Personalizing the duration of androgen-deprivation therapy use in the management of intermediate-risk prostate cancer.

Abstract

When considering a risk-based scheme for predicting outcomes following standards of practice for nonmetastatic, nodenegative prostate cancer (PC), men experience recurrence either infrequently or commonly, depending on their risk group, and some who recur will go on to die as a result of PC. However, men with intermediate-risk PC have been shown to have 5-year prostate-specific antigen (PSA) progression-free rates that range from 30% (a high-risk outcome) to 98% (a low-risk outcome). As a result of this variation, investigators have sought to identify clinical factors in addition to those on which intermediate risk is based (PSA 10 to 20 ng/mL, Gleason score 7, or clinical tumor category 2b or 2c) to stratify these men into lowand high-risk subgroups for PC-specific mortality (PCSM). Multiple studies exist to show that the percent of positive prostate biopsies (ppb), the number of intermediate-risk factors, and primary versus secondary biopsy Gleason grade 4 in men with Gleason score 7 PC can aid in making this stratification. On the basis of these data, investigators from Memorial Sloan Kettering proposed two intermediate-risk subgroups as follows: (1) Favorable—Gleason 3 4 or less and ppb not exceeding 50% and only one intermediate-risk factor excluding 4 3; and (2) Unfavorable— Gleason 4 3 or at least two intermediate-risk factors or at least one intermediate risk factor and ppb greater than 50%. The clinical significance of this subdivision relates to personalizing the use of androgen-deprivation therapy (ADT) for men with intermediate-risk PC. We know from prospective randomized controlled trials (RCTs) that when administering lowdose (approximately 70 Gy) external-beam radiation therapy (EBRT), the addition of 4 to 6 months of ADT as compared with no ADT prolongs overall survival in men with intermediate-risk PC. We do not know, however, which men with intermediate-risk PC benefited from the addition of ADT and whether that benefit would have existed if high-dose RT was administered. Specifically, several investigators have shown that using EBRT without ADT in men with favorable intermediate-risk PC results in low rates of PCSM after 10 years ( 2%), consistent with outcomes achieved for men with low-risk PC where ADT is not indicated because it has not been shown to prolong survival. Given that this is recent information, it was not incorporated into patient selection or prerandomization stratification schemes in RCTs such as in the RCT that accompanies this editorial which creates difficulty in interpreting the results. Specifically, if men with favorable intermediate-risk PC are enrolled on RCTs evaluating the impact of the duration of hormone therapy on PCSM, the power of the trial to observe a difference in the PCSM outcome, if one exists, decreases because PCSM will not be impacted by differing durations of ADT in men with favorable-intermediate-risk PC. Therefore, while a significant reduction in PCSM may exist for longerrather shorter-course ADT in men with unfavorable intermediate-risk PC, the overall study may be negative due to the dilution of any impact of the duration of ADT on PCSM from the inclusion of men with favorable intermediate-risk PC. This limitation is significant because data from nonrandomized institutional series show a significant association with a reduction in PCSM when 6 months of ADT are added to high-dose EBRT in men with unfavorable intermediate-risk PC. What has been lacking, however, are data from RCTs in which the potential for confounding is minimized so causality can be established. In the article that accompanies this editorial, Pisansky et al present the initial results of Radiation Therapy Oncology Group (RTOG) 9910 [Clinical Trial Registry Number (NCT) 00005044], a randomized trial evaluating the impact on PCSM of adding 4 months versus 9 months of ADT to 70.2-Gy EBRT. The study cohort consisted of 1,489 eligible men (median age, 71 years) with 84% intermediate-, 15% high-, and less than 1% having very high–risk PC. After a median follow-up of 9.4 years, 450 (30.2%) men died; 54 (12%) of PC. The hazard ratio for PCSM, the primary end point was 0.81 (95% CI, 0.48 to 1.39; P .45) and cumulative incidence estimates of PCSM at 10 years was 4% and 5% in the 9and 4-month ADT arms, respectively. The expected number of PC deaths on which the study was powered was 270 or five times the number observed. Ideally, given the new information regarding the possible benefit of ADT in men with unfavorable but not favorable intermediate-risk PC, a postrandomization hypothesis generating analysis evaluating the impact of the duration of ADT in these subgroups would be helpful. Since this categorization requires knowing the ppb, these data may not have been prospectively JOURNAL OF CLINICAL ONCOLOGY E D I T O R I A L VOLUME 33 NUMBER 4 FEBRUARY 1 2015