I've not catalogued many actual links, because I've listed my sources before
(e.g., Strum, Mayo, J-H, S-K, Walsh, Harvard Med, NCS, ASCO, the PC books).
But here are some snippets from my many pages of cut'n'pastes. I'm sure none
of them is new to you. I don't have the time to repeat my research to
provide exact sources again. The bottom line for me was that asymptomatic
adjuvant CAB offers a poor theraputic index (little if any benefit over ADT1
at a significant SE cost, plus greater threat of more aggressive
recurrence).

Quick finds include http://www.jco.org/cgi/content/full/22/14/2927 ,
From which this emerges:
Is Combined Androgen Blockade Better Than Castration Alone?
Summary. Three systematic reviews (two with meta-analyses of the
literature14,15 and one meta-analysis of individualized patient data16), one
large RCT,17 and one Markov model (decision-analysis) paper18 were available
to address this question. CAB confers a statistically significant but
questionable clinical improvement in survival over orchiectomy or LHRH
monotherapy. The benefit seems to be limited to patients taking nonsteroidal
antiandrogens and seems to appear only after 5 years of follow-up.
Gastrointestinal and ophthalmologic side effects are worse on combined
androgen blockade. LHRH and nonsteroidal antiandrogens have the highest
estimated quality-adjusted survival but have an incremental cost of over $1
million (US) per quality-adjusted life-year over orchiectomy alone.

For CAB as primary therapy, this link shows some benefit:

http://www.prostateinfo.com/professional/cab/

But this one offsets that with other studies showing no benefit of adjuvant
CAB:

http://www.pcaw.com/newsite/info_articles/articles/archives/tolerability.asp

General ADT guidance from http://www.prostate-help.org/pohtwhen.htm
CONCLUSIONS
We do not believe that controlled clinical trials demonstrate clear
advantages to hormonal therapy in the following scenarios: before RP, after
RP with LN-negative disease or at the time of PSA elevation, adjuvantly with
brachytherapy, or as primary treatment for localized or locally advanced
disease. The evidence supports the use of adjuvant hormonal therapy in
subsets of men undergoing external beam radiotherapy, but only in those with
locally advanced or high-grade tumors, bulky but lowgrade tumors, and LN+
disease. Limited data suggest a possibility of a survival benefit for
adjuvant hormonal therapy in the rare patient with LN+ disease after RP. The
appropriate timing of hormonal therapy for patients with asymptomatic
metastatic disease has not been determined. Much remains to be learned,
including the most effective and best-tolerated androgen deprivation
strategy and the appropriate timing of the institution of such therapy. In
all settings in which androgen deprivation is used, the benefits must be
weighed against the significant costs and toxicities, with particular
consideration given to the expected length of exposure to hormonal therapy.

And that's for ADT1, let alone the more traumatic ADT3.

From ASCO: Recommendations for the Initial Hormonal Management of
Androgen-Sensitive Metastatic, Recurrent, or Progressive Prostate Cancer

They define Early deprivation therapy as initiation of ADT at diagnosis of
metastatic or recurrent/progressive prostate cancer, deferred deprivation
therapy as withholding ADT until the presence of clinical signs or symptoms.
Overall survival (OS) was the primary outcome of interest; QOL hardly
considered.

Loblaw, Mendelson, Talcott, et.al.

PURPOSE: To develop a clinical practice guideline for the use, combinations,
and timing of various forms of ADT for the palliation of androgen-sensitive
metastatic, recurrent, or progressive prostate carcinoma. (Attributes of
good guidelines include validity, reliability, reproducibility, clinical
applicability, clinical flexibility, clarity, multidisciplinary process,
review of evidence, and documentation.)

METHODS: An expert panel and writing committee systematically reviewed the
literature, which included a search of online databases, bibliographic
review, and consultation with content experts. A priori criteria were used
to select studies for analysis and study authors were contacted when
necessary.

RESULTS: There were 10 randomized controlled trials, six systematic reviews,
and one Markov model available to inform the guidelines.

CONCLUSION: A full discussion between practitioner and patient should occur
to determine which therapy is best for the patient. Bilateral orchiectomy or
luteinizing hormone releasing hormone agonists are the recommended initial
treatments. Nonsteroidal antiandrogen (e.g., Casodex) therapy may be
discussed as an alternative, but steroidal antiandrogens (e.g., Cyproterone
acetate, aka Androcur) should not be offered as monotherapy. Patients
willing to accept the increased toxicity of combined androgen blockage for a
small benefit in survival should be offered nonsteroidal antiandrogen in
addition to castrate therapy. Until data from studies using modern medical
diagnostic/biochemical tests and standardized follow-up schedules become
available, no specific recommendations can be issued regarding the question
of early versus deferred ADT. A discussion about the pros and cons of early
versus deferred ADT should occur.

For patients with documented metastatic disease, or whose clinical
parameters suggest too small a chance for cure to justify the toxicity of
extirpative therapy, systemic ADT through surgical or pharmacologic
castration, antiandrogen therapy, or a combination, is the standard
first-line treatment. The goal of ADT is palliation. However, there are
considerable practice variations in the use of ADT in these situations. The
possible explanations are numerous and include: the effectiveness of ADT in
suppressing the PSA; the palliative nature of ADT; its cost and toxicity;
the prolonged treatment required (further extended by PSA surveillance for
biochemical recurrence); the highly variable time lag between initial PSA
rise and symptomatic metastatic disease; and the sometimes conflicting
results of clinical trials. Because the relative efficacy of alternative
approaches to ADT appears small and its toxicity is substantial, patients
may weigh the balance between the favorable and adverse consequences of
palliative ADT differently. Therefore, shared decision-making between
patients and their physicians is necessary for optimal use of ADT.

Metastudy: early ADT for treatment of advanced PC reduces disease
progression and complications. Early ADT MAY provide a small but
statistically significant improvement in overall survival at 10 years. There
was no statistically significant difference in PC-specific survival. But
there were more frequent treatment-related adverse effects with early
therapy. Additional studies are required to evaluate more definitively the
efficacy and adverse effects of early versus delayed androgen suppression in
men with prostate cancer. In particular trials should evaluate patients with
advanced prostate cancer diagnosed by PSA testing and men with persistent or
rising PSA levels following treatment options (e.g. radical prostatectomy,
radiation therapy or observation) for clinically localized disease.
Antiandrogen monotherapy is not recommended. Patients should be followed
clinically and started on ADT once symptoms of locally progressive or
metastatic disease present.

I.P.