To hear Dr. Norman Sharpless tell it, it’s time to open a new front on the war on cancer. That front, he says, will match the latest cancer treatment advances and lab breakthroughs with big data analytics to help determine the best treatment options for patients. According to Dr. Sharpless, oncologists and molecular biologists are drowning in data. The Smarter Planet blog caught up with him recently to discuss his ideas for overcoming the challenge.
Smarter Planet: The World Health Organization predicts that the number of new cancer cases will grow 70 percent within two decades. It’s no wonder many of us believe we’re losing the war on cancer.
Dr. Norman Sharpless: Cancer is not one disease. That misconception dates back to the Nixon administration. Maybe we should have said back then we’re launching a war on cancers. Each cancer requires different treatments. Each has different causes. And because every cancer is different you really can’t talk about a one-size-fits-all approach. What is needed is personalized care.
SP: How can personalized care be an option with such an explosion of new cases?
DNS: Part of the answer is big data. Another part is gene sequencing, which can give us that vital detailed personal information about specific genetic mutations. We’re going to have to process all that data with something. That’s why we are eager to have our petabytes of patient data analyzed by Watson.
SP: How did you first team with Watson?
DNS: Through a colleague, Lynda Chin, at MD Anderson. I’m on the external advisory board there. Lynda is the chief innovation officer and a big data wrangler. She showed us what they were doing with Watson, and I got it. Lynda was programming Watson to read and ingest disparate sources of data on a single patient. Watson would read the chest X-ray, the electrocardiogram, the physician notes from another hospital, pathology results too — images, PDFs, and some spoken word. Watson would make sense of it all and tie it up into something an oncologist could review really quickly. Watson would also generate evidence-based treatment options based on its understanding of the problem. So it had this sort of expert adviser capability. That is a real problem in oncology, because we’re drowning in information.
SP: At the World of Watson conference you proposed organizing a clinical trial for cancer treatment involving Watson. How would that work?
DNS: The trial would involve taking patients and assigning a therapy the old way. That means doctors implementing recommendations from the molecular tumor board, say, versus a new way. Which would be the same caliber of doctors but with help from Watson. Watson would suggest treatment options based on their training data set. This would be a really straightforward trial, and I think it would accrue, from a patient’s point of view, very little risk. In fact, I think it’d be very appealing to them to have a cognitive computer help their doctors.
SP: What do you hope to achieve from this type of clinical trial?
DNS: We’d like to develop an algorithm that uses genomic information, DNA and RNA, to make recommendations on therapy that are better than what a group of humans can. I think that’s an achievable goal. I’m not saying I want Watson to pick your chemotherapy. I want Watson to provide treatment options to help your doctor pick the optimal chemotherapy. I think it’s very doable.
SP: If your hypothesis is correct, what would that mean for how we treat cancer patients in the future?
DNS: Cancer is not going away, but John Q. Public might be able to have his cancer diagnosed by his physician earlier. He might be able to get more effective treatment and have a much better chance of being cured. We’re going to reduce toxicity. We’re going to save money. We’re going to save lives.
Dr. Sharpless’s story is one of several featured in IBM’s Wild Ducks series of podcasts and videos that celebrate the innovators of tomorrow, today. Check out the entire Wild Ducks series of podcasts and follow us on twitter @IBMWildDucks.
And be sure to read these other great Smarter Planet stories on the Wild Ducks series: