Once the virus is injected into the spinal fluid and makes its way into the nerve cells of the pain gate, it uses the host cells' machinery to churn out the opioid protein--which then goes to work blocking pain signals on their way to the brain. Normally, the gene is rarely activated. But the version used for therapy has no such limitations because the gene carried by the AAV has been modified to continuously produce the opioid chemical.

Cope says that using endogenous opioids is inherently superior to injecting synthetic opiate drugs directly into the spinal fluid, an approach that requires the installation of a pump in order to deliver the drugs over a long time period. "It's kind of a holy grail," she says. "If the body's own system for pain control were activated by genetic expression, that would be superior to an artificial medication."

In Beutler's study, which was published this week in PNAS, rats were surgically modified to have a stronger than usual response to pressure on their paws, mimicking the effects of so-called neuropathic pain. The gene-therapy treatment effectively restored the rats to a normal level of pain sensitivity. The team also tested a nonopioid gene, which produced comparable pain relief through an entirely different mechanism. But while the opioid gene's effects will likely extend to humans, who respond to opiates the same way rats do, the nonopioid's effects may be rat specific.

The Stockholm-based company Diamyd Medical has been developing a different approach to gene therapy for chronic pain that also bypasses the side effects of standard pain treatment. The approach uses a deactivated version of herpes simplex virus (HSV). HSV can be administered straight through the skin as it naturally finds and infects peripheral nerves and travels to the spinal cord on its own. Darren Wolfe of Diamyd says that this method is superior to spinal injection because it's safer and easier, and it can be administered repeatedly.

Because of these considerations, the HSV method may be preferable for treating localized pain. However, when chronic pain involves multiple areas of the body--as it often does with, for example, metastasized cancers--going straight to the pain gate could work more efficiently.

While both of these methods have proved effective in animal models of pain, their efficacy in human patients remains to be shown. Diamyd recently applied to the FDA to begin phase I clinical trials, and Beutler estimates that his approach could be tested on humans in as few as three years.