charitable Trust.
This work was performed as part of the Multicenter Acute Stroke Imaging Study and the Stroke Imaging Repository (STIR) Task Force 1 activities (https://stir.ninds.nih.gov/html/projects.html).
Jeffrey L. Saver, MD, was the Guest Editor for this paper.
The online-only Data Supplement is available at http://stroke.ahajournals.org/lookup/suppl/doi:10.1161/STROKEAHA.111.629923/-/DC1.
- Received June 20, 2011.
- Accepted July 7, 2011.
- © 2012 American Heart Association, Inc.
- Wintermark M,
- Albers GW,
- Alexandrov AV,
- Alger JR,
- Bammer R,
- Baron JC,
- et al
- Hacke W,
- Furlan AJ,
- Al-Rawi Y,
- Davalos A,
- Fiebach JB,
- Gruber F,
- et al
- Takasawa M,
- Jones PS,
- Guadagno JV,
- Christensen S,
- Fryer TD,
- Harding S,
- et al
- Ma H,
- Zavala JA,
- Teoh H,
- Churilov L,
- Gunawan M,
- Ly J,
- et al