One of the foundational principles I use to reason is "if there was a free-lunch (i.e. a higher-ROI solution—more reward for less risk), then anyone truly motivated to solve the problem would have switched to it, or at least added it to their offering as an extra hail-mary on top of their current approach. Given that they didn't, the free lunch must either not be free (i.e. no less risky), or not be much of a lunch (i.e. not rewarding.)"
Yes, in this case, you've got a system with mixed motivations; the FDA isn't motivated to solve problems, so there might be some beneficial drug that isn't on the market. But that doesn't mean that your doctor in particular—someone who is motivated primarily to solve your problem—won't still tell you about such drugs, and help you to acquire them through grey-market channels. Doctors do this all the time—they get patients into drug trials, order in drugs from the medical systems of other countries, acquire new-old stock batches of drugs no longer produced, etc. They also will just tell you to eat a certain food or take a certain supplement as your "prescription", if doing that is a higher-ROI solution than any drug.
If doctors are only offering you treatments with devastating side-effects, that's because those are the highest-ROI treatments that exist. Doctors always prioritize safety over benefit[1]; the only time they offer unsafe treatment is when there is no safe treatment—or at least no safe treatment that pays for the costs of going through it by actually having a non-zero likelihood of helping you.
[1] For example, doctors use known placebo-equivalents as the first-line treatments for many psychiatric conditions, because placebos have no potential risks; so even if the likelihood of remission from placebo is low, as long as it's higher than zero (which it usually is in psychiatric conditions), doctors want to first make sure you aren't someone who would go into remission from placebo, before moving onto trying to treat you with drugs that actually do something (and therefore have risks.)