Historically, locally advanced prostate cancer has been defined as a tumor that clearly extends outside the prostate as documented on digital rectal examination or imaging. Such patients are at high risk for disease recurrence following surgical or nonsurgical treatment. A number of studies have been conducted that correlate the presurgical physical examination, prostate biopsy tumor grade, and presurgical serum PSA findings with the pathological findings at the time of radical prostatectomy. These studies have revealed that a substantially greater number of patients than previously thought have cancer that is already outside the prostate at diagnosis, although it is not palpable or visible by routine imaging. Such data have been used to develop stratification criteria that attempts to establish a patient's "risk" of disease recurrence within a defined period of time (frequently 5-10 years) following prostate cancer therapy.Through this work, a contemporary method of stratifying patients into "low-risk," "intermediate-risk," and "high-risk" groupsfor tumor recurrence has been achieved. Not surprisingly, the likelihood of disease recurrence between such groups varies dramatically following unimodal therapy (radiation only or surgery only). Patients at low risk, intermediate risk, and high risk have a likelihood of biochemical recurrence at 5 years of approximately less than 30%, 40% to 50%, and 65% to 75%, respectively. For patients with intermediate- and high-risk disease, improved therapeutic approaches are needed.

Given the responsiveness of metastatic prostate cancer to androgen suppression therapy (AST), many investigators have integrated such an approach into the management of men with nonmetastatic disease in combination with radiation therapy (RT) in an attempt to improve treatment outcome. Randomized controlled trials have been conducted to test the benefit of adding AST to RT in the treatment of men with historically defined locally advanced prostate cancer and in some men with "high-risk" disease as defined earlier. These trials have varied in the exact entry criteria, the total dose of radiation used, the radiation fields treated, and how long AST was administered, making direct comparisons between trials difficult. Even so, each trial has shown a statistically significant advantage of the addition of AST to RT in improving local control, biochemical-free survival, and distant metastasis-free survival. Only 1 trial, however, revealed both a cause-specific and overall survival advantage by the addition of AST to RT compared with RT alone. In this trial, AST was administered in the neoadjuvant, concomitant, and adjuvant setting (for 3 years), sometimes referred to as "long-term" AST.

Taken together, data from these large multi-institutional trials strongly argue that patients with locally advanced and high-risk prostate cancer are appropriately treated with a combination of AST and RT and the benefits of such combination therapy are real and quantifiable. There is general consensus among most treating physicians on this point and widespread application of this approach. Several questions arose from this body of work and remain unanswered, including (1) Does the addition of AST to RT also provide a benefit to patients with clinically localized prostate cancer? and (2) How long should AST be administered? These questions are, in part, addressed by the study reported in this issue of JAMA by D'Amico and colleagues.

Some patients with clinically localized prostate cancer have a combination of poor prognostic features that render them at an intermediate to high risk for extraprostatic disease and at an increased risk for disease recurrence (eg, clinical stage T1b-T2b and a Gleason score 7 with multiple positive biopsy cores). Given this elevated risk and the clear benefit AST and RT provide in patients with locally advanced prostate cancer, D'Amico et al designed and conducted an important clinical trial in men with clinically localized prostate cancer to test whether the addition of a short course (6 months) of AST to RT provides an improvement in cause-specific and overall survival when compared with patients treated with RT alone. This prospective trial randomized patients to either 70 Gy RT alone or in combination with complete androgen blockade using a luteinizing hormone–releasing hormone agonist and a nonsteroidal anti-androgen administered for 2 months before RT, 2 months concurrent with RT, and 2 months adjuvant to RT.

The study by D'Amico et al is extremely important. This study is the first trial to demonstrate that patients with clinically localized disease who are treated with 6 months of AST and RT achieve a statistically significant increase in overall survival, cause-specific survival, and survival without salvage AST when compared with patients treated with RT alone. These are all relevant and meaningful end points, particularly for patients with clinically localized intermediate- to high-risk disease who have a definable rate of clinically significant recurrence from prostate cancer.These data confirm previously reported retrospective data that the addition of short-term AST with RT provides a relapse-free survival advantage in a similar group of patients.The results are also intriguing, as all but 1 previous study have failed to show a benefit in survival with the addition of AST to RT. Together, these results are critically important in the development of improved treatment strategies for prostate cancer management.

Like any complex clinical trial, particularly one involving a disease with a number of important clinical variables, several issues should be considered to put the data into perspective. First, most of the patients treated in the study by D'Amico et al appear to be eligible for enrollment by virtue of having a Gleason score of at least 7. Patients with a Gleason score of at least 8 have generally been found to be at higher risk for recurrence and exhibit a lower survival than those with a Gleason score of at least 7. This study also enrolled patients with a wide range of serum PSA level (<40 ng/mL) and also included patients with documented disease outside the prostate as determined by magnetic resonance imaging. As such, this study enrolled a mixed population of patients at varying risk of disease recurrence, thereby complicating the extrapolation of the results to any single risk group of patients.

The most important findings of this trial are the benefits in cause-specific and overall survival when AST is added to RT in the management of clinically localized prostate cancer. These data are exciting and provocative and as such, deserve close examination. The authors prospectively adopted a reasonable definition of PSA failure, wherein the serum PSA was required to be more than 1 ng/mL and increasing by more than 0.2 ng/mL on 2 consecutive measurements. Interestingly, the authors did not prospectively define when salvage AST was to be instituted following failure. This omission adds a level of variability that could have affected the time to recurrence and, possibly, time to death. However, because the median and interquartile ranges of time to initiation of salvage therapy were comparable between the 2 groups, this methodological flaw does not appear to explain the difference in survival.

Importantly, the study by D'Amico et al also sought to determine if the benefits of long-term AST (2-3 years)could be observed with a relative short course (6 months) of AST in patients with clinically localized prostate cancer. With all the caveats noted earlier and with the proviso that this was a relatively small study, the answer would appear to be yes. On the surface, these results might appear to conflict with previous data. For instance, the apparent overall survival benefit with short-term AST in the study by D'Amico et al was not observed in the previously conducted RTOG 9202 study that randomized patients to 4 months of AST plus RT vs 2 years and 4 months of AST plus RT. These differences are certainly complicated by the fact that the 2 studies did not enroll patients with identical risk features. The RTOG 9202 trial enrolled patients with both clinically localized as well as locally advanced disease, the latter being a group for whom long-term AST is likely required. The study by D'Amico et al did not allow patients with locally advanced disease to be enrolled.

The study by D'Amico et al also forces contemplation of several lingering questions: (1) Do patients with clinically localized and locally advanced disease receive an equivalent survival benefit from a short course of AST? (2) How short a course of AST can be used effectively? (3) When should AST be administered in combination with RT: neoadjuvantly, concomitantly, adjuvantly, or a combination of the 3? (4) Do the adverse effects associated with AST outweigh the apparent benefits? (5) What is the mechanism by which AST is benefiting patients? Although the answer to these questions will most certainly improve the quality and potentially quantity of life for patients at an increased risk of death from prostate cancer, no data are available to fully answer each of them.

The study by D'Amico et al does, however, begin to shed light on some of these questions. Specifically, the magnitude of the survival benefit observed with short-term AST in the patient groups treated was significant and similar to the magnitude of survival benefit achieved when long-term AST plus RT were administered. This suggests that short-term AST is a very reasonable option as it provides similar value for this patient group with less advanced disease. Results from RTOG 9408 and an European Organization for Research and Treatment of Cancer trial of short-term vs long-term AST should help further refine this answer. The study by D'Amico et al does not specifically study or provide data to suggest that patients with high-risk locally advanced prostate cancer can forgo long-term AST as has been previously reported beneficial.

Furthermore, on the question of AST-related toxicity, this trial did not specifically or rigorously evaluate this important end point. The authors did not use a validated quality of life instrument capable of accurately and completely accessing the effects of AST to make a determination about a risk/benefit ratio regarding its use. Moreover, the study by D'Amico et al was not specifically designed to ascertain the optimal timing of AST administration. That is, it did not seek to determine if the beneficial effects observed with the addition of AST to RT result from a direct interaction of the 2 modalities or are a result of the treatment of subclinical metastatic disease. Nonetheless, given that essentially every trial completed to date, including this one, has revealed an improvement in metastasis-free survival with the use of AST in combination with RT, these data argue that treatment of subclinical microscopic metastatic deposits is likely one of the primary benefits of AST when administered in this context. Data from the RTOG 9413 studysupport the contention that a direct interaction of the 2 modalities is possible and, thus, concurrently applied AST is certainly reasonable. Thus, the most efficacious duration of AST administration (short-term vs long-term) is an issue that remains complicated and unresolved.

Another aspect of contemporary prostate cancer management should be considered to place these data into context. Radiation dose escalation presumes that higher radiation doses administered to the prostate and periprostatic area are more likely to eradicate prostate cancer cells than lower doses of radiation, ultimately translating into improved control of disease.Techniques such as 3-dimensional conformal RT and intensity modulated RT allow for safe delivery of high doses of radiation to the prostate. A recent prospectively randomized trial of low-dose (70 Gy) vs high-dose (78 Gy) RT revealed that higher radiation doses result in improved freedom from failure (biochemical or clinical) in patients with intermediate- to high-risk prostate cancer. No overall survival advantage to this approach has been noted to date, with relatively short follow-up and increased adverse events. Interestingly, all patients in the study by D'Amico et al were treated with 70 Gy of radiation, a dose that matches the low-dose group of the prospective trial. The relatively low radiation dose used in the study by D'Amico et al could have resulted in a reduction in disease control in both treatment groups, perhaps impacting the RT-only group the most.