Pharma’s New R&D Models: Federating Innovation

Big pharma’s R&D productivity challenge is forcing a re-think in how to access and nurture innovation. New technologies are altering the kinds of partners and skills that pharma firms need.

Pharma groups are all changing, but they’re not doing so in the same way or at the same speed.In Vivo explores the approaches taken by J&J, Bayer, Roche and Novartis.
So what? The shifts within big pharma R&D are part of wider-ranging changes to pharma’s products, business models and its role in health care delivery more broadly.

Whatever numbers you use, big pharma’s return on its R&D investments is declining fast. New drug approvals are increasing, but they are coming from smaller companies: the FDA’s 51 novel drug approvals in 2017 were spread across 42 sponsors. Innovation is scattered around. Biotechs, smaller biopharma, and, increasingly, technology and digital health start-ups are eating big pharma’s lunch. “We face an existential challenge,” says Bayer AG  board member Kemal Malik, a 27-year pharma industry veteran.

Until recently, few senior executives would talk openly about the problem. For good reason: most companies were still living richly off the spoils from when big pharma R&D did work, and beautifully well. Today, many of those executives who have not jumped ship to biotechs or venture capital funds are up-front about the issue, and what they’re doing about it. “The data tell us, consistently, that we have to do something different,” says Malik, who is also in charge of innovation at Bayer.

Big pharmais changing but each firm is evolving differently, reflecting their respective leadership styles, therapy area priorities, culture and expertise. There is not a single right way to get better at discovering drugs (though there are wrong ways: see sidebar When Structures Speak Louder Than Science). These transformations are intertwined with rapid advances in digital technologies, data sources and data analytics tools.

Johnson & Johnson: Multi-Channel Externalization

Big pharma has long been licensing and acquiring R&D assets from outside. Dollar for dollar, smaller biotechs, with new technologies and creative, focused approaches (sometimes forced by limited funding), have proven better at discovering new drugs than large pharma.

Johnson & Johnson's R&D, under the stewardship of EVP and Chief Scientific Officer Paul Stoffels, is held up by some in the industry as the most radically externalized. Stoffels has set up a series of channels for accessing outside science: JLABS biotech incubators, in nine cities across North America, China and Europe, and idea-hunting, modality-agnostic Innovation Centers in San Francisco, Boston, London and Shanghai. These sit alongside more traditional corporate venture investments and partnerships. “We’re organized to be able to embrace great innovation wherever it arises,” sums up Richard Mason, head of the London Innovation Center.

The strength of the J&J model is not the share of external versus internal science, though. Instead, “the strength is the flexibility of the system,” explains Stoffels. There’s no single way, or time, to access external innovation. A molecule could be taken on earlier, or later, by partnership, investment, or acquisition, he said. Or it could be developed in-house. What matters is not who invented it, but “who brought the best solution forward.” The reward structures reflect that: there is no difference in compensation, whether something is internal or external. Therapy area heads, like Peter Lebowitz in oncology, are charged with getting transformative products to patients, fast, wherever they come from. Sometimes, that still means buying a bit of pipeline: in May 2018, Lebowitz’s group bought BeneVir Biopharm Inc., with its clutch of preclinical oncolytic virus candidates.

The therapy area heads are behaving a bit like venture capitalists, in other words – making capital allocation decisions across a portfolio of programs, whether internal or external (or a bit of both). Meanwhile, JLABS mirror high-end technology transfer set-ups: Johnson & Johnson provides state-of-the-art equipment and infrastructure for selected “strategically-relevant” start-ups, thereby reducing the capital required to achieve proof of concept. “We’re creating a community where it’s easy for young scientists to start-up,” says Stoffels, who co-founded his own biotech, Tibotec Group NV, in 1997 without such support. Tibotec was acquired by J&J in 2002.

It is not philanthropy, however, as for each dollar that J&J pays into the various JLABS, someone else – the start-up’s backers – pays $4. Yet “in our internal shop, we pay dollar for dollar,” says Stoffels. In other words, JLABS is a highly-leveraged external innovation engine, providing a second leg to the group’s overall activity, according to Stoffels.

If J&J can make each of its R&D dollars go five times further outside than inside, it could forget the in-house part, and become an innovation portfolio manager. Yet that would be to overlook the value of experience. “Pharmaceutical development know-how is built over a very long time. We fail much more than we succeed by experimenting. Having that knowledge in an organization is very important to improve success rates,” says Stoffels. You need experts to judge what is good, in other words.

Stoffels would not specify how many in-house experts are required for success. “You can’t do the maths on that,” he says. Besides, ‘internal’ versus ‘external’ is no longer binary; it is unclear, even to Stoffels, how many of J&J’s approximately 10,000 R&D employees are working on internal versus external projects, given the number and nature of collaborations across all areas, from toxicology to pharmacokinetics.  Discovering drugs, these days, is about having the right mixture of expertise: few drugs are found by a single team or organization; most are target-oriented, requiring a range of platform technologies.

J&J today is structured around half a dozen disease areas [cardiovascular & metabolism; immunology; infectious diseases & vaccines; neuroscience; oncology and pulmonary arterial hypertension] each with considerable autonomy in managing their portfolio, but with access to company-wide platform technologies (e.g. antibodies, or small molecule chemistry) and downstream resources across regulatory, access and commercial. But Stoffels talks less about structure than about people and culture. He repeats advice offered to him from his mentor, Paul Janssen, founder of what would become Janssen Pharmaceutica in 1956. “Janssen told me: first get the people, then build the organization around the people.” And when you’ve built it, make it work by following the principle that “you take the risk and I take the blame.” Companies must give people freedom to operate.

Pharmas must also be on top of what everyone else is doing in a given field in order to succeed, and have an efficient development and approval process for rapid market access. J&J’s Accelerando program is designed to accelerate certain life-saving drugs from development through to the market, by expediting data collection, analysis, report-writing and regulatory submissions.

But is J&J’s outreach model working? There are promising signs. Among its peers, it is third-best at getting products from Phase I to on-market success, according toIDEA Pharma’s Pharmaceutical Innovation Index, which ranks big pharma according to clinical, commercial and regulatory metrics. The group’s shares have risen by over 30% since early 2013. Yet JLABs and the Innovation Centers are only five years old. J&J is over 130 years old.  "This is a big experiment," cautions Stoffels. Results are measured at the end, not at the beginning.”

Bayer: Federating Innovation

Bayer is another long-established company at the start of a big experiment. Unlike J&J, letting innovation seep into its organization through increasingly porous boundaries, 150-year-old Bayer has ring-fenced its experiment outside company walls. The Bayer Leaps program is investing heavily to build separate companies around transformative technologies, such as cell-based therapies, gene editing and RNA activation. These breakthrough technologies are “almost orthogonal to what Bayer does” inside, mostly focused on small molecule chemistry, says Malik.

Why keep it all separate? Because these new approaches could be a competitive threat to an R&D model that many employees have built decades-long careers around – and that most see little need to tamper with. “In a company like Bayer, that has survived for as long as it has, the need for change is less apparent,” concedes Malik. That in turn means there’s less appetite for change. “Change challenges people, their role, their careers.”

Bayer Leaps – investing equity, not R&D dollars that would hit the P&L – looks rather like corporate VC. It’s not, Malik insists, because the sums are larger, the investments are fewer, and the time-frames are longer. Bayer Leaps is a handful of huge, long-term investments, free of the internal rate of return constraints and fixed-term cycles associated with most VCs. They are joint ventures with other investors, however: stem-cell focused BlueRock Therapeutics was created in late 2016 with Versant Ventures; the partners are providing $225 million over four years.Casebia Therapeutics is a joint venture with gene-editing firm CRISPR Therapeutics AG, created in mid-2016 with $300 million from Bayer, and $35 million from CRISPR, over five years. A third company, applying synthetic biology tools to the plant microbiome for agricultural applications, received $100 million in September 2017 from Bayer, Ginkgo BioWorks and Viking Global Investors. “These are some of the largest series A investments ever,” says Malik.

The idea is for Bayer Leaps companies to become best-in category, with the best programs and, critically, exclusive and robust intellectual property. Cambridge, MA-based Casebia will have IP exclusivity on CRISPR-related research programs covering cardiology, ophthalmology and hematology, according to Malik. These are Bayer’s traditional areas of focus, and will provide a foundation for future partnerships, in which Bayer will share some of its own IP and know-how. Meanwhile, Vertex Pharmaceuticals has rights to CRISPR-based therapies in certain blood disorders. “We’re federating IP,” sums up Malik. (Also see "Giant LEAPS For Mankind: Bayer's Malik On Breaking The Mold In R&D" - In Vivo, 11 Apr, 2018.)

Rather than putting all of Bayer’s innovation eggs in just a few baskets (the sums involved are a fraction of the €4.5 billion Bayer AG spent on R&D in 2017), Bayer Leaps envisions a future where “pharma can act as glue, bringing together multiple activities and technologies around a particular disease,” says Malik. For example, CRISPR Therapeutics and Casebia have already bought access to CureVac AG's messenger RNA technology for their liver disease programs.

Although Bayer Leaps companies are deliberately sealed off – financially, physically, and operationally – from Bayer, the hope is that some of themodus operandi of these start-ups will eventually percolate into the mothership.  Joerg Moeller, head of a combined discovery and development operation at Bayer Pharma since January 2018 (following the departure of discovery chief Andreas Busch) sits on the board of Bayer Leaps, along with Bayer Leaps chief Axel Bouchon. So, the person responsible for traditional R&D can see close-up how small companies progress – typically more rapidly, with fewer people. “It will be a bit of a wake-up call. These are strong subliminal messages going into the organization. They won’t always be welcome ones,” warns Malik. Bayer R&D, shaken already by Busch’s departure, may be facing further job cuts, according to recent reports in the German media.

Malik is optimistic about pharma’s continued survival. But the business model will be “fundamentally different”, he says. “We’ll continue to do development, commercial and manufacturing. But we’ll pivot to a new model for sourcing innovation, collaborating with those who are good at it. We federate. “

Roche: Data Curator

Roche is collaborating and amalgamating, too – around data. Through acquisitions and partnerships – most notably, the $1.9 billion purchase of Flatiron Health Inc. in April 2018 – Roche is transforming its R&D by turning into a data curator. It’s accessing and analyzing vast volumes of data that are key to developing and delivering new, effective therapies. Indeed, “over the next five years, we’ll have three major legs to the company: pharmaceuticals, diagnostics and data/data analytics,” says pharma CEO Daniel O’Day.

O’Day claims that Flatiron data – electronic health records for over a million patients across 250 oncology practices – has already helped at least one medicine gain approval in 20 countries over a year faster than it would otherwise.Alecensa (alectinib) targets ALK-mutated cancers that occur in about 40% of lung cancer patients. Flatiron’s database included enough patient data, of sufficiently high quality, to make up a control group using standards of care (SOC) regimens appropriate to local payers and regulators. “The Flatiron database provides us with data that gives regulators and payers what they need….to accelerate the drug” onto the market, says O’Day. It helps solve developers’ challenge of running clinical trials to cover all of the growing number of SOC regimens used in cancer.

O’Day hopes that in future, Flatiron’s structured, high-quality data may substitute for SOC arms in prospective clinical trials. “We looked at data from the chemotherapy SOC arm in a trial of [bladder and lung cancer drug] Tecentriq (atezolizumab), and compared that with Flatiron data from patients on the same SOC regimen. The curves overlapped,” he says. Tapping into a ready-made source of control arm data could significantly reduce both the time and costs of getting a drug to market. “Real world data is now mimicking randomized control trial data. This is game-changing,” says O’Day.

The Flatiron data is just one (big) piece of the puzzle (see Exhibit 1). Roche’s majority stake in Foundation Medicine Inc., which collects genomic sequencing data from cancer patients, is another. Combining both sources to create a full, genotypic-phenotypic picture of patients’ cancer over time will be the Holy Grail, according to O’Day. He expects that the overlap between the two sources, currently about 30,000 to 40-000 patients, will rapidly expand into the 100,000s, uncovering rare cancer-causing mutations and consequently ready-validated targets for new therapies.

Flatiron and Foundation Medicine are both free to pursue relationships with other pharmaceutical firms, as well as providers, practices and patients. But the data and analytics skills Roche is acquiring cannot sit outside the main organisation, Genentech-style; they need to merge with traditional R&D to accelerate and inform all stages of the process.

And they are, according to O’Day. “We’re systematically bringing a new discipline into the organization,” he says. But the challenge is not firing up researchers to believe in the transformative power of data. Nor is it marrying a technology company with a pharma company; even though the groups do things differently, “we’re both focused on patients’ needs.” The challenge is organizing and harnessing all the various types of data, analytics and related technology platforms. “It takes a great deal of my time,” says O’Day.

Novartis: Digital Development

Novartis AG’s C-suite has also spent a lot of time on data and analytics, believing this to be key to transforming R&D efficiency. CEO Vas Narasimhan, during his previous role as chief medical officer, started to digitalize the company’s vast clinical development operations – by far the biggest R&D line item. Last autumn, the company rolled out its advanced analytics platform, Nerve Live, which gathers vast volumes of data from past and current trials, linking it into multiple daily systems used in development. It closely monitors the progress of the hundreds of trials ongoing at any one time across the globe, and provides predictions and insights based on trial objectives. One goal is “to predict failures before they happen,” says Head, Global Development Operations, Badhri Srinivasan. Others include designing and implementing trials more efficiently. The company expects Nerve Live to help reduce patient enrollment time by 15%. The Nerve Live Study Operation Center in Basel (portrayed in some media outlets as a Star-Trek-style Mission Control) certainly appears to take Novartis firmly into the digital era. But Luca Finelli, Head of Predictive Analytics and Design, strikes a note of caution. “We must be wary of claiming benefits too early,” he says. When it comes to capturing, analyzing and using all the data now more readily available, “we’re just scratching the surface,” he says.

But Finelli highlights another benefit from the digital revolution, beyond efficiency: transparency. Having integrated data systems that are designed to be accessible means that everyone, from the senior director to the assistant laboratory scientist, sees the same data picture. So “instead of spending energy telling each other stories about the pieces of data one does have access too, everyone can combine forces to address the strategic questions that emerge from a full, shared perspective. It’s a powerful change,” he says.

It is also likely to be a disruptive change. It challenges not only work methods and processes, but also existing hierarchies, power-systems and mind-sets.

The cultural side of pharma’s tech-driven R&D revolution may be the toughest. Sure, some of the technologies are complex. The data volumes are unimaginable. But embracing and trusting new kinds of experts, flatter, more porous organizational structures, multiple kinds of partnerships and data systems, and calling into question all the traditional R&D processes in place for decades: that will be more difficult. “The technology itself is usually the simplest part. Figuring out what to do with the technology, ensuring people feel supported in using it, and working out how to actually deliver solutions is the hard part,” says Craig Lipset, Head of Clinical Innovation at Pfizer Inc. “That’s our biggest challenge as an industry.”

C-suite compensation is another challenge facing big pharma in its transformation. Individual leaders’ appetite for change is evident, but the right incentives are needed. Senior executive pay should be linked less to sales or even profits (since these can, and often do, arise from innovation-free price rises) and more to innovative and accessible products, priced to reflect their outcomes, not what the market will bear. “There is a climate for change. But to effect it, you need to address compensation, and tell investors why you’re doing that,” says Jeremy Levin, CEO of rare disease-focused Ovid Therapeutics Inc., and formerly CEO of Teva Pharmaceutical Industries Ltd. and SVP, Strategy and Alliances at Bristol-Myers Squibb Co.

Meanwhile, biotech investors are making hay. They are benefiting from the experience of a growing number of ex-big pharma executives who have jumped across to where most of the innovation is happening. Their portfolio companies are benefiting from big pharma’s increasingly well-funded outreach formulae. They are attracting new backers as scientific innovations continue to hit the mainstream press – and as technologies make their way into regulated health care. And  they are also testing new investment models of their own – not only funds devoted to, for instance, later-stage companies that are not typically within a venture capital company’s firing line, but also more flexible, open-ended financing approaches, including alongside big and mid-sized pharma (see sidebar Aggregating Ideas).

The shifts in how big pharma and other investors access and nurture innovations are mirrored by changes in pharma’s role in health care delivery. Large pharma firms are no longer just selling drugs (or even digital pharmaceuticals) to payers and other intermediaries. Some firms are quietly testing the waters as co-providers of care. Witness Roche’s deal with GE Healthcare, providing clinical decision support software for oncologists. Or AstraZeneca PLC's pilot deal in the UK with online doctor and nurse appointment-provider Babylon Health, in part to “see what NHS thinks” about drug manufacturers getting into the game. Or any of the sponsors of diabetes management support systems, such as the virtual clinic launched by theSanofi-Verily Life Sciences LLC  joint venture Onduo, and multiple other (mostly local or regional) schemes.

This potentially much broader role for drug firms opens up further opportunities to innovate and partner – and calls even more loudly for the kinds of open, flexible structures and attitudes that many are trying to embrace. “Those [companies that will] survive will be those most willing to adapt,” says Bayer’s Malik.