It's always good to hear about an older compound that may be doing good things that we didn't realize. The current example is metformin, the diabetes drug known to many by its brand name Glucophage, but a generic compound for some years now. Evidence has recently been accumulating that patients taking it over the long term may well have lower incidence of several types of cancer, which is a refreshing change from the usual creeping realizations in this business. (There's a reason for that - the opportunities to mess something up inside a cell, something you probably didn't even know was there, are far, far, more numerous than the opportunities to make something work the way you want).
A new paper may well have tracked down a mechanism for this effect, which adds to the sense that it's real. Here's a summary of the work - it looks like an mTOR-driven process, which is plausible. Specifically, it seems to inhibit the TORC1 pathway, though at least two different mechanisms. That's an important player in nutrient sensing and cellular growth, among a bewildering variety of other things, and the whole mTOR area has been the subject of oncology research for quite a long time now.
Metformin (and related biguanides) might be acting on it in a very useful way. Mice exposed to a known lung-cancer agent were substantially protected by pretreatment with the drug. What's not clear yet is if that direct TORC1 effect is the reason, or if it's a more general downstream effect having to do with metformin's effects on glucose levels and insulin signaling. If it's the latter, there are tumor lines that should (unfortunately) be able to evade the problem, specifically ones that have their PI3K signaling cranked up already, so it's going to be quite interesting to see how metformin does protecting against those. As has been noted many times, nutrient sensing, insulin signaling pathways, carcinogenesis, and mechanisms of aging are tangled together in ways that it's very much in our interest to unravel. (mTOR specifically is right in the middle of it, apparently).
These results (both the new mechanistic study in mice and the retrospective clinical observations) would seem to strongly suggest trying metformin out in patients with a high risk of developing various sorts of cancer. It also suggests that, other things being equal, Type II diabetics might want to use metformin to take advantage of its apparent side benefits. A protective effect would be very welcome news indeed - it's terribly difficult to do anything about most tumors once they've occurred, and the best thing would be for them not to appear in the first place.
Oh, and one more thing. If everyone had followed Sidney Wolfe's advice when metformin first came out - not to use it - we wouldn't have found out about these effects at all. Would we?