Stuart R. Gallant, MD, PhD

Within pharmaceutical development, there are two types of go-no-go decisions:  1) clinical and 2) financial.  These decisions are linked together by the drug development plan.  This post touches on both types of decisions but focusses particularly on the financial.

Clinical Decision Making

The drug development timeline is marked with milestones in the form of clinical decisions.  Some key clinical decision points are listed below:

Event	Decision	Fundable Event
Pre-Clinical Development	Is there sufficient data to proceed with pre-clinical development?	Filing provisional patent.  Publication of laboratory data.
Filing IND	Is there sufficient data to safely administer the drug product to humans?	Completion of non-clinical studies, including GLP toxicology studies.
Clinical Proof of Concept	Have the mechanism, toxicity, and dose of the drug product been demonstrated?	Completion of Phase 2.  Investors will be looking for early signs of efficacy.
Pivotal Clinical Study	What resources are required to demonstrate efficacy (number of patients, end points, length of treatment)?	Successful End-of-Phase 2/Pre-Phase 3 meeting.
Marketing Application	Has efficacy been demonstrated?	Completion of Phase3.

In each case, the clinical decision is based around the presence or absence of data to support advancement of the clinical program.  For instance, prior to filing the IND, the team is asking whether preclinical development has provided “sufficient evidence to safely administer the drug product to humans.”  In the absence of this data, the program does not proceed to file an IND.

Clinical decisions and clinical development are strongly coupled to the fund-raising program.  Each clinical milestone becomes an opportunity to raise money, a “fundable event.”  For instance, completion of non-clinical studies, including GLP toxicology studies is an event that can be presented to potential investors as documentation that the program is moving forward in a timely manner.

The precise method that fundable events are used is part of the art of fundraising.  For instance, the end of the clinical Phase 2 study may be presented in two different ways:  1) “we are about to complete Phase 2” or 2) “we just completed Phase 2.”  Each approach has its own advantages.  The first approach (fundraising prior to the event) has the effect of de-risking the event itself.  Since no clinical data is available at the time of solicitation, a subsequent failure of the clinical study has no effect on money which has already been raised.  The second approach (fundraising after the event), provided that the clinical data is positive, will undoubtably raise more money.

In general, managing risk by raising funds prior to a fundable event is a prudent strategy, particularly for one-product companies.  Having a little money in the bank at the time when bad clinical data is returned can be the difference between being able to pivot to a new indication or new delivery strategy and being forced to close shop and liquidate assets.

Financial Decision Making

Financial decision making starts with a model of value which includes:

• Clinical risk
• Time value of money
• Cost and timing of development program
• Market value of treatment

The easiest way to incorporate all this information is using Monte Carlo simulation [1].  To discuss the topic of financial decision making, I am going to use a simplified example which can be represented in Excel.  In spite of these simplifications, important lessons can be learned from this example.  The basic assumptions of the model are tabulated below:

The likelihood of clinical success was taken from Wong [2] and Hingorani [3].  For cost per phase, the 2003 estimates from DiMasi were inflated to 2022 values [4].  Baedecker [5] provided the market value, and Wang [6] provided the market exclusivity period.  The time per stage was taken from Martin [7], and I used my own experience for pre-clinical costs and timing.  Cost of goods was taken from Ledley [8].  And, the inflation rate and discount rate were based on long-term economic experience, though I recognize that recently these indices have increased.

Of course, the use of average numbers requires a disclaimer.  An appropriate model to support an individual clinical program requires careful parameterization that accounts for indication, trial design, and myriad other factors.  Because average parameters are being used, the conclusions from calculations in this post represent average results which would not be appropriate for use in any specific drug development program.  They are, however, useful for thinking about drug development in general.

A copy of the mode is provided below:

Let’s consider what we can learn from this model:

• I began this simulation with a set of 25 imaginary companies.  Using the “Likelihood to Advance” values, I calculated how many companies would survive through each stage of clinical development.  25 begin in pre-clinical development.  11 survive into Phase 1; 4 survive into Phase 2, 2 survive into Phase 3, and 1 reaches licensure.  This represents a graphical demonstration of the law of the jungle on display in clinical pharmaceutical development.
• The companies are color coded based on where they fail in the development process.  Companies 1-14 (shaded in beige) fail in preclinical development.  Companies 15-21 (shaded in green) fail in Phase 1.  Companies 22-23 (shaded in blue) fail in Phase 2.  Company 24 (shaded in grey) fails in Phase 3.  Only Company 25 survives to licensure.  The inflation adjusted costs (or profits) for each year appear in columns to the right of each company’s number:

• It is assumed that the costs are the same for each company in a given year.  So, the present value of the yearly costs only needs to be calculated for Company 25, and those same numbers can be used for the other companies:

• There’s a lot going on here, so let’s take things one step at a time to learn as much as possible.
  • Consider Row 64 (Year 11), this is a pivotal year for this group of companies.  At the beginning of this year 2 companies remain (24 and 25).  During the course of Year 11, Company 24 receives news that it failed in Phase 3, and Company 25 receives news that it succeeded in Phase 3.  So, in Year 12, Company 25 is the only surviving company from the original group, and it begins to receive revenue which grows over the next 2 years as sales ramp up and in subsequent years due to inflation.
  • Starting at the left of Row 64 (Year 11), appears the sunk cost for Company 25.  (In this model, sunk cost is always from a single company’s perspective.)  The investors have spent $278.5M dollars (i.e., inflation adjusted year 11 dollars) to reach their accomplishment of a successful Phase 3 trial.  The present value (in inflation adjusted year 11 dollars) for the expected revenue stream from Company 25’s product is $13.6B.  (This number has been adjusted to account for cost of good, so it represents profit alone.)
  • Before we start to draw lessons, let’s look at one more row.  Row 61 (Year 8) is the year that Companies 22 to 25 receive their Phase 2 clinical data.  Two companies (22 and 23) fail and cease to exist.  Two companies (24 and 25) advance to Phase 3.  Starting at the left of Row 61 (Year 8), the sunk costs for any one of the companies is $84M (in inflation adjusted Year 8 dollars).  The present value of the expected revenue stream (in inflation adjusted Year 8 dollars) for company 24 and 25 together is $6B.  Why company 24 and 25 together?  In year 8, we know that companies 22 and 23 ceased to exist, so they don’t contribute going forward, but we don’t know that 24 will fail in Phase 3 and 25 will succeed, so we have to go by averages.  We take the present value of the revenue streams for the 2 companies together and divide by 2 (the number of surviving companies in Year 8).
• Let’s think about what we can learn from this model.  First the investor perspective:  Imagine that we are in Year 8 (Row 61 of the model).  We are the investors in Company 24 or 25 (in Year 8 it doesn’t matter which).  We have invested $84M (in inflation adjusted Year 8 dollars), and we want to get something out of this investment—it’s been 8 years and they want a payoff.  But, that $84M really understates our investment.  To be fair, you have to imagine our investors as participating in all of the companies (1 to 25) through the first 8 years.  (That’s a number around $627M—I didn’t develop a way to handle that number from an accounting point of view, so the fair number might be a little bigger or smaller, but it’s much more than $84M.)  As we know from experience with venture capital, investors are looking for a 10x return.  The reason they need a 10x return is that with such a high failure rate, they’ll be out of business in short order if their successes aren’t home runs.
• Now let’s consider the buyer’s perspective in Year 8 (Row 61).  A large pharma company is looking for additions to its portfolio.  After a successful Phase 2, a drug has about a 50/50 shot of licensure (actually 59% according to our model).  The present value of any given drug is $6B (in Year 11 inflation adjusted dollars).  To purchase this business, this large pharma company will have to put up perhaps $840M (to give the investors their 10x return).  To increase its chances of success, it could buy Company 24 and 25 for at total of $1.6B; the present value of the combination of two drugs would be $12B.

Using the model predictions, it is possible to calculate the fractional increase in value due to pre-clinical and clinical studies (i.e., expected value divided by sunk cost).  If that fraction is plotted on the year each clinical trial data is announced, a graph is obtained like the one below:

A moderate trend downward in expected value per sunk cost is seen.  This indicates that, although the absolute profits are less for early investors (angels) compared to late investors (venture capitalists and pharmaceutical companies), returns can be attractive for early development projects if costs are managed effectively.  This lesson teaches against current trends in investment where large investors are seeking late phase opportunities.  There is still a business in early development for the canny investor.

Portfolio Management

A mature diversified pharmaceutical company (or even a small start-up) will use Monte Carlo techniques (similar to this type of modeling, but far more sophisticated) to manage its development portfolio.  This type of modeling allows:

• The return on investment of a given project to be managed continuously throughout the investment lifecycle.
• Evaluation of which projects should proceed, be sold, or be shelved in an unbiased, net-present-value-oriented way.
• Establishment data-based negotiation positions for purchases and sales of pharmaceutical development projects.
• Linking of marketing and clinical data to evaluate the effect of clinical efficacy data on the project’s economics.

Considering this post in light of an earlier post on Financial Metrics of Pharmaceutical Companies [9], it is possible to see how the economics of mature diversified pharmaceutical companies (MDPCs) arrive similar levels of peer S&P 500 companies (as was shown in that post).  You will recall that price to earnings ratio of the MDPCs fell in the 2^nd Quartile, and the PEG ratio fell in the 3^rd Quartile—indicating investment performance more toward the middle than the top of the S&P 500.

What MDPCs are doing as an economic strategy is continuously adding pharmaceutical projects to diversify their portfolio and manage risk.  Clearly, they would like every project to be a block buster, but that is unrealistic for numerous reasons (for example, imperfect scientific knowledge and competition for the most attractive markets).  So, they add projects that are on the margin (small market or limited efficacy) to provide benefits to the company related to risk management and strategy (entry into new markets or protection of existing market).

Coincidentally, this approach of picking up every development project that can make a case for itself economically maximizes the number of drugs in development, supporting the public health goal of having a rich diverse group of pharmaceutical treatments for diseases that affect the public.

However, there is a weakness of this approach which is created, not by the economics, but by the scientific demands of the process.  With our current level of scientific knowledge, it is necessary to use expensive and time-consuming animal and human clinical trial testing.  This testing of both the successful drugs (Company 25) and the unsuccessful drugs (Company 1 to 24) drives the cost of development of a new drug to the billion dollar range.  Until we are able to more easily and cheaply distinguish the effective from the ineffective drugs, goals like treating every orphan disease will remain distant.

[1] Chang, M.  Monte Carlo Simulation for the Pharmaceutical Industry, CRC Press (2011).

[2] Wong, et al. “Estimation of clinical trial success rates and related parameters,” Biostatistics (2019) 20, 2, pp. 273–286.

[3] Hingorani, et al.  “Improving the odds of drug development success through,” Nature Scientific Reports (2019) 9:18911, https://doi.org/10.1038/s41598-019-54849-w.

[4] DiMasi, et al.  “The price of innovation: new estimates of drug development costs,” Journal of Health Economics 22 (2003) 151–185.

[5] Baedeker, et al.  “Value of 2019 FDA approvals,” doi.org/10.1038/d41573-020-00002-6.

[6] Wang, et al.  “Variations in Time of Market Exclusivity Among Top-Selling Prescription Drugs in the United States,” JAMA Internal Medicine April 2015 Volume 175, Number 4 635-637.

[7] Martin, L.  “Clinical trial cycle times continue to increase despite industry efforts,” Nature Reviews:  Drug Discovery 16, March 2017, 157.

[8] Ledley, et al.  “Profitability of Large Pharmaceutical Companies Compared With Other Large Public Companies,” JAMA March 3, 2020 Volume 323, Number 9, 834-843.

[9] pharmatopo.com/index.php/2022/03/06/financial-metrics-of-pharmaceutical-companies-2-of-2/

Disclaimer:  PharmaTopo^TM provides commentary on topics related to drugs.  The content on this website does not constitute technical, medical, legal, or financial advice.  Consult an appropriately skilled professional, such as an engineer, doctor, lawyer, or investment counselor, prior to undertaking any action related to the topics discussed on PharmaTopo.com.