> If there is no therapy that is known to be effective for a particular illness, then placebo control remains appropriate.
For cancer there's almost always some kind of combination that gives some results (even if not great). So your statement proves what I said.
> Use of placebo controls may be justified in these situations:
> To prove effectiveness of a new treatment for diseases with high placebo response rates
> In conditions that alternately become worse or better, have spontaneous remissions (the disappearance of the signs and symptoms of cancer, but not necessarily the entire disease), or have an uncertain and unpredictable course
> When existing therapies are minimally effective or have serious side effects
> In the absence of any effective therapy
