In what one expert called “a tour de force,” and another “probably the best preclinical neuroprotection study I've ever seen,” researchers have shown that inhibition of a glutamate-related signaling receptor can halt brain damage after stroke onset, reducing infarct volume and preserving behavioral function. They did it in macaque monkeys, whose brains are anatomically much more similar to humans' than those of rodents or marmosets, the standard subjects of most neuroprotection trials in stroke. And there is an early indication it may work in people as well.
The stroke field has become deeply skeptical of neuroprotection in recent years, as drug after drug has looked solid in the lab, but gone on to fail in the clinic. In retrospect, many stroke researchers think that a chief problem has been that the animal studies have not been conducted with the same type of rigor applied to clinical trials.
“In fact, all of the animal trials have failed to meet all the STAIR criteria,” a set of guidelines designed to reduce the chances of a false positive in preclinical studies, including randomization, blinding, and testing in higher primates, said Michael Tymianski, MD, PhD, lead author of the new study in the Feb. 29 issue of Nature. [See “The STAIR Criteria” for more on the guidelines.]
In testing his new compound, Dr. Tymianski, director of the Neuroprotection Laboratory at the Toronto Western Hospital Research Institute in Ontario, committed to this rigorous approach. The drug, called Tat-NR2B9c (also called NA-1), is a peptide that disrupts a neuronal protein called PSD-95 (post-synaptic density protein 95), which links NMDA [N-methyl-D-aspartic acid] receptors to neurotoxic signaling pathways. In this way, glutamate signaling can be preserved while glutamate-induced toxicity can be reduced. Dr. Tymianski developed the drug with this in mind, and has spent the past decade bringing it forward in the lab. The current study is the culmination of that effort.
“Given that the appetite for rat models as predictors of success in human trials is pretty low, we knew that nothing short of evidence in brains that are very similar to those of humans would do,” he said, hence the choice of macaques as the experimental model.
Dr. Tymianski induced a severe middle cerebral artery occlusion (MCA) in 20 macaques by clamping off the artery distal to the orbito-frontal branch of the MCA, isolating the occluded tissue from collateral circulation and causing a rapidly evolving stroke with a small penumbra.
The animals were then given active drug or placebo intravenously 60 minutes later, and then reperfused at 90 minutes by unclamping the artery. MRI was performed within 15 minutes of stroke, at 24 hours, and at 30 days. Stroke severity scores were assessed at 8 hours and 30 days, and behavioral tests were performed at 7 days and 30 days.
Four animals receiving placebo died due to stroke, along with three receiving active treatment, whose deaths were due to complications unrelated to stroke or drug, Dr. Tymianski said. All were included in an intent-to-treat analysis, and the worst neurologic scores and infarct volumes were used when data were missing. While this tends to diminish the potential of finding a treatment effect, he said, it is the most rigorous approach, and similar to that in human trials.
Both groups had similar volumes of tissue at risk, as determined by neuroimaging, immediately after stroke. But active treatment reduced the infarct volume compared with placebo by 37 to 44 percent after 24 hours, depending on the imaging technique used. Differences of similar magnitude remained at 30 days.
A battery of sensorimotor tasks was administered blindly throughout the follow-up period. Animals receiving active treatment performed significantly better than those receiving placebo at all time points, and at the end of 30 days, they were only half as behaviorally impaired. Gene expression analysis from the ischemic penumbra in the acute post-stroke period indicated that active treatment was associated with a much smaller perturbation of normal gene expression, including pathways associated with response to ischemia and cell stress.
Dr. Tymianski then performed a second trial of severe stroke, but this time with treatment at 60 minutes and reperfusion at 4.5 hours after onset, which he said more closely modeled the typical clinical severe stroke. There were no mortalities. Again, active treatment was superior to placebo over a seven-day follow-up. The results, he said, suggest that early treatment with Tat-NR2B9c “may increase the window in which reperfusion may have functional benefits,” even in severe strokes.
Finally, he induced a less severe stroke by occluding the MCA proximal to the orbito-frontal branch, creating a slower-evolving stroke and a larger penumbra potentially available for neuroprotection. He administered treatment at three hours and reperfusion at 3.5 hours. Here again, “despite the prolonged ischemic interval and the delayed treatment” with the drug, those receiving active treatment did significantly better than those receiving placebo, both on imaging and behavioral tests.
“This is not the first time neuroprotective drugs have been studied in this species, but I've never seen a study in primates that had such comprehensive assessments. This is a tour de force,” said Sean Savitz, MD, a stroke researcher and associate professor of Neurology at the University of Texas Medical School at Houston.
“I am definitely guardedly optimistic. We've been sobered by so many of the failures of the past. But this is the closest I think you can get to proving that something is going to have the chance to work in patients without actually doing it in patients.”
Marc Fisher, MD, professor of neurology at University of Massachusetts Medical School in Worcester, and editor of the journal Stroke, agreed. “It's probably the best preclinical neuroprotection study I've ever seen.” Earlier studies with positive results in rodents or marmosets, including Dr. Fisher's own work with a growth factor, have universally failed in the clinic, leading him to doubt the predictive utility of these models. “The whole field of neuroprotection is about to die. If we are going to go forward, we need positive primate data,” he said, and in his opinion, this trial meets that standard.
For Thomas Carmichael, MD, PhD, professor of neurology at the Brain Research Institute at the University of California, Los Angeles, the study's strength was the “great pains” it took to use a clinically relevant design. They did a number of things right, he said, including delivering the drug at different time points, and carefully measuring behavioral outcomes as well as tissue outcomes. After all this, “it appears ready for clinical trials.”
And those trials have already begun. After a successful safety study, Dr. Tymianski helped oversee a phase 2 trial in 185 patients receiving endovascular coiling repair for either ruptured or unruptured aneurysms. Such patients typically suffer very small strokes as a result of treatment, and the trial was designed to assess the ability of the drug to reduce stroke burden on MRI.
“The drug was effective in doing so,” he said. Complete details will be forthcoming in a paper later this year. “We now have data to show that this ‘monkey business' does translate to humans who have similar demographics to stroke victims and have verifiable strokes. We are trying to forge a path that has failed every single time it has been trodden before, but it's a great adventure and we plan to take it all the way.”
THE STAIR CRITERIA
In 1999, the STAIR (Stroke Therapy Academic Industry Roundtable) group published criteria in the journal Stroke for advancing candidate therapies for the treatment of stroke into clinical trials. The criteria include adequate dose-response and serum concentration measurements; confirmation of efficacy in relevant time windows; the use of physiological and behavioral outcome measures in animal studies instead of only infarct size; use of multiple types of stroke models; trials in larger species; and reproducibility of preclinical results by independent laboratories.