Finally, here it is. Sorry it took so damn long. Between being busy and needing some downtime, I confess I lacked not just time but motivation. Because depression.
Anyway. Ketamine's anti-depressant effect has been known for decades now, but the first academic study about it came only in 2000 (first file), and confirmed the anecdotal evidence, deeming that K's anti-depressant effect worked by "antagonizing NMDAR", whatever that means. This has one interesting implication that I'll elaborate on later. From there on there were a lot of K studies on animals and people, from which I got that stat that it works on 70-80% of people with treatment resistant depression. Notice that this number is probably quite higher if you were to include the remainder of MDD cases. You can find these studies with Google and sites like Pubmed. Also drug communities are full of anecdotes of all sorts. I recommend Drugs-Forum.com, Bluelight.org and Reddit (search with Google's site: function, it's far better than Reddit's shit search engine).
Someone could self-medicate with K, but of course, there is the one big problem of risking a trip to the American gulag when buying it, and another two are possible minor brain damage (like a lot of drugs) and guaranteed bladder damage. Reports on the latter are conflicting; just about everyone says only people consuming massive amounts develop noticeable bladder problems, one person I know claims he started to notice them even with small, therapeutic doses. Speaking of which, the anti-depressant dose is thankfully much lower than the recreative one, so you wouldn't need to trip balls if you don't want to. This also means cost isn't really a problem.
In 2012, a group of scholars did unpublished research that indicated the metabolite HNK was the actual antidepressant, by "increasing the function of mTOR". This reinforces the "interesting thing" I mentioned before, I'll get to it in a minute. Anyway, in 2013 (second file) they patented the use of HNK and one of its metabolites to fight depression and neuropathic pain (capitalism!) before finally releasing their findings in 2014 (third file). From them on, there have been some studies of direct administration of HNK in animals, but none in humans yet as far as I know.
By now, it's already known that HNK has no recreative effects and can cause no brain damage, but the jury is still out on bladder damage. The good news is, being a K metabolite, direct use of HNK can't possibly cause more bladder damage than a K dose that would produce the same amount of HNK as the direct use dose. By the same token, it should have a success rate at least the same as K's, possibly higher, as we don't know yet why some people don't respond to K; if it's because their bodies can't create HNK, then direct HNK doses might work at 100% success rate, but if their bodies do make HNK but still they don't respond, then at least we still have the 70-80% baseline. One of the drawbacks, depending on the person, is that you can't trip balls. But the real drawback is that, despite a synthesis reaction being known since 1986, no one manufactures the fucking thing because it was thought to be completely useless.
Which brings to possible ways to acquire HNK. There are laboratories that provide a service called custom synthesis, which is exactly what it sounds like. It costs an arm and a leg – US$5000-7000 quotes are common – because there's no known economical, efficient way to synthetize your substance (if there were, it means it would already be produced and thus you wouldn't need custom synthesis in the first place, would you?), so they essentially have to create an industrial process from scratch. This means that there are several risks. It usually takes a few months, you might get a lot of impurities in your shit, you might get a minuscule amount of the desired substance amidst a lot of impurities, you might get a completely bunk result; 10g is a treasure trove, 50g is winning the lottery, but I think the average is 1-5g. Mercifully, this allows for a lot of HNK doses, which probably aren't higher than 15mg. As you might be guessing, these labs don't offer refunds, but some offer purity checks or resynthesis or other niceties. Also, further batches of the same substance from the same lab should be increasingly cheaper, as they refine the process, and being a loyal customer might help too. Unsurprisingly, labs in developed countries tend to be better but more expensive, and the reverse for labs in poorer countrie. Emphasis on China, and to a lesser degree India, whose quality varies absurdly. You might get the best and cheapest service in the world, or you might just plain get scammed. Eastern Europe seems to be a good compromise of price and quality. Thanks, Gorbachev.