EXOME

all the information, none of the junk | biotech • healthcare • life sciences

Genentech Redux? Roche Buys Majority Stake in Foundation Medicine

Xconomy Boston — 

In a deal that shakes up the full-steam-ahead world of cancer diagnostics, the world’s top oncology drug company, Roche, said today it’s spending more than $1 billion for a majority stake in Foundation Medicine and creating a powerful drug-diagnostic collaboration.

It’s a significant gamble on the as-yet-unrealized potential of tests that aim for broad genetic profiles of cancer patients’ tumors, one that Roche hopes can not only accelerate the impact of its broad range of cancer-fighting drug programs, but also extend the reach of its global diagnostics business.

The deal has three parts. Roche will buy about 15.6 million Foundation (NASDAQ: FMI) shares at $50 apiece—more than double the roughly $24 at Friday’s market close—for about $780 million through a tender offer. It’ll also acquire another 5 million newly issued shares at the same price in a separate transaction, giving the Cambridge, MA-based Foundation another $250 million—and Roche, as much as a 56.3 percent stake in the company.

All told, that’s more than $1 billion in cash up front for Foundation.

The third piece of the deal is what the companies are calling a “broad R&D collaboration” of at least five years, through which Foundation could pocket over $150 million in additional cash from Roche. The initial focus of that alliance is to develop genomic profile tests both for cancer immunotherapies and blood-based monitoring.

Roche will also get international rights to Foundation’s current products, as well as future offerings the two companies develop together. Foundation CEO Mike Pellini (pictured above) said in an interview that the company will still see some revenue from those products, but declined to offer specifics.

Foundation’s top three shareholders, Kleiner Perkins Caulfield & Byers, Third Rock Ventures, and Google Ventures have agreed to support the deal and tender “at least a majority” of their shares. They hold a combined 31 percent of the company. The deal is expected to close in second quarter.

Foundation will add four new people to its board: three Roche designees, including chief operating officer Daniel O’Day, and one new independent director.

“These deals are done once every 20 years. And in the life sciences arena there’s only been one company that’s done that, and that’s Roche—with Genentech, and then Chugai [Pharmaceutical],” Pellini says.

Pellini says Foundation and Roche have been working together in “a very limited way” over the past few years, but Roche broached the idea of a broad collaboration complemented by an investment in Foundation, and a majority stake.

Pellini says it was “paramount” in these talks that Foundation remain independent after any deal, and that Roche supported that notion “from day one.”

Although Roche is now Foundation’s majority owner—and has in the past, at least with Genentech, used such leverage to engineer a buyout—the Cambridge company says it’s no given this will end in a similar fashion. “There’s no such thing as forever, but there are components built into the agreement that allow us to stay independent, at least for the near and medium term,” Pellini says. He added that some of those details would come to light on Monday when regulatory documents about the deal are filed.

Pellini is remaining Foundation’s president and CEO, and the rest of its executive team will stay in place.

Foundation will hold a conference call Monday morning to discuss the deal.

The attraction here is obvious. Roche is the biggest maker of cancer drugs in the world. But it also has a sprawling international diagnostics business, and is constantly adding pieces to it. In the past months alone, for instance, it dipped into the world of non-invasive prenatal tests and scooped up Ariosa Diagnostics, and then turned around and bought Bina Technologies, a startup that helps manage the data coming out of next-generation sequencers.

The Foundation deal, however, is Roche’s biggest foray in either diagnostics or genomics since it famously tried and failed to engineer a hostile takeover of San Diego-based Illumina (NASDAQ: ILMN). Investing heavily in Foundation is risky for Roche on two fronts. First, it’s a bet that broad molecular profiling will overcome sizeable reimbursement challenges to become part of the standard of care for cancer patients. Right now, it’s more of a desperate measure to help people who have failed a bunch of therapies, as well as a tool for researchers to find new drug targets.

Second, it’s a bet that Foundation, not one of its competitors, is in the best position to bring this new technology into the mainstream of cancer treatment. The need to bring deeper genetic insights to oncology is hard to dispute. How to do it, and with whose technology, is more of a question. “There’s nothing besides the data and the insights you generate that are proprietary,” Johnson & Johnson chief scientific officer Paul Stoffels said, speaking generally, not specifically about Roche and Foundation. “The technology is not proprietary. With deep sequencing from Illumina, soon everyone can do it.”

With the help of some of the world’s richest people, Foundation has built from scratch what is essentially a one-stop shop for tumor profiling. It has developed two tests: FoundationOne, a diagnostic for solid tumors, and FoundationOne Heme, for blood cancers. With these tests, Foundation looks for hundreds of cancer-related genetic alterations in patients’ tumors and uses that information to suggest treatment options to doctors, who read the reports through an Internet portal. Foundation’s database currently contains about 27,000 samples, up from about 4,000 as of May 2013.

The services aren’t cheap. FoundationOne has a list price of $5,800; FoundationOne Heme, $7,200. So part of the company’s challenge is to make them important enough—say, by finding treatment options for those who seemingly have none, or unearthing a tumor-driving gene that has a corresponding therapy that narrower tests would’ve missed—that insurance companies will pay for them.

Making these tests part of routine clinical practice would be the ultimate goal, and Foundation wants Roche to help. Pellini says Roche should help clinicians and pathologists “understand the value of… comprehensive genomic profiling.”

Cancer diagnostics, of course, already exist. Most are focused on single genes known to drive cancer, like HER2 (for breast cancer) or ALK (for lung cancer); or “hotspots” of the genome, as in dozens of the most relevant genes. Foundation is trying to prove that the future lies in gathering much more information about a tumor, and that patients will ultimately benefit as a result.

There is a second part of the Foundation business, however, which is where Roche comes in.

Its tests have their usefuless in the pharmaceutical business, both for researchers who want insights for new targets, and for clinical developers who want to identify the right patients for their trials. To that end, Roche says it’ll use Foundation’s technology to “standardize” its clinical trials, meaning, Pellini says, that it’ll use Foundation’s tests across all of its studies.

Some cancers will take longer to decode than others. For example, multiple myeloma is the second most prevalent blood cancer, but it fragments into about 50 genetic subtypes. Foundation’s Heme test has been of limited use in myeloma, said Multiple Myeloma Research Foundation CEO Walter Capone, but MMRF and Foundation have been working together to improve it. “The near term practical benefit for patients [of broad tumor profiling] is somewhat limited because drug discovery hasn’t caught up yet,” said Capone. “The biology is still about a decade ahead of where drug development is.”

Foundation has partnerships with a number of pharma companies in place. Pellini insists that the Roche deal could “enhance” rather than impact those alliances because its capabilities should increase with the Swiss company’s backing.

Foundation isn’t the only company purporting to offer comprehensive tumor profiling. Others with different types of approaches include Caris Life Sciences (which has been around much longer than Foundation) and Helomics (previously known as Precision Therapeutics). But Foundation is the only publicly-traded company in the group, and the only one to get this level of validation from a large pharmaceutical company. With the support of big-named backers like Bill Gates and Yuri Milner, Foundation had an easy time raising private money, then pulled off one of the rare successful IPOs in the diagnostics field last year.

From there, however, Foundation really hadn’t really taken off until today. Shares on Friday closed just $6 over its $18 debut price in the fall of 2013.

What’s more, the company still loses money, and for all its technological accomplishments, it has major hurdles to clear. It must convince major insurers to cover its tests—chiefly Medicare, whose patients account for some 30 percent of the FoundationOne tests ordered to date. And it must convince oncologists to use its tests to guide their treatment decisions. These types of challenges form the mountain that every diagnostics startup must climb to succeed.

“There are a lot of companies that have come up with great tests, but when they try to bring them to market, they’ve been killed in the process,” says Thomas McLain, a veteran of the industry and the CEO of Cambridge-based Exosome Diagnostics, which is developing a type of liquid biopsy technology.

Getting insurers to pay for tests like Foundation’s is a long arduous slog, with the ultimate goal of “coverage decisions”— insurance-speak for making reimbursement automatic. But because all payers have different requirements, there isn’t one specific piece of evidence or another Foundation can use to win over all payers. So Foundation, as Pellini told me late last year, is “casting a wide net.” It’s running one clinical trial after another, publishing papers in peer-reviewed journals, releasing posters at scientific meetings, trying to demonstrate the value of its tests.

“Our position has been to err on the side of doing as much as humanly possible and hope that’s enough,” he said at the time.

Until Foundation gets the much-coveted automatic reimbursement, getting paid is an uncertain negotiation, with the list price as a starting point. A claim is submitted, and Foundation basically waits to see when the check comes in and how big it is. While Foundation isn’t getting paid for any of its Medicare tests as of yet, Pellini said it does for the “significant majority” of the other 70 percent of its ordered tests. How fast Foundation is reimbursed, and how much, varies, but the typical cash back per test is around $3,500, Pellini said. Foundation has told Wall Street that its reimbursement challenge to get the major payers on board is a “five year effort,” with about three years to go.

There have been positive steps. Last year, Ann Arbor, MI-based Priority Health began reimbursing Foundation’s tests for its 600,000 members. And Google began offering Foundation’s tests to its employees (the company’s VC arm, Google Ventures was a key early investor in Foundation and remains one of its largest shareholders, with an 8.7 percent stake as of an April proxy filing). The amount of tests ordered have steadily increased each quarter. The 6,428 in the past quarter were its highest, up from the 5,908 the prior quarter and 4,702 the quarter before that.

These are small steps for sure. Foundation also needs to see adoption among academic and community oncologists, and there are a few positive signs on that front, as well. Ramp up has been strong so far—as of its last quarterly report, 2,100 doctors from large academic centers and community practices had ordered FoundationOne since its June 2012 launch.

In addition, the National Comprehensive Cancer Network, a non-profit alliance that helps establish practice guidelines for oncologists, endorsed broad molecular profiling (generally, not Foundation’s products specifically) for the majority of patients diagnosed with Stage IV non-small cell lung cancer.

But the company has its share of critics who believe that broadly profiling tumors with an expensive test doesn’t accumulate more useful signals, just more noise.

Here are some thoughts on Foundation, for instance, from Dr. Bradley Monk, a director of gynecological oncology at St. Joseph’s Hospital in Phoenix, who does drug development and translational science and contracts with Foundation for some of the trials he runs.

“I think it’s a research tool,” he says of the company’s test. “The way I use Foundation is if I’m trying to investigate a drug. I’ll treat a whole bunch of patients with it, and I’ll require a biopsy that’s sent to Foundation, and then we’ll look for a molecular marker that predicts response. I don’t use them very often to treat patients.”

The reason? “Although you can identify what you call actionable mutations, I’m not sure that those actionable mutations are ‘driving’ the cancer. I think they’re passengers—in other words they’re just there,” he says.

Monk, instead, views Foundation’s tests as more applicable in “desperate” situations: “We don’t profile [a tumor] for [hundreds] of genes, we profile it for what’s relevant to that tumor type and to those approved agents. And all of that other noise, I don’t know what it means.”

This signal-versus-noise issue is one Foundation is trying to counter. It has given presentations at medical meetings containing anecdotes, like when FoundationOne spotted a type of “ALK fusion” mutation that other tests had missed in a 19-year-old female lung cancer patient. Her prior treatment had failed and she was basically preparing to die in a hospice. Instead, she was given the ALK-targeting drug crizotinib (Xalkori) and was discharged from hospice in five days.

At this point, however, the anecdotes need to turn into data.

“While we do believe providers (and payers) increasingly want more comprehensive evidence of practical utility, we expect this evidence to build over time,” Leerink Partners analyst Dan Leonard wrote after a major medical conference last year.

What type of evidence would sway a skeptic like Monk? “A randomized clinical trial,” he says.

Foundation began running such a trial with the University of Texas MD Anderson Cancer Center in August. The goal: to see whether targeted drugs, identified through broad molecular profiling, will provide patients a clinical benefit. Specifically, Foundation and MD Anderson hope to keep tumors in check longer (what’s known as progression-free survival) in patients with advanced disease. It’s also hoping that Roche will help sway the skeptics through its relationships with the medical community.

But Foundation also isn’t actually providing anything inherently proprietary, making Roche’s zeal to forge a massive alliance with the company both risky, and a huge endorsement in the value of its early-mover advantage and ability to build on it going forward.

It’s unclear how Foundation will defend its business if, say, Illumina continues its push into molecular profiling and throws some of its billions of dollars behind the pan-cancer tests that Foundation offers. The threat is real. Illumina has already teamed with AstraZeneca, Sanofi, and Johnson & Johnson to develop what it calls a “comprehensive tool for precision medicine”—a diagnostic used to test people for dozens of genetic drivers for cancer.

Here’s another scenario: What if billionaire Patrick Soon-Shiong’s NantWorks achieves its ambitions of essentially bringing the entire tumor profiling and analysis process into the point-of-care setting?

Pellini counters that so-called “hotspot panels” (such as the one Illumina is developing with the three pharma companies, which target groups of the most relevant genetic abnormalities) aren’t actually direct competitors—he views them as “complementary.” He argues, for instance, a hotspot test would work well in early-stage disease, when a patient has more time, and a less expensive, narrower test would be more reasonable.

Foundation isn’t focusing there. Rather, it’s mapped out six areas where it thinks its tests are most viable. Those are: metastatic non-small cell lung cancer (“all the different pieces you would find that you would not pick up with a hotspot-based test,” Pellini said); cancer of unknown primary, where patients already undergo a bunch of tests to find out where their cancer originated; rare to uncommon tumors, for which there is typically no standard molecular testing, and patients are often subjected to broad chemotherapies; more aggressive cancers (pancreatic, for one), where there isn’t much time to do more than one test; patients with metastatic disease who have failed multiple treatments; and when there isn’t enough tissue from a sample to do the appropriate testing.

All told, Pellini said these groups add up to anywhere from 800,000 to 1 million patients a year in the U.S. alone.

And while Pellini said it’s “reasonable” that a competitor could have technology similar to FoundationOne on the market in the next few years, he believes it will be too late. “At that point, we’ll have been in the market for five years. Think of the database we will have, think of the relationships we will have with the community oncologists as well as the academic oncologists,” he said. “Don’t you still have to dislodge the relationship Foundation has with its clients?”

Pellini added that the Foundation of three years from now will be a different company, with a bigger database, more products, and more information incorporated into the tests it already offers.

He noted efforts to look into liquid biopsies, or apply what Foundation is doing to cell-free DNA or circulating tumor cells; or to incorporate markers for immunotherapies into FoundationOne.

Roche today is clearly betting Pellini is right. By acquiring shares at such a premium, the Swiss giant is not only betting on Foundation’s utility as a research tool—its partnerships with pharmaceutical companies, after all, do account for some 60 percent of its revenue—but that it really can make a dent on the clinical side, and become a mainstay in cancer care. Can it get there? Roche’s endorsement goes a long way, but payers, and doctors on the front lines, will ultimately decide.

“By combining [Foundation’s] pioneering approach to genomics and molecular information with Roche’s expertise in the field of oncology, we can bring personalized healthcare in oncology to the next level,” Roche’s O’Day said in a statement.

Alex Lash contributed to this report.

By posting a comment, you agree to our terms and conditions.

  • Mineman

    It would ot surprise me if Roche/Genentec expanded their relationship with Samgamo Biosciences. A clear association between sequencing DNA and modifying DNA.