| The goals of magnetic resonance imaging (MRI) of cerebral ischemia are deceptively modest: (a) to reveal ischemic injury at early time points, and (b) to quantify the degree of damage that either has occurred or may develop. Prolonged 1H T2 relaxation is the MR gold-standard for cerebral infarction, and decreased 1H apparent diffusion coefficient (ADC) is an early marker for ischemic injury. However, one should be cautious in applying these findings. These signals can vary over time, and the proper interpretation often depends on contextual variables such as the duration and elapsed time since ischemia.; We hypothesized that increased MR signal from 23Na is an unambiguous marker for dead tissue. To test this idea we used a rat model of stroke involving 30 minutes of focal ischemia. 1H T2 maps, ADC maps, and 23Na images were acquired up to the 14 th post-surgical day and compared with triphenyltetrazolium chloride (TTC)-stained sections. From these data, we confirmed that the volume of tissue with abnormally elevated 23Na signal correlated well with the TTC-defined infarct at all time points. We also found that (a) 1H T2 transiently normalized after infarction, suggesting that it can be an unreliable means for the detection of infarction, and (b) 1H ADC appeared normal up to 12 hours after ischemic injury, indicating that it does not always identify injured tissue at early time points.; In the clinical literature, transient T2 normalization is known as the “MR fogging effect.” Since our animal model exhibited this phenomenon, we used it to test three of the proposed mechanisms: (a) hemorrhage, (b) normalization of water content, and (c) normalization of magnetization transfer (MT). No evidence for hemorrhage was found with T2*-weighted images, hematoxylin-eosin stained sections, or Prussian blue stained sections. There was a trend towards normalization of both water content and MT detected by wet-to-dry weight measurements and MT-weighted images, respectively. However, changes in water content are expected to also cause changes in MT. Therefore, the simplest explanation is that the MR fogging effect is driven by a normalization of water content to which 23Na images are apparently insensitive. |
