Innovation as Measured by the New Molecular Entity – it’s the First but Not Necessarily the Best

The term New Molecular Entity (NME) owes its derivation to the Hatch-Waxman Act of 1984 which distinguished first in kind FDA approval applications from secondary applications for improved formulations or other indications.  New Molecular Entity has taken on added meaning as a measure of innovation:  the more New Molecular Entities approved by the FDA in a given year, the more innovation in the Industry.  Unfortunately, the plot of NMEs per year shows anything but growth, especially when plotted along with the R&D spend per year, Figure 1. (Ref 1, 2, 3)

Figure 1, NMEs 1, 2 per year versus total R&D spend in Billions of Dollars (US$ 2008) (prior to 1984 first NCE Approvals). 3

We explore the implications of this figure in the following section, NME Output versus R&D Expense – Perhaps there is an explanation. In this section we discuss the implications of NME as a measure of innovation.

New Molecular Entity is reserved for small molecule drugs.  While the term could have been meant to apply to protein-based drugs, the first Biologic License Application (BLA) is called a “new BLA”. 4  McCormick suggests New Medical Entity as an all encompassing term. 5  Figure 2 shows the new BLAs do help pull the numbers up. Ref 6  Because the biotechs showed that high prices could be charged for protein-based drugs, Big pharma have taken an intense interest in protein-based drugs. (Ref 4)

Figure 2. NMEs plus new BLAs per Year. 6

Note that when NMEs per year are plotted from 1996, as in Figure 2, it would appear that productivity is decreasing dramatically, but when plotted since the 70s, as in Figure 1, it would appear that productivity spiked in the late 90s but has returned to an earlier level of productivity in recent years.   The NMEs in the 90s were, in fact, an anomaly related to another act of Congress, PDUFA, as discussed in NME Output versus R&D Expense.

While NME is perhaps the simplest measure of innovation or productivity, since it removes the subsequent filings and approvals on a drug, NME rarely defines the most valuable version (indication and/or formulation) of an asset.  Often the formulation and indication that results in the first filing and approval may not represent the most desirable formulation or indication for a drug.  In fact it may not be intended to be the best.  Better formulations and more lucrative indications may be further back in the pipeline.  Nonetheless NME is the measure of innovation or productivity most employed by industry analysts, as evidenced by this partial list of references. (Ref 6, 7)

Why Is It that the Best Formulation and Most Lucrative Indication may not be Embodied in the NME?

One may ask how it is that better formulations and more lucrative indications may be found in subsequent Post-NME Approvals, i.e. further back in the pipeline.  Why aren’t the best formulation and most lucrative indication contained in the first filing with the FDA?   The answer lies in the projectization of work in Development.  And the issue starts back in Discovery.

The Dilemma Originates In the Lead Optimization Stage of Discovery

Back in Discovery, when a project team in Lead Optimization puts forward a clinical candidate for admission into the Development pipeline, there is some opinion about the best formulation and best indication and there may be data to support such opinions.  But the focus at candidate selection is the molecule – is it the best one to propose from a small set of other clinical candidates?  The clinical candidate will have been studied in one or more animal disease models, and will have shown sufficient efficacy in at least one disease model to suggest that it may show similar efficacy in humans suffering from the disease.   There will also be some preliminary studies by formulation experts who will develop their opinion as to whether they see any hurdles towards creating a formulation that will be useful in the disease markets potentially addressable by the candidate.  But the true formulation work comes later.

The Optimal, Most Lucrative Indication.  At the point of Candidate selection, the Optimal Indication for the candidate is a hypothesis. There may be other hypotheses about other diseases in which the candidate may show benefit, which would bring far greater sales to the company than the diseases implicated in the one to few disease models studied in Lead Optimization, but there are three scenarios by which such data may not be available at candidate selection.

A. The candidate didn’t show efficacy in the animal model of the more lucrative disease.  It may be subsequently found that the lack of efficacy was due to an issue with orthologue selectivity, improper formulation or dosing, or issues with the animal model.  Often these issues get resolved with further investigation.

B. The animal model may not exist for the disease.  For example, animal models have successfully led to the discovery of rheumatoid arthritis drugs, but not osteoarthritis. 8

C. The physiological chain of evidence linking the mechanism of action of the clinical candidate to the more lucrative disease may not have been established at the time of candidate selection.  Years of genetics, biology, and biochemistry (now days called Translational Research) go into the chain of evidence that a drug with a particular mechanism of action might influence a particular disease.  It is not uncommon for a company to develop a drug for a particular disease only to find subsequently that a drug with the same mechanism of action will work in a more lucrative disease, an activity now called Repurposing. 9   The now classic example is that of the PDE-V inhibitors which were initially developed to work in hypertension, but were subsequently observed in patients to work in erectile dysfunction.   Understandably, each disease indication requires a different set of clinical trials.  Each indication is a different project with a different set of stage completion dates.

The Optimal Formulation.  For safety studies in Preclinical Evaluation and in Phase I, a buffered aqueous solution of the drug is sufficient.  If there is a market for the drug as an intravenous solution further optimization may not be necessary.  If there is also a market for an oral formulation, then studies around oral formulations may begin in Lead Optimization, but are likely to continue in Preclinical Evaluation, and later until the optimal formulation is discovered.  Issues with early formulations may appear requiring further work on the formulation well into clinical studies.    Formulation optimization is an iterative process that is often fueled by unforeseen issues with current formulations.

Different formulations often revolve around different salt forms of the drug.   Each salt form is considered a different drug and must go through the same set of preclinical studies and Phase I dose ranging studies as earlier formulations.  Consequently, each formulation is a different project with a different set of stage completion dates.

The Real Benefit May Come After the NME with Post-NMEs as Incremental Improvements

So for the reasons just discussed, the first project to be approved for preclinical evaluation and first in human dosing is NOT likely to reach approval as the ultimate indication or the ultimate formulation.   These will come later via subsequent projects that will necessarily follow those of the first project that gives rise to the NME.  A subsequent approval has been called “Non-NME”.  We prefer the term Post-NME.

For tactical reasons it may be best to advance the indication and formulation that is most likely to succeed first and leverage the confidence gained in the first approval to build a case for bringing along subsequent indications and formulations that may be more challenging or more difficult with respect to clinical trials and with respect to regulatory approval.

Thus, it should come as no surprise that Post-NMEs comprise between 40-50% of the late stage (Phase III+) portfolios of the top 12 biopharmaceutical companies in January 2010. 10

The Congressional Budget Office estimates that “on average, only about one-third of new-drug applications submitted to the FDA are for new molecular entities.  Most of the rest are either for reformulations or incremental modifications of existing drugs or for new “on-label” uses (additional health conditions for which an existing drug can be prescribed)”. 11

Based on the reasons just provided, this plethora of Post-NME approvals is not necessarily a bad thing.  They may better more lucrative formulations and indications than their NME counterparts.

This author struggled to find any report comparing NME approvals to Post-NME approvals, so Figure 1 and 2 were constructed from the NME approvals in a Tufts CSDD article (Ref 3) and the Post-NME approvals in the CBO study (Ref 11).

Figure 3.   NME Approvals per Year (Ref 2) versus Post-NME Approvals (“Non-NMEs”) per Year compared to total Approvals.

Figure 3 is interesting in that we can see in the thirty years of approvals shown in the figure, Post-NMEs (“Non-NMEs”) ran higher in the ten year period of 1975-1984 than in subsequent 10 year periods:  1975-1984: 79.5%, 1985-1994: 68.7%; 1995-2004: 68%.  Figure 3 also shows that Post-NME (“Non-NME”) approvals were just as high as NME approvals around 1996.

Figure 4.  NME Approvals per Year (Ref 3) versus Post-NME Approvals (“Non-NMEs”) per Year as a Percent of Total Approvals. (Ref 2,Ref 3), 12

Looking at averages over five year periods we see again that the 90s had fewer Post-NMEs (“Non-NMEs”) but the last five year period 2000-2004 the 5-year average, 72%, was essentially the same as in 1985-1989, 73.1% Table 1.














Post-NMEs versus NMEs








Table 1, Five-year averages of NME Approvals per Year (Ref 3) versus Post-NME Approvals per Year as a Percent of Total Approvals.

Without understanding the value contributed to the assets by the Post-NME approvals one cannot say whether this is a good or a bad thing.  Berndt et al stated that “neglecting to account for the contribution of incremental or follow-on innovation in the form of approvals obtained for new indications and formulations of drugs that are already in use probably results in a substantial underestimate of innovative output in biopharmaceuticals.” 13

If it were really critical to a biopharmaceutical company to provide the most lucrative indication and best formulation in the first application to the FDA, the company would wait until that information was known.  Of course, if the company can make some money from the first approvable application it will file that application at that point rather than wait.

Challenges to the Post-NME Strategy

There are challenges to the Post-NME strategy, such as the potential for third party payor unwillingness to pay for what may be seen as only incremental improvements.  Supplemental applications must be supported by additional clinical trials, the duration of which may push the asset beyond its period of exclusivity.   Dunn has discussed the relationship between the date of patent filing and the date of market authorization, which shows that if market authorization occurs twelve years or more after patent filing the potential for achieving the maximal 14 year period of exclusivity is rarely achievable. 14


In the current hypercritical climate, the large number of second to Nth NDAs is taken as proof that the Industry has lost its ability to find new drugs, blockbusters in particular, and is only squeezing water from a stone to get more profit from old drugs.  But as this discussion shows, such comments ignore or overlook a more complex reality that has to do with the way the Industry has always discovered and developed drugs and will continue to need to do so.

  1. F J Cohen “Macrotrends in pharmaceutical innovation” Nature Rev. Drug Disc. 2005, 4, p78-84
  2. K J Kaitin “Deconstructing the Drug Development Process”, Clin Pharm Ther 2010 87 p356-361
  3. PHRMA “Pharmaceutical Industry Profile 2009”,
  4. See section Small Molecule, Petides and Protein-Based Drugs – Differences and Similarities
  6. Asher Mullard, “2010 FDA Drug Approvals”, Nature Reviews Drug Discovery, 2011, 10, p82-85.
  7. See references in NME Output versus R&D Expense – Perhaps there is an explanation
  8. “Animal Models of Arthritis: Relevance to Human Disease”, Alison Bendele, Jennifer Mccomb, TY Gould, Tracy Mcabee, Gina Sennello, Elizabeth Chlipala and Michael Guy, Toxicol. Pathol., 1999 27, p134-142.
  9. S. Sleigh, C. Barton , “Repurposing Strategies for Therapeutics”, Pharm. Med., 2010, 24, p. 151-159
  10. Deloitte Thompson Reuters 2009 “Is R&D Earning its Investment?”…/UK_LS_RD_ROI.pdf
  11. See P. 14 in “A CBO Study – Research and Development in the Pharmaceutical Industry”, Congressional Budget Office, October 2006, Pub. No. 2589, p. 1-55.
  12. The FDA provides raw data in the form of lists of approvals at
  13. See P. 81, E. Berndt, I. Cockburn, and K Grépin, “The Impact of Incremental Innovation in Biopharmaceuticals. Drug Utilisation in Original and Supplemental Indications”, PharmacoEconomics, 2006, vol. 24, no. Suppl. 2, pp. 69-86
  14. M. Dunn, “Timing of patent filing and market exclusivity”, Nat. Rev. Drug Disc. 2011, 10, p.487-488.