Saturday, April 12, 2014

Drug Prices and the Pharmaceutical Market


In recent weeks a fair amount of discussion in the news has focused on the high price of drugs. Some of that is attributable to the high price Gilead Pharmaceuticals set for Solvadi, as mentioned in my last post. Last week, the Massachusetts Biotechnology Council "MassBio" issued a report which warned that the increasing pressure on drug prices along with the overall push to contain healthcare costs could threaten the future growth of the biotechnology industry and the rate of innovation in the pharmaceutical industry.1 For many years, the major pharmaceutical companies and their trade association, now known as PhRMA,2 have been engaged in a reasonably successful effort to convince the American public and their elected representatives that the high cost of many drugs is the result of the very high costs of drug development, which is frequently estimated at $1.2 billion to bring a new drug to market. I don't want to use this post to debate that $1.2 billion dollar figure. Drug discovery and development is a very expensive process, with high costs and many failures, which may well bring the total costs per new drug approval to $1.2 billion. However, I begin my course on FDA Law by asking students "What determines the high cost of drugs?" The success of the PhRMA public relations effort is reflected in the very large percentage of students who do indeed answer, "The high cost of developing new drugs." I then proceed to give them their first lesson on pharmaceutical policy-- which is that the market for pharmaceuticals is like almost every market in our essentially free market U.S. economy. This means that sellers set the price of their goods at what the market will bear.


This does not mean that the high cost of drug development has no relationship to the price of drugs, just that it does so in a relatively indirect way: pharmaceutical companies will not embark on a drug development program unless their analysis shows that their ultimate sales revenues will provide a reasonable return on investment on their very sizable investment in drug development overall. In other words, major pharmaceutical companies are generally not interested in working on drug development programs that, if successful in reaching the market, will generate less than a $1 billion in annual sales. A billion-dollar drug represented the threshold for the "blockbuster"drug category in the eighties and most of the nineties, and every biotech company executive knew that big pharma would not be interested in hearing about drugs with smaller expected revenues. In the seventies and eighties, drug sales revenues generally achieved blockbuster levels by selling a "modestly" expensive drug (for example a new $1500 per patient per year antihypertensive) to a very large number of patients (often a million or more patients). So the earlier antihypertensives, the first statins, and the first SSRI antidepressants, all sold for prices that were very modest by today's standards, but were sold to very large numbers of patients. However, it has become increasingly difficult to develop drugs for these kinds of large patient populations, so the second method of achieving enormous annual revenues, pioneered to a significant extent by Genzyme, is to sell a drug at unheard of high prices to even a small patient population. A drug that sells for $150,000 per patient per year can be enormously profitable even when the market is 10,000 patients. With the increasing difficulty of developing drugs for very large patient indications, this second model of drug pricing has become increasingly dominant, and with it has come more and more scrutiny of drug pricing and drug company profits.

These first two aspects of drug pricing and the pharmaceutical marketplace are relatively uncomplicated. First, drug prices are set in the marketplace. Second, the high cost of drug development has made it unattractive to develop drugs that do not have enormous potential revenues. Neither of these two assertions would be a surprise to any pharmaceutical company executive nor is either partricularly controversial. And there is not any real debate over the logical conclusion: the revenues required in order for a major pharmaceutical company to greenlight a drug development program can only be achieved by very high prices if the target population is not very large. The third major aspect of drug pricing and the pharmaceutical marketplace, which will be discussed next, is more complicated to explain and more controversial: drug companies do not compete extensively, if at all, on price.

The prices of brand name pharmaceuticals in a class of drugs, such as the statins for cholesterol reduction, or the proton pump inhibitors for gastroesophageal reflux disease (heartburn), do not come down much, if at all, when one of the other pharmaceuticals in that class becomes available as a generic at a small fraction of the cost. A 30-day supply of Crestor (a potent statin) costs just under $200 at Costco (one of the lowest cost retail pharmacies in the U.S.), while the Costco price for a 30-day supply of Pfizer's branded Lipitor varies from $165 to $235 depending on the dose.3 At the same time, Costco sells a 30 day supply of the generic version of Lipitor, atorvastatin, for between $15.59 and $20.33, again depending on the dosage. This generic price, like most generic prices, is less than 10% of the cost of the brand name drug, in this case higher doses of Lipitor. So Pfizer is clearly not competing with the generic versions of its Lipitor; nor has Astra Zeneca's Crestor come down substantially in price in response to the introduction of generic Lipitor. Crestor continues to command a high price even since Lipitor became available as a generic. This is despite the fact that a heavily funded Astra Zeneca SATURN study failed to show that Astra Zeneca's drug was superior to Lipitor on Astra Zeneca's carefully chosen surrogate endpoint of percent atheroma volume (PAV) measured by intravascular ultrasound.4 Similarly, Costco sells Nexium for much more than 10 times the price of Omeprazole, the generic version of Prilosec, which has the same active ingredient in a lower dose. The high prices of Crestor and Nexium in the face of similarly effective generic compeetition clearly illustrate the fact that the pharmaceutical marketplace is not one in which price is the major point of competition (although there is some competition on pricing at the margin in deals between pharmaceutical companies and the pharmaceutical benefits managers of the major health insurors).

If pharmaceutical companies are not competing on price, and pharmaceutical prices for branded drugs are not determined by the costs of developing or producing the drug, what is the nature of competition in the pharmaceutical marketplace? It would be nice if the answer to that question was that drug companies are competing on the basis of their absolute contributions to patients' health; but, in the great majority of cases that isn't true either. In the example of Crestor and Lipitor, discussed above, both drugs were marketed heavily by promoting their ability to reduce levels of LDL (low density lipoproteins, or "bad" cholesterol), which is a primary endpoint for evaluating such drugs. And when Astra Zeneca funded the SATURN trial in an attempt to show that the slightly greater reduction in LDL achieved by Crestor would translate into a greater reduction in the amount of plaque inside arteries, the trial showed that it did not. And, even more important for purposes of this discussion, the relationship between that endpoint and the "real" endpoint of a reduction in cardiovascular morbidity and mortality was not tested, almost certainly because any difference in that endpoint would be very, very small between the two drugs, and almost impossible to demonstrate in a trial of even several thousand patients.

Another rather interesting illustration of the strange and complex way the pharmaceutical marketplace focuses on factors other than price and effecticacy, is provided by the competition between three high-cost rheumatoid arthritis (RA) drugs--Humira, Remicade, and Enbrel--which all work on the same pharmaceutical target in the body. These drugs work by blocking or reducing the activity of TNF (tumor necrosis factor), a circulating protein that is a key driver of the inflammation pathway. Inflammation is, of course, the underlying pathology in a number of diseases, from rheumatoid arthritis to inflammatory bowel diseases such as Crohn's disease. A recent study found NO difference between these drugs in terms of their safety and effectiveness in treating RA.5 That's right-- there is NO difference in how well these three drugs work or their relative safety. And the prices of all three drugs are high. One industry source estimates the average per patient cost of all three drugs at approximately $20,000 per year.6 While all three drugs are priced very similarly, Humira and Remicade each had total 2012 sales of just under $2 billion per year (for the RA market only, not total sales for all uses of those drugs), while Enbrel's 2012 sales in RA lagged well behind at about $938 million.7

So how are these drugs competing? One answer seems to be that they are competing on patient convenience (Humira wins on that) and reimbursement (Remicade has the advantage, at least for Medicare patients). Humira is packaged in two different dosages in pre-filled syringes for patient self-injection subcutaneously. Remicade is administered in physicians' offices or clinical settings, as an intravenous infusion, which is not convenient for patients, but which means that Medicare reimburses the drug's cost for all Medicare patients.8 Enbrel, like Humira, is a subcutaneous injectable drug. However it is not pre-packaged in convenient doses, but must be properly stored and then mixed prior to injection, which can be done by some patients after appropriate instruction. This is obviously much less convenient than Humira, and because it is not presumptively administered by a physician, as is Remicade, Medicare reimbursement is not automatic. The result is that Enbrel's sales lag well behind its two main competitors, despite its equal safety and efficacy.

The pharmaceutical marketplace is a complicated one, where patents provide a signficant measure of protection against competition based largely on price, where data on the real benefit of competing drugs in terms of patients' health is often lacking, and where drugs sometimes compete on other factors such as convenience and reimbursement. When all drugs are equally convenient, as is the case of the statins and most other once-a-day pills, the competition is driven by marketing, often focusing on non-primary health endpoints, such as marginal incremental reductions in LDL levels. Despite the lack of any evidence whatsoever that Crestor is "better" for patients' health than atorvastatin, the generic Lipitor, Astra Zeneca managed to sell over $5 billion dollars worth of Crestor in 2013, the first full year after generic Lipitor became available! No wonder marketing budgets for pharmaceutical companies are even greater than their R&D costs.9 So to return to the question of why pharmaceutical companies set such high prices for drugs, the answer is that it is a bit complicated, but mostly because they can.

1 MassBio, Impact 2020: Advancing Massachusetts Leadership in the Life Sciences for Patients, online at http://www.massbio.org/news/450massbio_issues_impact_2020_report_outlining_the/news_detail
(visited April 11, 2014).
2 Pharmaceutical Research and Manufacturers of America- the "R" was added with an eye to public relations decades ago when the organization was simply PMA.
3 All Costco pharmacy prices quoted in this post were obtained online on April 11, 2014 at http://www2.costco.com/Pharmacy/DrugInformation.aspx?p=1 .
4 http://clinicaltrials.gov/ct2/show/study/NCT00620542 . See also Nicholls et al., Effect of Two Intensive Statin Regimens on Progression of Coronary Disease, 365 New Eng J of Med 2078, December 1, 2011.
Michael O'Riordan, Rosuvostatin and Atorvastatin Equally Regress Atherosclerosis, Medscape Pharmacists Heartwire, November 15, 2011 at http://www.medscape.com/viewarticle/753603 visited April 11th, 2014.
5Jeffrey D Greenberg et al, A Comparative Effectiveness Study of Adalimumab, Etanercept and Infliximab in Biologically Naive and Switched Rheumatoid Arthritis Patients, 71(7) Annals of Rheumatic Diseases, 1134-1142 (2012)
6 Peter Geschek, Fierce Competition in the RA Market, Seeking Alpha, June 1, 2012 (online at http://seekingalpha.com/article/632311-fierce-competition-in-the-ra-market).
7 Id.
8 Department of Health & Human Services (DHHS), Self-Administered Drug Exclusion Lists, CMS Manual, Pub 100-02 Medicare Benefit Policy, June 20, 2008.

9 Science Daily, York University, Big Pharma Spends More On Advertising Than Research And Development, Study Finds, January 7, 2008, available at http://www.sciencedaily.com/releases/2008/01/080105140107.htm

3 comments:

  1. My paper, which criticizes the FDA's decision to withdraw approval of original OxyContin, points to the potential for higher-priced drugs as one of many adverse consequences of the decision (based in part on Purdue's continued monopoly). I suppose I'll have to cite you (contra)!

    ReplyDelete
  2. Let us not discount the theory of high cost of drug development. But, to give patent and make it exclusive for 20 years is just mind numbing and amounts to exploitation. Everyone knows that the cost of drug development is recovered in short period may be two or three years. so, giving patent for five years for any new drug entity would be justified. Moreover, the pricing of the drugs ought to be cost base. i.e. material cost+research cost+conversion cost + packing + marketing cost. It may be review every two year and increased to meet the rising costs.
    Our bad luck is that there are several politicians who are on payroll of Pharma companies (of course not directly). That skews the Pharmaceutical Policy of every country.
    It is absolutely necessary to amend International Patent Laws and Treaties to restrict patent period for five years only. At least in healthcare and pharmaceutical sector.
    Jayendra Pandya, Mumbai India

    ReplyDelete
  3. Jayendra,
    I appreciate your reading of the post and your input. I think that the market incentives for drug development aren't perfect, but can be improved without tinkering much with patent term. The effective term of patents in pharmaceuticals is now about 14 years, and has increased in the past decade or so from 12 years. I think treating inventions for life saving drugs LESS favorably then cell phone inventions, for example, would do more harm then good.

    ReplyDelete