A lot of (possibly most of?) the money spent treating cancer is spent in cases that have very little chance for long term success. I don't think it really is, but it almost seems like a deliberate attempt to extract all the wealth from the patient before he dies.
Medical services are an inelastic good. If someone has a problem that's a serious threat to life or happiness, they will spend every penny they have and some they don't on the slightest chance to avoid it. Competition, for various reasons[2], doesn't work well.
[1] In some cases where the intervention has slight mean/median benefit. In other cases, the outcomes are no better than control/placebo, even on average, but studies funded by drug companies or medical suppliers suggest that the intervention outcomes are better than controls, due to scientific mistakes or even intentional fraud (starting with bad experimental design, then cherry-picking data and cherry-picking studies).
[2] Comparison shopping works well for choosing suppliers of standardized drugs and services, in non-emergencies, but even that's usually torpedoed by what your insurance will cover; If you have the luxury, you can compare various brands of drugs that have generic versions. You can comparison shop for prices for blood draws and scans (CT/MRI). Beyond simple things like that, the treatments are not identical so comparing costs is difficult. Also, patents. Nobody will raise the price of a drug 100% in a day, but companies build in fatter and fatter margins because they have a patent monopoly, and everyone else is doing it to. There is no price control through competition except where there are genuine equivalent goods. Since healthcare is largely price-inelastic, without strong competition on the supply side, you get runaway costs. This shouldn't be a surprise.
Citation needed.
But really, in this case shouldn't we ask if the 42% who loose their life savings had an easily curable cancer or not?
It's the rich that spend most money, and they don't manager to spend everything.