Approval Rating

Many readers of this column will have seen the recent analysis, performed by Matthew Herper and Scott DeCarlo at Forbes, where they simply went back over the last 15 years of annual reports from the large drug companies, added up the amounts charged to R&D, and divided that by the number of drugs approved over that span. The idea was to determine the real cost of a new drug, by the most direct measure possible.

That’s not exactly the most complicated financial model in the world, but its simplicity does have some virtues. There’s money spent near the end of the period that hasn’t had time to pay off, but then, there are drugs from the beginning of it that get counted, even though their costs aren’t in there completely. Then there are acquisitions, but since the R&D expenditures behind such drugs aren’t fully accounted for as such by the acquiring company, those might tend to make the figures look better.

Better than what, though? In case you haven’t seen them, the results were . . . well, pretty grim. Amgen got the most drug-for-the-buck, with about $4 billion of R&D per approval, while AstraZeneca got the least, with an eye-watering $12 billion per. I believe we can all agree that no matter how the numbers get emulsified, $12 flippin’ billion (!) per FDA approval is not a sustainable figure. I know that we just set a record recently, with $11 billion paid for a drug in Phase II (Gilead/Pharmasset), but a lot of people seem to be hoping that that’s an aberration. Just consider the size of the various markets, and the years left on most patent expirations: things don’t add up. The question is, at what point does a drug portfolio break even — and how many companies are on track to land on the right side of the line? 

After spending a while staring at the money column of that article, though, my attention wandered over to the number of approved drugs. That got me to thinking: how many of you, I wonder, have worked at a drug company where (at some point) the Almighty Top Management has proclaimed the company’s powerful, newfound ability to bring X new chemical entities to market per year? X, in these cases, is some impressive-sounding integer, meant to dazzle the investors with visions of gold. But when you look at the table of approvals during the last 15 years, X is almost always “less than one.” In no case has X ever reached the vertigo-inducing heights of “two.” Not even close. One-and-a-half is the upper limit, reached by Novartis. (And congratulations to them for it, by the way — they didn’t do it for the least amount of money per drug, but they at least got more to come out the other end of the pipe than the rest of us did).

That’s something to think about, given the number of organizations that have announced, at one point or another, their ability to crank out new drugs and their targets for doing so.

Remember those brave statements? Two approvals per year, three per year . . . all of it about to kick in just a few years down the road, once all those pesky little problems are ironed out. It’s never happened. There are explanations for what went wrong, of course — there are always explanations. But I think all these tales of unexpected toxicity, regulatory setbacks, licensing deals that went awry, and screening campaigns that went nowhere can be summed up very briefly: what happened to all these predictions was reality. But since the reality of drug discovery has never been very pretty, it’s been a lot more appealing to look to something more enjoyable, which is a fantasy of how things could be if we knew more about what we’re doing.

Investors have been more than willing to go along with that, a human tendency that’s been noticed for long enough to be described in Latin (mundus vult decipi — “the world wants to be deceived”). That saying sometimes has the continuation, “So let it be deceived,” which is a pretty long-standing trend, too.

The problem is, investors don’t really want to hear about the true difficulties of this work, not when there’s someone else down the street still willing to sell them on the fantasy. (The economist-minded may recognized another variation of the “tragedy of the commons,” except this time, what’s being over-grazed isn’t the village green, but accurate information). The evidence since the mid-1990s is that one drug approval a year is a very good pace indeed — hardly any company has managed to beat that. But try telling that to Wall Street! Your single drug per year had better be a multi-billion dollar monster . . . but we know that they’re not, not always.

So that brings us to the broader topic of the other things that we tell the Street. One of those, which fits right in with the fantasy of regular drug approvals, is the fantasy of predictable earnings. Now, the only thing that investors like more than predictable earnings are upside surprises, but an R&D-based business full of uncertainty and wasting assets (like ours) has trouble delivering those. Then there are the marketing estimates in general, which are notorious for missing the mark. But that’s a team effort: both the companies and the analysts get together for that one. I treasure an analyst’s report in my files from back when Pfizer was getting Exubera (their inhaled insulin product) onto the market. The luckless author pointed out that the company kept saying that it expected a billion dollars in sales from the product, but this guy wasn’t having any of that. No, he was sure of two billion; he thought that Pfizer was lowballing everyone in order to produce some of those upside earnings surprises. The earnings were a surprise, at that. A few tens of millions of dollars in revenues (nowhere even close to a fraction of the development costs), and Exubera was no more. How is a world where something like this can happen supposed to fit neatly into a predictable earnings pattern?

But what would we tell the analysts if we didn’t have these stories? I realize that the truth is supposed to set you free, but mostly what the truth seems to do is tick people off. Merck chief executive officer Ken Frazier tried to withdraw earnings guidance a while back, saying that the company needed to concentrate on R&D, and look where that got him: earnings guidance and a multi-billion-dollar stock buyback, to boot. It’ll be a while before anyone tries that line again, which is a pity, since it’s accurate.

So it’s a kind of symbiosis. We agree to keep the investors interested with tales of new clinical trials, new results, and new discoveries, and not dwell too much on patent expirations, Phase III failures, FDA holdups, and other depressing subjects. We try to give the impression that we have things under control, even when it’s clear to the most casual observer that we don’t (and that no one could), but people agree to act as if, this time, we really do. My worry is that a relationship founded on telling whoppers to each other is not necessarily fitted out for the long term.
 

Derek B. Lowe has been employed since 1989 in pharmaceutical drug discovery in several therapeutic areas. His blog, In the Pipeline, is an awfully good read. He can be reached at derekb.lowe@gmail.com.