Analyzing Cancer Cells to Choose Treatments

Continued from page 1

Analyzing tumor cells in blood presents two major challenges. Tumor cells are found at very low concentrations in the blood--about one in ten million cells--making it difficult to isolate them. And the small numbers of cells must be analyzed in very low volumes. In the last year, Sloan scientists and others have developed ways to capture these cells using antibodies that detect a molecular marker present only in cancer cells.

In the new Sloan study, scientists face an even more challenging problem--they must detect differences in gene expression, rather than a specific genetic mutation, such as the mutation linked to Iressa responsiveness in lung cancer. Scher and collaborators will use a microfluidics chip made by Fluidigm, a South San Francisco, CA- based company . DNA from each cell is filtered into one of 96 tiny channels on one side of the chip, while reagents flow in from 96 channels on the other side. A precise plumbing system then combines the molecules in different combinations, generating about 9,000 simultaneous reactions. Each reaction takes a volume of just nanoliters--about the size of a period--rather than the microliter volume typical of most commercial fluidics devices. The chip, which costs about $300, "can detect differences in gene expression that are as subtle as twofold with very good accuracy," says Gajus Worthington, Fluidigm's president, CEO, and co-founder.

Researchers plan to analyze levels of about 30 genes in each patient, including genes involved in production of testosterone and in cell signaling. Expression of these genes has been shown in animal models to predict how well a tumor will respond to a drug called dasatinib, which is approved for treatment of chronic myelogenous leukemia and in late stage clinical trials for prostate cancer.

The microfluidics technology could also be used to examine other properties of tumor cells. Scientists might look for changes in gene expression that suggest a cancer has metastasized, or whether a tumor has evolved specific mutations that make it resistant to specific drugs.