Loading...
For the initial diagnosis and prognosis in suspected acute stroke, increasing use of magnetic resonance diffusion-weighted imaging (DWI) and perfusion-weighted imaging (PWI) instead of computed tomography (CT) is driven by the increasing use of thrombolysis and concern about radiation safety. This evidence-based review included research published between 1966 and January 2008 on DWI, PWI, or both, performed within 24 hours after stroke onset, excluding thrombolysis recipients, with the aim of determining the following:
1. Diagnostic sensitivity and specificity of DWI and PWI within 12 hours after symptom onset compared with other imaging techniques, with follow-up imaging, with clinical follow-up, and with discharge diagnosis
2. Predictive value of initial DWI and PWI abnormality volumes for initial clinical stroke severity, final infarct size on imaging, and late outcome
Compared with CT, one Class-I and three Class-II studies of DWI for acute stroke diagnosis revealed much higher inter-reader reproducibility, diagnostic sensitivity (roughly 77%–100% vs. 16%–70%) and accuracy (86%–91% vs. 55%–61%) and near-equal specificity (96% vs. 97%). Only Class-IV studies were found addressing PWI in acute stroke diagnosis. One Class-I, two Class-II, one Class-III, and four Class-IV studies collectively suggested that baseline DWI volume likely predicts initial clinical severity, final infarct size, and NIH Stroke Scale and Barthel Index measures of outcome in anterior-circulation but not posterior-circulation stroke. One Class-II study suggested that baseline PWI volume predicts clinical stroke severity.
The authors conclude that (1) DWI is superior to CT for acute stroke diagnosis, and (2) DWI and PWI volumes can predict initial stroke severity and late clinical outcome in anterior- but not posterior-circulation syndromes. They suggest that the imperfect (estimated 80%–90%) sensitivity of DWI might reflect small infarcts below the spatial resolution of current DWI technique and the earlier onset of clinical impairment than of brain infarction. They also suggest that semiquantitative and quantitative PWI, and DWI–PWI combined with magnetic resonance angiography and clinical metrics, might further improve the diagnostic and prognostic values of MRI.
Schellinger PD et al. Evidence-based guideline: The role of diffusion and perfusion MRI for the diagnosis of acute ischemic stroke. Report of the Therapeutics and Technology Assessment Subcommittee of the American Academy of Neurology. Neurology 2010 Jul 13; 75:177.
Comment
Because small brainstem strokes may produce severe impairment and large cerebellar strokes minimal impairment, better prediction of stroke severity and outcome with DWI volume may require subdividing the posterior circulation category, developing targeted posterior-circulation clinical metrics, and improving posterior-fossa DWI technique. Studies of PWI in combination with DWI more closely mirror clinical practice and may more fully exploit PWI information than studying PWI alone for diagnosis or prediction.