Supercomputers Tackle Cancer

Four years ago, Howard Young was diagnosed with pancreatic cancer, which is notoriously tough to beat.

"It was really a devastating blow, but I was blessed to have a great support group with my wife and children," Young says.

He had surgery, followed by chemotherapy and radiation. But then, as CBS Evening News anchor Katie Couric reports, Young's doctors at Scottsdale Healthcare tried something revolutionary: a supercomputer that could analyze his tumor at warp speed, 2 trillion calculations per second.

"I was excited there was more that could be done because I wanted to do everything I could," Young says.

Instantaneously, the supercomputer discovered that Young's pancreatic tumor was the result of a specific gene mutation that could be treated with two targeted therapies — Iressa, a treatment normally used for lung cancer, and Urbitux, which treats colon cancer. That was two years ago.

"Without the treatment, he probably would have recurred," Dr. Daniel Von Hoff says.

But that's just the beginning. Supercomputers are taking it a step further, figuring out how a drug will affect an individual's specific form of cancer — even before the drug is tried on the patient.

"Already we are seeing people using targets and using drugs much more affectively," says Dr. Anna Barker of the National Cancer Institute.

Watch more of Katie Couric's interview with Dr. Anna Barker of the National Cancer Institute.

Dr. Brian Druker of the Oregon Health & Science Univ. Cancer Institute discusses the new supercomputer helping to diagnose cancer.

The beauty of this merger between technology and medicine is that doctors can determine therapies that will be effective before they're administered. For so long, cancer therapy was about throwing something against a wall and seeing if it stuck.

"If we know, for example, that you weren't going to respond to a therapy, it would be of great value for you," Barker says.

Meanwhile, a tiny drop of blood is giving doctors at Cedars-Sinai Hospital in Los Angeles a sea of information.

"We get 60 gigabytes of data from one drop of blood," Dr. David Agus says. "That is the picture of all the proteins in the body, the proteins involved in the cancer."

Agus successfully treated Jonathan Roth's prostate cancer. But the story behind his disease lives on, stored in a massive data bank, so other prostate cancer patients who have a similar protein profile can be given the same treatment.

"My blood samples could also be used to help other people because those differences and similarities help the medical community figure out how to pinpoint with laser accuracy a treatment," Roth says.

The NCI wants doctors throughout the country to have access to this information so patients can benefit from those who have come before them. Before long, a supercomputer may predict whether cancer is in your future.

Barker says it won't be long before computers not only will be able to give the best course of action for a patient who's already been diagnosed, but will be able to determine if someone is at risk.

"It's going to be much faster than we think. And I think we're living in an era which is absolutely unprecedented," Barker says. "So I think, just fasten your seatbelts."

You can read more at the National Cancer Institute's Web site at

Click here for more information about the Translational Genomics Research Institute. The Institute is focused on developing earlier diagnostics and smarter treatments.

Click here to read more about supercomputers from the Advanced Biomedical Computing Center at the National Cancer Institute.

Click here to read more about how Cedars-Sinai is using supercomputers. And here for more information about prostate cancer treatment from the Cedars-Sinai Prostate Cancer Center.