The free market in Tamiflu

It was inevitable. Roche is now planning to cut production of its antiviral Tamiflu because, they say, supply is exceeding demand:

Swiss pharmaceutical group Roche said Thursday it would trim production of the frontline bird flu drug Tamiflu unless demand picked up, but warned that the world was still not ready for a pandemic.

Manufacturing capacity for the antiviral treatment has reached 400 million treatment courses a year and is outstripping demand, William Burns, the chief executive of the group's pharmaceuticals division told journalists.

Roche has received orders for 215 million treatments from governments and private companies, and just 40 million still have to be fulfilled, he added.

"Today we can satisfy significant additional orders from governments and corporations, and unless the demand picks up Roche will be tailoring its production schedule accordingly," Burns said.

"The question is how much do we need to activate," he added.

The drug is now the pharmaceutical giant's fourth best seller. (AFP)

400 million treatment courses -- a year? I wouldn't exactly say that meets global demand. Now, I realize there is demand and there is "demand." Let's forget the word "demand," for a second, and substitute the word, "need." Is it true that Roche's annual production of 400 million treatment courses exceeds global need? At the moment, the need is almost zero, at least when it comes to human avian influenza. The need is "in case" need. In case there is a pandemic with a strain sensitive to H5N1, then there isn't nearly enough of this stuff to go around. Not even close. When Roche uses the word "demand," they are talking about "supply and demand" as in a theoretical free market. Currently buyers aren't willing to pay Roche's price, so Roche has two choices: reduce the price or reduce the supply. We see what choice roche decided to make. The "market" isn't the right regulator, here.

Burns said: "We've acted responsibly but we do need a partner in government with demand."

"Capacity constraint is no longer the issue. The issue is demand and how prepared governments want to be. We can't just keep building the inventory with nobody as a partner," he added.

The group says it could halt production now and still satisfy current demand.

It will maintain a buffer stock at all times and remain in close contact with manufacturing partners "to respond speedily to a surge in demand," Burns added.

How long would it take them to restore full capacity? Four months (AFP). So this isn't "speedy," and even when running full blast, supply wouldn't meet need. In such circumstances supply has to greatly exceed demand. It's not economically efficient. Neither is a pandemic. So we need to go outside the market mechanism. Just like the vaccine situation.

In a related story, the the director of the Center for Biosecurity and Public Health Preparedness in Houston has called on the Texas State legislature to purchase enough antiviral drug to cover 25% of the state's population, roughly 2.3 million courses at the Federally discounted price (which, as I understand it, is not discounted by Roche but subsidized by DHHS). The 25% figure is a CDC recommendation, but Texas is one of 17 states not to have reached that level of preparedness (press release, Center for Biosecurity and Public Health Preparedness). That's not too bad for the US, but much of the developing world has nothing.

And it doesn't sound like the Free Market is going to ride to the rescue.

More like this

It's not exactly clear that the company should pick up the tab for world preparedness for avian flu, either, right? Who should pick up the tab, and why aren't they?

For the poorest countries, the fact is, they have enough current crises that they probably don't have a lot of budget for preparing for possible disasters that may never happen.

Even for rich countries, there are plenty of things they might spend their money on. Stockpiling Tamiflu for their own citizens is only one of them, and not necessarily the top one.

By albatross (not verified) on 27 Apr 2007 #permalink

I belive that the cost for Tamiflu for the U.S. federal government as well as the states is only a fraction of the normal retail price for Tamiflu (~25 - 30% of retail price). The 25% that the federal government pays to subsidize Tamiflu for the states is 25% of the discounted price, not the retail.

We also need to keep in mind that most government stockpiles (including the U.S), are based on a 5-day treatment regimen of Tamiflu. (based on seasonal flu). Most of the evidence to date (including both animal studies, as well as treatment of infected humans) indicate that the dosage needs to be higher for the H5N1 strain (potentially either a 10 day treatment course and/or potentially at twice the normal dosage). Past reports from WHO, including the WHO clinicians meeting (clinicians treating H5N1 patients) report from a couple of weeks ago also recommended increasing the treatment to 10 days.

Many of the current patients being treated are receiving at least 10 days (unless they die before that time)

No official directive has been given by WHO yet (still studying the issue), but it is likely that most stockpile amounts would only be half of what is actually needed.

I agree that we need to try to keep the production capacity as high as possible, and that even the 400 million/year treatment course production would be significantly low if we have a H5N1 pandemic outbreak.

Compare this to a world population of 6.5 billion and a 30% infection rate - 2.1 billion people would potentially need Tamiflu for treatment (if you ignore any prophylactic use). If you then multiply this by 2 (10 day treatment, rather than 5 --> 4.2 billion courses needed), we would only be able to currently produce enough Tamiflu for about 10% of the world's population in a year. (we had been at 1%, back when production capacity was 40 million/year).

Economic market demand is the reality of how our businesses operate. It was the economic demand (along with some federal government arm twisting), that got the production up in the first place.

I personally believe that stockpiling of Tamiflu for governments, businesses and individuals is a good thing - because it is this demand that keeps production high, and will better enable us to respond to a pandemic, if one develops.

Stockpiling is all well and good, but given that the shelf life of Tamiflu is only 36 months, and much of the worlds existing stockpile is at 12 months (some longer, some less), it would seem maintaining existing stockpiles would require these governments to place orders in advance of their stocks expiration (unless thay plan to use Tamiflu in a pandemic which is past it's expiration date). This would help Roche ensure that they maintain adequate production capacity, and that governments, etc have the stock they deem necessary. Hopefully they are not assuming that if a pandemic does not happen within the expiration of their stock, it will not happen.

The longer H5N1 is around before going pandemic, if it does go pandemic, more likely than not it would be Tamiflu resistant. But an influenza pandemic, H5N1 or other, is inevitable at some point, so having stockpiles of the antivirals makes sense.

By Paul Todd (not verified) on 28 Apr 2007 #permalink

Paul: Tamiflu resistance, even if rare, will spread rapidly, but the mdoeling (see antiviral modeling in the categories on the sidebar) suggests we are still better off using it. And there is the question of fitness cost, which we don't have an answer to at the moment. So use of antivirals is indicated, even with spread of resistance. There are now cheaper and quicker ways to make it and it could be done at cost outside the market mechanism.

Revere, any idea how long it would last before the resistance level went up to the unusable column. Is there a graph or one thats suggested from the use of antivirals from previous stuff that shows us where line will be crossed. It seems to me that since we only have had what pushing 300 cases and its already showing large resistance to it, that it wouldnt be long. I agree with you though that it should be used until the fat lady develops pneumonia.

By M. Randolph Kruger (not verified) on 28 Apr 2007 #permalink

I don't think anyone can say this with any confidence. The modeling shows that combining it with other measures greatly enhances the effect of each. My guess is that we would get at least one good wave of benefit out of it, maybe two. This would buy time for figuring out how to cope with it, including possible vaccine production or as yet unknown measures. Or maybe not. But we don't have many arrows in our quiver and antivirals are currently one.

The shelf life for Tamiflu is actually closer to 10 years, and possibly much longer if it does not get wet.

The original packaging for Tamiflu said 3 years, then was updated to 5 years. Discussions with physicians at the CDC as well as with Roche have indicated that the shelf life is actually much longer.