New medicines cost tonnes, take ages, and there are hardly any of them

I was reading through the literature on technology and came across some papers on research and development of new medicines. I had a general idea that they were expensive, but the papers here and here present data that has new medicines produced in the United States costing almost a billion dollars each, taking over a decade, and numbering only a few dozen per year.

Many anti-malarial medicines are chemically similar to quinine, and it is feasible that new medicines could be found by changing chemical structures by relatively small amounts. In this case, it would be easy to produce new medicines based on old ones. If such small modifications account for many of the new medicines, then the number of genuinely new medicine pathways is tiny. Some new medicines are derived from plants, so the rate of finding new laboratory-derived pathways seems tiny.