February 15, 2017 (Vol. 37, No. 4)
Flatiron Health Aggregates Cancer Treatment Data from Clinical Trials and the “Real World”
Just 4% of cancer patients enroll in clinical trials. The other 96% are in the “real world,” receiving the standard of care. All of these patients experience treatment outcomes that should interest cancer researchers, drug developers, and clinicians. Yet the 4% receive a disproportionate amount of attention. While the 96% have their experiences collected in electronic medical health records (EMRs), these experiences are rarely included in accessible cancer databases.
If we continue to neglect the experiences of the 96%, we may end up with a narrow view of cancer. A broader view, however, is possible—provided outcome data from both clinical-trial and real-world cancer patients can be standardized, integrated, and analyzed on something like an equal footing.
Enter Flatiron Health, a data company that supports cancer care providers and life science companies through its OncologyCloud™ software suite. OncologyCloud components include a cloud-based EMR for oncology, advanced analytics, a patient portal, and integrated billing management.
“Our goal is to learn from every patient’s experience—not just those in clinical trials,” declares Zach Weinberg, president and COO of Flatiron Health. Weinberg co-founded Flatiron with Nat Turner, the company’s CEO. Together, Weinberg and Turner assembled the team that developed the OncologyCloud platform, an infrastructure to organize, track, and analyze data from cancer patients regardless of whether they are enrolled in clinical trials or not.
“There were two different challenges involved in creating this database,” Weinberg recalls. “First, we had to build a network of cancer centers that wanted to work with us to use their data more effectively.” Nearly five years after Flatiron was founded, more than 260 cancer centers throughout the U.S. share information with Flatiron’s network.
The clinico-genomic database developed with Foundation Medicine lets researchers delve even deeper into the data, because it includes tumor sequencing data for 20,000 cancer patients. “Relatively few cancer patients receive tumor sequencing,” Weinberg points out. Yet these patients provide data that is integral to understanding the correlations among tumors, therapies, and outcomes. Because it incorporates this data, the clinico-genomic database will be especially enlightening.
“The second challenge in creating this infrastructure was to curate the data,” Weinberg continues. Curation is vital because the underlying data is, as he says, “messy.” It includes structured and unstructured data in a variety of formats, and often such important data fields as stage of disease, smoking history, and histology are missing. “There are no standards for the data.” Therefore, Flatiron had to normalize the data to make it easily usable.
Atop this infrastructure, Flatiron developed a suite of products for clinicians, patients, and researchers. They include OncoAnalytics® for deep clinical and business insights, OncoEMR® for EMR and workflow software, OncoBilling® for claims filing, and SeeYourChart® for sharing laboratory tests and other data directly with patients. Last November, Flatiron was certified by the Health Information Trust (HITRUST) Alliance, an information protection framework supporting patient data privacy.
“We collaborate with biopharmas and research groups, and we use real-world evidence to help them understand how their therapies are performing in the real world,” Weinberg says. Clients include 11 of the top 13 life sciences companies, with Bristol-Myers Squibb, Roche, and Merck among them.
Flatiron also is working with the FDA to determine how real-world information gained outside clinical trials can be used for regulatory decision-making, particularly in the areas of safety and efficacy. This particular collaboration analyzes the use of immunotherapies to treat advanced non-small cell lung cancer (aNSCLC) with the goal of determining the most effective analytic approaches, clinically relevant endpoints, and safety assessment methods.
Identifying best practices and statistical methods doesn’t depend entirely on the outcome of the FDA study, however. “We work daily with the cancer clinics and centers in our network to provide analytics to make data from EHRs useful for physicians, researchers, and patients, and to help providers resolve operational challenges, such as better ways to identify patients for clinical trials,” informs Weinberg.
By networking with large numbers of community and academic oncology centers, Flatiron can answer questions smaller databases can’t address, notes Weinberg: “In rare diseases, for example, it may take 200 cancer centers to find 10 to 20 patients.”
Making Humans More Efficient
“There’s a misconception that compiling and curating databases like this is purely an informatics problem,” Weinberg remarks. “There’s a tremendous amount of software infrastructure involved around tools, analytics, data processing, and curation, but we’re not trying to automate human beings. Instead, we’re building software to make humans more efficient.”
To do that, Flatiron has attracted a diverse staff that includes biostatisticians and community and academic oncologists, as well as IT specialists from powerhouses such as Google, Microsoft, and Facebook. Consequently, Weinberg suggests, Flatiron can sustain and reap the benefits of collaborative efforts: “Projects that normally would take six months and 20 people to accomplish, we can do in less than one month and with only a few people. If we had just healthcare experts or just software experts at Flatiron, this wouldn’t be possible.”
“At our core, we are a software engineering company,” Weinberg insists. As an angel investor and serial entrepreneur, Weinberg founded an advertising analytics company in 2007 that was acquired by Google. After folding the company into the fabric of the tech giant, he and Turner wanted to apply their skills to something that was more “impactful” socially.
“We spent 18 months travelling. We met experts in healthcare and oncology and took 300 or so meetings. It was an educational process,” recalls Weinberg. “There was no ‘Ah ha!’ moment. It was, rather, a series of steps that led to the formation of Flatiron.”
As the company grows, Weinberg and Turner plan to increase the number of cancer centers in its network, expand the type of oncology research in which it participates, and continue building its team.
“We’re going to stay focused on oncology,” Weinberg stresses. “The more we can learn from patient experiences, germline or tumor sequencing, or other data, the better, faster, and cheaper it becomes to do research—so everybody wins.”
Location: 200 Fifth Avenue, 8th Floor, New York, NY 10010
Phone: (202) 840-6964
Principal: Nat Turner, CEO
Number of Employees: 400
Focus: Flatiron Health creates infrastructure that can organize, track, and analyze anonymized data from cancer patients, enabling researchers to learn from real-world treatment experience. The company’s curated database includes data from 1.5 million patients.