Leukemia overpowers drug in two ways.

The medicine known commercially as Gleevec serves as a powerful weapon for people fighting the blood cancer called chronic myelogenous leukemia, or CML. Although the drug appears to cure many patients, it usually provides only fleeting improvement for those who have entered the crisis stage of the lethal disease.

A new finding could help scientists patch this weakness in the drug's otherwise potent assault on CML. In an upcoming issue of SCIENCE, researchers at the University of California, Los Angeles (UCLA) reveal how this cancer rebounds.

The leukemia originates when pieces of chromosomes 9 and 22 fuse, forming a hybrid gene called Bcr-Abl (SN: 12/11/99, p. 372). This mutation encodes an enzyme, Bcr-Abl tyrosine kinase, that causes white blood cells to proliferate. Without Gleevec treatment, CML would smolder for years. Eventually, it would explode into a crisis stage in which white blood cells multiply rapidly and crowd out healthy cells in the bone marrow.

The oral drug, also known as STI-571, works by binding to Bcr-Abl tyrosine kinase on CML cells, thereby disabling them. The researchers find that in patients in the crisis stage who relapse despite treatment, this action is subverted or the drug is simply overwhelmed.

Recent studies have revealed some of the biochemistry underlying the leukemia. In CML cells, Bcr-Abl tyrosine kinase adds phosphate groups to a protein called Crkl. This phosphorylated protein, in turn, binds to the kinase and links it to other proteins in a chain reaction that triggers white blood cell proliferation. Bcr-Abl tyrosine kinase is difficult to track in the blood, however, so the team monitored Crkl to gauge the enzyme's activity.

In four untreated CML patients, the UCLA group found that the rate of Crkl phosphorylation in cancerous bone marrow cells was roughly three times that of the disabled cancer cells in eight patients treated effectively with STI-571. Tests of CML cells in 11 other patients, who had responded to STI-571 initially but later relapsed, showed more than double the rate of Crkl phosphorylation seen in effectively treated patients, says study coauthor Charles L. Sawyers of UCLA's Jonsson Cancer Center.

Sawyers and his colleagues conducted further tests on nine of these relapsed patients and found that the Bcr-Abl gene had changed in six of them. The combination gene had mutated further to encode a Bcr-Abl tyrosine kinase that's impervious to STI-571, Sawyers says. The genetic change is minor, but the alterations in the protein that the gene encodes are enough to prevent STI-571 from binding to the CML cells.

In the three other relapsed patients, there was no secondary mutation. Instead, the cells had gone into overdrive, mass-producing Bcr-Abl tyrosine kinase. This seemed to outgun STI-571, causing a CML recurrence, Sawyers says.

"This work is very exciting," says John Groffen of Children's Hospital and the University of Southern California, both in Los Angeles. Crkl protein provides a reliable marker for gauging Bcr-Abl activity and the effectiveness of STI-571, he says.

Scientists still don't know precisely how CML spirals into crisis. There's evidence that the cancerous cells acquire mutations in genes other than Bcr-Abl.

"The chromosomes are crazy-looking," Sawyers says, noting that during the crisis stage, missing or rearranged genes might contribute to the steep decline in a patient's health. CML patients often survive 5 years or longer, even with treatments other than STI-571, but those in the crisis stage usually die within months.

Understanding Bcr-Abl behavior in relapsed patients will help researchers find or design a compound to complement STI-571, Sawyers predicts.