How much does it cost to develop a drug?
We still don’t know.
The most recent analysis on this subject was published earlier this year.
In an excellent analysis, Sertkaya et al. analyzed data from 2000-2018, using a broad blend of data sources.
We will not steal their thunder. Indeed, if you are at all familiar or interested with this literature, we recommend taking a look at this study.
A few of their conclusions were interesting.
So how much does it cost?
The short answer is that we still don’t know.
The authors estimated that actual cash outlays for drug development ranged from $173 million for genitourinary drugs to nearly $300 million for pain and anesthesia drugs.
But this ignores the cost of failures and capital.
When these additional costs are included, the authors estimated costs of nearly $380 million for anti-infectives, all the way to ~$1800 million for pain and anesthesia.
To arrive at these numbers, the authors had to make a number of estimates and assumptions, largely because we may never know, for example, exactly how long it takes to get through Preclinical (what does “Preclinical” even mean, anyway?).
PoS and LOA
PoS is the probability of a drug candidate successfully being promoted from one clinical phase to another. For example, if there are 100 candidates in Phase I, but only 30 make it into Phase II, then the PoS for that particular indication (or modality or therapeutic area) is 30%.
LOA is the likelihood of approval, typically estimated by calculating the progressive PoS from phase to phase.
Now we realize that these figures are, by their nature, very broad estimates. And yet, they are quite important in calculating project valuations and understanding relative risk.
Fortunately, Tables 1 and 2 in Sertkaya includes fresh calculations of PoS and LOA for a range of therapeutic areas. These are good complimentary data to previous studies, such as Hay, 2014.
Importantly, this paper calculates LOA from Preclinical to Approval, whereas most papers in this field calculate LOA from Phase I to Approval. Obviously, it depends on what we mean by “Preclinical.” But, regardless, the numbers in Table 2 look about right to us.
The lowest LOA is….oncology, with an LOA of 4.1%.
So one way to think about this number is that oncology is the “riskiest” therapeutic area in their dataset. And yet, we see record numbers of both dollars and deals in this space. Indeed, oncology was the second most expensive therapeutic area for R&D spending.
We will leave it up to our readers to interpret this reality as he/she sees fit.
It should be noted that our collective PoS and LOA have been flat to declining over time, as reported by IQVIA. This is somewhat understandable, in the sense that we are pursuing more challenging targets with more complex modalities.
But it speaks to our collective skill as an industry to overcome these biological and chemical challenges. In other words, are there management teams and scientific personnel with “above average” PoS and LOA track records?
Or, would a team of all star drug developers be hampered by the difficult realities associated with drug development?
What does this all mean?
As usual, it is difficult to make sweeping generalizations from a study like this, even a good one.
After all, each situation is different. Even within a single indication, there can be two wildly different development program lengths, timelines, costs, and risks.
So, how much does it cost?
We still don’t know for sure. And we will never know.
And, frankly, it doesn’t matter whether or not we know exactly how much it costs to develop a new drug product.
But what does matter, and what we can say for sure, is that we as an industry simply must do what we can to bring development costs down.
But How?
Obviously that’s the billion dollar question…a question which defies an easy answer.
Adaptive clinical studies…better biomarkers…correlative animal models…artificial intelligence…these are all pieces of the puzzle that are already being deployed across our industry.
Given our historical industry PoS and LOA performance, it is clear that “better management” is not the long-term solution.