"Sometimes the BCR-ABL protein itself has a mutation, which makes it less susceptible to imatinib," said Li. "But in some patients who do not respond to imatinib, there's no imatinib-resistant mutations detected and the disease still progresses. No one understood how that happens when the known disease cascade is inhibited."

In research published in online in the Proceedings of the National Academy of Sciences, Li's team discovered is that imatinib does not inactivate all BCR-ABL signaling pathways in the cascade. Part of the cascade, proteins called SRC kinases, are still activated by BCR-ABL in imatinib-treated mouse leukemic cells. When Li treated the mice with a compound, dasatinib, that inhibits the SRC proteins as well as BCR-ABL, he found that not only was it more effective for CML, but it also led to complete B-ALL remission.

While these results are encouraging, a small population of leukemic cells (fewer than 1 percent) persisted through treatment with imatinib or dasatinib and led to a recurrence of the leukemia. These cells, leukemic stem cells, present an additional challenge.

"These results show that clinicians need to address SRC kinase activity as well as BCR-ABL to get the best outcomes with Ph+ leukemia," said Li, "but it's not a real cure. The keys to a cure for Ph+ leukemia are the leukemic stem cells, and we've now isolated them in the mouse for the first time. We will be working hard to figure out how to target and eradicate them."