T he determination of the extent of brain tissue that is ischemic, but still viable and therefore under risk for irreversible injury (i.e., penumbra), is one of the major goals in acute stroke imaging (1). Advances in technology and the availability of perfusion imaging, either by computed tomography (CT) or magnetic resonance imaging (MRI), are major steps in this regard, and multi-modal imaging by CT or MRI is gaining importance as a tool in the development of both prognostic and therapeutic algorithms in acute stroke (2). Despite these developments, there are certain roadblocks in the utilization of perfusion imaging, with MRI in particular, in acute stroke patients. These include the unavailability of resources, primarily MRI scanners, in emergency departments; the need for technical and medical personnel for the postprocessing and interpretation of perfusion images, which can partially be overcome by automated imaging softwares; and the use of intravenous contrast agents, which are highly critical in patients with renal failure.Disregarding perfusion imaging by arterial spin labeling, other imaging signatures of compromised tissue perfusion that can be detected by MRI sequences that do not necessitate the use of contrast administration might provide clinicians with alternative and more practical tools in this setting. One of these signatures is the hyperintense vessel sign (HVS) observed on fluid-attenuated inversion recovery (FLAIR) imaging, which is considered to reflect slow and compensatory blood flow within leptomeningeal collaterals distal to the proximal occlusion (3). Prior studies have shown that patients with HVS evident on FLAIR are more like-
PURPOSEProminent hypointense cerebral vessels on susceptibility-weighted imaging (SWI) and the hyperintense vessel sign (HVS) on fluid-attenuated inversion recovery (FLAIR) imaging are considered as markers of compromised tissue perfusion in cerebral ischemia. In this study, we aimed to identify the correlation between HVS on FLAIR and hypointense vessels on SWI, and to determine whether these imaging features provide independent prognostic information in patients with ischemic stroke.
METHODSWe retrospectively analyzed consecutive ischemic stroke patients with proximal middle cerebral artery (MCA) occlusion who underwent SWI and FLAIR within 24 h of symptom onset. The presence of hypointense vessels on SWI and hyperintense vessels on FLAIR in >4 of 10 slices encompassing the MCA territory were considered to represent prominent hypoperfusion.
RESULTSAmong 50 patients, 62% had a prominent HVS on FLAIR and 68% had prominent hypointense vessels on SWI. There was a moderate but significant correlation between the number of slices with HVS on FLAIR and prominent hypointense vessels on SWI (r=0.425, P = 0.002). In multivariate analyses, the prominence of hypointense vessels on SWI, but not HVS on FLAIR, was significantly associated with a higher discharge NIHSS score (P = 0.027), mRS score (P = 0.021), and lesion growth (P = 0.050).
CONCLUSIONThe significant, al...