Assigning Credit and Blame to Biopharma R&D Chiefs

5/29/13

There’s been an awful lot of turnover lately among Big Biopharma R&D chiefs, with departures from Merck, Amgen, Bristol-Myers Squibb, AstraZeneca, Roche, and AbbVie. Their average tenure in this position was less than 7 years; the range was from less than 2 to 12. That’s similar to the average CEO tenure, which is only 8.4 years across all industries and is even shorter in the pharmaceutical sector. Compare these tenure numbers to the lengthy time frame (widely reported as 10 to 15 years) it takes to move a new drug from the lab to the marketplace. These numbers imply that most R&D chiefs will not be in their jobs long enough to see drugs discovered on their watch ever win FDA approval. This raises an interesting series of questions: who gets the credit if one of their current portfolio drugs makes it to market? Who will be blamed if it doesn’t? Is it even possible to figure this out?

Many of the folks hired to replace these departing R&D executives will be joining organizations that already have an existing pipeline of potential drugs at various stages of development. Once they come on board they will move quickly to differentiate themselves from their predecessors, who may or may not have left under good terms. The keys to their success, given what may turn out to be a brief tenure, will not be found at the early end of the pipeline. In fact, their achievements will be tied most closely to their acquisitions, as detailed below. This skewed focus towards near term results even in R&D may help explain why Big Biopharma has developed the reputation of buying, but not creating, innovative medicines.

The task of tracking down the details of who did what (and when) to properly assign credit and/or blame in R&D could give morphine a headache. It also likely requires insider knowledge to parse the fine details. Let me illustrate this by focusing on just a couple of these recent, high profile personnel moves. Peter Kim, the former head of Merck’s R&D group, saw 20 products launched during his 12 years in that position. Mercks’s website lists some 98 marketed prescription medicines, which is not as impressive a number as it appears. For example, the list includes three distinct versions each of four different drugs. You also can’t tell by perusing it that generic versions of some of their marketed medicines are widely sold by competitors. What really matters (at least to the Wall Street crowd) are the top financial drivers in the group. Put another way, it’s not the number of drugs that you sell; it’s how much money they collectively bring in. One blockbuster is generally worth way more than a handful of second-tier medicines.

Kim’s predecessor at Merck, Ed Scolnick, likely chose a substantial number of the medicines that won FDA approval in Kim’s first few years after taking the top R&D job. Longtime Merck watchers (which I’m not) may have some feel for the relative contributions of Kim and Scolnick, but it’s likely that both their fingerprints will be found on a number of Merck molecules. Who deserves the lion’s share of the credit for the recent top earners in this portfolio (e.g. montelukast [Singulair; FDA approved in 2005] and sitagliptin [Januvia; FDA approved in 2006]), and who made the key decisions to develop and continue to market the problem children like extended-release niacin/laropiprant [Tredaptive] and rofecoxib [Vioxx; FDA approved in1999]?

Merck doesn’t bother to include any drugs in Phase I on their most recent pipeline list, but they have 23 drugs in Phase II and 15 drugs in Phase III. They also list six drugs as being up for review by the FDA. Oversight for this extensive collection of drugs in development has now been shifted to Kim’s replacement at Merck, Roger Perlmutter. He returns to Merck (where he had an earlier stint) having spent the previous 11 years as the head of Amgen’s R&D group. We know, based on published numbers, that only about half (on average) of the 23 drugs in Phase II are likely to move forward, along with only about two-thirds of the 15 drugs in Phase III. Trying to significantly improve these percentages is a challenging assignment. It requires calculating which molecules have the least attractive chances for clinical and/or marketplace success, moving them to the back burner, and possibly purchasing some near-term replacements.

Amgen and Merck have distinctly different biopharmaceutical business models. Amgen’s marketed drugs are primarily protein based, whereas Merck’s historical focus has been on small molecules made through chemical synthesis. As a result, Amgen’s portfolio of drugs has been much less susceptible than Merck’s to generic competition. Substantially changing the makeup of a pharma pipeline is akin to altering the course of an oil tanker navigating dangerous shoals; it can be done, but it takes a long time and a wrong decision may be disastrous. Perlmutter’s other major task during his tenure will be stocking the pipeline for the next person to hold the job. It’s been suggested that one of the major reasons for Merck selecting Perlmutter was to shift the balance in their R&D efforts more towards protein based drugs. The rationale for this is clear: recombinant protein drugs accounted for 4 of the top 5 best selling drugs in 2012, and the final rules for gaining FDA approval of generic versions of these drugs (called biosimilars) in the U.S. have still not been released.

What legacy did Perlmutter leave behind at Amgen? According to Forbes he oversaw the launch of 8 drugs during his 11 years there. A separate bio of Perlmutter, however, indicates that he was “responsible for the registration of 10 significant new drugs”. Parsing these numbers can be tricky, and it’s unclear to me exactly how these calculations were done. Amgen’s website lists only 9 currently marketed products. So if 8 or 10 drugs were added to the Amgen lineup during Perlmutter’s reign, why are there only 9 being sold now? Let’s review the current Amgen drug portfolio in detail to figure this out:

—Amgen was selling three of these drugs—epoetin alfa (Epogen), filgrastim (Neupogen), and pegfilgrastim (Neulasta)—which all achieved blockbuster status prior to Perlmutter joining the company.

—Another blockbuster, darbepoetin alfa (Aranesp), won FDA approval in September 2001, eight months after Perlmutter joined Amgen.

—Another four of Amgen’s marketed products were fully or partially obtained via the acquisition of Immunex in 2002. Etanercept (Enbrel) (yet another blockbuster) has been on the market since November 1998. Seattle-based Immunex and Fremont, CA-based Abgenix had been jointly developing panitumumab (Vectibix) since July of 2000; it won FDA approval in September 2006. Both Amgen and Immunex had active bone biology research programs since the mid 90s. Their combined efforts led to the creation of Amgen’s drug denosumab, which is sold for two different clinical conditions (as Prolia, which won FDA approval in June 2010 for postmenopausal osteoporosis, and as Xgeva, which came on the market in November 2010 for the prevention of bone fractures in cancer patients).

—Amgen acquired the rights in 1996 from NPS Pharmaceuticals to cinacalcet (Sensipar), a drug for secondary hyperparathyroidism. It won FDA approval in March 2004.

—Romiplostim (Nplate) won FDA approved in August 2008 for a rare platelet disorder. Amgen had originally acquired the rights to thrombopoietin in 1995 from Novo Nordisk (whose Zymogenetics group won the race to clone this molecule). Nplate was developed as an alternative when difficulties arose in the clinical development of thrombopoietin itself.

By my count, Amgen is currently selling five or six drugs (depending on how you count denosumab) that were approved on Perlmutter’s watch. Work on most of them either began before his arrival or arrived via acquisitions. So what happened to the other FDA approved molecules that are no longer being sold? Amgen unloaded the poor selling drugs ancestrim (Stemgen, which I don’t think ever sought FDA approval in the U.S. but was sold in Canada, Australia, and New Zealand), palifermin (Kepivance; FDA approved in December 2004), and anakinra (Kineret; approved in November 2001) to Biovitrum of Sweden in 2008. Combined sales of the three drugs in 2007 were a less-than-stellar $70 million. Amgen also won approval of interferon alfacon-1 (Infergen) in 1997, but the rights to this were out licensed to InterMune in June 2001.

A look at Amgen’s current pipeline shows 14 Phase III trials (although most of these are clinical extensions of already marketed drugs; only six represent novel agents, two of which are also in Phase II for other indications). Similarly, they have 13 drugs in Phase II trials (representing 7 new agents), and 19 new drugs in Phase I. Sean Harper, Amgen’s new executive vice president of R&D (and former company chief medical officer), expressed confidence that many of these pipeline drugs will be approved. Could you imagine him sharing any other viewpoint with the media? Time will tell as to what pipeline revisions he makes going forward.

Given their short tenures, the best opportunity for R&D chiefs to achieve success appears to reside in acquisitions. It’s the only way to quickly make a mark in an industry burdened with such a lengthy product development cycle. Perlmutter’s Amgen tenure is likely to be remembered for the numerous deals that took place on his watch, including the acquisitions of Immunex (2002), Tularik (2004), Abgenix (2005), Avidia (2006), Ilypsa (2007), Alantos Pharmaceuticals (2007), and BioVex (2011). He may have also have helped lay the groundwork for companies acquired after he announced his planned departure, including Micromet (2012), Kai Pharmaceuticals (2012), and Decode Genetics (2012). Without acquisitions you are mostly finishing off the work of your predecessor, leaving your successor to complete the projects you started. Judging success within this paradigm would be a very difficult exercise indeed.

Achievements in biopharma can be measured in a number of different ways, and we know that corporate accomplishments are greatly influenced by the performance of CEOs and other corporate bigwigs. Xconomy’s Wade Roush nicely summarized Princeton psychologist Daniel Kahneman’s description of outcome bias as a “tendency to reward or blame decision makers for the performance of their organizations, even though the correlation between leadership quality and corporate performance is generally low.” Outcome bias seems to be well in play in assigning credit or blame to drug pipeline results.

The bottom line on who gets most of the credit or blame in biopharma R&D ultimately comes down to the bottom line of the company’s financial statement. Stock analysts don’t really care how many drugs are approved under your watch or how many have been added to the pipeline by the time you depart. Profitability is the name of the game. Investors want to know: Did company revenues increase during your time on the job? Financial returns are the net result of a large number of events, including drug launches (and withdrawals), acquisitions, spinoffs, dividend payments, stock buybacks, lawsuits, settlements, etc. In Peter Kim’s case, Merck’s revenues went from $47.7 billion in 2001 to $47.3 billion in 2012. With Roger Perlmutter, Amgen’s revenues rose from $4 billion in 2001 to $17.3 billion in 2012. Care to hazard a guess as to which of these two guys is considered more successful?

Stewart Lyman is Owner and Manager of Lyman BioPharma Consulting LLC in Seattle. He provides strategic advice to clients on their research programs, collaboration management issues, as well as preclinical data reviews. Follow @

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

  • Kyle Serikawa

    Stewart, nice commentary on the problems of short term incentives versus long term thinking, and the difficulty of assessing leadership quality in an industry that has such a long development cycle and that is subject to so much random variation. I’m reminded of a similar example in Leonard Mlodinow’s “The Drunkard’s Walk,” discussing how studio heads also are often assessed on current successes of movies even when those movies were greenlighted and developed under a previous studio head’s tenure.

    I think an interesting question, given the low correlation between leadership quality (however that is measured) and performance, is how much effort organizations should be putting toward finding, assessing and hiring quality at the top. If, let’s say, the difference between the best CEO and the worst only affects probability of success by 5% due to the host of uncontrollable factors influencing drug development (I’m making that figure up. I’d love to know how to calculate that more accurately), how do companies factor that into their job search?

  • Stewart Lyman

    For those of you who want to track Roger Perlmutter’s transformation of Merck’s R&D group, check out the update at http://online.wsj.com/article/SB10001424127887323734304578545612704171132.html