I can't muster too much surprise that our brave, efficient, world of private medical testing, where IRBs are responsible for serving the customer by approving the study, would fall egregiously flat on ethics.
I am somewhat surprised that these sorts of studies have been allowed to pass as evidence for drug approval purposes. Yes, poor powerless homeless people are cheap and unlikely to make a fuss; but that's a study population almost certainly loaded with confounding preexisting conditions of various sorts, a relatively difficult one to observe over the long term, and so on.
At least from what I've been told by friends involved in the area, you couldn't get away with doing a mouse study if your test population was recruited at random from whatever wild types you managed to trap in the building the biology department is housed in. You'd think that people would want their drugs tested more systematically.