Until the day he collapsed, John Kelly, 57--an exceptionally fit, nonsmoking, retired soldier--had never given a moment's thought to strokes. His cholesterol was low; he worked out six days a week; no one in his family had ever had one. One morning in January he sat down on the edge of his bed, bent over to tie his shoes, and, he says, "kept right on going" until he toppled to the floor. A blood clot had cut circulation to a large region on the right side of his brain, which instantly began shutting down; muscles on the left side of his body, with no input from the nerves, went limp. Inside his head, a biochemical riot had begun, which if unchecked would lead in a few hours to a massive cell die-off, leaving him an invalid at best. Like many stroke victims, Kelly seemed barely aware of what was happening to him: the right side of the brain controls muscles on the left side of the body. But the speech center is in the left hemisphere, so Kelly could still talk. He asked his wife to help him finish dressing so he could go to work. She ran for the phone instead...

It will happen to 20 million people this year, two thirds of them 60 or older but some just in their 20s, and nothing, short of a massive coronary, kills as quickly. Of those who survive the initial attack--roughly 75 percent--nine in 10 will have long-term impairment of movement, sensation, memory or reasoning, ranging from slight to devastating.

The prevalence of strokes, which began leveling off around five years ago in the United States and Europe after a three-decade drop, may soon be on the rise, a result of an aging population and immigration. Strokes are particularly high among Hispanics and Africans. Rates are also expected to rise in many developing countries as they adopt Western lifestyles. And there is one other factor, says Dr. John Marler, associate director of the U.S. National Institute of Neurological Disorders and Stroke (NINDS): as we get better at treating heart attacks, "more people are surviving a myocardial infarction and going on to have a stroke."

The two share identical risk factors, including high blood pressure, excess cholesterol and diabetes; the basic mechanism--blockage of a crucial artery--is usually the same. But progress against coronaries has not, until now, led to new treatments for stroke. While the last decade has seen big advances in long-term rehab for stroke victims, for most people who show up in the emergency room there is exactly one drug doctors can use to get blood flowing into their brains again--and only a few patients get even that. Still, promising research into how blood vessels function and how neurons die may lead to new treatments that can save the lives, and brains, of people like John Kelly. Says Dr. Joe Broderick, head of neurology at the University of Cincinnati: "We're five to 10 years behind cardiology, but we're going in the same direction."

Get your unlimited Newsweek trial >

tPA is a powerful drug that can increase the chance of a full recovery by as much as 33 percent. But it must be given within three hours from the onset of symptoms; if it is given later, the risk of further damage outweighs the benefits. But first the patient has to get to the hospital. Unlike a heart attack, whose pain generally sends people to the hospital immediately, strokes can be subtle, especially if they strike the parts of the brain dealing with memory or cognition rather than movement or speech. The U.S. National Stroke Association says the average stroke patient waits more than 12 hours before going to the emergency room; some never get there.

And like any drug that counteracts clotting, tPA has the drawback that it can promote bleeding. It is useful only on the type of stroke Kelly had ("ischemic"), in which a clot blocks one or more of the arteries taking blood to the brain. But about 15 percent of strokes are "hemorrhagic," resulting from a burst artery bleeding into the brain--in which case tPA is precisely the worst thing you can do. (There are no drugs to treat acute hemorrhagic stroke, which is fatal in up to 50 percent of cases.) Unfortunately, the symptoms of the two types of stroke can be identical. Only a brain scan can distinguish with certainty between them--so by the time doctors are ready to start treatment with tPA the three-hour time limit may have already passed. "You're always looking at the clock on the wall," says Dr. Anthony Furlan, a leading stroke researcher at the Cleveland Clinic. Rather than risk potentially fatal bleeding into the brain, doctors tend to err on the side of caution--which means fewer than 5 percent of stroke victims even get tPA at all.

Humans aren't the only creatures with an interest in keeping blood liquid; they share it with the vampire bat, Desmodus rotundus. The thumb-size bats land on sleeping animals and make tiny cuts in their skin, lapping up the blood as it oozes out; an enzyme in their saliva keeps the wound from clotting. That suggests a possible new therapy. At a meeting of the American Stroke Association earlier this year, a group of international researchers reported on a study in which stroke victims were treated with a genetically engineered version of the bat's active ingredient; they found that it could be used safely up to nine hours after a stroke began, and appeared to cause less collateral bleeding than tPA. "I'm not sure this will be the silver bullet," says Furlan, who is leading a similar study in the United States. "But we're optimistic."

Even when given in time, clot-busting drugs don't always work. Are there ways to enhance their effectiveness? Dr. Andrei Alexandrov, a neurologist at the University of Texas-Houston Medical School, made an unexpected discovery in 1999 while using diagnostic ultrasound to investigate how tPA restores blood flow to the brain. What he found was that the ultrasound machine, which was not intended as a treatment, helped break up the clots and made a dramatic difference in how patients recovered.

Get your unlimited Newsweek trial >

High-energy ultrasound has been used for years to bust up kidney stones, but nobody was crazy enough to use it anywhere near the brain. At the much gentler levels used for diagnosis, though, it seems to create just enough hydraulic pressure to get the tPA inside and around the clot, where it can go to work. "We've had cases where patients who couldn't move suddenly regain the use of their body," says Alexandrov. "They reach up with the arm that was paralyzed and try to remove the ultrasound device, saying they're done and they're going home." In a study of 126 patients reported at the stroke conference, the combination of tPA and ultrasound restored blood flow in almost half the patients, compared with just 30 percent of those treated with tPA alone.

The most straightforward way to open an artery, of course, is to reach inside and clear it out. Also earlier this year, another team announced preliminary results from the use of the MERCI retriever. This is a corkscrew-shape device inside a catheter that can be maneuvered through the blood vessels to the site of a blockage; emerging from its sheath, the retriever grapples the clot and pulls it clear. In a clinical trial of 114 patients, the device successfully unblocked arteries in 61 of them--and of those, 20 recovered dramatically, many of them right on the table. The U.S. Food and Drug Administration is currently considering whether to approve this still-experimental device.

Brain cells are voracious consumers of oxygen and glucose, both carried by the blood; when totally deprived of circulation they suffer irreversible damage in as little as five minutes. But a stroke usually cuts off blood flow to only a relatively small area of the brain, surrounded by a much larger region with diminished circulation. There, cells die more slowly. "Up to 24 hours, and even beyond that, damage continues in these marginal areas," says Thomas Jacobs, the stroke program director at NINDS. The holy grail of stroke research over the last decade has been a means of slowing this secondary damage, keeping the brain as intact as possible until circulation can be restored. "We know the brain can survive a long time without oxygen," says Dr. Patrick Lyden, director of the stroke center at the University of California, San Diego. "We see it every winter when some kid falls through the ice and is fished out 45 minutes later, and once he thaws out he's fine."

Without knowing precisely how cold protects brain cells, researchers have been trying for years to harness its benefits. But it isn't easy to duplicate the effects of falling into a frozen pond, and the results have been mixed. In one experiment, Dr. Derk Krieger of the Cleveland Clinic used external cooling to reduce the body temperature of a group of stroke patients by nine degrees Fahrenheit for two days; it triggered violent shivering, which had to be counteracted by inducing paralysis, which required putting them on ventilators. On the whole, the group did better than the control patients who were not cooled, but the differences were small. Newer research is focused on endovascular cooling, using small catheters that can apply cooling directly to the blood vessels leading to the brain.

Perhaps the most important insight of recent years is that cells in the brain don't just "starve to death" when blood flow is interrupted. Cell death is a process, a cascade of biochemical changes that lasts for hours and can, theoretically, be interrupted at several points in time to save the patient's brain. "The analogy I use with patients is, you hit your hand with a hammer, and the injury is done, but you look at your hand and it looks the same, even if it hurts. But 48 hours later it's black and blue and swollen," says Dr. Lawrence M. Brass, a neurologist at the Yale School of Medicine in New Haven, Connecticut. You can alleviate some of that by applying ice; researchers are still looking for the best way to slow damage to the brain.

Better still would be a way to prevent strokes in the first place--beyond the universal prescription to stop smoking, to exercise and to control cholesterol, that is. Dr. John Hallenbeck, chief of the stroke branch of NINDS, is taking a unique approach, attempting to "immunize" blood vessels against stroke by focusing not on blood coagulation but on the artery walls themselves. Artery segments, for reasons still unclear, periodically become "activated"--a process akin to inflammation that promotes the formation of clots at those locations. Working with rats, Hallenbeck found that a naturally occurring molecule called E-selectin seems to dampen inflammation specifically in the lining of blood vessels, without affecting normal immune-system function. He called his results "startlingly successful" so far, but warns that the technique has not yet been tried in humans.

Another possible approach is to identify people at high risk for stroke, who could then be monitored and treated prophylactically. A classic warning sign for stroke risk, often overlooked, are "transient ischemic attacks," or "mini-strokes"--episodes of diminished blood flow to the brain that don't last long enough to cause permanent damage. The symptoms, which can last from a few seconds to as long as 24 hours, are the same as a stroke, including sudden numbness or weakness on one side of the body, trouble in seeing or speaking and loss of coordination. (Counterintuitively, Dr. Roger Simon, head of the Dow Neurobiology Laboratories in Portland, Oregon, has found that TIAs, although they may signal an impending stroke, also seem to play a protective role; animals with artificially induced TIAs tended to survive induced strokes with less brain damage.)

Another significant risk factor is atrial fibrillation, a disruption of the heart rhythm that affects one in 10 people over the age of 65. The condition has few obvious symptoms and often goes undetected, but can cause clots that can travel to the brain or other parts of the body--which appears to have been the case with Kelly, as his doctors discovered, some weeks later.