Published online before print June 19, 2003, doi: 10.1161/01.STR.0000082062.93106.23
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
(Stroke. 2003;34:1683.)
© 2003 American Heart Association, Inc.
Original Contributions |
Editorial Comment—Mismatch or Misconception?
Department of Radiology, Massachusetts General Hospital, Harvard Medical School, Charlestown, Massachusetts
The clinical utility of diffusion weighted MRI for assessing acute stroke patients has been known for some time.1 The combination of diffusion- and perfusion-weighted imaging (using dynamic MRI to follow an intravenous bolus injection of contrast agent) has been shown to be valuable for predicting infarct growth by defining the mismatch between the lesion volume defined by diffusion imaging and that defined by abnormal blood flow or volume (the so-called diffusion-perfusion mismatch2). There have been numerous published studies, both in human patients and in various animal stroke models, supporting the use of diffusion- and perfusion-weighted MRI for characterizing acute stroke lesions and predicting lesion progression and eventual stroke outcome. As a result, diffusion/perfusion MRI has already become a part of the clinical decision-making process for acute stroke at many sites, and has also made its way into clinical trial design. With this in mind, it is rather surprising that little work has been done to measure the reliability of these techniques in the clinical setting. This article by Coutts et al addresses this issue by comparing estimates of the diffusion/perfusion mismatch, as determined by 6 individuals, in a group of acute stroke patients.
Coutts et al sound an important note of caution when interpreting such MRI data. They have shown that operator-defined measurements of the diffusion/perfusion mismatch in acute stroke patients are quite reproducible for any given individual but that interobserver reliability is rather poor. As they point out, this has serious implications when such measurements are determining the application of a potentially dangerous therapy or evaluating the efficacy of a new drug. Their study was performed using a commercially available 3T MRI scanner and imaging methods also available as standard on the system. For analysis they have used manufacturer-supplied software to calculate maps of the diffusion and perfusion parameters and used a commercially available analysis workstation for viewing and measuring them. As such, this approach is perhaps the simplest and easiest to implement, but it is also that used by many clinical sites and is the approach perhaps most likely to be adopted by clinically oriented departments wanting to begin diffusion/perfusion stroke imaging.
There are several open questions arising from Coutts et al’s study; perhaps the most important is the reason why the intraobserver reliability is poor. Although acute stroke lesions are generally bright and well defined on diffusion-weighted images, the calculated maps of relative blood flow, volume, time-to-peak, or mean transit time are often noisy, which, combined with variable and graded blood flow around stroke lesions, often makes it difficult to accurately judge the extent of the perfusion lesion. Thus different observers may well come to different conclusions when visually interpreting these images. Furthermore, in the present study there was no "standard" perfusion measurement with which to compare the MRI data. Such a standard may not really exist for humans, although several groups have used the more established PET or SPECT technology to validate MRI perfusion data. A good correlation has been shown between MRI and SPECT-derived perfusion lesions in acute stroke patients,3 while a good correspondence between blood flow and volume measured with MRI and PET was found in a pig stroke model.4 However, a study in human volunteers comparing quantitative MRI and PET perfusion in repeated measurements on the same subjects showed that PET was still much more reliable.5 These other studies used the more accurate arterial deconvolution approach for processing the MRI perfusion data6,7 (which still involves some human intervention). Interestingly, reliability of the MRI perfusion measurements was improved (although still inferior to PET) by removing artifacts from large vessels.5 The great majority of clinical PWI studies have used the gradient-echo echo-planar imaging method to follow the contrast bolus injection. This has the greatest sensitivity to the magnetic susceptibility effects of the contrast agent, but larger vessels often show up very bright. In principle, spin-echo EPI has greater sensitivity to microvascular perfusion and can avoid large vessel artifacts8; however, this approach usually requires a double dose of contrast agent, and a good empirical comparison between gradient- and spin-echo PWI in stroke has, to my knowledge, yet to be published.
One additional consideration in Coutts et al’s study is their use of a 3T MRI scanner, compared with the more standard 1.5T field strength. Although this probably does not affect the reliability of their measurements, there has been some debate about the possible advantages of 3T in the acute stroke setting. One might expect, very roughly, at least a 2-fold increase in signal-to-noise (SNR) at 3T and this has driven interest in 3T for functional MRI, anatomical imaging, etc. However, a clear advantage for 3T in acute stroke MRI has yet to be demonstrated due to several confounding factors, including increased tissue T1 (reducing SNR for PWI), increased susceptibility artifacts in EPI, and increased RF power deposition (SAR).
Although improved perfusion MRI data processing algorithms and new MR contrast agents can yield better perfusion images, advances in automated or semi-automated definition of the lesion, and approaches combining all available MRI data,9,10 will allow a more accurate and reproducible definition of the extent of the stroke lesion and estimates of the eventual stroke outcome. However, in the meantime, while we wait for advances in imaging methodology and automated lesion classification to make their way into commercially available (and clinically useful) tools, perhaps the best approach is to use a very small group of carefully trained individuals for measuring all the perfusion and diffusion maps at a given site, or for a given trial, and to consider constraining the viewing conditions to reduce interobserver variations.
In short, the message from this study is to be cautious when using diffusion/perfusion data to guide acute stroke therapy. In my view, this should be combined with optimism that various methodological improvements could significantly increase the reliability of this MRI approach. Most recently it was shown, using 2 human operators to manually define lesions on perfusion diffusion and FLAIR images, that acute PWI data could accurately differentiate core and "penumbral" brain tissue prior to recanalization.11 This again demonstrates the power of perfusion/diffusion MRI and its potential to eventually become perhaps the most valuable determinant of acute stroke therapy options.
 |
References |
|---|
 Top References |
|---|
- Lutsep HL, Albers GW, De Crespigny A, Kamat GN, Marks MP, Moseley ME. Clinical utility of diffusion-weighted magnetic resonance imaging in the assessment of ischemic stroke. Ann Neurol. 1997; 41: 574–580.[CrossRef][Medline] [Order article via Infotrieve]
- Sorensen AG, Copen WA, Ostergaard L, Buonanno FS, Gonzalez RG, Rordorf G, Rosen BR, Schwamm LH, Weisskoff RM, Koroshetz WJ. Hyperacute stroke: simultaneous measurement of relative cerebral blood volume, relative cerebral blood flow, and mean tissue transit time. Radiology. 1999; 210: 519–527.
[Abstract/Free Full Text] - Karonen JO, Vanninen RL, Liu Y, Ostergaard L, Kuikka JT, Nuutinen J, Vanninen EJ, Partanen PL, Vainio PA, Korhonen K, et al. Combined diffusion and perfusion MRI with correlation to single-photon emission CT in acute ischemic stroke: ischemic penumbra predicts infarct growth. Stroke. 1999; 30: 1583–1590.
[Abstract/Free Full Text] - Sakoh M, Rohl L, Gyldensted C, Gjedde A, Ostergaard L. Cerebral blood flow and blood volume measured by magnetic resonance imaging bolus tracking after acute stroke in pigs: comparison with [(15)O]H(2)O positron emission tomography. Stroke. 2000; 31: 1958–1964.
[Abstract/Free Full Text] - Carroll TJ, Teneggi V, Jobin M, Squassante L, Treyer V, Hany TF, Burger C, Wang L, Bye A, Von Schulthess GK, Buck A. Absolute quantification of cerebral blood flow with magnetic resonance, reproducibility of the method, and comparison with H2(15)O positron emission tomography. J Cereb Blood Flow Metab. 2002; 22: 1149–1156.[CrossRef][Medline] [Order article via Infotrieve]
- Ostergaard L, Weisskoff RM, Chesler DA, Gyldensted C, Rosen BR. High resolution measurement of cerebral blood flow using intravascular tracer bolus passages, I: mathematical approach and statistical analysis. Magn Reson Med. 1996; 36: 715–725.[Medline] [Order article via Infotrieve]
- Wirestam R, Andersson L, Ostergaard L, Bolling M, Aunola JP, Lindgren A, Geijer B, Holtas S, Stahlberg F. Assessment of regional cerebral blood flow by dynamic susceptibility contrast MRI using different deconvolution techniques. Magn Reson Med. 2000; 43: 691–700.[CrossRef][Medline] [Order article via Infotrieve]
- Weisskoff RM, Zuo CS, Boxerman JL, Rosen BR. Microscopic susceptibility variation and transverse relaxation: theory and experiment. Magn Reson Med. 1994; 31: 601–610.[Medline] [Order article via Infotrieve]
- Welch KM, Nagesh V, D’Olhaberriague LD, Zhang ZG, Boska MD, Patel S, Windham JP. Automated three-dimensional signature model for assessing brain injury in emergent stroke. Cerebrovasc Dis. 2001; 11 (suppl 1): 9–14.[Medline] [Order article via Infotrieve]
- Wu O, Koroshetz WJ, Ostergaard L, Buonanno FS, Copen WA, Gonzalez RG, Rordorf G, Rosen BR, Schwamm LH, Weisskoff RM, Sorensen AG. Predicting tissue outcome in acute human cerebral ischemia using combined diffusion- and perfusion-weighted MR imaging. Stroke. 2001; 32: 933–942.
[Abstract/Free Full Text] - Shih LC, Saver JL, Alger JR, Starkman S, Leary MC, Vinuela F, Duckwiler G, Gobin YP, Jahan R, Villablanca JP, Vespa PM, Kidwell CS. Perfusion-weighted magnetic resonance imaging thresholds identifying core, irreversibly infarcted tissue. Stroke. 2003; 34: 1425–1430.
[Abstract/Free Full Text]
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||