Sept. 9, 2002 | LIKE MANY CANCER DRUGS, mercaptopurine can have harsh side effects. For more than 50 years, doctors had to

William Evans pioneered pharmacogenomic diagnostics at St. Jude's Children's Hospital

accept that some patients developed serious bone marrow-toxicity, leaving them at risk of infection and possibly death. But after a pair of leukemia patients at St. Jude's Children's Hospital in Memphis, Tenn., reacted badly to the drug, William Evans decided to do something about it.

"This was a profound effect," says Evans, chairman of pharmaceutical sciences at St. Jude's. "These children had to go to the ICU because they were getting so sick." As pharmacogenomics and leukemia were areas of particular interest for Evans, he initiated an National Institutes of Health (NIH)-supported study spanning five years, to the tune of approximately $1 million. One advantage St. Jude's had was direct access to patients, but they didn't have the rapid-fire genotyping tools available today.

Evans' team compared DNA sequences from patients who suffered the toxic reaction with those who did not. They found three SNPs that could cause the problem. "Those same SNPs show up in Asian populations, as well as in Europeans and others," Evans says. Any of these SNPs result in the deactivation of the enzyme thiopurine methyl transferase (TPMT), which is needed to metabolize mercaptopurine. About 10 percent of patients are heterozygous for such a mutation (that is, they carry one errant copy of the TMPT gene), which can still cause problems. But the most serious effects are seen in the 1 in 300 people who are homozygous for the faulty gene. "In these patients you have to decrease the dose down to 5 to 10 percent of what is normally given," says Evans. "That's not the kind of adjustment you'd normally start with."

As word of the St. Jude's findings spread in the late 1990s, other groups validated their results. "When we published this, people from around the country started to contact us, and we would try to help them," says Evans. "In 70 percent of the cases we saw like that, the inherited defect was the problem."

Best of all, the test made it to the clinic. "We were delighted when a couple of national reference labs decided to make it available as a clinical diagnostic so physicians could order it, just like a blood glucose test," he says. "TPMT became the first pharmacogenomic test that went all the way to CLIA [Clinical Laboratory Improvement Amendments] certification."

The progress of the test illustrates what many other pharmacogenomic projects are lacking, says the University of Louisville's Mark Linder. "To get momentum, the tests have to be driven by centers of excellence where they have a particular interest in the problem, like St. Jude's has."

A potential diagnostic DNA array that would detect genes influencing a patient's response to chemotherapy for acute lymphoblastic leukemia, including genes that determine drug metabolism, disease sensitivity, and the risk of side effects (cardiac or endocrine toxicities, infections, etc.).

St. Jude's is unusual in many ways. They are even building a GMP pharmaceutical manufacturing facility to make their own small molecules, vaccines, and gene therapies. More pharmacogenomic tests may also be forthcoming. For example, Variagenics has just licensed a patent application from St. Jude's that covers genotyping methods and diagnostic kits for the cytochrome CYP3A5 drug metabolizing enzyme. Variations in CYP3A can influence the metabolism of more than 50 percent of all cancer agents, including colon-cancer drug irinotecan.

Evans' group is now part of the NIH Pharmacogenetics Network. "We are doing more genome-wide investigations now," says Evans. "But our focus is on a candidate gene strategy."

—Malorye Branca