Gene chips aid drug search in rare cancers

When Kimberly Stegmaier was a pediatric oncology fellow at Dana-Farber and Children’s Hospital Boston six years ago, she says,”I was struck by how poorly our young patients with AML (acute myeloid leukemia) do.” Long-term survival is about 50 percent, compared with 80 percent from those with the more prevalent ALL (acute lymphoblastic leukemia).

Clearly, better drugs could save lives, but AML is relatively rare and not a likely priority for pharmaceutical companies. Moreover, scientists haven’t identified a specific molecular defect common to all AML cells that could be corrected with an existing or “designer” drug. If such a target were known, it would simplify testing a large number of possible drugs (a process known as “screening”) to identify compounds that cause a desired change in the target.

At the time, the young pediatric oncologist was a member of the laboratory of Todd Golub of Dana-Farber and the Broad Institute in Cambridge. Golub, among other things, has pioneered the use of gene chips – or microarrays – to classify different types of cancer on the basis of their distinctive genetic activity patterns, or gene expression “profiles.”

AML cells, which proliferate uncontrollably and take over the normal bone marrow, are in a state of arrested development. They are immature, dividing abnormally, and unable to become specialized blood cells. Unfortunately, the abnormal proteins or molecular pathways responsible for failed maturation are poorly understood.

Stegmaier and Golub came up with a new approach for screening. First, they’d capture the gene expression profile, or signature, of an immature AML cell and that of a mature, normal blood cell, and identify genes that were expressed differently between them. The next step was to expose the cells in the laboratory to thousands of drugs and chemicals, observing which of the compounds changed the cancerous gene signature to a normal one. Those chemicals merited further testing as potential AML treatments.

After screening more than 1,700 chemicals, the scientists got 13 novel “hits” – compounds that altered the gene profiles of the cancerous cells, and of those, eight reliably induced the gene signature of mature cells. One of the compounds is already being tested in a Phase II clinical trial at Dana-Farber for AML, says Stegmaier.