Why did Roche pay $2B for an oncology healthcare start-up and what does it mean for oncologists?

Authors: Bobby Daly, MD, MBA; Josyula Sowmya, MD, MPH

On April 6, the Swiss pharmaceutical giant, Roche, announced that it had completed the acquisition of Flatiron Health for $1.9 billion. Based in New York City’s flatiron district, the company was started in 2012 by serial entrepreneurs Nat Turner and Zach Weinberg. They had sold their last venture Invite Media, a demand side platform for banner ads, to Google for a reported $81 MM, and Flatiron was born from the challenges one of Turner’s family members had in securing the right information for the treatment of pediatric acute myeloid leukemia. Its founding mission was “organizing the world’s cancer data and making it easy to analyze.” To accomplish this goal, it acquired an oncology-specific electronic health record (EHR) platform and built a suite of software products designed to aggregate and analyze patient data and cull insights that could drive improved outcomes.  In a 2017 ASCO Quality Symposium presentation, Flatiron’s CMO, Amy Abernethy, detailed the company’s “high-efficiency,” “technology-enabled” abstraction of both structured and unstructured EHR fields using a combination of human review by trained medical abstractors and machine learning to collect research-grade data (see figure below, and full JOP article here). This clinical data is then combined with other data sources including claims data, genomic data, and mortality data to create an asset that is of value to patients, providers, payers, pharmaceutical companies, and regulators. These data underlie the value proposition for Roche through three mechanisms:

  1. Accelerate the drug development and approval process: Flatiron partners with 265 community cancer clinics and six major academic research centers and its organization and analysis of their patient data has the promise of identifying the right patient cohorts for clinical trials. The market and the FDA have already seen value in this as Flatiron reports relationships with 14 out of the top 15 therapeutic oncology companies, and the FDA has collaborated with Flatiron on studies using real-world data , adding credibility to their data curation process. As David Shaywitz writes in Forbes, the acquisition offers the “possibility of obtaining regulatory-worthy data with unprecedented ease, potentially saving significant money both from clinical study costs and by delivering the relevant data with the speed of a database query .” The real-world data thus provides a fast track to the development, evaluation, and approval of new therapies, potentially saving significant costs and moving revenue streams earlier in the product lifecycle. Empowering analytics to prioritize which products to bring to market for which patient cohorts provides a possible strategic advantage to Roche in drug development.

  2. Value-based pricing: With the increasing cost of cancer drugs, there has been a push for value based pricing by organizations and institutions such as ASCO and ESMO, and it also appears to be of interest to the Trump White House. The concept is tying the price of a drug to outcomes, and adjusting the price based on a patient’s disease response. Flatiron’s data allows for the assessment of drug performance outside the controlled environment of a clinical trial in patients with comorbidities or other factors that might make them different from the population in which the drug was initially evaluated. This could allow for a deeper understanding of the efficacy of a therapeutic and enable the head-to-head comparison of drugs that might not otherwise be assessed in a trial setting (such as different PD-1 inhibitors) to allow for more nuanced pricing discussions. As Roche’s head of pricing, Jens Grueger stated, “It gives us better insight into the potential for our product. But if also creates flexibility in how we create value contracts in the future.”     

  3. Increase asset value through data enrichment: As reported by CNBC, Roche also has investments and partnerships in Foundation Medicine, a genomic testing company, and 23andme, the personal genetics company. By integrating this genomic data with Flatiron’s data on patients and practice patterns, the company can amplify the clinical value and, hence, commercial value of this asset. This data enrichment will also make it a fiercer competitor against other oncology clinical data companies like Cota and Tempus, which recently became a licensee of de-identified, aggregated data from ASCO’s CancerLinQ.

What are the potential concerns to oncologists about this deal? First, though Roche has affirmed that Flatiron will maintain its autonomy as a wholly owned subsidiary, it raises the specter that Roche will have access to Flatiron’s clinician customers for more personalized marketing of its products. Second, the value of Flatiron is in the data provided by this oncology network. If it loses providers because of this acquisition and concern about the integrity of patients’ data, then the value of its product is eroded limiting its ability to generate clinically meaningful insights. Finally, if Roche moves to curb Flatiron’s relationships with other pharmaceutical firms it could hinder oncology drug development. The indication from public statements is that the aim of the acquisition is “to accelerate industry-wide development and delivery of breakthrough medicines for patients with cancer” but the burden of proof will be on Roche to offer transparency as this integration unfolds. 

Given the above potential return, the $1.9 billion price tag might be a very good deal for Roche and, as one financial analyst pointed out if the value of this deal can be unlocked, the price for future companies in this oncology data and analytics space is likely to soar. This could include oncology clinical pathway companies as well who are increasingly trying to link individual patient treatments to clinical and resource-utilization outcomes. Only time will tell though as to who will reap the returns of this transaction. However, if Roche and Flatiron can collaborate to use these data to create a learning healthcare system to identify the right treatment, either on or off protocol, for the right patient at the point of care it could be of tremendous benefit to patients and providers.

Flatiron’s CMO, Amy Abernethy, detailed the company’s “high-efficiency,” “technology-enabled” abstraction of both structured and unstructured EHR fields using a combination of human review by trained medical abstractors and machine learning to collect research-grade data.

Flatiron’s CMO, Amy Abernethy, detailed the company’s “high-efficiency,” “technology-enabled” abstraction of both structured and unstructured EHR fields using a combination of human review by trained medical abstractors and machine learning to collect research-grade data.