SAN ANTONIO, TEX—Neuroprotection in human stroke does work. This pronouncement, which comes after years of promising animal studies, disappointing human trial results, and subsequent attempts to explain what went wrong with neuroprotective drugs, was made by Steven Warach, MD, PhD, at the 2002 International Stroke Conference. The key to proving that these drugs work is selecting the right patients by using modern imaging methods, he said.

“What we see when we use MRI is evidence of early brain damage leading to stroke. It’s so early that it’s potentially reversible,” said Dr. Warach. His findings showed that MRI revealed brain recovery in patients with stroke following treatment with citicoline. “Before, we would just take a buckshot—treat all patients regardless of the ischemic pathology—and hope that something hit the right target. The potential of MRI is to target the right therapy to the right patient,” he said. Dr. Warach is Section Chief of Stroke Diagnostics and Therapeutics, National Institute of Neurological Disorders and Stroke, Bethesda, Maryland.

Dr. Warach’s optimism was shared by Jeffrey Saver, MD, who reported encouraging neuroprotective findings of his own through a meta-analysis of choline precursors and with a trial in which paramedics administered magnesium sulfate to stroke patients en route to the hospital. “What we’re trying to do is treat patients who still have brain left to save, whose stroke isn’t yet finished,” said Dr. Saver. “These drugs usually only work in animals within one to two hours at the start of a stroke,” he continued. “But in neuroprotective trials to date, drugs have been given from four to 48 hours after onset of stroke when there may be very little brain left to save and even an effective drug is not going to be able to show a benefit.” Dr. Saver is Associate Professor of Neurology, UCLA School of Medicine and Neurology Director, UCLA Stroke Center, Los Angeles.

BETTER METHODS, BETTER TIMING, BETTER RESULTS

Findings such as those reported by Drs. Warach and Saver are a breakthrough in the overall disappointment surrounding neuroprotection research. As yet, no clinical trial has proven definitive efficacy in neuroprotective drugs, despite the fact that millions of dollars have been spent in research and development. Thus, the field has been going through a “period of soul searching,” said Dr. Warach. “One reaction to that is to say, ‘Well, maybe this whole concept of neuroprotection works well in rats; it just doesn’t work in human beings.’ Another take on that is maybe we need to get smarter about how to do clinical trials of neuroprotectants in human beings.”

Until now, said Dr. Warach, trial inclusion data and diagnosis have been based purely on a clinical bedside impression instead of on a radiologic or pathologic confirmation. Now, with MRI, changes in stroke are visible almost immediately. “Before you enroll the patient in a trial, you can confirm, like you could with an EKG if it was a heart attack, like you could with a biopsy if it was a tumor, objectively that the patient has the diagnosis that the drug is intended to treat.... The idea is this could target the right drug to the right patient in a clinical trial or regimen of practice but also could as closely as possible replicate the preclinical data, the laboratory data that led the drug company into the clinical trial.”

CITICOLINE REVISITED

“I think these data are really the most convincing we’ve seen that in human beings there is a dose-dependent effect of a neuroprotective drug that replicates what we might see in an animal model or laboratory model,” said Dr. Warach. He added that the trial sample size required to show an effect on the brain is probably “fivefold or tenfold fewer patients” than is needed to show an effect on stroke scales. The reason, said Dr. Warach, is that “the lesion volume is a direct measure of the infarct, whereas the stroke scales are relatively subjective and not quite as reliable.” However, an effective therapy must ultimately prove itself on those clinical end points. He hopes that larger, definitive citicoline trials will confirm these findings.

TRIAL BY META-ANALYSIS

After reviewing all the published literature, the Cochrane Stroke Review Group found eight trials using choline precursors in acute stroke, and all of them turned out to be examples of citicoline. “These eight trials individually have shown somewhat conflicting results and have been received in the medical community as showing no definite conclusion as to whether citicoline is beneficial or not. In the meta-analysis, we formally combined all the trials and found that when you put them together, there does seem to be actually quite a substantial beneficial treatment effect of citicoline when given within the first 14 days of stroke onset,” he said. According to Dr. Saver, 67% of individuals given placebo treatments were dead or disabled at the end of the trial. Of patients who got active choline precursors, 57% were dead or disabled.

“Some small-sized positive trials were published, and there’s an absence of small-sized negative trials in the literature, which makes you think there might be a little bit of a publication bias—that people tend to publish positive results when a trial is a home run and they tend to not bother to write up a negative trial. We corrected for that somewhat by only looking at large trials that had enough momentum to get published no matter what the results,” he said.

MAGNESIUM AS NEUROPROTECTANT

He speculated that one way to try to get an experimental neuroprotective drug started earlier would be to have paramedics administer the drug in the field immediately after stroke onset. Although t-PA requires a computed tomography scan before it is safe to administer, most neuroprotective drugs are safe to give to patients with brain hemorrhage as well as to patients with brain infarction. “Unlike t-PA, most neuroprotective drugs are not going to harm hemorrhage patients and may even be beneficial for hemorrhage patients as well as for infarction patients,” he said.

Dr. Saver and colleagues chose magnesium sulfate, because it has both direct neuroprotective and beneficial effects in dilating blood vessels and increasing blood flow to the brain. It has also been shown to be effective in reducing stroke size in several animal models and in early randomized clinical trials using standard in-hospital start-up therapy in humans and is a very safe drug. “It’s been used for treatment of eclampsia in pregnant women for 70 years in medical practice, so it’s a good drug to try and go out in the field with and have paramedics administer.” In addition, he said, “We did train paramedics to understand the basics of magnesium physiology [and its] complications. In some emergency medical systems magnesium is already part of the regimen that paramedics are allowed to give.”

For the 20 patients, the on-scene to needle time was 23 minutes. “In comparison, for patients at our own center, who had been enrolled in acute neuroprotective trials using standard consent and enrollment after hospital arrival, the average time from paramedic arrival on scene to start of drug running was 141 minutes. So we’re able to get drug in when there’s still more brain to save.”

For the 20,000 patients who had been enrolled in neuroprotective trials in the past, essentially none had ever gotten their drug within one hour, and very few received their drug within two hours, said Dr. Saver. “So we have achieved a goal of being able to get drugs to patients in the same time that they worked in animal models and, therefore, we think when they would have a chance of being effective in humans. We saw no safety complications from having paramedics start the drug rather than having the drug started in a hospital, so there were no serious adverse events.” In addition, he said, several patients dramatically improved after getting the drug, and about 25% of the infarct patients got better from the time the drug started in the ambulance to the time the patients arrived at the hospital.

NR