The role of hormonal therapy in the management of clinically localized prostate cancer is controversial. Extensive questions remain about how to implement hormonal therapy and what would be considered optimal. Part of the confusion comes from the different roles that the urologist and radiation oncologist play, both independently and when treating the disease as a team. It is a fair statement to suggest that, at times, different schools of thought affect how patients are ultimately treated. To close this theoretical gap, it is best to rely on prospective randomized studies and extrapolate applicable data as needed. In early randomized surgical series reported by Labrie et al,[1] patients were treated with several months of neoadjuvant and concurrent hormonal therapy prior to radical prostatectomy. This series showed an approximate 90% response rate, defined as decreasing the size of both the prostate and the tumor. More recent series, such as those reported by Soloway et al[2] and Aus et al,[3] randomized patients to radical prostatectomy vs neoadjuvant hormonal therapy plus radical prostatectomy. These studies demonstrated a statistically significant decrease in the incidence of extracapsular extension and positive surgical margins. Neoadjuvant hormonal therapy did not, however, improve freedom from biochemical failure or survival at 5 years. In contrast, intermediate- and high-risk patients who have received hormonal therapy and external-beam radiotherapy (EBRT) appeared to benefit when compared to the same patients treated with EBRT alone. A good explanation for the benefit of neoadjuvant hormonal therapy in the setting of EBRT but not radical prostatectomy has only recently become clear and will be discussed. Several prospective randomized studies addressed the role of neoadjuvant and adjuvant hormonal therapy with EBRT. These trials have shown an improvement in local control, disease- free survival, and overall survival. The role of hormonal therapy and permanent prostate implants is less established. Before discussing hormonal therapy plus permanent prostate implants, it is essential to establish which prostate patients benefit from hormonal therapy. The role of hormonal therapy and EBRT will be discussed initially as a springboard for how hormonal therapy should be applied in conjunction with permanent prostate implants. What Is the Role of Hormonal Therapy in Patients Receiving EBRT? Based on the Radiation Therapy Oncology Group (RTOG) 8610[4] and 9202[5] trials, different subpopulations of patients have emerged as groups likely to benefit from short- and long-term hormonal therapy. RTOG 8610 included patients with locally advanced bulky prostate cancer who were evaluated for the benefit of neoadjuvant hormonal therapy and EBRT or EBRT alone. RTOG 9202 assessed 4 months of neoadjuvant hormonal therapy with or without 2 years of adjuvant hormonal therapy. The results of these studies and other phase III prospective randomized studies have led to answers to classic questions addressing whether there is a biologic interaction between hormonal therapy and EBRT, the timing of hormonal therapy, the optimal duration of hormonal therapy, and the volume to be irradiated. It is now clear that the benefits experienced by different subgroups receiving short- or long-term hormonal therapy appear to depend on the risk that patients have for dying of prostate cancer. Defining Risk of Death From Prostate Cancer Randomized studies have shown that patients with low-risk disease do not benefit from hormonal therapy, whereas patients with intermediateand high-risk disease do. Although most experts would agree on how to define low risk (T1-2, Gleason score [GS] < 6, and prostate-specific antigen [PSA] values < 10-20 ng/mL), how to define intermediate- and high-risk subgroups remains controversial. Several classification schemes are employed by physicians across the country. The rationale for using a risk-group classification scheme is that it can help to determine prognosis. Ideally, such a scheme should help determine which patients are appropriate candidates for a particular type of therapy. Different institutions subscribe to different riskgroup stratification, but most of the commonly used schemes do not provide insights into how patients should be selected for hormonal therapy. An exception is the RTOG risk-group scheme, so for the sake of simplicity, the RTOG risk groups will be used as a frame of reference. Table 1 shows the RTOG riskgroup classification system, which was developed to predict overall and disease-specific survival, and has been validated to predict PSA failure in contemporary cases.[6,7] Using this risk-group scheme, the value of hormonal therapy has been studied in a meta-analysis. Based on the data, it appears that low-risk patients did not benefit from hormonal therapy. Intermediate- risk patients appeared to benefit from short-term hormonal therapy, and high-risk patients (groups 3 and 4) were found to have an overall survival benefit with the addition of longterm hormonal therapy. Is There a Biologic Interaction Between Hormonal Therapy and EBRT? How exactly does hormonal therapy affect EBRT? In theory, the major issue in low-risk patients is local control because such patients are at low risk for regional disease and at very low risk for distant disease. Intermediate- risk patients might benefit from local and regional control because they have a significantly higher risk of lymph node involvement. In contrast, high-risk patients are at a substantially greater risk for distant as well as local regional failure and likely to benefit from therapy that addresses distant metastasis. Mechanism of Interaction
The exact mechanism by which EBRT and hormonal therapy interact is not known. Based on assessments of the prostate itself, it appears that androgen deprivation induces apoptosis and thereby reduces the number of tumor cells. The technique shifts cells that are actively dividing into quiescence.[8] Using the Shionogi mouse in vivo tumor system, Zietman et al showed that the dose of radiation to the tumor plus hormonal therapy could be halved (Figure 1).[9] Their study suggested that maximal androgen suppression prior to radiation was the most effective strategy for controlling these implanted tumors. Based on these in vivo data, most radiation oncologists assumed that, in humans, the most favorable interactions between androgen deprivation and radiation therapy would be sequence dependent, with the greatest response following maximal androgen suppression. It was assumed that local control might be achieved with lower doses of radiation and that there were synergistic interactions between hormonal therapy and EBRT. This concept was supported by the initial reports of findings from RTOG 8610. However, in vitro data failed to demonstrate evidence of synergistic interactions between hormonal therapy and EBRT. This observation challenged the notion that we would see improved local control.[10] Similarly, recent data suggest that biopsy status after hormonal therapy and radiotherapy may not be reliable end points for predicting outcomes when neoadjuvant hormonal therapy is added to EBRT. For example, although Laverdiere et al[11,12] demonstrated that the positive biopsy rate was reduced with 9 (vs 3) months of neoadjuvant and concurrent hormonal therapy, longer follow-up showed no difference in the incidence of biochemical failure between the two treatment arms. According to other recent data, at least one type of favorable interaction between neoadjuvant hormonal therapy and EBRT occurs in the lymph nodes. It remains to be determined whether this has something to do with the shape of the radiation doseresponse curve. This hypothesis is well illustrated in Figure 1, which shows a plateau of local control at doses of approximately 80 Gy in animal models. It might be, for example, that in humans the plateau of the dose-response curve might also actually occur at 80 Gy, such that no additional benefit is seen when hormonal therapy is added to doses above this level. If this is the case, however, a favorable interaction might be observed in the lymph nodes because the dose of radiation given is still on the steep part of the radiation doseresponse curve at 50 Gy. An alternative (and the most provocative) explanation of the apparent benefits of pelvic radiotherapy is that it is mediated via a combined hormonal immunologic mechanism. Space limitations do not allow this mechanism to be elaborated on in detail, but suffice it to say, there are compelling preliminary supporting data (Roach, personal communication, 2004). Optimal Timing of Hormonal Therapy Should hormonal therapy be given adjuvantly or neoadjuvantly? RTOG 9413 was the first phase III prospective randomized trial to stratify patients by PSA, Gleason score, and TNM stage, and to use progressionfree survival (biochemical failure and clinical failure) as a primary end point. Using hormonal therapy in the form of androgen blockade, the trial randomized patients to treatment arms that included whole-pelvic radiotherapy and neoadjuvant/concurrent hormonal therapy, prostate-only radiotherapy and neoadjuvant/concurrent hormonal therapy, whole-pelvic radiotherapy plus adjuvant hormonal therapy, and prostate-only radiotherapy plus adjuvant hormonal therapy.[ 13] RTOG 9413 not only addressed the timing of hormonal therapy, it also demonstrated the importance of whole-pelvic radiotherapy when using hormonal therapy. Prostate-only radiotherapy plus neoadjuant/concurrent hormonal therapy vs adjuvant hormonal therapy plus prostate-only radiotherapy did not show any difference in biologic interaction, despite the 2-month advantage in the adjuvant arm (as time to failure was measured from the randomization date). This finding indicates that there is no difference in the interaction between neoadjuvant and adjuvant hormonal therapy with prostate-only radiotherapy. Whole-pelvic radiotherapy plus neoadjuant/concurrent hormonal therapy vs whole-pelvic radiotherapy plus adjuvant hormonal therapy also showed a 2-month bias in the adjuvant arm, but the adjuvant arm was inferior (Figure 2), proving that there is a sequence-dependent interaction that is occurring in the lymph nodes and not in the prostate. In Figure 3, with both neoadjuant/ concurrent hormonal therapy curves, the bias is eliminated and the difference in the curves is more apparent. Evaluation of disease progression favored whole-pelvic radiotherapy plus neoadjuant/concurrent hormonal therapy. Assessment of death and PSA failure demonstrated a trend toward overall survival benefit, but at this time follow-up is too short to expect differences to be apparent. RTOG 9413 eliminates the freedom from biologic failure bias seen in other studies that have compared patients receiving radiation to those receiving radiation plus hormonal therapy. Previously, an inherent bias was seen in hormonal therapy arms because there is a delay in the time for a rise in PSA. In RTOG 9413, all arms received a similar duration of hormonal therapy, thereby avoiding this bias discrepancy in the definition of PSA failure. Because of the impact that a rising testosterone level has on PSA, the American Society for Therapeutic Radiology and Oncology (ASTRO) consensus definition of three consecutive increases is also problematic. Of interest, the definition of PSA failure for RTOG 9413 was very similar to one of the four definitions shown to have a higher sensitivity and specificity than the ASTRO definition.[14] Optimal Duration of Hormonal Therapy Prospective randomized trials such as the "Bolla Study," RTOG 8531, and RTOG 9202 demonstrated improved overall survival using longterm hormonal therapy in patients with high-risk disease.[15] RTOG 8610 established the role of neoadjuvant hormonal therapy in intermediate-risk patients.[16] A meta-analysis of RTOG trials[ 17] suggested that neoadjuvant hormonal therapy showed a benefit in patients with GS7, T1/2 or GS6, T3. Short-term neoadjuvant hormonal therapy did not appear to benefit patients with GS7, T3 or GS8-10; however, this risk group benefited from long-term adjuvant hormonal therapy. Role of Hormonal Therapy With Permanent Prostate Implants Controversy surrounds what type of radiation therapy is most beneficial for treating prostate cancer, but permanent prostate implants offer an excellent strategy in this setting. A number of earlier studies, including those by D'Amico et al[18] and Beyer and Brachman,[19] have concluded that EBRT is better than permanent prostate implants when treating intermediate- or high-risk patients. A study by King et al[20] compared permanent prostate implants, radical prostatectomy, and EBRT, and concluded that permanent prostate implants and radical prostatectomy produced superior results when compared to EBRT. However, the EBRT dose was inadequate (66 Gy), and the EBRT patients were worse candidates at baseline, compared to patients receiving the other modalities. One of the difficulties in interpreting these studies was the "PSA blip" seen after permanent prostate implants. This phenomenon was not well recognized at the time these studies were conducted, and many of these cases may have been mistakenly considered biochemical failures. First described in 1997, the "blip" occurred after brachytherapy in approximately 25% to 30% of patients.[21] That is, patients were found to have a transient rise in PSA followed by a decline (Figure 4). Studies that have biopsied patients with "PSA blips" have occasionally found histologic evidence of cancer on repeat biopsy.[22] However, it has been well documented that with further follow-up, positive biopsies can become negative due to slow cancer involution.[23] Previous Retrospective Studies
Before 1995, hormonal therapy was mostly used for cytoreduction. Investigators from Memorial Sloan-Kettering Cancer Center (MSKCC)[24] assessed the prognostic significance of Gleason score in patients treated with permanent prostate implants. They made treatment distinctions based on Gleason score to ascertain who were appropriate candidates for monotherapy with permanent prostate implants. Patients with GS 4+3 had significantly lower 7-year biochemical freedom from recurrence rates compared to those with GS 3+4. Unfortunately, what causes this difference is unclear. This finding may suggest that with increasing risk, permanent prostate implants alone may be inadequate therapy. Conflicting applicable data suggest that there may or may not be a significant benefit to a combination of EBRT and permanent prostate implants.[ 25,26] Perhaps, in theory, this is a population that would benefit from adjuvant hormonal therapy. The MSKCC investigators also assessed 263 patients between 1992 and 1997 who had prostates weighing more than 60 g and were given neoadjuvant hormonal therapy for cytoreduction (Table 2). A retrospective matched-pair analysis was unable to show any benefit with neoadjuvant hormonal therapy plus permanent prostate implants, compared with permanent prostate implants alone.[27]