How pharmacogenomics might help addiction treatment

Several weeks before Christmas, 2005, Jean Valentine, a 60-year old grandmother from Arlington, Virginia, read an ad in the Washington Post placed by doctors at the National Institutes of Health recruiting for a study on alcoholics. By then, Valentine had drunk for 40 years, and "it was getting worse and worse, and I knew I was killing myself," she says. Valentine was skilled at hiding her problem from friends and family, "but every night I drank half a bottle of bourbon."

The final straw came one night when she was babysitting. She suddenly realized that she was considering leaving her granddaughter alone sleeping so she could get more alcohol. She didn't leave her granddaughter alone, but she decided it was time to get help. Valentine responded to the ad and qualified for a study that was designing protocols for inpatient and outpatient treatments. Valentine spent several weeks detoxing as an inpatient, and when she went home she began taking naltrexone, an opioid receptor antagonist originally indicated for opiate dependence.

The idea is that naltrexone blocks opioid receptors from the effects of ?-endorphin release stimulated by alcohol. "You eliminate the pleasurable effects of alcohol, and therefore naltrexone works. Beautiful story," says Markus Heilig, the clinical director at the National Institute of Alcohol Abuse and Alcoholism (NIAAA). In 1994 Heilig started a naltrexone clinic in Sweden. "The experience was amazing. Some people would come back and say, 'Doc, this turned my life around.'" For Valentine, the once-a-day pill dampened the intense cravings that would hit her every afternoon. "I think it really did help me those few months," she says.

But naltrexone didn't work for all alcoholic patients. "Most people didn't tell that story at all," Heilig says. "Most people, in fact, would say this doesn't do anything for them."

It turns out there are two forms of ?-opioid receptors, one of which has a single nucleotide polymorphism (SNP) that makes ?-endorphin bind to the receptor with higher affinity.¹ In 2003 David Oslin and Charles O'Brien at the University of Pennsylvania's Center for Addiction Treatment and their colleagues reported an association between this SNP and how well patients responded to naltrexone.² "We discovered that people with this allele do much better when randomly assigned to naltrexone than they do when they get placebo," says O'Brien.

O'Brien would like to include genetic data in addiction trials. "I think we ought to be genotyping patients...in all our clinical trials," he says. There are numerous genes whose polymorphisms might predict a person's risk for addiction and response to therapies, O'Brien adds. David Goldman at NIAAA says that understanding them might also reveal the subtype of addiction a person has. For example, SNPs in the gene coding for the GABA-? receptor have been linked with addiction, and Goldman has shown that one haplotype of the GABA-? receptor was most frequent among alcoholics with high anxiety.³ "We showed that although [this haplotype] is linked to alcoholism, it's through the anxiety dimension of behavior," he says. Goldman contends that diagnosis and treatment will be much improved once the genes particular to addiction subtypes are understood. "There will come a time when we have the same sorts of pieces of hard data for the addictions like we do for things like anemia," Goldman says. "There are a hundred types of anemia and there was a time when we said you're anemic or not.... It's the sort of specificity that is everything toward advancing treatment and perhaps prevention."

The study Valentine participated in was not designed to include genotyping, so she hasn't been tested to determine whether she carries the allele that improves naltrexone performance. She says she would hesitate to be tested in case it would exclude her from treatment. "I think when you're struggling with the addiction you're just about ready to try anything."