Drug Truths

A site devoted to teaching about drug discovery and development.

Pharma R&D Productivity: Have They Suddenly Gotten Smarter?

with 3 comments

For the past few years, critics of the pharmaceutical industry have been very negative about the productivity of these companies.  The declining number of FDA approvals has only supported this view.  From 1990 – 1999, the FDA approved an average of 31 drugs per year.  In the next ten years, this number dropped to 24.  The outlook didn’t get any better in 2010, which saw only 21 FDA approvals.  Thus, yesterday’s front page story in The Wall Street Journal is stunning: “Drug Makers Refill Parched Pipelines.”

Huh? Have authors Rockoff and Winslow been sampling hallucinogens?

Not quite.  The data they present in the article is certainly encouraging.  The story has its roots in last week’s testimony by FDA Drug Division Director Dr. Janet Woodcock, who told Congress that the FDA has already approved 20 innovative medicines this year that “work differently or better than existing drugs or tackle ailments lacking good treatments.”  She went on to say: “We’re seeing a lot of innovation, much more than in recent memory.”  Dr. Woodcock is talking about the same pharmaceutical industry R&D engine that impatient critics have deemed to be broken.  According to a graph in The Wall Street Journal article, Rockoff and Winslow predict that drug approvals in the 2010 – 2019 period will exceed anything that the industry has ever produced.  If this is true, what a blessing this news would be for patients around the world.

Unfortunately, The Wall Street Journal authors are off in their assessment of the causes behind this productivity jump.   To quote:  “Today’s new drug output appears to mark the beginnings of a payoff from a research reorientation the industry began undertaking several years ago.”  Actually, the productivity surge, if it follows the projected path, is the result of research done in biotech and pharmaceutical laboratories in the 1990s.

To support their argument, Rockoff and Winslow cite 12 compounds in a table labeled, “Novel drugs recently approved and in the pipeline.”  The very first compound in the table, Benlysta (for treating Lupus), was discovered by Human Genome Sciences (HGS) and jointly developed with GlaxoSmithKline (GSK).  HGS started the discovery program that led to Benlysta in 1996, started clinical trials with this drug in 2001, and finally got approval this year.  The discovery program that uncovered tofactinib, Pfizer’s breakthrough drug for rheumatoid arthritis, started in 1994, and the New Drug Application (NDA) for this drug will be filed this year.  Assuming this is approved in 2012, this R&D program will have taken 18 years! While I cannot attest to each of the compounds enumerated by The Wall Street Journal, my guess is that, based on the length of time it takes to get a drug approved, most or all of them had their roots in programs that commenced in the late 1990s, not, as the authors suggest, in “research reorientation the industry began undertaking several years ago.”

If this is the case, what led to the surge that we are now seeing?  Two seismic changes occurred roughly 10 years ago that greatly impacted the industry’s productivity.  The first involved answering the question: “What value does this medicine bring over existing therapies?”  This answer was being demanded not just by payers but also by regulatory agencies.  Before 2000, these studies were generally done after a drug was approved – so-called Phase 4 studies.  However, in order to achieve a reasonable price for a new medicine, these studies needed to be included in the initial filing for approval.  The importance and costs of such studies cannot be underestimated.  In many cases, studies that measure the performance of a new drug in a real life situation are necessary.  For example, it was no longer important to show that a new medicine just lowered bad cholesterol.  It had to be proven that it could also reduce heart attacks and strokes.  These studies alone add 3 – 5 years to the clinical development program and literally hundreds of millions of dollars in costs.  Since many clinical programs were already underway in the mid-2000s, they had to be adjusted and so the programs took longer than originally planned.

In addition, by the mid-2000s, far more safety data was being required by the FDA than ever before.   For example, the older NSAID pain relievers like naproxen and ibuprofen had little long-term patient exposure at the time their respective NDAs were approved.  To put it into perspective, all that was required in the past was a study showing safety in patients exposed to a drug for 90 days.  Now, the FDA won’t approve any pain reliever without patients being treated for at least a year (and more likely three years) with a study that also measures impact on overall health outcomes.

Are these important changes?  Absolutely, as any drug that can get through these hurdles will, as Dr. Woodcock said, “work better than existing treatments.” However, companies had to adjust to these changes in the last decade, and this drove up costs, caused development programs to take longer and also resulted in more late-stage failures as compounds that were safe and effective might not have been as effective as existing, cheaper treatments and thus not commercially viable.

Given all of this information, I hope that the industry has adapted, development programs and timelines have been adjusted and we can now expect a steady stream of new medicines.  Yet, three issues temper my hope.  First, the industry has seen such an increase in productivity before.  This happened in the mid-90s when Congress enacted the Prescription Drug User Fee Act (PDUFA).  The FDA was grossly understaffed in the 1990s and, as a result, many compounds languished awaiting approval while they were being approved in Europe, oftentimes years before they would be available in the US.  Congress was outraged by this.  When the FDA showed data that indicated how understaffed it was, Congress’s solution was the PDUFA, which essentially charged a company a fee when it filed an NDA, then used the revenues generated to hire more FDA reviewers.  This action resulted in the FDA being able to review dossiers more rapidly which led to the removal of the logjam and more medicines being approved than any time before or since.  Has the adaption of the industry to the new NDA expectations led to a similar increase of compounds being approved? We won’t know for five years.

Second, the consolidation of the industry in the last 15 years has raised havoc with R&D organizations.  It is my firm belief that mergers are particularly difficult for R&D as starting/stopping research programs takes time.  I once heard a Nobel Prize winner in Medicine say that it takes at least three years for a professor, who leaves one university for another, to get his laboratory back running at full speed.  It is a lot quicker in the industry, but it still takes time.

Finally, the cuts that have been made across the board in R&D in many companies will have an impact going forward.  This effect will not be seen in the short term.  As was detailed above, discovery-development programs for successful new medicines take over a decade.  However, the turmoil in R&D of the recent past (mergers, reorganizations, site-closures, new business models, etc.) will be felt in the next decade.

I am really hoping that Rockoff and Winslow are correct and we are about to see the approval of hundreds of new medicines in this decade.  But the question remains: is this a five-year blip or a sustainable trend?


Written by johnlamattina

July 12, 2011 at 1:54 pm

3 Responses

Subscribe to comments with RSS.

  1. John –

    Excellent points, especially regarding the requirement for value being proved within the NDA, and not Phase IV.
    Does this analysis I did recently ( http://www.randdreturns.com/innovation-upturn-new-medical-entities-increasing/ ) suggest any other insights? I also believe that firms have been working to improve R&D in many ways for over a decade and wonder if these efforts also contribute to the shift commented by Dr. Woodcock, and the shift my data seem to suggest. As you say, we have seen artifacts in the past.


    July 13, 2011 at 5:31 pm

    • Terry,
      First of all, you article certainly caused people like Herper to take note that NDA approvals were likely to start increasing. Companies have been making this point for awhile, but articles like this and Woodcock’s testimony are cause analysts to reassess things. So thanks for putting it forward. As you state, this is a very complex topic. The fact is that current NDA approvals are based on work started 10 – 15 years ago. So one may correctly say that the decline of the past decade was due to a lack of discovery productivity in the 1990 – 1996 timeframe. To be honest, a lot of companies were making changes then in terms of making use of new technologies, trying to improve candidate output into the clinic, and asking hard questions about the commercial viability of second and third generation products. We are seeing the fruits of those changes now.


      July 13, 2011 at 6:08 pm

  2. […] last week.  The first came from Jonathan Rockoff and Ron Winslow in The Wall Street Journal, an article highlighted in this blog last week.  Although Rockoff and Winslow focused on the increase in FDA approvals likely to occur over the […]

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: