publication date: May. 4, 2018

Flatiron, BMS form collaboration to curate regulatory-grade real-world data

By Matthew Bin Han Ong

Bristol-Myers Squibb and Flatiron Health announced a three-year collaboration to advance methodologies for the curation of real-world data for cancer research and to develop regulatory-grade information.

BMS will use Flatiron data to analyze real-world evidence across a range of tumors and will collaborate on the development and validation of real-world endpoints within Flatiron’s longitudinal datasets.

Flatiron was acquired by Roche earlier this year for $1.9 billion (The Cancer Letter, March 2).

The BMS-Flatiron collaboration includes broadened access to Flatiron and Foundation Medicine’s jointly established Clinico-Genomic Database, which contains research-grade longitudinal data on about 30,000 patients.

In addition to the research agreement, the collaboration extends BMS’s existing work with Flatiron by another three years.

“From our perspective, we have just begun to reveal the potential of real-world data,” said Saurabh Saha, senior vice president of research and development and global head of translational medicine at BMS. “Five years ago we could not have fathomed that we would have this enormous amount of data aggregated across a specific disease area from which we can extract quality information.

“At BMS, we have discrete goals to use digital health for translational, regulatory and commercial purposes,” Saha said to The Cancer Letter. “It’s imperative that the field continue to evolve, and we’re really optimistic about the future.”

A conversation with Saha appears here.

The BMS collaboration will also focus on similar goals of using real-world data to gain insight into how the pharmaceutical company’s cancer portfolio is performing in the real-world setting, said Amy Abernethy, chief medical officer, chief scientific officer, and senior vice president of oncology at Flatiron Health.

“Like Roche, BMS is really interested in what we need to do advance the scientific methods to make sure that this work is done in a credible and consistent fashion,” Abernethy said to The Cancer Letter. “There’s a lot of detailed work that has to go into getting that right—the validation experiments, the third-party review, the getting information into the peer reviewed literature. Similarly, the use of real world data to replicate clinical trials and understand when do real-world data look and act like clinical trials data. What you’re going to see through this partnership with BMS is a set of activities that are intended to drive the methods forward.”

In a recent case study that was presented at the ASCO Gastrointestinal Cancers Symposium in January, BMS used Flatiron data to evaluate survival in U.S. real-world standard of care cohorts and examining outcomes for patients with advanced/metastatic gastric/GEJ cancer who were treated with nivolumab.

The study was conducted by researchers at Memoral Sloan Kettering Cancer Center, Johns Hopkins University Sidney Kimmel Comprehensive Cancer Center, Dana-Farber Cancer Institute, BMS, and the NHS Foundation Trust.

The study matched the U.S. real-world standard of care cohorts to patients with advanced, metastatic gastric, and GEJ cancer who received two or more prior lines of therapy in the ATTRACTION-2 placebo and CheckMate 032 nivolumab monotherapy arms.

After a two-step matching process, US RW SOC cohorts with similar characteristics to the ATTRACTION-2 placebo and CheckMate 032 nivolumab monotherapy arms were identified.

The data suggest favorable outcomes with nivolumab monotherapy vs. SOC in both Asian and Western patients with third-line advanced/metastatic GC/GEJC, researchers concluded.

“The breadth of things that need to be done to advance our understanding of the science of real-world evidence and how to ensure credibility and accurate evidence generation—the to-do list and the methods development is massive,” Flatiron’s Abernethy said. “The first thing to emphasize is that what this really does is get more brains working on the problem. That’s actually, I think, really important.

“So, Roche acquired us because they really want to advance the whole space of real-world evidence.

“The BMS relationship is about regulatory decision-making, and they’re pushing this forward because they also want to see how many ways can they help to advance this conversation. Roche and BMS are taking leadership positions to put regulatory use cases in front of the FDA. The FDA has the opportunity to comment on them and provide advice as to when the data are credible, what kinds of questions can be answered, what use cases such as label expansion can be accommodated.”

Utilization data compiled by Flatiron and made available to The Cancer Letter in 2017 marked a fundamental shift in how cancer researchers can now understand and interpret clinical data in real-time (The Cancer Letter, June 2, June 9, 2017).

Nearly a year later, researchers are moving beyond utilization to studying drug effectiveness in the real-world, as well as dosing, outcomes, and adverse events.

“The case studies that we’re starting to develop now, we’re getting progressively more complex,” Abernethy said. “They’re calling upon us to create statistical analysis plans beforehand and be very clear about what’s the question to be answered and how we’re going to go about doing that. And then, essentially in an a priori fashion, sometimes asking, drug A versus drug B or one dose versus another dose.

“Sometimes they’re asking about patients who otherwise would have been excluded from clinical trials—how do they perform in certain scenarios? So, the list of the kinds of use cases that we’re seeing now or case examples are getting more complex quickly, and these are the kinds of things that the FDA needs to be able to see so that they can then provide feedback.”

A comprehensive real-world evidence framework would improve the way oncologists care for patients who often differ from participants that qualify for a clinical trial.

“And as I see it, especially in oncology, the person who’s got heart disease and is sitting in the chair in front you, that person was never in the clinical trials, but you know, ultimately you got to make a decision on how to care for that person or the person who’s 82,” Abernethy said. “And so, all of these bodies of work help to refine how to take care of real patients and how to make decisions every day.”

Flatiron has been working with FDA to expand the role of real-world evidence in drug development. The 21st Century Cures Act requires the agency to establish, within two years from Dec. 13, 2017, a draft framework for combining real-world data and regulatory science.

“When you bring an example to FDA of comparing situation A versus situation B, they need the following things: clarification about how the dataset was created, clarification about cohort selection, clarification about how you analyze these data, clarification about how you weighted and matched the two different comparison groups in order to make sure that they were as equal as possible,” Abernethy said. “Those kinds of feedback statements from the FDA become critical in deciding what are the main tasks that have to be performed in order to have a reliable study to be brought to the FDA.”

Flatiron continues to maintain autonomy in its partnerships, and the Roche acquisition does not present a conflict of interest for the company or Roche’s competitors, Abernethy said.

“We’ve taken incredible precaution to put firewalls and other mechanisms in place between our clients and between even our clients and our owner, to make sure that the confidential information is preserved within the walls of the appropriate people who need to know,” Abernethy said. “There are similar parallels in other industries. For example, in strategy consulting, there’s lots of need to wall off confidential information between clients.

“We worked hard as Flatiron to make sure that those rules are in place. I think that this BMS announcement actually helps to clarify that that’s a real commitment from us.”

Copyright (c) 2018 The Cancer Letter Inc.