I am not a health economist, nor a financial analyst, and that makes me a very happy man. I am sure my head would spin if I had to explain drug pricing. I looked at some recent FDA drug approvals and came up with (by Googling and a little back-of-envelope calculating) the following pricing:
- Vandetanib at $10,454/month
- Yervoy (ipilimumab) at $120,000 for 4 treatments
- Zytiga (abiraterone) at $60,000 per year
- Provenge (sipulucel-T) at $93,000 for a course of treatment
- Halaven (eribulin mesylate) at about $75,000 per year.
The cost of brand name prescriptions overall rose by 6.9% in 2010, according to Barclay's Capital, which analyzed the top 130 drugs by sales.
These of course are not the “real” costs of the drugs, at least not to payers and patients. Most patients don't remain on Zytiga or Halaven for a year, so their average cost is less in absolute terms (a good or a bad thing, depending on whether you are a payer or a patient).
On the other side of the equation, the true cost of a drug goes far beyond what Eisai or BMS or AZ or Centocor or Dendreon charges. Patients and payers are charged for chair time and for the costs of administration. The agents are often co-administered with antiemetics, growth factor support, steroids and antihistamines. Routine tests appropriate to specific drugs are obtained. Patients have side effects that require hospitalization or at least management by a doctor or nurse. Patients spend time away from work, often accompanied by family members who also spend time away from work.
The “real” cost of a new cancer drug to the health care system is not easily calculable, not in any real sense, but it is almost always significantly more than the cost of the compound itself, and the add-on costs of cancer drugs are almost certainly more than the add-on costs for antihypertensives or cholesterol-lowering drugs.
But regardless of the “real” cost, the proxy cost of what the pharmacy is charged by the company is still pretty impressive. The agents listed above improve overall survival, on average, by a few months. The cost per quality-adjusted life year (QALY), the standard means of looking at this, is through the roof for almost all of them. The words “obscenely high” come to mind.
So why do new cancer drugs cost so much? Again, I am no health economist, so I only know what I read, and what I read suggests that economists disagree significantly on this. The companies themselves are mum on how they calculate price, and if there is a standard way of measuring drug pricing I am unaware of it. It certainly isn't solely the cost of manufacture and distribution; it is rare for this to exceed 10% of what is actually charged.
Are New Drugs Really Costly to Bring to Market?
The standard answer is that new drugs are costly to bring to market. But is this really the case? One widely quoted figure is $802 million (in 2000 dollars; adjusted for inflation now over $1.3 billion) to bring a new drug to market, but a more recent analysis said this was nonsensically high and should be put at a median net R&D cost of $59.4 million (for the more recent figure, and a review of this literature, see a fascinating article by Light and Warburton in BioSocieties (2011;6:34-50 [March]); for a counterblast on this subject, read Derek Lowe's blog posts “The Costs of Drug Research: Beginning a Rebuttal” and “That $43 Million R&D Figure” at http://bit.ly/otUTuk and http://bit.ly/osUn3z respectively).
I suspect the actual number is somewhere in between, but the truth is that we really don't know.
R&D Costs: Murky
What makes up the R&D costs? Murky here, as well, at least to my aging eyes. How much of the target selection and validation occurs in NIH-funded laboratories as opposed to company labs? And once the target has been selected, then what are the costs?
Preclinical studies have gone through a revolution in recent years, as computational chemistry on enormous drug libraries identifies inhibitors for specific molecular targets in days or weeks. Costs escalate as one gets closer to approval: phase III trials cost more than phase II trials cost more than phase I studies cost more than animal toxicology costs more than animal model studies cost more than cell line studies.
Clinical Trial Costs
The clinical trial costs are full of self-inflicted wounds: two prime suspects are the piles of regulatory paperwork that add little to either patient safety or a better understanding of efficacy, and the recent faddish love for independent radiology review that adds nothing other than expense.
But a major and continuing cost in oncology (as opposed, say, to hypertension or hypercholesteremia) is the testing of new agents in untargeted populations, often simply because these populations represent “unmet medical needs” (for which, read: great opportunity to charge lots of money in an ecological drug niche with little competition).
The flip side, and clearly the way forward, is represented by agents such as vemurafenib for metastatic melanoma or crizotinib for lung cancer, where trials based on targeting a very specific molecular lesion required relatively smaller numbers of patients, and less flailing about. Plexxikon's laser-like focus on the BRAF V600E mutation, where it co-developed a reliable molecular assay along with a highly targeted agent, and did so economically both in terms of time and cost, is a model of this approach.
‘What the Market Will Bear'
But back to the initial question: why do these drugs cost so much? If the costs of production represent a minor part of the price, and if the R&D costs have been overstated (an arguable point), then the real answer is that companies are charging what the market will bear.
Part of the problem is that Medicare, as a matter of law, is not allowed to directly negotiate with companies on drug costs. In any other aspect of life, if you buy a million widgets rather than ten you get a better per unit price. But not when you are the US government buying drugs. The end result, to quote Bill Clinton, is likely to be that companies charge what they do “for the worst possible reason — just because I could.”
Oncology drugs, like Hollywood films, can be great successes or horrid failures. Avastin makes Genentech/Roche several billions of dollars per year, a pretty good return on investment for a drug brought to market largely on the backs of federally funded cooperative oncology group studies.
But most new drugs are not Avastin, and the companies are well aware of this fact. They are also aware, if they are honest with themselves, that they are pretty lousy at picking winners. Do they even trust their in-house scientists to do early drug development anymore? I suspect not: in recent years just about every big pharma company has shelled out large amounts either for a biotech start-up or just the rights to an unapproved agent. And these purchases—in essence, outsourcing of early drug development—are reflected in the cost of new drugs.
Oncology divisions in big pharma companies also compete with other divisions in the same companies for resources, and have to justify their existence at a time when they are producing boutique drugs (AKA: targeted therapy) for small markets (melanoma deaths = 48,000 worldwide; the prevalence of hypercholesteremia in the US is >50% of adults). “Boutique drug vs mass market drug” must be part of the price equation; so must the fact that several big pharma companies are facing expiration of patents for blockbuster products.
All of which to say that drug pricing is very complicated.
Companies exist to make money for their shareholders, who have invested their capital in the company and expect a return on investment. I see nothing wrong with this. No one works for free, and financially healthy companies are more likely to invest in R&D. But when I look at the new agents, with their break-the-bank prices for a reeling health care system relying on the tenuous support of a Congress that is schizophrenic on the issue of cost, I do wonder how long this crazy party will last.
More from George Sledge on His OT Blog! Recent Posts:
- Lawyers and Genes
- On Amy Winehouse