The cost of prescription drugs is a perennial subject of heated debate, as advocates on one side argue that drug companies make windfall profits and overcharge health care systems and advocates on the other side argue that higher drug prices simply reflect a higher cost of doing business and a need for companies to make a return commensurate with the risks they take on. Although I have no delusions that I'll change the minds of those who believe drugs are too expensive and that drug companies are abusing patients and insurance companies, the impact of rising costs of drug development can't be ignored.
Costs Keep Climbing
According to a study done by Tufts University, roughly 40 years ago (1975) it only cost $100 million in 2005 dollars to develop one drug from the lab to FDA approval. By 1987, that figure had tripled and by 2000 the cost was up to $800 million. In 2005, the per-drug cost of a successful approved drug had reached a whopping $1.3 billion.
It's not hard to see why costs are rising. Above and beyond the “everything's just more expensive now” argument, drug companies have had to significantly expand the size and complexity of their studies. The number of procedures performed as part of drug trial (that is, particular tests, treatments, and other such steps) has increased by 50% to 65% just in the last ten years, and as anyone who has seen a recent bill from a clinic or hospital can attest, every test comes at a cost. At the same time, drug trials have become larger and larger, and drug companies are required to study the effect of drugs in larger and more diverse populations than ever before.
There are at least three notable factors playing a role in these larger, more complex drug trials. First, the FDA has steadily increased its expectations and requirements for approval – over the last decade, for instance, it has become a nearly established requirement that companies looking for approval of new diabetes drugs must include a cardiovascular outcomes study. With a growing burden on establishing the safety of new drugs, then, there are more required tests throughout the clinical trial process.
On a related note, larger drug companies are increasingly looking to larger, more complex studies as a way of de-risking the approval process. In most cases it is cheaper to conduct a larger, more complex Phase 3 study than to face FDA rejection and a requirement to go back and do another study to address a perceived deficiency in the drug's data package. Last and not least, drug companies have seen more pressure from their insurers to conduct more thorough studies – with the cost of legal settlements also rapidly rising, insurance companies are demanding more thorough (and expensive) safety testing as a way of reducing product liability payouts.
Success Is Far From Assured
At its core, drug development is still a high-risk endeavor. Only about 19% to 21% of drugs in Phase 1 trials will ultimately see approval. Phase 1 studies are not particularly demanding, though, and roughly 70% of drugs go on to Phase 2 where the odds of ultimate approval are still quite low (roughly 28%).
Once a drug makes it into Phase 3 testing, the odds significantly improve (58% to 62% make it through to approval), but the costs also go up considerably – Phase 3 studies are about one-third more expensive on a per-patient basis and Phase 3 studies are almost always considerably larger (6x to 30x the number of patients versus Phase 2). Of those drugs that are ultimately approved, as much as 90% of the development spending can occur in Phase 3 (though 65% to 75% is likely a closer estimate for the typical drug).
Only about 32% of drugs put into human studies will it make it to Phase 3, and those failures (nearly seven out of every 10 drugs) will cost around $40 million each on a median basis, but with some pre-Phase 3 programs costing as much as $100 million. Once those drugs make it into Phase 3, trials can cost anywhere from hundreds of millions of dollars on average to more than $2 billion depending upon the disease in question (cardiovascular disease drugs, for instance, tend to require much larger phase 3 studies).
My calculations suggest that, on the basis of past company-reported surveys regarding per-patient trial costs and using data from ClinicalTrials.gov to estimate median trial sizes, for the two approved drugs out of 10, a drug company may spend around $800 million, while the cost of the eight failures will total to almost $750 million. Clearly these numbers are lower that the $1.3 billion/per drug figure cited earlier in the Tufts study, and at least some of this has to do with how to allocate expenses like preclinical development and “basic research”, licensing fees, and so on, and some of it also the difference between median and mean spending.
There's a sizable difference between mean and median costs, as certain kinds of drugs (particularly cardiovascular, metabolic, and autoimmune/inflammatory) require much larger and more expensive trials. All things considered, independent biotechs tend to focus on developing drugs with more modest Phase 3 requirements, while the largest trials are often only run by the big drug companies.
Arena's (Nasdaq:ARNA) Phase 3 BLOOM study (which was just one of three Phase 3 studies for the obesity drug lorcaserin) enrolled almost 3,200 patients and likely cost upwards of $200 million. By comparison, Medivation's (Nasdaq:MDVN) Phase 3 AFFIRM study for prostate cancer drug Xtandi enrolled 1,200 patients, while a single Phase 3 study of Eliquis, an oral Factor Xa inhibitor marketed by Pfizer (NYSE:PFE) and Bristol-Myers (NYSE:BMY), enrolled more than 18,000 patients and the total Phase 3 development program enrolled more than 60,000 patients across multiple studies. Even if Pfizer and Bristol-Myers somehow managed to do those trials at about only half the cost of a normal Phase 3 study, that's still upwards of $1.5 billion in Phase 3 development costs.
Earning Back Those Costs
Not only do approved drugs have to pay for themselves, but they also have to subsidize the company's failures and still leave something for shareholders. While multi-billion dollar blockbusters like Pfizer's Lipitor, AbbVie's (Nasdaq:ABBV) Humira, and Sanofi (NYSE:SNY)/Bristol-Myers' Plavix certainly have more than paid their way, they really are the exception.
The average approved drug generates about $300 million per year in revenue over its lifetime. Excluding R&D costs, the average Big Pharma drug will generate a margin of approximately 50% and will have a productive revenue-generating life of about 15 years. Looking again at the averages (and excluding taxes or the cost of capital), those two-of-10 approved drugs will earn back their cost of development in about three years and pay for their failed siblings in another two and a half years.
That means roughly one-third of the productive life of an average drug goes to paying back the cost of drug development. It also explains why drug companies are so willing to spend on marketing and follow-on label extension studies and so fierce in protecting their patents – each extra year of branded drug sales has a significant impact on the net present value of the drug in question.
Using the prior inputs, that portfolio of 10 drug candidates could be expected to produce an IRR of about 13.25%. Just one extra year of exclusivity on the back-end for both drugs would improve the overall IRR by about a quarter-point, and the development of blockbuster drugs like Lipitor can lead to huge IRRs. By the same token, though, companies that hit dry spells in their clinical development and/or have major disappointments and late-stage trial failures can see their portfolio IRRs quickly plunge.
The Bottom Line
At the end of the day, any analysis of the cost of drug pricing comes down to a question of fairness. While some advocates and commentators choose to focus on what's fair for patients and payers, I think it's also worth considering what is fair the drug companies as well.
I will admit that the mix of medians, averages, and estimates makes the conclusions here only rough at best, but is a 13% internal rate of return excessive when the odds of success are pretty poor and large sums of money have to be invested with no expectation of any returns for five to 10 years? Looking at it differently and using an NPV approach, a 10% discount rate on those cash flows would lead to a net present value of about $270 million for that fictional 10-drug portfolio. While it's certainly true that the occasional blockbuster drug can lead to substantially higher returns for pharmaceutical companies, it's likewise true that a dry spell in the clinic and/or expensive late-stage failures can lead to periods of very poor returns.
Again, “fair” is a very subjective concept. I see nothing wrong with pharmaceutical companies earning a good return given the large amounts of capital and high risks that go with drug development. Perhaps just as much to the point, if the public continues to demand new drugs that allow them to live longer and/or better and that are proven to be relatively safe, they need to realize that they're going to have to pay for them to keep the companies motivated to take on the development risks.
Disclosure - At the time of writing, the author did not own shares of any company mentioned in this article.