Patients depend on the biotechnology industry for a steady stream of cutting edge medical products to drive an ever-increasing overall quality of care. This paper explores the parameters that influence whether individual prospective medical products reach the market. It then examines how these parameters can be adjusted to expand and accelerate the pipeline of medical products reaching patients in the clinic, resulting in improved health outcomes for society as a whole.

Section 1

Overview

The greatest driver of innovation in healthcare around the world is the biotechnology industry, which discovers and brings novel medical products (drugs, devices, etc.) to market that can be used by medical practitioners and patients to prevent, treat and cure diseases in ways that they were unable to do before.

At the same time, compared to previous times in history, the modern biotechnology industry has greatly stagnated. The cost per new drug brought to market has skyrocketed. And the number of new drugs approved per year has not gone up as is certainly possible. The same is true for medical devices. Simply put, modern biotech is not producing new medicines at a frequency and quality that is as high as what we should expect as patients and consumers.

Given that the biotechnology industry is both the greatest driver of healthcare innovation and could be substantially more impactful to our health and our lives, we can conclude that we have an incredible opportunity and even obligation to accelerate and expand the medical products pipeline so that we may bring more transformative products to the clinic and make patients substantially healthier than ever before.

In this paper, we will first identify the most impactful levers that states can pull to expand and accelerate the pipeline of medical products in their country. Next, we will formalize all of these levers into a unified model of pharmaceutical value. Last, we will consider how regulators of medical products can utilize these levers to help the biotechnology industry bring medical products to patients in a way that is faster and cheaper, with higher quality products and a greater overall volume of medicines available.

Section 2

Value is All You Need

We will begin by introducing a model of biotechnology that explains the conditions under which a medical product is brought to market.

Simply put, in order for a medical product to be brought to market, a sponsor needs to invest in doing so, with the hope that the product will yield positive returns. The sponsor can be an existing company or a new company formed around the opportunity to bring product to market.

We can conclude that for a given opportunity to bring a new product to market, if zero companies identify it as being valuable to pursue, then the product will not come to exist, as no company will be willing to take the risk to shepherd it through clinical trials and to market.

We can also conclude that for a given opportunity to bring a new product to market, if one or more companies identify it as having value, one of these companies may proceed to bring the product to market. It doesn't necessarily follow that a company will, but some company may.

When we say “valuable to pursue”, we mean that the “risk-adjusted net present value”, or rNPV, is positive. Here, “net present” refers to the smaller amount of money one would accept to receive the value today rather than waiting to receive payments over time in the future, and “risk-adjusted” means an adjustment of value that reflects the risk of failure due to clinical trials or regulator disapprovals.

From these considerations we can conclude that a positive risk-adjusted net present value (rNPV) of a medical product is a necessary but not sufficient condition for that product to exist in the future.

Of course, at some point, a product's risk-adjusted net present value can become so large that it is almost inevitable that some company will bring that product to market, as the market opportunity will be so large and tempting that it will be nearly impossible for all players in the market to avoid pursuing it.

Thus, we can expand our conclusion to say that while a positive risk-adjusted net present value (rNPV) of a product is a necessary but not sufficient condition for the product to be brought to market, it can be said that a sufficiently positive risk-adjusted net present value (rNPV) is both a necessary and sufficient condition for the product to exist in the market.

Therefore, we can equate the condition of a medical product existing or not existing to ensuring that the value of bringing the product to market is sufficiently large and sufficiently attractive to find a sponsor.

Now that we have established this relationship, we will focus the future conversation on (a) how we can increase the value or rNPV of medical products and (b) what the levers are that go into increasing medical product value.

Section 3

Modeling Value for Drugs and Devices

The value of a medical product is equal to the value of the total discounted cashflows over the lifetime of the product, from the first year that the product is pursued until the last year of the lifetime of the product. We can model this value by using the generic formula for risk-adjusted net present value, or rNPV:

\mathrm{rNPV} \;=\; \sum_{t=0}^{\infty} \frac{CF_{t}\,p_{t}}{(1 + r)^{t}}

In this formula, t is time in years, \infty is infinity, CF_t is the projected cashflows for the product in year (assuming project success), p_t is the probability of success of the project in year t, r is the weight-adjusted cost of capital, and (1+r)^t becomes the time-based discount applied to the cashflows in any given year. The \mathrm{rNPV}, then, is the summation of the risk-adjusted, time-discounted cashflows across all years modeled (from time t = 0 to \infty).

It's important to note that the cashflows of a product during years of operation (defined as the years up until regulatory approval) can and should be modeled quite differently from the cashflows of a product during years of development (defined as the years after regulatory approval).

To model this accordingly, we can define an equation that expresses that the overall rNPV of a product as the sum of the cashflows in the years of development and the cashflows in the years of years of operations.

\mathrm{rNPV} \;=\; \mathrm{rNPV}_{\mathrm{dev}} \;+\; \mathrm{rNPV}_{\mathrm{op}}

Now that we've broken these periods of cashflows into separate terms, we can acknowledge that each period has its own characteristics and so can have its own modified formula derived for simplicity and applicability.

For the value of a product's cashflows during the years of development (\mathrm{rNPV}_{\mathrm{dev}}), we derive the following formula:

\mathrm{rNPV}_{\mathrm{dev}} = \sum_{s = \mathrm{S}_{\mathrm{Current}}}^{\mathrm{S}_{\mathrm{Launch}}} \frac{(R_s - C_s) \cdot \left( \prod_{j = \mathrm{S}_{\mathrm{Current}}}^{s} p_j \right)} {(1 + r_s)^{T_s}}

With this formula, we make a few changes to the generic rNPV formula. First, rather than summing the cashflows across years of development, we sum it across development stages (preclinical, phase 1, phase 2, phase 3, and application). Second, we redefine the cashflows of the project as the revenues minus the costs (note that in the vast majority of cases the revenues attributed to a product during development will be zero). Third, we define the probability of reaching a given stage as the cumulative probability of each stage up until that point, calculated as the product of the probabilities of each of these stages \prod_{j = \mathrm{S}_{\mathrm{Current}}}^{s} p_j. And last, we establish that each phase has its own discount rate r_s and define the cashflow discounting term as (1 + r_s)^{T_s} where T_s is the number of years from present until the cashflow for the stage occurs. Of course, a stage can take up several years and this could be fairly complex to model. For simplicity, therefore, the standard way to model this is to define the cashflows for a given stage as being realized in the midpoint of the stage. Therefore T_s can be understood as the number of years from the present until the midpoint of the given stage.

Meanwhile, for we derive the following formula:

\mathrm{rNPV}_{\mathrm{op}} = \sum_{t = T_{\mathrm{Launch}}}^{\infty} \frac{(R_t - C_t) \cdot \left( \prod_{j = \mathrm{S}_{\mathrm{Current}}}^{\mathrm{S}_{\mathrm{Launch}}} p_j \right)} {\prod_{\tau = 1}^{t} (1 + r_\tau)}

For the formula for the years of operation, we first update the summation so that it starts at the first year of launch and it ends with the last year that the company is modeled (an arbitrary point that is up to the modeler). Second, we expand out CF_t to R_t - C_t. Third, we establish that each year in market has a unique discount rate, and define the cashflow discounting term as \textstyle\prod_{\tau = 1}^{t} (1 + r_\tau). And last, we replace the term p_t with \textstyle\prod_{j = \mathrm{S}_{\mathrm{Current}}}^{\mathrm{S}_{\mathrm{Launch}}} p_j , or the probability that the project clears the application stage and reaches the market.

Now that we have these expressions for the value of the product in development and operations, we can combine them into an expanded formula for the value of the product.

\mathrm{rNPV} = \underbrace{ \sum_{s = \mathrm{S}_{\mathrm{Current}}}^{\mathrm{S}_{\mathrm{Launch}}} \frac{(R_s - C_s) \cdot \left( \prod_{j = \mathrm{S}_{\mathrm{Current}}}^{s} p_j \right)} {(1 + r_s)^{T_s}}} _{\text{development period}} + \underbrace{ \sum_{t = T_{\mathrm{Launch}}}^{\infty} \frac{(R_t - C_t) \cdot \left( \prod_{j = \mathrm{S}_{\mathrm{Current}}}^{\mathrm{S}_{\mathrm{Launch}}} p_j \right)} {\prod_{\tau = 1}^{t}(1 + r_\tau)}} _{\text{operational period}}

Which we can simplify to the following:

\mathrm{rNPV} = \underbrace{ \sum_{s = \mathrm{S}_{\mathrm{Current}}}^{\mathrm{S}_{\mathrm{Launch}}} \frac{(R_s - C_s) \cdot P_s}{(1 + r_s)^{T_s}}}_{\text{development period}} + \underbrace{ \sum_{t = T_{\mathrm{Launch}}}^{\infty} \frac{(R_t - C_t) \cdot P_\mathrm{Launch}} {\prod_{\tau = 1}^{t}(1 + r_\tau)}}_{\text{operational period}}

Compared to the original equation for risk-adjusted net present value that uses a single term across all years of the product, this form of the equation is useful because during a product's development, we will typically have quality information about the average costs, probabilities, timelines, and cost of capital for products across stages, but it is more difficult and less useful to have information about a product across various years of development.

Optionally, if we are ok with somewhat of a decrease in accuracy, we can (1) ignore any possible revenues during development (2) use a consistent discount rate r for all years of operation (3) stop modeling the revenues for a medical product in market at a given year T_\mathrm{Max}. This simplifies the equation further as follows:

\mathrm{rNPV} = \underbrace{ \sum_{s = \mathrm{S}_{\mathrm{Current}}}^{\mathrm{S}_{\mathrm{Launch}}} \frac{-C_s \cdot P_s} {(1 + r_s)^{T_s}}} _{\text{development period}} + \underbrace{ \sum_{t = T_{\mathrm{Launch}}}^{T_\mathrm{Max}} \frac{(R_t - C_t) \cdot P_\mathrm{Launch}} {(1 + r)^{t}}} _{\text{operational period}}

It's important to note, however, that (1) discount rates can vary fairly widely from year to year (2) some medical products end up booking revenues before approvals, and (3) even after revenues decline significantly many years in the future, there is still value that can be realized. Thus, for more accurate modeling one should aim to use the more precise equation.

Section 4

Factors of Biotech Success

We can use the rNPV equation for a medical product to inform us on the various factors that go into the success or failure of a prospective medical product. These factors can be thought of in turn as the various levers that one can pull as either a sponsor or a regulator to expand and accelerate the medical product pipeline.

First, we see from the equation that rNPV is proportional to the revenue of a product over its lifetime. If the sponsor of a candidate for a medical product expects to make more money once the product is in the market, the candidate is more attractive as a potential future medical product and there is a greater tolerance for costs and risk. The sponsor is therefore more likely to take on the enormous costs and risk to shepherd the product to market.

Second, we can see that rNPV is inversely proportional to the total costs across the lifetime of the product. This includes costs during the various stages of development, as well as costs during years of operation. This makes sense as the greater the costs of a product in development and operations, the less valuable the product is to sponsors and thus the less attractive it is to pursue.

Third, the equation shows us that rNPV is proportional to the probability of success across each stage of the product's development. At any given stage, if the probability of reaching market goes up, the expected costs of course rise, but the expected revenues increase to a much greater extent, resulting in a much larger overall rNPV for the product.

Fourth, rNPV is inversely proportional to the cost of capital. As the cost of capital rises, the value of the denominators in the rNPV equation increases, decreasing the discounted present value of the cashflows and thus the overall rNPV.

Fifth, rNPV is inversely proportional to development time. As the expected time to go through each stage goes up, the size of the discounting term in increases, raising the denominator of the development portion of the equation and decreasing the overall value of the cashflows.

We can summarize the aforementioned factors of biotech success, or levers for accelerating biotech, as follows:

1. Revenues
2. Costs
3. Probability of success
4. Cost of capital
5. Development time

Section 5

Interventions for Accelerating Biotech

Policymakers and regulators can utilize this model of biotechnology success in order to inform the various ways in which they can adjust their policies to improve the pipeline of medical products into the market.

First and foremost, regulators can help medical product sponsors decrease their time to market.

Perhaps the most direct way to do this is to decrease the amount of time that it takes for sponsors to hear back from regulators in all of their communications and submissions. This can represent a substantial amount of time, including the 10 months on average that it takes for sponsors to hear back for New Drug Applications (NDAs), Biologics License Applications (BLAs), and Premarket Approvals (PMAs).

Beyond this, however, regulators can make targeted adjustments to the rules around how clinical trials must be performed and how they will be reviewed. Through reform in these areas, the average time it takes for medical products to go through each stage of trials can be reduced.

Second, regulators can help sponsors reduce the costs of clinical trials and overall costs in various stages of development. They can (1) remove regulations that are keeping the costs of clinical trials higher than needed (2) promote advanced forms of clinical trial design such as adaptive trials, bayesian trials, platform trials, and basket trials, as well as update and clarify the rules around them (3) promote the use of challenge trials for certain low severity infectious diseases as an option to reduce the time and cost of trials for these indications.

Third, regulators can help companies increase their probabilty of success of clinical trials. They can (1) de-emphasize pre-clinical animal models in favor of pre-clinical in-vitro models of human tissue and in-silico models of human biology (2) promote advanced forms of clinical trial design like the ones mentioned above (3) provide more guidance and early communication to sponsors to ensure they're more likely to pursue a path that the regulators will approve of and less likely to pursue a path that will inevitably be denied.

Fourth, regulators can help companies increase the revenues of medical products. They can (1) remove regulations that restrict how drugs may be priced and paid for, as some drugs like gene therapies that are administered once and cure a disease may not be viable unless they can reach the same lifetime revenue as a mere treatment that needs to be taken for a lifetime and produces worse results (2) support the process of market turnover and "creative destruction" whereby superior medical products are more likely to actually reach market and redirect the revenue from existing products over to them, resulting in increased revenues to new products without increasing the overall expenditure that is borne by insurance companies and consumers.

Last, regulators can reduce the cost of capital for medical product sponsors. When all of the other factors are improved, when probability success and revenues go up while out-of-pocket costs and time to market go down, the cost of capital for such companies goes down as well. Thus the efforts on the other factors can have a compounding effect beyond their initial direct effects. This is the most straightforward and sustainable way to reduce cost of capital, without shifting costs to other areas of society. Of course, beyond this the cost of capital can be further reduced through (1) government grants and tax credits (2) public-private partnerships (3) provide government-enabled loans at reduced rates for biotech sponsors as a form of industrial policy (4) allow biotech companies to book revenue before approval, potentially as a part of a right-to-try program.