Ad-Tech Entrepreneurs Build Cancer Database

In In The News by Barbara Jacoby

Thumbnail for 6867By: Timothy Hay


Nat Turner was working as a Google Inc. GOOGL +0.77% product manager when he learned that his seven-year-old cousin, Brennan, had been diagnosed with a rare form of leukemia.

He observed the boy’s dad traveling all over the country interviewing different cancer specialists about his son’s case, all the while carrying stacks of DVDs and medical files, because medical specialists couldn’t easily share information with one another.

Brennan’s treatment entailed initial misdiagnoses, and eventually four bone-marrow transplants that drove the disease into remission. Mr. Turner and his former University of Pennsylvania classmate Zach Weinberg, with whom he had founded Invite Media, an ad tech firm acquired by Google in 2010 for $81 million, began building a new business gathering data on cancer treatments and outcomes over time. Vast amounts of data sit in doctors’ and nurses’ notes, unused, that might lead to insights about treatment efficacy, they figured.

But they faced a fearsome learning curve.

At first, they set out to create a second-opinion service that would make it possible for patients like Brennan and their families to pull up treatments and survival statistics. But after months of research, the two-time entrepreneurs concluded that their New York-based startup, Flatiron Health, would have greater impact if they changed the focus of their platform to connect doctors and cancer centers to one another, for the sake of sharing data that might help them make more informed treatment decisions. Their goal was finding and presenting the data. It would be up to oncologists to analyze it and make treatment decisions.

The duo spent longer than they expected convincing oncology chiefs and the facilities’ chief information officers to give up some patient information. Some doctors and administrators, Mr. Turner says, weren’t comfortable with the idea of sharing case information, which they felt was proprietary.

All information about actual patients and treatments is protected by strict federal privacy laws, so Flatiron spent a full year building a secure architecture for anonymized data, and hired a full-time privacy officer to deal with the many privacy audits that cancer centers subject Flatiron to.

Most treatments for cancer are based on protocols developed from clinical trials. Outcome reporting can lag for a while until papers are published. Flatiron gathers data that weren’t previously available and shareable, the 96% of cases where the patient didn’t participate in a clinical trial.

Revenue comes to two-year-old Flatiron from charging cancer centers for access to its database and by joining with life-science companies on their research. The centers contribute specific information about actual cases, with patient names taken off. This includes the course of treatment prescribed, and the outcome. Every cancer center has lots of such information: If they contribute theirs, they get access to the whole Flatiron database, and their doctors can see a much wider data set as they prescribe courses of treatment.

Today, Flatiron Health has 105 employees, and half of those are engineers. More than 200 cancer centers in the U.S. either use or are testing its database, which holds information about more than 550,000 cancer cases. This spring, it received Google Ventures’ largest-ever investment in a medical-software company, $130 million. Mr. Turner, age 28, recently spoke to The Wall Street Journal about his struggles so far to build the business, which he expects to break even within a few years. Edited excerpts:

WSJ: Why is there a need for the kind of database Flatiron provides?

Mr. Turner: My cousin lives in a remote part of Georgia, and it seemed unfair that there was no way to connect him with specialists in other places who understand his situation. It felt like we should be able to use the Internet to connect doctors with patients.

Our idea changed when we came across a statistic that just boggled our minds. Only 4% of cancer patients over the age of 18 participate in clinical trials.

A lymphoma doctor I met in Boston told me: ‘I see 200 lymphoma patients each year, but there are more than 5,000 lymphoma cases in the U.S. each year. I would really like to see what’s happening with the treatment of these cases.’

We had a lot of trouble with the politics of cancer centers and getting them to agree to participate and add data. The first three or four said yes, but then the next five said, ‘Why would we give you that information? That’s our intellectual property. That’s a security risk.’ We needed a better strategy.

We started investing in the software, because we had to have something that we could show to cancer centers. ‘Give us your data, and our platform will make it valuable.’ This wasn’t easy. We spent a year and a half building OncoAnalytics, the data platform. The database grew slowly for the first year. We just stayed busy building a high-fidelity demo, to get the centers to agree to sign on. We did a lot of evangelizing.

WSJ: In what ways does your tech background help or hurt you?

Mr. Turner: Surprisingly, the technology architecture is similar. At Invite Media, we made sense of disparate sources of advertising inventory. Google, Facebook, AOL, Yahoo—they all have ad space, but they all sell it different ways. The bidding process is all different, company by company. We allowed marketers to place one universal campaign in our system, and our system normalized it and standardized it. An ad is an ad, right?

In cancer care, the disease is similar between one facility and another but each facility has different ways of capturing and storing information about it. So, we’re normalizing cancer data.

WSJ: As two-time entrepreneurs, what’s harder, and what’s easier?

Mr. Turner: Developing a good business model might be harder the second time, as you constantly second-guess yourself. It’s still long hours, but you don’t stress as much about little things. Fundraising and recruiting get easier, once you’ve made a bunch of connections.

Invite Media was considered a good acquisition for Google, so that gave us credibility. Most doctors use Gmail and YouTube and Android phones, and they love these things. They remember Google Health, and they love Google. The academic centers, they feared that we could go bankrupt, because we’re just a little company. But Google is not going bankrupt, they realized.