Cost of developing new drugs may be lower than industry claims: trial

The cost of manufacturing the drug is only relevant if the drug can be discovered and proved effective by the Magic Drug Fairy.

Here in the real world, that manufacturing cost is largely irrelevant.

This reminds me of P=NP. It's hard to find a sha256 with a given number of leading zeros, but it's super easy to verify. It's hard to make Microsoft Office, or a Pharmaceutical Drug, but easy to see that they work. The confusion is at the root of most economic fallacies.

> the Magic Drug Fairy.

Thats the government funded universities.

They do not, in fact, do the hard work of drug development.

You say all this, while the literal title of the article you're commenting on is "The cost of developing new drugs may be far lower than industry claims".

I think this comment perfectly proves that this is relevant.

Manufacturing is quite relevant. If the actual cost of "discovered and proved effective" is far lower than claimed, then manufacturability becomes a concern. If the cost of discovery and proven effectiveness is borne by governments and universities, then the manufacturing is the only cost borne by the pharmaceutical companies.

The article appears to imply that the direct cost of developing a successful drug is the relevant metric. It isn't; the cost must include the indirect costs of development on drugs that didn't pan out. And the vast majority of drug candidates don't pan out, often in Phase 3 trials where much of the cost has already been sunk.

Put it another way: if making new drugs was so cheap and easy, drug companies would be doing that in competition with each other, and prices would be low as a result of that competition.

The universities find compounds that work in cell cultures or mice. They don't currently optimize candidate compounds for use in humans or perform actual human clinical trials. In theory, there's no reason that a government couldn't fund all phases of drug development through final approval and then manufacture new drugs without seeking a profit. But most compounds that work in mice fail in clinical trials, and human testing is a lot more expensive than mouse testing.

There would need to be consistent support for such a drug development program even though most attempts will fail and "breakthrough" drugs will arrive irregularly, separated by many years. It is politically difficult to maintain that sort of long term support in a system where voters and representatives are swayed by short-term arguments to cut programs that don't show results.

Yeah, I was discussing this with my wife the other day. She spent about 10 minutes looking for a doodad and when she eventually found it, I was quite surprised. It took mere seconds to open that last drawer where she found it. She could have just opened it first and saved herself almost all of the 10 minutes.

It was a learning experience: always look in the place where the thing is. It saves a lot of time. Her problem was that she first checked all the places it wasn't.

I bet the manufacturing cost is the largely relevant factor once the patents expire.

Which from the perspective of generic manufacturers, the drugs are discovered and proved effective by the Magic Drug Fairy.

Not particularly, but there really isn't much difference between a generic drug plant and a branded factory. Same machines, same kinds of people, basically all the same.

Yes, that is why once regulatory exclusivity is removed, the market price tends to be based more or less on the manufacturing cost.

It’s always the last place you look, because you’d need to have lost your mind to keep looking after you’d found it.

This is true even if you find it the first place you look.

A big part of the issue is too readily discounting or assuming where it isn’t too early. It’s often better to methodically search sections and areas at a time in a thorough process of elimination even of places you know it isn’t, just as a matter of practicality, and so the process of elimination is actually fully eliminating possibilities, and not just confirming biases.

> if making new drugs was so cheap and easy, drug companies would be doing that in competition with each other

They are. There is intense competition and lots of venture capital and research grants to fund research.

> prices would be low as a result of that competition.

Why? Patents end competition as soon as the research is successful. They have to compete to research the drug - and we see a vibrant competitive market for research - but they don't have to compete to sell the drug. We see this even when there are multiple brands competing to sell generic versions of drugs they didn't discover or spend R&D on.

There is simply no market forces to lower medically necessary drug prices. The market is captive (people will die without certain medication), and thanks to patents, there is often no competition for years to allow people to price shop.

I can buy ibuprofen or acetaminophen or aspirin for cheap at any drug store, because there is a competitive market for OTC pain meds. There is no OTC market for insulin, and people will die without it. Unsurprisingly, insulin is much more expensive, despite being discovered by a university and being over 100 years old.

PreP (prevents HIV) research was entirely funded by US taxes, and the pills costs $2000/mo - despite being available from multiple companies including available as a generic. Why are companies that didn't pay to research the drug (and didn't take US Gov money to research it) charging $2000/mo? The only cost they incurred was manufacturing.

What if your friend visited your house, and on the way inside, you told them "if you're ever looking for a doodad, it is in the third drawer on the left". What's a reasonable amount of time for your friend to spend looking for a doodad, and how would you treat a friend that asked for an hour to retrieve it?

I assume this is all an analogy for the cost of research, and how not every research opportunity pans out?

I think it's a cute analogy, but we have plenty of drugs that are decades old, research costs long paid off, and people die because they can't afford them. This isn't an economics 101 essay prompt, the costs of medicine have life-or-death implications.

As the go-to example, insulin was discovered in a university more than 100 years ago. The patent was famously sold for $1, so the research "cost" to the drug companies is negligible. The price didn't decrease in the US until 2023, when the government capped the price at $35/mo - an 70+% reduction. The cost in other countries is still less than 30% of the US price. This is basically the equivalent to someone telling your wife where the doodad was, and her pretending it takes 10 minutes to find.

PreP, which prevents HIV, was entirely funded by the US Government - both research and manufacturing costs - and its manufactured by several companies who weren't involved in the research. The average prescription cost was >2000/mo from every source until the government stepped in. That's more than some people's monthly salary. This is basically like someone handing the doodad to your wife, and her pretending it is a full-time job to find.

Just so I know, is the insulin thing rhetoric or do you not know what happened with insulin? If it's rhetoric, I'm happy to let you have the floor. I'm not one really to kick folks off the pulpit.

> Why? Patents end competition as soon as the research is successful.

One can get a patent on a drug, but not on other drugs treating the same condition. We see many similar drugs developed that don't infringe each others' patents. If they are all effective, then the drug makers will be competing with each other, even if the drugs are under patent protection.

Oh, well everything I said is documented and public, so if there's something I'm missing I'd love to be cued in. I think the rest of my point stands, of course.

But I'd love for you to tell me the truth about insulin, because I'm sure that the law(s) that attempted to cap prices had nothing to do with the prices being lowered to that exact amount at the exact same time?

Are you familiar with the history of insulin from porcine to using recombinant DNA technology? I think we’ll be wading in the shallows a little without that. The Wikipedia articles should be quite thorough.

Perhaps a quick self test would be to hypothesize the cost of Novolog before this law and then look up the price in old archives.

I usually do these self-tests to check whether my belief systems are accurate. But if you don’t want to, I’m not one to chastise. I just wouldn’t want to converse.

I don't like conversing with people who spend more effort condescendingly implying they're correct than making a point. But I’m not one to chastise if that's how you like to talk.

Novolog in 2024 is cheaper than almost any point in the history of (that version of) the drug. "Strangely", Novolog and competition (eg. Humalog) increase prices almost in perfect lockstep, for decades, dramatically outpacing inflation, which is not typically how prices look in a competitive market. Decades after the "cost of research" should be priced in. The obvious reason for this, is that the drug is a medical necessity and consumers can't shop around, so typical market dynamics don't work, allowing companies to raise prices without repercussions. Furthermore, drug companies voluntarily stopping production of cheaper, older, off-patent insulin products - not a single one has been stopped for safety reasons, and clinic benefits of new products since the 1921 patent are minimal.

In the 2020s, US politicians started publicly criticizing drug companies for their prices of insulin. Then they introduced a slew of bills intending to cap the price at $35/mo - a 70% reduction. Several bills failed to be passed, but one managed to pass. In that time, all major insulin manufacturers announced they were "voluntarily" reducing their prices to $35/mo (or lower) - set to take affect the same time as the above mentioned law.

Not sure how you think the history of insulin technology really relates to this, other than the 1982 recombinant DNA insulin was promised to be cheaper? It doesn't work meaningfully better for most patients. When recombinant DNA insulin ("human insulin") was first sold in the US in 1982, it was priced at $0.5/day (1.5 in 2024), while traditional "animal" insulin was closer to $0.3/day (0.95 in 2024). It was special because it promised unlimited supply, with minimal risk of shortages. Eli Lilly said, in 1982, that they expect the price to drop over time. The first time that it was cheaper (inflation adjusted) was 2024 when the new US law went into affect. Oh, this only applies to the US, elsewhere the prices have been low and dropping consistently, even from Eli Lilly.

What's truly damning, is that the clinical benefits of the insulin invented since that first University of Toronto $1 patent version is pretty minimal compared to the UT version from 1921. You can make snide comments of "wading in the shallows" discussing new technology, but the original patent would still have provided comparable medical outcomes to the millions of people who need insulin in 2024.

Is it possible you're defending the actions of a giant corporation that raised the prices of a medically necessary drug out of greed, and now you're re-writing history for them to justify your defense? Maybe you need to devise a new self-test to check your belief systems.

Citations:

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10718803/

Inflation Calculator: https://www.minneapolisfed.org/about-us/monetary-policy/infl...

Archival News from 1982 and

https://www.washingtonpost.com/archive/politics/1982/10/30/f...

https://www.nytimes.com/1982/10/30/us/a-new-insulin-given-ap...

Generic insulin has been available since 2019.

https://investor.lilly.com/news-releases/news-release-detail...

Since long before that.

It depends on exactly what you mean by "insulin".