In a new clinical trial for prostate cancer, scientists will capture rare tumor cells circulating in patients' blood, analyze them using a specialized microchip, and use the results to try to predict how well the patient will respond to a drug. The trial reflects a new phase of personalized medicine for cancer, enabled by microfluidics technologies that can isolate scarce cancer cells and detect very small changes in gene expression. Physicians ultimately hope these chips can become a routine part of clinical care for cancer. "We need to be able to profile the tumor at the time we are considering treatment," says Howard Scher, chief of the Genitourinary Oncology Service at Memorial Sloan-Kettering Cancer Center, where the trial will take place.

The study will focus on men with a difficult to treat form of prostate cancer that has failed to respond to other therapies. Changes in gene expression might help determine whether a specific drug will be effective--for example, if a patient has high levels of a receptor for androgen hormones, a drug that inhibits signaling of that receptor is more likely to work well. "We want to know why they don't respond to therapy and what therapies would be best for them," says Martin Fleisher, chairman of the Department of Clinical Laboratories at Sloan. "We collect tumor cells from blood, and do a gene analysis to find out what genes are overexpressed and whether or not they would be candidates for certain types of targeted therapies that would beat down their cancer."

The effectiveness of different cancer drugs can vary based on the molecular characteristics of the cancer, such as the presence of a certain hormone or genetic mutation. Physicians already do some molecular analysis of cancer tissue to select the best drug for a patient. Herceptin, for example, is used to treat breast cancer in women with a particular protein in their tumors. And lung cancer patients with a mutation in the gene for the epidermal growth factor receptor are more likely to respond to a drug called Iressa than patients without it. But these treatments are chosen based on analysis from tumor biopsies, which isn't always possible.