This is an extremely controversial figure within the pharmacoeconomic literature.
It comes from a publication by the Tufts centre for drug development, who are heavily pharma-funded, and it is repeated verbatim by those with vested interests in the pharma industry.
More on this here: https://www.washingtonpost.com/news/wonk/wp/2014/11/18/does-...
The problem is that there's a big range, which mostly reflects different routes of drug discovery. To take two contrasting examples:
1. A target and potential molecule are discovered and developed in an academic lab, following which a small lean biotech is 'spun out' to develop the molecule further. This biotech accesses funding, and spends an amount of money --usually in the low hundred-millions, but sometimes less-- running the studies and collecting the data necessary to approach approval. With luck and a following wind, they succeed, and their overall cost is low for the industry. (See the lower end of the table in [1]). If they fail, the biotech sinks without a trace.
2. A large pharmaceutical company need to continue developing new drugs to maintain its existence. It has a drug discovery operation or two, trying to find new targets and develop new drugs for them. There's significant attrition of these molecules here. Any promising molecules are taken into early-phase trials, where further significant attrition of these molecules takes place. Finally, if those studies are successful, the molecule heads to expensive phase III trials. Here, it's worth noting that the big players in pharma will have a combined R&D budget (encompassing all that we've mentioned here, plus some more) of multiple billions per year.
So for small single-molecule biotechs... Firstly the low end of the estimated development cost range ignores a lot of the discovery costs as they were borne by the academic lab. Secondly, the low estimates don't account for failure - as failed biotechs wouldn't be included in the estimates.
In contrast, the estimates for big Pharma do include their discovery costs, and can (in a decent analysis) also include the huge cost of multiple drug failures along the way to getting one drug approved.
Have a look at the linked analysis [1]. Rather than trying to estimate the cost of a single drug, it looks at the overall money spent on R&D versus new drug approvals in that time. (The methodology isn't perfect; but overall, it's probably a better picture of the cost of drug development across the board than a single-molecule estimate.)
[1] https://www.forbes.com/sites/matthewherper/2013/08/11/the-co...
Don't they spend a good deal on simply trying to make a drug to work around someone else's IP, rather than making drugs where there is a new benefit -- that might account for most of the R&D for all I can tell, it doesn't appear to be included in this analysis?
Marketing is essential for recovering development costs for drugs the same as it is for software. If marketing did not pay for itself and contribute positive revenue, nobody would do it. Drug companies have a finite amount of time to recover their very high development costs.
https://www.washingtonpost.com/news/wonk/wp/2015/02/11/big-p...
HBM did an analysis of novel FDA approved drugs from 2008-2018 which showed over 60% of FDA approved drugs were originally developed by small companies, and just 12% iirc by big pharma
And most drug development efforts these days focus on medicines that can be truly impactful to patients. There is a correlation between clinical value and price, and many markets are winner take most, and winners are chosen by prescribers and payers based on their benefits to patients
One data point to illustrate the industry's increased focus on innovation is that historically, nearly all drug development programs were small molecules and biologics -- decades-old tech. Of the companies that IPOd in the last 5 quarters, there are more gene and cell therapy programs than small molecule
Could you share the link to this article, please?
Mostly, yes. (This data[1] is from 2015, but likely not too far off what's happening today.).
> Don't they spend a good deal on simply trying to make a drug to work around someone else's IP, rather than making drugs where there is a new benefit -- that might account for most of the R&D for all I can tell, it doesn't appear to be included in this analysis?
Honestly, not so much. The big money these days is in drugs which offer a transformational benefit over the current standard of care - whether that SoC is another marketed drug (either from the same company, or another) or a not very good treatment (or treatments) that no-one has improved upon yet.
[1] https://www.washingtonpost.com/news/wonk/wp/2015/02/11/big-p...
- startups that license potential molecules from academia often do a lot of discovery work. Academics generally do some of the target validation / biology, but dont usually produce optimized drug-like molecules. Most of the expense of drug discovery is actually screening compounds against a validated target (and validating the target), filtering those for potential leads, and then extensive chemistry work to get a development candidate). In the OP, drug discovery is target-to-hit + hit-to-lead + lead optimization. Most of this is done by startups, not academia
- the cost of capital is higher for startups than big pharma. big pharma companies generally have cost of capital around 10-12%, whereas for VC backed startups, cost of capital is like 20%+. this has a huge effect on valuation, and on the price companies must eventually charge for drugs to generate a return that attracts investors
Part of the reason I made this tool is so people could input their own assumptions rather than just relying on somewhat opaque studies. I personally was surprised to see how big the effect of discount rate was -- it made me reconsider my position on whether more drug dev should be publicly funded. I think the effect of cost of capital on drug price is under-discussed in some of these studies