There are economic forces at play that actively discourage pharma from trying to identify who will benefit most from the drug. First of all, there is a huge pri

There are economic forces at play that actively discourage pharma from trying to identify who will benefit most from the drug. First of all, there is a huge priority on being first to market. They stand to make hundreds of millions if their drug is approved before their competitors, and they can be the sole option for even 6 momths. Doing prolonged preclinical research to delay registration trials and FDA approval makes no sense to them. Secondly, the fact is that they make more money if we just give the drug to everyone. It takes 3 doses or so before doctors can work out if a drug works or not. So if we give 100% of cancer patients at least 3 doses, they profit more than if we give it to just the 30 - 60% that could benefit the most. So although you are right that encouraging or forcing companies to come up with companion biomarkers that predict efficacy is a good way to reduce cost, in reality it is not so simple. The hype about these latest immunotherapy drugs is so massive (and much of it is justified) that pharma has everyone over the barrel, they will charge the earth and get away with it.