Much like Yanai, Borlongan found that injecting stem cells into the compromised rats' brains reversed some of the behavioral deficits seen before treatment. For example, the treated rats could balance for longer time periods on a rotating rod.

To bring this kind of therapy closer to clinical tests in humans, Borlongan has experimented with administering the stem cells intravenously. Last July, in the online version of the Journal of Cerebral Blood Flow and Metabolism, he and his colleagues announced that transplanted stem cells produced the same result in rats regardless of whether they were given intravenously or injected directly into the brain.

Yanai has had similar success with intravenous administration in his heroin-exposure model, which he plans to announce at this year's annual meeting of the International Society for Stem Cell Research, in Barcelona.

The injected stem cells are able to migrate from the bloodstream to the brain for two reasons, says Borlongan. First, the injured brain sends out chemical signals that recruit the cells. And second, brain damage can compromise the blood-brain barrier, which normally regulates which substances can cross the threshold into the brain.

Not everyone is enthusiastic about the intravenous approach, however. Darwin Prockop, director of the Institute for Regenerative Medicine at Texas A&M Health Science Center College of Medicine, cautions that the injected cells can lodge in other organs--particularly the lungs--causing unwanted and even deadly side effects. And according to Evan Snyder, it may be unnecessary to go in through the bloodstream; his group has not seen any major risks associated with direct brain injection, a route that he considers to be clinically feasible in humans.

But all of these therapies involve introducing foreign cells into the body, and therefore, run the risk of provoking a potentially dangerous immune response. In most studies to date, the treated rodents are dosed with powerful immunosuppressants. Yanai is currently exploring personalized treatments to circumvent this issue: cells are extracted from the animal to be treated, coaxed to return to a stem-cell-like state, and then transplanted. Because they originate in the treated animal, the cells are recognized as "self" and ignored by the immune system.

Recently, Borlongan has found that immunosuppressants are unnecessary in the infant-stroke model. Because he treats the rodents at a very young age, their still-immature immune systems appear relatively unfazed by the transplanted stem cells. Borlongan notes that a low-level immune response may actually be useful: by cutting down on the number of cells that survive in the long term, it may reduce the chance that injected cells will reproduce uncontrollably and form tumors.

Nonetheless, according to Prockop, the risk of tumors is a serious concern with any stem-cell-based therapy. And while he is optimistic about the future of cell therapies for treating a wide variety of diseases, he urges caution and conscience when considering severe birth defects. "The big danger is that you can take a child who may be doomed to die in a few years, and make that child a lifelong invalid who needs continuing nursing care," he says. "So the prospects, if you think about them hard, are extremely worrisome. If you don't get a complete cure, you may be causing more harm than good."