Oottamasathien and Crawford advance a hypothesis with which I heartily agree-androgen deprivation/antagonist (AD/A) strategies should be considered in many more patients than urologists and oncologists traditionally teach. However, I think the authors could substantially sharpen their message. I would like to make five specific points, and then offer a few nitpicking comments. Redefine Advanced Disease?
The authors suggest that we need to redefine what we consider to be advanced disease. I disagree with this suggestion. Redefining "advanced" disease may be emotionally and subjectively counterproductive-do we really want to tell our patients that their disease is worse? I think not. We desperately need to apply the risk assessment tools currently available and recognize that the risk of prostate- specific antigen failure equals neither the risk of detectable metastatic disease nor the risk of prostate mortality. We must use these tools to define patients in whom randomized, adequately powered studies of AD/A are conducted. Then we must convey the concept of risk and the areas in which we lack knowledge to our patients. It is instructive to remember that our colleagues who care for women with breast cancer did this 40 years ago-systemic adjunctive studies of estrogen deprivation/antagonists (ED/A) as well as cytotoxic agents were begun in the late 1960s in women with breast cancer. Risk assessment to judge entry in these studies employed nodal status, tumor size, menopausal status, and, more recently, biochemical characteristics (estrogen and progesterone(Drug information on progesterone) receptors, HER2/neu). Tens of thousands of women have entered studies comparing ED/A strategies vs placebo, cytotoxic therapy with or without ED/A, and biologic agents (trastuzumab [Herceptin]) and alternative forms of ED/A. What have we done in terms of comparable studies in men with prostate cancer? Very little. Message of the VA Cooperative Group Studies
In my view, the Veterans Affairs (VA) Cooperative Group studies, conducted in patients staged only with plain radiographs, acid phosphatase, and physical exam, and in which crossover to AD/A was unregulated or monitored, represent the first-and perhaps the best-trials we have done in prostate cancer. They are the best because they entered large numbers of men in a randomized setting with state-of-the-art trial designs. Given methodologic improvements that have occurred over the past 40 years, the failure of the VA Cooperative Urological Research Group (VACURG) studies to prove a long-term survival advantage with AD/A strategies should not be assumed to mean that these approaches are of no use. Best Approach to AD/A From the Patient's Perspective
The authors present a logical inconsistency in our current approach. Cassileth et al (Oottamasathien and Crawford's reference 13) suggest that men prefer luteinizing hormone- releasing hormone (LHRH) analogs to orchiectomy (in a study funded by an LHRH analog manufacturer), but Potosky et al (their reference 14) report more sexual dysfunction with LHRH analogs than orchiectomy. This issue should be reassessed, and we should recognize that:

• No substantiated data have yet proven the superiority of intermittent vs continuous AD/A. Therefore, the merit of the reversability of LHRH analog-induced castration is uncertain.


• It takes a considerable amount of time to really discuss surgical castration compared to LHRH analogs.


• There are incentives other than our patient's psychological and sexual well-being that have an impact on the choice of AD/A strategies.

Can Studies of Irradiation ± AD/A Be Challenged for the Lack of an AD/A-Only Control Arm?
I think not. That design would address a different issue than the one at hand. The studies of the European Organization for Research and Treatment of Cancer (EORTC) and Radiation Therapy Oncology Group (RTOG) asked a straightforward and carefully framed question-does adjunctive AD/A improve outcome in high-risk localized prostate cancer? The answer is an unqualified yes. At the recent American Society of Clinical Oncology meeting, Pilepich reported the more mature results of RTOG 85-31.[1] Survival and prostate cancer-specific survival were superior for irradiation (P < .001) with LHRH analog compared to irradiation only. In certain high-risk groups, AD/A improves survival obtained with irradiation. The role of AD/A alone and more precise definition of the risk subgroups who benefit will require other studies. But to this reviewer's eye, the overall answer is clear. Does AD/A Cause Osteoporosis?
The authors assert that "multiple groups argue that men receiving hormonal therapy do not have significantly lower bone mineral densities than age-matched controls" (an assertion supported by a single reference to an abstract). I heartily disagree. To the contrary, there are multiple groups who have reported clear differences in bone mineral density (BMD) in men who have undergone AD/A.[2] The Mayo Clinic group recently reported a substantial increase in fracture risk among castrated men with prostate cancer compared to age-matched controls.[ 3] Intermittent AD/A has been shown to have a lesser impact on BMD than continuous AD/A, and bisphosphonates modulate AD/A-mediated bone loss. AD/A use is clearly associated with an increased risk of osteoporosis, and osteoporosis-at least in women, in whom it has been adequately studied=is associated with increased fracture risk. Hip fractures in elderly patients are associated with a 25% 30-day mortality. We must consider this risk as we decide in whom and for how long to employ adjunctive AD/A. Additional Considerations
Several other points merit attention; the reader is referred to the appropriate literature for a more complete and balanced discussion[4-7]:

• The role of high-dose antiandrogen monotherapy and the blockade of gynecomastia by antiestrogens.


• Why was the outcome of the LHRH analog with or without antiandrogen trial different from that of the orchiectomy with or without antiandrogen trial? Was it the "flare," or was it the fact that the LHRH analog was to be given by the patient daily?

My biggest concern for this entire area of research is that we are performing in a shameful manner, as far as entry of patients in randomized trials of important and well-tolerated adjunctive strategies. We can assess risk. We have well-tolerated and safe adjunctive strategies, but we have barely scratched the surface when it comes to asking and answering the critical questions. We vitally need to initiate the right trials. We should look to the record achieved in the adjunctive therapy of breast cancer, not to mention colorectal and most recently lung cancer, for guidance. As Pogo said, "We has met the enemy....and he is us."