Imprecision Medicine: Eric Dishman Kicked Off the Converged IT Summit

By Allison Proffitt

September 15, 2015 | After a series of fainting spells in college, Eric Dishman was given a bleak prognosis: 2-3 years to live thanks to kidney disease. Of course that was a long time ago.

Today an Intel Fellow and general manager of the Health and Life Sciences for the Data Center Group, Dishman shared his own 1% story of successful precision medicine—“I like ‘personal’ better, but I’ll use ‘precision’ to be part of the hype.”—at the inaugural Converged IT Summit* co-hosted by the BioTeam and Cambridge Healthtech Institute, setting the tone for a two-day look at trends from the trenches of IT and life sciences discovery.

Dishman spent 23 years being treated for a rare kidney disease racking up $6 million in medical costs, $380,000 of which he was personally responsible for paying. After having whole genome sequencing and about three months of analysis time, Dishman began taking a pancreatic cancer drug that targets the same pathways that seemed to be responsible for his disease. For the first time in 23 years, he was disease-free.

“We have to admit we’re doing imprecision medicine,” Dishman challenged the audience. Most doctors, he argued, are at best practicing trial-and-error medicine.

Intel is pushing toward “all in a day” precision medicine; a goal Dishman hopes to reach by 2020. He acknowledged that it’s a big problem. A diagnosis isn’t just a needle in a haystack he said. It’s a needle in a distributed haystack with different people having different levels of ownership of the haystack.

Yet Dishman still envisions taking patients from diagnosis based on ‘omics, clinical and environmental data to a targeted treatment plan in one day.

Data management will be a huge stumbling block, Dishman conceded. If we start just by sequencing the new cancer patients each year at a terabyte of sequence data each, we will generate four exabtyes of new data annually, data that will need to be stored, analyzed, and securely shared.

Still, “We are going to get there technically before we have the business model and workforce to enable precision medicine,” he predicted.

For Intel, the business model question is a research question, Dishman said. As a first step, Intel launched the Collaborative Cancer Cloud earlier this month (see, Intel Joins Up with Cancer Centers to Build Secure Networks for Patient Data Sharing). Intel has been focusing on design pain points, Dishman said: how to move analytics, but not data; how to build an open and secure trust model (much of the tools associated with the Collaborative Cancer Cloud are open source); and how to scale the solution for any researcher.

Dishman also stressed the role that government has to play in the changing landscape. Dishman sits on a Precision Medicine Initiative Working Group and he praised the effort for being bipartisan and bicameral so far. His working group will soon report its recommendations for data infrastructure and what data to collect.

Achieving “all in a day” medicine is possible, Dishman believes, if the industry can shift its thinking: Reduce the cost and complexity of real solutions. Quit focusing on monetizing patient data (what Dishman calls a naïve gold rush era land grab). Tool providers must stop trying to make their tool the preeminent tool, and instead embrace and use the common tools. And the industry should drive and adopt standards.

Intel is doing its part to focus on the business model, workflow, and workforce challenges, Dishman said, and design for scale. “Let’s skate to where the puck is going… We are working on problems that are not even yet problems.”

*The Converged IT Summit was held in San Francisco, September 9-10, 2015.