However, when you spend a bit of time thinking and discussing the problems you realize that there are no easy solutions. It is daunting to try to find and safely implement really useful therapies. We may blame drug companies, the FDA or conservative doctors. All of those problems may be there but under all that the process is complex and not something that glib answers will solve.
I'm a cancer researcher in the US, have been involved in the therapy development process, and have authored a publication that led to formation of an NIH consortium to deal with certain aspects of preclinical research - so I'm not being 'glib'. I've had plenty of first-hand, and frustrating, experience with the nature of the drug approval process. I also wasn't referring to alternative medicines, but rather drug approval in general (hence the title).
There is medicine which works and that which doesn't. When something is shown to work it is used.
If only that were true. Unfortunately it is not.
It's not that MDs don't want to help their patients, or that preclinical and clinical researchers aren't coming up with viable therapies - a major hurdle lies between these two in the form of an outdated approval system.
For example, though it is now slowly evolving, the FDA definition of cancer remains by tissue site. "Breast cancer" is considered a single disease, as is "colon cancer", which is absolutely preposterous given the knowledge about the variability of cancer. In other words, "colon cancer" represents hundreds of different pathologies, which can be categorized by underlying molecular defects.
Increasingly, targeted therapies are being developed that can essentially cure a small, predictable subset of tumors. Imagine a drug that can predictably cure 5-10% of lung cancer cases, with absolutely no side effects. Such a drug currently exists, but isn't seeing use beyond clinical trials because it doesn't beat the cure rate for the standard of lung cancer treatment (with sometimes lethal side-effects) when given to a
randomized population of lung tumors. However, the drug was never meant to be a standard of treatment for all lung tumors, only a specific, diagnosable subset of lung tumors - the approval system isn't equipped to deal with this situation, and so advancement of such disease profiling and customized treatment is floundering.
In slightly different vein, therapies are pulled out of trials at the slightest hint of adverse toxicity, even if the current standard of treatment is worse, and even if toxicity susceptibility of the new drug is predictable. In other words, a possible cure is banned if 0.1% of people taking the drug will die, even if a simple test could be produced to predict that 0.1%.
Appeals to the current complexity of the approval system are incorrect, because the actual problem with the approval system is that is too damn simple. It is so simple it is virtually blocking all new cancer therapies from mainstream use, because the nature of the new therapy is more complex than the system to approve it.
It is high time the bureaucracy caught up with the medical science, and there really is no excuse at this point.