On the other hand, cerebral adrenoleukodystrophy would be expected to kill a significant fraction of those 67 and rob the rest of their cognitive capacity within a few years.

There are times where tolerating treatment-related mortality is the best alternative available.

Which is why these dangerous experimental treatments are used on patients with these classes of terrible diseases first. If it doesn't work or fails horribly it's not much different than what would happen naturally.

> Also the blood cancer has already killed one of the patients

This isn’t a fair interpretation of the death - the treatment of the drug-linked cancer bears a risk of multiple complications. One of these complications led to a patient’s death.