There is a mechanism to get your hands on an experimental drug ("compassionate use" in America and something like "named person use" (can't remember the term of art exactly) in Europe), in which a person with no other options's doctor can request a specific experimental drug. This reasonable sounding alternative is less cut-and-dry as it might seem on face, because the thing you really, really don't want in evaluating a drug's effectiveness is anecdotal evidence.

E.g., Patient A takes miracledrug and get better against all odds. Patient B hears about patient A's experience on the internet and also demands miracledrug. Patient B dies horribly because the drug is actually quite toxic and Patient A would have gotten better anyway but falsely attributed their recovery to miracledrug. Something like that. Or on the other side Patient A dies and all of a sudden miracledrug's manufacturer has a PR storm on their hands that threatens to sink the drug even though Patient A would have died anyway and miracledrug played no role.

There are also many issues relating to informed consent that are too long and boring to write about. For these reasons compassionate use is an ethical and business minefield that the FDA and drug companies shy away from. One patient isn't a study, and the efficacy and safety of any drug should always be evaluated within the confines of an appropriately powered double blind clinical study unless there's a really compelling reason not to do that.

To your other point, I think that biology as a school subject is really boring because of how it's taught, but it's taught that way because that's how many biologists think. I think that biologists think that way, because the type of people who are attracted to that style of learning are boring people (rote memorization). I and mk and our mutual mentor were all trained in physics, so that's how we look at the world. We fortunately have the ability to ask "why not?"