Intel's Eric Dishman: Let's make genomic sequencing scalable

BOSTON—Being a cancer patient for 23 years has given Eric Dishman, an Intel Fellow and general manager of the company’s Health & Life Sciences Group, a unique perspective on care delivery.

Dishman was diagnosed with cancer as a sophomore in college, he explained to the audience at the 2015 Connected Health Symposium held by Partners HealthCare. He “somehow kept surviving the well-intentioned guesswork of care” he received despite the fact that lot of his treatments might as well have been decided with a flip of a coin, he said.

Dishman estimated the cost of his care over the years at more than $6 million. “We can’t afford the cost of imprecise care.”

Dishman had his genome sequenced five years ago. Doctors told him that 92 percent of the medications they had put him on over the years were destined not to work. But, they had spent six months making sense of his genomic sequencing which led to the recommendation to try a drug for pancreatic cancer. Six months later, he was healthy.

Now, he is on a mission to make genomic sequencing scalable to everyone. The computing time for his sequencing five years ago was three months and another four months of manual collaboration. “That’s unacceptable. That’s not going to scale.” Less than 1 percent of people are sequenced today and most patients are in trial-and-error treatment.

Annually, newly diagnosed cancer patients in the U.S. generate more than 1 terabyte of data each and more than 4 exabytes total. It would take 1,500 million server hours and a $53 million power bill.

That doesn’t include the time it takes to do secondary analysis and find other similar people to see what worked for them. Or to walk through existing databases to find failed drugs that might have a chemistry appropriate for someone else. “It’s like looking for a needle in a haystack that in not all in the same place. It’s not scalable or sustainable.”

Dishman’s goal is to accomplish all this analysis and discovery in one day by 2020. With today’s rate of progress it will take until 2035 but “if the ecosystem worked together around standardizing workflow and tools we could get there much faster.”

Only 4 percent of data about cancer is currently available to research, Dishman said. The rest is stuck in private cancer centers and is too large to move and is protected by privacy laws and business models.

To combat these challenges, Dishman helped launched the Intel Collaborative Cancer Cloud, a precision medicine analytics platform that allows institutions to securely share patient genomic, imaging and clinical data for potentially lifesaving discoveries. The center is working to help organizations share their private clinical and research data with each other to generate larger datasets to benefit research and discover more tailored treatment.

"Cancer is never really done with us,” Dishman said, noting that he still has 12 clinicians. Because of this ongoing burden, “we need to get out of attitudes that are keeping us from sharing data.”

Systems are still not interoperable and sometimes it’s intentional, he said. Some organizations are hiding behind HIPAA. “That’s appalling. We need to connect the players and connect the patients.”