I was fortunate to stumble into an ongoing research project by inadvertently ruining the first attempt at scanning a volunteer. I began my fellowship at Wake Forest in 2006 and took an interest in physiologic imaging of brain tumors, and I became interested in DTI, spectroscopy, and perfusion, beginning work on finding a gadolinium-based perfusion protocol to run on our scanners. After a month or so, I “borrowed” a perfusion sequence from our radiation oncology researchers. I was going to run the first DSC perfusion in a clinical patient with a known glioblastoma multiforme. I got my perfusion data and, then, when I showed it to my attending, Dr. Joe Maldjian, he was upset (almost livid, if I remember correctly). It just so happened that very same patient was the first clinical volunteer for the arterial spin-labeling sequence. I hadn’t known this and acquired my gadolinium perfusion data first. Acquiring the gadolinium perfusion first, as I later found out, completely ruins the ASL perfusion signal. We had a long discussion and realized we were working for the same goal, getting perfusion data in more clinical patients. I was asked to join the research team and quickly began the most productive period of my career. The ASL sequence went into clinical release, and we rapidly gathered hundreds, and soon, thousands, of cases. We saw every pathology I could imagine, but consistently, the most striking were the cases of global hyperperfusion.

Our method of ASL at Wake Forest was quantitative, which meant we could see global and regional perfusion changes. There was nothing more memorable than to enter the reading room and see a white-hot brain from 20 feet away. Not just a little hyperperfused, these brains were off the chart. We started asking why. We had an automated method to calculate a mean global gray matter perfusion value. I began my part of the research by creating a massive spreadsheet, and quickly identified a significant number of patients with these global hyperperfusion changes. They had the same clinical history over and over… Found down. Anoxic injury. Drowning. Coded…. Some cases had obvious diffusion changes, but others had no infarction. This hyperperfused patient database became the

beginning of this paper. We identified all the patients who were coded, or resuscitated even for a few minutes, and found they all had the same global perfusion changes on ASL. We hypothesized that these global changes were the result of loss of autoregulation during the anoxic episode.

After my fellowship I moved to OHSU in Portland, Oregon. We also have ASL here and run it in all our stroke protocols, but it is not quantitative. We still encounter patients with anoxic injury, but the perfusion changes are regional (we can't detect global changes as easily with qualitative perfusion). The patients with anoxic injury we see now still show hyperperfusion, but, generally, these areas of hyperperfusion are confined to the regions of diffusion restriction. This finding can be seen in adult and pediatric patients.

Since the publication we have identified and published other causes of global hyperperfusion, including patient age (pediatrics) and hypercapnia, both of which offer other fascinating possibilities to use ASL in quantitative research projects.

I think the global hyperpefusion finding on ASL reveals something about the pathogenesis of anoxic brain injury, besides giving an obvious radiographic finding to suggest anoxic injury in the differential for even the most subtle cases of diffusion restriction, or cases with no identifiable infarct. In larger, future studies the hyperperfusion may have prognostic or therapeutic implications. I think, like many things, the ASL finding spawns more questions than answers, especially with regard to why. Why does the brain show regional and global hyperperfusion on ASL after an anoxic injury? Can this be confirmed with other perfusion methods? Is there a loss of autoregulation? Is the hyperperfusion real, or is it a manifestation of a blood-brain barrier injury, or a combination of both? I think in the future the research will be done to answer those physiologic questions, with the hope of finding ways to improve patient outcome after these devastating events.

Maybe one day I, too, will be able to answer those questions. Until then, I'll try to inadvertently stumble into my next research project.

 

Read this article at AJNR.org . . .