About studies over 20 years with human patients: http://apache2.pum.edu.pl/~fasting/bur_ab.doc

There are a few small-scale case studies done on human subjects, but they're fairly recent. I was just reading one late last year on ~10 subjects. The experimental group (5-6 patients) had equal or better response to chemo and experienced fewer side effects than the control.

For obvious ethical reasons, though, I think they can only do this with patients with a terminal prognosis. Nevertheless, in my experience, you'd get a lot of volunteers from anybody undergoing chemo.

Results will be available in ~90 years.