Waselenko and Dawson provide a summary of the extensive experience in the management of metastatic prostate cancer. Their article follows a traditional descriptive format and is quite informative. The part that is missing is a general discussion of the various biological aspects involved in the complex process of prostate cancer progression, which has been the focus of major research over the past few years. Undoubtedly, this emerging body of knowledge will provide the background for the design and development of new treatments. There are a few issues, however, that deserve more emphasis.
The introduction of the serum prostate-specific antigen (PSA) test has profoundly affected our ability to diagnose and stage the disease and has influenced treatment patterns rather strongly. Rising serum PSA levels in patients with prostate cancer following local treatment, such as surgery or radiation therapy, is frequently the first and only evidence of disease activity. Despite the facts that other objective evidence of disease activity frequently occurs much later (years) and that survival is usually fairly long, androgen deprivation is commonly initiated in these patients. Following progression to androgen deprivation treatment, which is an expected occurrence in virtually all patients, a rising serum PSA level may, again, be the only evidence of disease activity.
Obviously, one would expect that these patients are both biologically and clinically distinct from those who present with stage D2 disease and subsequently do not respond to conventional androgen deprivation treatment. Presently, there are no organized clinical trials that provide a precise characterization of this group of patients with regard to their clinical course, and there is no consensus regarding the most appropriate therapeutic intervention. Furthermore, because PSA is the only manifestation of disease activity, new drug development becomes a major methodologic challenge.
Most of the review focuses on second-line hormonal approaches and the androgen withdrawal syndrome, despite the fact that first-line treatment with androgen deprivation represents, arguably, the most effective systemic palliative treatment known for epithelial tumors. Much has been learned about prognostic factors in this group of patients. Patients with hormone-naive metastatic prostate cancer who are in more favorable prognostic categories usually demonstrate remarkably durable responses to androgen deprivation treatment, with median survivals of approximately 5 years according to observations derived from large prospectively randomized studies.[3,4] This is in contrast to patients defined as being at high risk, usually those with evidence of more extensive disease, who demonstrate a lower magnitude of response to androgen deprivation treatment and who have substantially shorter progression-free and overall survival.
The authors focus on various aspects of the different methods of applying endocrine treatments in this disease (including intermittent androgen deprivation, sequential endocrine approaches, and, along the same lines, the emerging antiandrogen withdrawal syndromes). While their review of the data on these issues is quite appropriate and informative, there is an equally substantial and important body of data dealing with the critical biological information about the mechanisms of prostate cancer tumor progression. These data could shed further light on the heterogeneity of the D2 population and provide an excellent opportunity for meaningful translational research.
One of the fundamental aspects of the translational approaches is molecular characterization of the various clinically distinct prognostic groups. This may provide important insights into the translation of knowledge about the biology of the disease and help identify pertinent therapeutic targets that could influence choice of treatment. In view of the relatively unchanged pattern of outcome for prostate cancer patients treated with various hormonal approaches over the past 3 to 4 decades, I feel that emphasis should be placed on studying nonhormonal approaches in the early phases of disease progression. It is perfectly justifiable to pursue this avenue prior to the application of second- and third-line hormonal approaches, or even earlier in combination with androgen-deprivation treatment.
Chemotherapy for Hormone-Refractory Disease
The evolving data on chemotherapy in this disease, indeed, suggest evidence of antitumor activity with various regimens. An example are those regimens based on an antimicrotubule mechanistic approach. In view of the fact that most such regimens include one or more hormonal compounds, such as estramustine(Drug information on estramustine) phosphate (Emcyt) and corticosteroids (in combinations containing taxanes), it becomes imperative to define the patient population with regard to prior therapy, extent of disease, method of evaluation, and other factors, because a hormonal effect may confound the evaluation of efficacy.
There is extensive literature on prognostic factors in patients with metastatic prostate cancer[5-7] that highlights the heterogeneity of this patient population. Similarly, there appears to be a growing sentiment that treatment may have a differential effect in subsets of patients with distinct clinical and pathologic features. As an example, I would mention those patients who present with significant soft-tissue involvement, usually have an anaplastic histology, and have a disproportionately low PSA compared to their tumor burden. These patients appear to respond to multidrug chemotherapy regimens that are relatively ineffective in the most conventional patients with metastatic disease predominantly involving the bones. While much controversy still exists regarding the appropriate definition of this clinical entity, it clearly represents a separate subset of patients with a different biology of disease. This suggests that an across the board evaluation of the therapeutic benefits of a given regimen without consideration of subtypes of patients may not appropriately define the efficacy of the regimen.
PSA and Quality of Life As Response Criteria
The authors seem very confident that PSA level is an appropriate surrogate marker in clinical trials in this disease; however, the information thus far, derived from variably sized, mostly heterogeneous data bases, in which clinical and laboratory parameters, including PSA, may predict for survival in multivariate analyses. Results of such analyses have not been consistent.[4,5] Furthermore, and perhaps most importantly, such multivariate analyses should be viewed as hypothesis-generating observations to be tested further in prospective randomized trials specifically designed to address the issue. So far, prospective randomized studies do not validate PSA as a surrogate for response or as a predictor of progression-free and overall survival. In fact, evolving data would appear to refute the preliminary observations.
For example, a large prospective clinical trial comparing bilateral orchiectomy with or without flutamide(Drug information on flutamide) (Eulexin) resulted in a major difference in the incidence of PSA normalization (much more than a 50% decline in serum levels). This was highly significant statistically, and yet there was no difference in clinical benefits or in progression-free and overall survival. Prior to the results of the randomized study, a multivariate analysis originating from the data base of the study population, performed within the context of a landmark analysis, suggested a strong association between PSA response and survival. Similar results were reported with mitoxantrone(Drug information on mitoxantrone) (Novantrone) in the second randomized trial included in Waselenko and Dawsons review. The issue has been confounded further by observations that various drugs are able to affect, independently, various mechanisms of PSA expression, without a proportional change in cell growth.
Finally, the effects of treatment on quality of life have been a major focus of clinical trials in prostate cancer and have resulted in the approval of one new compound, mitoxantrone, for the symptomatic treatment of this disease. The addition of mitoxantrone to prednisone(Drug information on prednisone) led to an improvement in pain control without any convincing evidence of a differential impact on other objective parameters of response and without prolongation of survival. While there is certainly a role for such quality-of-life end points in any disease in which pain and other symptoms can be debilitating and available treatment is insufficiently effective, investigators should be cautious about construing such observations as evidence of conventional antitumor activity and therefore routinely employing this regimen even in asymptomatic patients, in whom evidence of a benefit is lacking at this point.