>>if a drug is safe, it doesn't have to have any indication of efficacy in order to get the go-ahead for a phIII trial from the FDA?
Pretty much. I mean if there was not much evidence of efficacy, the FDA would certainly grumble about it and recommend that the company do some more work on identifying a good population or dosage, but the company doesn't have to follow the FDA guidance on that.
For a good example, read about Burzynski and his "antineoplaston" therapy. Basically a money-making scam, but the FDA allowed him to open a Phase III - I believe a hold was placed on it though before any patients were admitted because of a death in a patient under his care.