I don't think that he can legally prescribe it, but almost every man with
PCa has some arthritis.

Prescribing drugs "off-label", meaning for conditions that are
not approved by the FDA, is perfectly legal.  Whether it is
desirable or represents good medicine, is another story.

According to:
http://www.nlm.nih.gov/medlineplus/news/fullstory_33273.html
a study in 2001 looked at 160 prescription drugs and found 21% of
the prescriptions written for them were off-label.  Some other
estimates are higher than that.

Drugs may be prescribed off label for any of the following reasons:

1. There is evidence of the drug's efficacy for the particular
  off-label use, but it was never submitted to the FDA, which
  requires two rigorously conducted clinical trials, with
  expensive operating and reporting costs.

2.a. There is no "on-label" drug for the condition.  The doctor
  is trying to do something for the patient and the only option
  is something that is, in a real sense, experimental.  The
  doctor either reasons that this may help, or has heard from
  others that it may help, and feels that he has try to do
  something.

2.b. On-label drugs haven't worked.  The doctor tried other drugs
  without success, presumably including the drugs that are
  recommended for this condition.  He wants to help.  So he
  tries something else.

3. The patient demands the drug.  This actually occurs a lot.
  Patients see ads on TV for some drug for some condition.  The
  doctor says, "We don't have any evidence that you have that
  condition, there's no reason to think this will help you."
  But the patient says, "Well I think it might help and I want
  it and if you won't give it to me I'll find another doctor who
  will." Doctors get tired of hearing that and often just give
  in.

The first two reasons seem to be ethical as well as legal.
However the doctor really needs to read the research on the drug
and have at least some good reason to believe the drug will do
more good than harm.  I think it is unethical if the doctor
hasn't read up on the drug and satisfied himself that it is
unlikely to harm his patient and may do some good.

The third reason seems to me to be highly questionable.  I
understand why it happens.  Doctors have a great many ignorant
and/or unreasonable patients.  Most people in the U.S. and
elsewhere have only a hazy idea of what science is even about.
[I have even met a doctor like that.]

On the face of it, rising PSA after failed primary treatment
certainly falls into my category "2a" or "2b".  If ADT is also
failing, or even if it's not, we don't have a good treatment that
will prevent death in the long run.  If there's evidence for
off-label use of drugs, I think I'd be inclined to go for it.  If
I were a doctor I might be willing to prescribe it.

As for Celebrex or statins for PCa, I suggest that anyone
considering this should look it up in Pubmed and anyplace else
you can find.  Read the research.  If it looks positive, print it
out and take it to your doctor to discuss it.  Don't just print
out the postive reports.  If you find a credible negative report,
print that too.

This won't always be of much use.  Some doctors won't read
printouts handed to them by a patient.  Some won't read research
at all - they just do what they've always done.  But it is
definitely a good idea to discuss any off label use with a
doctor, listen to what he says, and not make demands if he gives
reasonable arguments against doing what you want to do.

   Alan