Matt Herper in Forbes’ “the Medicine Show” posts a point of view by Bernard Munos on what firms have done best in terms of approved drugs since 1995. The link is here (http://onforb.es/zb6xgu), as are my comments on this provocative topic. The table and implied results can confuse or misdirect people on achieving the best performance in R&D, so I repeat my concerns here:
TJMC: … I just feel that the number of drugs approved is NOT a reasonable measure of R&D performance. I helped on a number of “me too” projects with marginal profile advantages (and value to society or the bottom line) … the results of such strategies speak for themselves. The 21 approvals cited in the article for Novartis is a simplistic extract from the FDA public database, but it includes quite a few of those marginal innovations. I posted similar analyses here, but removing such a “double counts” of “me-too’s”, etc.: http://www.randdreturns.com/innovation-upturn-new-medical-entities-increasing/
To be sure, Novartis still comes out high in the “league tables” with seventeen truly novel “New Medical Entities” from 1995 to 2010. But the pattern then changes dramatically changes – Look at where Pfizer sits – summing all of its acquisitions’ innovations (I know, shocking…) *
Such “novel innovation” is a more reasonable measure of R&D, but the even better measure is the WORTH of those innovations. Yes Lipitor was an incrementally innovative statin, but it hit society’s sweet spot and they were willing to pay a huge premium for that drug. Still, many approvals included in Bernards’ numbers are products with marginal market success, and a number failed to pay back their cost of development.
Measures of apples-to-apples financial ROI for R&D have been done internally in major firms for decades, with increasing sophistication. The problem is that for tax, competitve advantage reasons and the like, cross-industry data in depth is NOT available. Still, we should not create these misimpressions when we know better. The true cost is not $12B, nor is it 150MM. Somewhere in the $1.5-2B range seems plausible, but again you have to check your divisor – true innovative drugs? Or approvals of me-too follow-ons, or salt variations?
Finally, if you net out the cost per drug, then DO THE SAME FOR THE PROFITS THEY CREATE. The Net Present Value (NPV) of Lipitor over ten years of exclusivity is a LOT larger than $12B. Typical drugs that will have seven years exclusivity but have a more modest $1B per year peak sales has a NPV of about $2.5B (8% discount.) Seems convenient or coincidental that is about what most think the average R&D cost is. Of course, if you launch it a year sooner by better R&D operations, or launch a biologic with 12 years exclusivity, the values go up to $3.3B and $5B, respectively. (So who wants to do synthetic patented R&D?) BTW – Lipitor was on sale in 1997. Losing exclusivity in 2011, it still is projected to sell $2-4B/year for several years forward. A staggering NPV of when launched of over $50B.* Even using Bernards’ numbers, one should add together Pfizer + Pharmacia + Wyeth numbers. After all, GSK and Novartis benefitted from the R&D innovations of their parents/acquisitions. So the “ranking” is a bit different – that way Pfizer would have 36 approvals. I am NOT saying that alone is a good thing in itself. Apart from Lipitor, …