“Nearly one-third of the participants’ MoCA scores showed declines between three months and one year out from their stroke,” Buckwalter said.
Each participant’s mass cytometry results and MoCA scores were fed into an algorithm developed in Aghaeepour’s laboratory. This algorithm was specifically designed for its ability to reduce massively complicated collections of highly correlated data to more manageable data sets.
The investigators identified three distinct phases — recognizable at post-stroke days 2, 5 and 90 — through which the immune system progresses before returning to normal about a year after the event. Each phase is marked by its own unique set of deviations in relative numbers and activation levels of a small number of immune cell types in the blood.
It was the first, or acute, phase, peaking just two days after the stroke, that grabbed the researchers’ attention. The more exaggerated the deviations among the handful of immune cells identified as defining that phase, the more likely a participant was to suffer a drop in performance on the mental test between three months and one year out from the event, the study found.
The correlation, which was big enough to achieve statistical significance in this small study, was independent of age, sex, body mass index, location of the stroke or type of treatment given when the participants initially reached the hospital. Remarkably, cognitive loss was also unrelated to the initial size of the stroke.
“We didn’t know what to expect,” Aghaeepour said. “This was an exploratory study looking to understand how cells function after stroke. We had no idea we’d find a correlation with long-term cognitive outcomes. It was an exciting accident.”
Predictive power of acute phase
Intermediate- and late-phase analysis added nothing to the acute phase’s predictive capacity. The investigators speculate that maybe later on, whatever immune cells are disrupting a patient’s cognitive apparatus are acting within the brain itself, so their activity might not get picked up in screenings of blood circulating peripherally.
“A lot more needs to be done to differentiate the people who will go on to have post-stroke dementia from those who won’t,” Buckwalter said. But with the key cellular players predictive of long-term dementia at least tentatively identified, the job gets much simpler.
“Being able to detect something in the bloodstream would be much easier than doing a brain biopsy,” Gaudilliere said. “Mass cytometry lets you measure the entire haystack in order to find the needle. Once you have the needle, you don’t need the haystack anymore. You can focus on just a few features of a few cell types,” opening the door to clinical applications.