Discussion: (0 comments)
There are no comments available.
View related content: Health Care
AstraZeneca’s revitalization strategy, announced this week, follows the same well-worn playbook used by so many in the industry, employing approaches vividly familiar from my consulting days: cut headcount, externalize R&D, focus on select therapeutic areas, push biologics, and explore an interesting flyer (in this case, technology from Moderna, a Cambridge, MA-based company developing novel mRNA therapeutics – see Luke Timmerman’s recent Xconomy post).
While not offering profound solutions, these restructuring activities, through cost cutting and distraction, are likely to buy AZ at least a little bit of breathing room from the increasingly critical analysts that have massed at the company’s gates.
I’d like to review what may be driving these changes in the industry, and conclude with several alternative strategies big pharma might consider.
Pharma’s Underlying Challenge
The fundamental problem the industry is wrestling with is this: car companies know how to make a car, soft drink companies know how to make soda, yet drug companies really have no reliable way of knowing where their next products are going to come from, and in a sense, have to start from scratch each time – at least if they want to make radically new, “first-in-class” products that offer unprecedented, dramatically better benefits to patients.
The problem is, these products are incredibly difficult to come by. Disease remains very complicated, and it’s exceptionally hard to devise a new molecule that durably interferes with a pathological process yet leaves the rest of the body alone; the technical risk, as it’s called, is ridiculously high.
Not surprisingly, strategies that involve tweaking existing products, or reformulating them in a new way (e.g. liquid Ritalin, as Bruce Booth has discussed), remain popular because they at least reduce the technical risk, and may offer an incrementally – and often meaningfully – better option for patients (see here). However, an increasingly difficult payor environment is likely to make this approach ever more challenging, materially elevating the commercial risk. Proving an incrementally better product enhances value can be expensive (because it takes many patients to demonstrate a small difference in an active comparator trial), and of course, risky as well.
Not only is R&D uncertain and expensive, but no one seems to have demonstrated a consistent aptitude for this, at least in the context of new mechanisms. Data suggest neither big pharmas nor small biotechs have been especially productive, although (as Gary Pisano has shown) pharmas appear to have worse statistics because they have to eat the cost of failures, while small companies tend to disappear, and thus their costs are often not included in productivity calculations. While it’s always possible to force-rank companies (a favorite technique of consultants), it’s far less certain whether the R&D shops that seem to be performing the best at any given moment are actually good (as the consultants would have it) versus just lucky (as simple statistics might suggest). I agree it’s possible to be especially bad at novel drug discovery, but I think it’s quite difficult to be reliably good (or consistently successful) at discovering and successfully developing drugs that utilize unprecedented mechanisms.
To improve the odds and the economics, many pharmas are now embracing what some are calling a “pick the winners” approach, which (as a consultant explained to me) is better described as “kill the losers.” The basic idea is that the astronomically high cost of drug development is driven by the high failure rate, and especially by the staggering cost of late failures – failures in Phase 2 and especially in Phase 3. Ruthlessly kill bad programs early, the logic goes, and you’ll save money. The problem, of course, is how to do this.
There are two basic impediments to early kills: scientific (figuring out what should be killed) and organizational (actually doing the killing). I’ve discussed the organizational challenges previously – please see here and here; see also this Drug Baron post, and references therein. To address the scientific challenges, some companies have looked to genetics and biomarkers to increase their probability of success or conversely, let them know when they’re failing. Typically, these are motivated by a conspicuous success, a development program that was driven by the strength of beautiful molecular data, and the canonical example these days is PCSK9, a target which seems to feature an almost unprecedented alignment of human genetics, biomarker, and animal model data. Clearly, Amgen (one of two companies leading the PCSK9 pursuit) hopes that through the acquisition of DeCode Genetics, additional high-probability targets will be revealed; other companies have gone down this route without conspicuous success, and it will be interesting to see whether (as my colleague Matt Herper has argued) this time will be different. I hope he’s right, of course, but I’m somewhat less sanguine.
The challenges of R&D have led many large drug companies to ask whether it makes more sense to let the early-stage innovation happen outside of big pharma, and then in-license or acquire the rare development programs that seem adequately promising. Rather than drop early R&D entirely (as the sharp industry analyst Andrew Baum provocatively proposed in 2010), most big pharmas are managing their exodus from basic research by initiating a series of high-profile alliances with academic centers at the same time they reduce R&D headcount. Depending on your degree of cynicism, this approach represents either the embrace of open innovation or covering fire used to obscure a massive intellectual retreat.
Given the sad state of early-stage life science financing (for pretty much the same reasons – almost everything fails, and investors have now figured this out), it’s unclear whether there will always be enough promising companies and products for the big companies to pick off, at least at reasonable valuations. To this end, many pharmas are using in-house venture groups to support early-stage companies, and as Atlas VC Bruce Booth reports, life-science financing these days (especially early-stage) increasingly involves at least one corporate VC. While such funding is unquestionably useful, it’s hardly enough to support a vibrant range of early-stage activities.
Pharmas are also continuing their strategic embrace of biologics, a move driven by two key considerations. First, most biologics target serious conditions, for which a high level of reimbursement is expected. Second, biologics are really difficult to copy, and thus even after the patents expire, knock-off biologics (biosimilars) are far less a threat than knock-off small molecules (generics). This makes sense: biologics are notoriously tricky to manufacture, and both their safety and efficacy depends not only on the specific protein sequence of the molecule, but on the highly specific culture conditions used to manufacture it. With a few exceptions, I would readily prescribe, use, and recommend a generic drug, but I imagine I’d be reluctant to consider a biosimilar in place of a required biologic.
So long as they have enough cash to purchase and complete the development of promising products – especially biologics – most big pharmas can (and likely will) use the strategies discussed above to stay in the game. That said, the megatrends remain concerning; while drug costs represent only a small fraction of the overall healthcare equation, they are likely to remain (like other healthcare costs) under intense pressure, and it will be increasing important to demonstrate that the benefits of innovative new products justify their high costs.
Two prominent threats facing big pharmas are: (1) payors may figure out more effective ways to reign in the cost of specialty drugs, particularly those offering only modest incremental benefit; (2) the early-stage innovation ecosystem may sputter to the point where there aren’t enough products to purchase, and those available command prohibitively high valuations.
I suspect a more likely outcome is that the industry will continue to slowly consolidate and shrink its footprint, and that we’ll see a few more mega-mergers in the years ahead.
Three Alternative Approaches
The far more interesting question is whether there are alternative, creative options pharma companies can consider as they look to an increasingly uncertain future. Three considerations are: analytics; phenotype; and risk.
Analytics: While most pharmas employ talented biostatisticians, the companies’ analytics capabilities tend not to be exceptional; drug companies are more likely to be consumers of analytics services than they are to be developers of powerful new platforms (i.e. they are approaching data like Romney’s campaign rather than Obama’s). I’m intrigued (as I’ve noted before) by the possibility of a pharma company that decides to differentiate by focusing on brilliant analytics, analytics that would be the envy of the industry – and of others. Imagine if you had a team comprised of the sort of people who now work at Google, Facebook, Palantir, and turned them lose on all the different sorts of data pharma companies deal with, from basic science to marketing. It’s hard not to envision this would be radically transformative.
Phenotype: The explosion of sensors and mobile technologies, and of digital health more generally, has dramatically increased our ability to understand a patient’s experience of disease, providing the opportunity for continuous versus episodic assessment, and understand phenotype at a far more granular level. Beyond the obvious immediate challenges – figuring out what to measure, figuring out how to measure as passively and unobtrusively as possible, figuring out how to capture and incorporate the data in the patient’s record – there’s the even more difficult task of making sense of it all, and (asRecon Strategy’s Tory Wolff and I recently discussed), turning this data into insight, and the insight into value. There are at least two reasons for pharma companies to jump into this space: (1) as reimbursement barriers get higher, being able to demonstrate your new drug meaningfully impacts patients will be increasingly important; a far more detailed understanding of the manifestation of disease in the patients you’re trying to help should enable better demonstrations of efficacy and value; (2) longer term, the ability to combine phenotypic and genomic data may reveal different subgroups, that respond distinctly to therapies, or that reflect a common biology that could be understood and perhaps targeted.
Risk*: As the healthcare system increasingly focuses on delivering outcomes, it might make sense for some medical products companies to own part of this risk. This could be done in collaboration with providers: either you could jointly assume risk (for example, splitting risk 50/50) or apportion out different risks to each other depending on who can most impact outcomes. Alternatively, and perhaps more boldly, it might be done directly – for example, by owning the surgical centers, contracting the care of hip repair patients, and making money by managing the entire process well and delivering the best results. Or you could run diabetes clinics and make your money by contracting the care of diabetic patients, incentivizing the delivery of quality outcomes rather the prescription of expensive drugs. While the margins would be lower, the business model might be more durable, and better aligned with the needs of patients. This would also be expected to drive further value-oriented innovation. (*Hat-tip to Tory Wolff for useful discussions.)
Bottom Line: Medicines, in many cases, represent remarkably effective solutions to vitally important problems. Unfortunately, they’re also incredibly difficult to devise, validate, and economically justify. The aggressive application of emerging analytic and digital health technologies, while currently viewed as “nice to have” at best, are likely to emerge as table stakes in a world focused on the more precise identification and delineation of value. Medical product companies that figure out how to embrace and most effectively apply these emerging technologies, and think creatively about new risk-sharing business models, will be best positioned to deliver impactful medicines to patients, durable health to populations, and long-term returns to stockholders.
There are no comments available.
1150 17th Street, N.W. Washington, D.C. 20036
© 2016 American Enterprise Institute for Public Policy Research