DUBLIN – Liquid biopsy developer Freenome Inc. has – literally – capitalized on its recent proof-of-concept AI-Emerge colorectal cancer (CRC) screening study by raising $270 million in a series C round that will enable the South San Francisco firm to complete its ongoing Preempt CRC registration study and to file for pre-market approval from the FDA.
Timelines are, unavoidably, somewhat hazy at present but the company hopes to minimize the disruption to its schedule by expanding the study to practically any U.S. zip code. Its original, pre-COVID-19 goal was to complete enrollment by the end of 2021 and report data in the first half of 2022. “We’re hopeful we’ll hit that timeline,” Freenome co-founder and CEO Gabe Otte told BioWorld. “But things are changing on a weekly basis.”
The registration study, which it launched in May, has a recruitment target of 14,000 average-risk individuals aged between 45 years and 85 years who are undergoing routine colonoscopy screening for CRC. It builds on the prospective AI-Emerge study in 3,500 individuals, from which Freenome reported its multiomics blood test achieved 94% sensitivity and 94% specificity for early stage (I/II) colorectal adenocarcinoma. “It was unprecedented,” Otte said. “That trial was designed almost exactly like our pivotal trial – it was just smaller.”
Freenome’s closest competitor, Guardant Health Inc., of Redwood City, Calif., is also conducting a pivotal trial of a CRC screening test, Lunar-2, which analyses ctDNA genomic and methylation data. It reported 90% sensitivity and 94% specificity for detecting early-stage CRC at the American Association for Cancer Research annual meeting in June. Guardant’s pivotal Eclipse study is recruiting 10,000 participants.
Freenome has adopted a distinctive scientific and commercial strategy with respect to most of its peers. The liquid biopsy field was ignited by the emergence of Illumina Inc. spin-out Grail Inc., which raised $900 million in 2017 to develop a multi-cancer test based on next-generation sequencing of circulating tumor DNA (ctDNA). The Menlo Park, Calif.-based company, which has raised more than $1.9 billion in total, subsequently abandoned this strategy in favor of analyzing methylation patterns of ctDNA, on the basis that this approach offered a greater wealth of information relevant to cancer diagnosis.
Freenome has from the outset pursued a multi-analyte but single-indication approach. Its multiomics test, which employs genomics, transcriptomics, methylomics and proteomics, examines signatures derived from any resident CRC tumor and from the immune system, which is constantly sampling its environment in order to detect transformed cells, as well as invading pathogens.
It employs machine learning to crunch through the resulting volumes of multi-dimensional data to detect relevant biological signals. Training that algorithm to focus on relevant rather than confounding information has taken years of effort. “Earlier on, we had huge issues because the machine learning [algorithm] was picking up on all the confounding differences between the patients,” Otte said. But the machine learning model employed in the AI-Emerge trial was locked. A similar outcome to the Preempt CRC study would herald a new era in CRC screening.
Freenome’s commercial strategy is based on questions of clinical utility, as well as accessibility. Payers may be unwilling to support multi-cancer testing if early cancer detection is of no obvious benefit to patients. A positive test may lead to a diagnostic odyssey, starting with a full-body PET scan, which takes three hours to perform. The value proposition of a blood test for CRC is much easier to understand, given the presence of an existing colonoscopy screening test against which it can be compared in terms of performance, cost and accessibility. “We made an intentional decision to go after the colorectal cancer market,” Otte said. A blood test would make CRC screening “less racist,” Otte said, as it would improve accessibility for low-paid workers who are often unable to miss two days work to prepare for and undergo a colonoscopy. Many African Americans, who are twice as likely to die from CRC as Caucasians are, fall into this category. “It’s a practical impossibility for these people to get screened for CRC,” Otte said.
A blood test would also be more financially attractive to gastrointestinal (GI) physicians, as it would enable them conduct more interventional colonoscopy procedures, which offer better reimbursement than screening tests. No less importantly, it would also save time. “There’s not a single GI doctor in the United States who would say ‘I have extra time in my schedule’,” Otte said.
Its strategy has gained significant moral as well as financial support. Its investors include two cancer organizations, the American Cancer Society’s Brightedge Ventures arm and the non-for-profit Colorectal Cancer Alliance. Lead investors on the current round were Bain Life Sciences, a new investor, and existing investor Perceptive Advisors. Other new investors included: Fidelity Management & Research Co. LLC; Janus Henderson Investors; Farallon Capital Management; Rock Springs Capital; Cormorant Asset Management; EcoR1 Capital LLC; Catalio Capital Management and the Colorectal Cancer Alliance. Freenome has now raised more than $500 million since its formation in 2014.