There are several papers that use alkaline hydrolysis to delipidate or "detoxify" the endotoxin for use as a vaccine (1, 2, 3). The point of many of these papers is to develop a detoxified LPS (dLPS) vaccine to protect against sepsis in the pathogenesis and treatment of gram-negative infections. The detoxification is important, as the lipid-A portion of the LPS contributes significantly to sepsis. In this case you would be targeting the O-antigen or core oligosaccharide.
So one of the problems that I can see is that there are many bacterial serotypes, comprised of many different O- and core groups. I'm not aware of any vaccines with that sort of breadth. Issue two is you get an infection and you need to administer these vaccines quite fast, if not prophylactically, as adaptive immunity takes a number of days to come to speed.
It's also not clear to me whether a prophylactic vaccine to E. coli is useful to the populace at large, epidemiologically (just as an example).
The field also seems to lack clinical data on tolerability and efficacy, which is what I'd be more interested in, e.g. is it safe for humans and does it work?