Contemporary Management of Prostate Cancer With Lethal Potential

The article by Davis et al is important for several reasons. First, it reminds us about the most lethal phenotype in patients with apparently localized prostate cancer. This subgroup is easily forgotten in today's era of PSA screening because the majority of patients now diagnosed with prostate cancer are classified as low risk. Second, there have been few, if any, good reviews that define the issues, including the definition of high-risk disease, the effectiveness of the major treatments (ie, radical prostatectomy, radiation therapy, and their neoadjuvant or adjuvant supplemental therapies), and the current gaps in our knowledge of these issues. Defining High-Risk Disease
The authors nicely review the major factors that most experts identify as important in definitions of high-risk disease in the setting of apparently localized prostate cancer, when treated either by radical prostatectomy or radiotherapy. As a start, a Gleason score > 7, prostate-specific antigen (PSA) levels > 20 ng/mL, and clinical stage T2b define high risk as it relates to pathologic stage (eg, Partin tables[1]) or PSA recurrence (eg, D'Amico tables[ 2]), with the latter probably being more important to both the patient and physician. This high-risk category as a whole yields PSA recurrence rates averaging about 40%. Yet, this group is obviously heterogeneous, and the risk of PSA recurrence is best viewed as a multiparameter gradient. Thus, nomograms such as those developed by Kattan, Scardino, and associates provide a better definition.[3] This article also reminds us that including biopsy characteristics (eg, percentage of cores positive or percentage of cancer within the cores) can further delineate risk, and these factors are gradually being added as the nomograms and tables are revised. Other groups have incorporated many additional routine parameters (eg, age) into analytic tools such as neural networks to further clarify and individualize risk.[4] Finally, many groups using a variety of molecular techniques such as cDNA microassays are discovering various distinguishing tissue markers that may add prognostic power when incorporated into existing nomograms.[5] I have no doubt that soon we will be able to much more sharply delineate various groups at high risk for PSA recurrence. There are two important caveats to consider when using these predictive tools. One is that PSA recurrence is probably a good surrogate marker for cure but not a good marker for cancer- specific survival. These predictive tools will be more useful when survival can become the end point in the datasets used for these nomograms. Second, a patient's PSA outcome is not a percentage; it is more of a "yes or no" parameter. As such, in my opinion, these predictive tools are not especially helpful in deciding the course of therapy, although they are extremely useful in designing clinical trials and in interpreting results. These tools are also helpful to the treating physician in executing therapy for the individual patient. For example, in patients who are truly at high risk, I perform wider lymphadenectomies and local resections. Radical Prostatectomy for High-Risk Disease
This article nicely reviews the many studies detailing the fate of high-risk patients treated with radical prostatectomy. Because they are all retrospective and mostly single-institution efforts, the results are somewhat confusing. As expected, once the pathologic stage is known after surgery, a variety of pathologic parameters (eg, seminal vesicle invasion, lymph node status, margin, and extracapsular status) increase predictability. Yet in the PSA era, when the disease is presumably detected early, significant numbers of patients can be cured. Thus, while a minority of patients with Gleason scores of 8 or above have organconfined disease, the cure rate in this subgroup is high. Of course, cure declines as more adverse high-risk parameters are added, and large numbers of high-risk patients need to be considered for neoadjuvant or androgen deprivation therapies. As the authors point out, retrospective studies convincingly demonstrate that local control is improved with adjuvant radiation therapy. However, an overall survival benefit is still questionable and must await the results of the phase III Southwest Oncology Group trial that is now closed but still blinded. Salvage radiation therapy also contributes to long-term biochemical disease-free status, if not cure. A recent report[6] convincingly showed that in patients with good characteristics (eg, preradiotherapy PSA < 2 ng/mL, Gleason score < 8, PSA doubling time > 10 months), the freedom from biochemical recurrence rate at 3 years was 50% to 60%. Even among patients at high risk (eg, Gleason score > 8), long-term PSA suppression was an acceptable 20% to 30%. I use adjuvant radiation therapy when I am sure that residual disease remains, but wait and use salvage radiation therapy when I am not sure (but start it early after PSA recurrence is realized). The data on salvage radiotherapy show that local persistent disease may well be a more important factor in the metastatic process than has been assumed. Androgen deprivation therapy is important in the management of prostate cancer. However, as neoadjuvant therapy after radical prostatectomy, this strategy does not seem to have great value so far. In patients with positive nodes after surgery, immediate androgen deprivation therapy seems to increase survival when used as continuous therapy, and I use this supplement often in that context. I also use it as intermittent therapy if the patient prefers, after I emphasize that this approach is not yet proven to be as good as continuous therapy. Radiation Therapy in High-Risk Disease
The authors provide a very good detailed review of the various radiotherapy approaches and their effect on patients with high-risk parameters. Without dwelling on the details, several salient factors are worth highlighting. For example, dose escalation to > 70 Gy is beneficial especially when coupled with modern focusing technologies (eg, intensity-modulated radiation therapy). Also, whole pelvic irradiation appears to be beneficial in patients at high risk for positive nodes, although I would like to know more about its long-term morbidities. Brachytherapy can be used along with external-beam irradiation in high-risk patients, but I have yet to be convinced that this combination is better or more convenient than externalbeam irradiation alone. Finally, many studies have indicated that androgen deprivation therapy plus radiotherapy is superior to radiotherapy alone in high-risk patients, especially when the adjuvant deprivation therapy is administered before and also after irradiation for at least 2 years. In addition, a trial of androgen deprivation therapy vs androgen deprivation therapy and radiotherapy is under way, and I believe radiotherapy will prove additive to androgen deprivation therapy for survival in high-risk disease. Of course, it would be particularly interesting to see the results of a trial comparing radiotherapy plus androgen deprivation therapy vs radical prostatectomy plus androgen deprivation therapy and adjuvant radiotherapy as needed. Regrettably, the collective will to conduct such a study does not exist in the United States. Adjuvant Chemotherapy
This article ends with brief descriptions of phase II adjuvant chemother- apy trials. The results of these trials will be important to watch; hopefully, as in breast cancer, the current chemotherapies that produce a modest benefit in advanced disease will be much more efficacious when used in an adjuvant setting for high-risk localized prostate cancer. Neoadjuvant chemotherapy protocols for high-risk patients before radical prostatectomy are also under way at many centers and, in addition to the clinical results, may provide important clues about benefit through routine and molecular analysis of tissue before and after surgical therapy. Conclusions
In summary, this excellent article leaves us with several important facts. The current definitions of highrisk disease in the setting of apparently localized prostate cancer are useful for analysis of therapies but soon will be more accurately defined in a gradient fashion. Radical prostatectomy achieves cure with significant frequency, and this cure rate may be increased when accompanied by radiotherapy as adjuvant or salvage therapy. Permanent androgen deprivation therapy after radical prostatectomy seems to produce a survival benefit in men with positive lymph nodes. Modern radiation therapy given at adequate doses, especially when coupled with adjuvant androgen deprivation therapy, produces cure rates similar to those seen with radical prostatectomy. Nevertheless, the failure rates with all these treatments are high, and newer agents or newer combinations of established agents (eg, chemotherapy) need to be tested. This then is the central message of the article by Davis et al: We must make it a high priority to design, organize, and execute controlled clinical trials in this most important class of patients because they probably represent the largest group that suffers prostate cancer mortality.