Source: online.wsj.com
Author: Ron Winslow

In a fresh advance for the burgeoning field of personalized medicine, researchers have developed a blood test based on the DNA of tumors that could help tailor treatment for individual cancer patients.

The report, announced Thursday, represents one of the most tangible examples yet of how the ability to sequence a person’s entire genetic code could have a direct impact on patient care. There have been a flurry of reports on new sequencing technology that is yielding enormous amounts of information about genetics and disease, but that has yet to deliver much in the way of new treatment strategies.

“For cancer patients there hasn’t been much utility so far. This may prove to be one of the first useful approaches,” said Victor Velculescu, co-director of the cancer biology program at Johns Hopkins University’s Kimmel Cancer Center and senior author of the new study.

Much research involving whole-genome sequencing is aimed at finding differences in the individual letters that make up the genetic code. The belief is that those small alterations will point to molecular pathways that regulate disease, which would be potential targets for drug therapies.

The Hopkins researchers, writing in the journal Science Translational Medicine, took a different approach. They scanned the DNA of tumors taken from six patients with breast or colon cancer, looking not for tiny DNA changes, but what they call rearrangements in large sections of the genome of tumor cells. The DNA of tumors varies genetically from that of normal tissue, and the rearrangements are essentially a fingerprint of the cancer.

The findings suggest that by testing blood for this fingerprint, doctors will be able to learn whether a patient treated for cancer is free of disease or needs additional or more aggressive care.

“It’s a very clever use of the [sequencing] technology,” said Jeffery Schloss, program director for genome technology at the National Human Genome Research Institute, a division of the National Institutes of Health.

Dr. Schloss, who wasn’t involved with the Hopkins study, likened the approach to drawing a map. Sequencing the letters of the genetic code would be akin to plotting every house in a large neighborhood. The Hopkins team was looking only for neighborhoods—in particular, neighborhoods out of place compared with where they would be in normal tissue.

The uniqueness of such “neighborhood” rearrangements is what makes the DNA a good biomarker for the status of tumors, Dr. Velculescu said. A blood test can detect evidence of even minuscule levels of such alterations. That can indicate, for instance, whether a tumor may be progressing after initially responding to treatment or whether any residual disease is present after surgery to remove a tumor. That could help a doctor and patient decide what other treatment options to pursue.

In the study, researchers tracked for two years one patient whose disease initially responded to treatment and then recurred. Levels of the DNA marker in the patient’s blood accurately reflected the course of the disease, researchers said. The test wouldn’t help predict which drugs a patient is likely to respond to—that likely would require a more detailed DNA scan, researchers said.

This advance comes as the cost of sequencing a patient’s entire genome has fallen sharply—to less than $10,000 now from about $1 million three years ago, according to companies that make sequencing machines. Experts predict that the cost will soon get to about $1,000, potentially making it affordable for medical centers to routinely run the genome of patients with cancer and other diseases.

Currently, cancer doctors often use CT scans to check on the status of a tumor—at a cost of about $1,500, the Hopkins researchers said. They estimated the current cost of the DNA test at about $5,000 but suggest the cost will soon be comparable to a CT scan.

Hopkins has filed patents for the technology. Dr. Velculescu and two other co-authors of the paper are entitled to a share of royalties on sales of any products related to the research if it is commercialized. Further research is necessary to validate the approach with more patients and more cancer types, but researchers said such a test could be ready to use within two years and widely available “within several years.”