br Sensitivity analyses were performed for the
Sensitivity analyses were performed for the 16 830 men who were treated to 66.6–86.4 Gy and with known information on whether pelvic Dynasore nodes were also treated. In these regression models, information on treatment volume (whole-pelvis, prostate-only), use of dose escalation (no [<75.6 Gy], yes [ 75.6 Gy]), and the modality of radiotherapy (3D conformal radiation therapy, intensity-modulated radiation therapy, proton therapy, other) was[7T$DIF] also incorporated.
Adjusted Kaplan-Meier curves for OS were generated using the mean of covariates method based on the multivariable Cox regression model . All statistical testing was two-sided with significance set at p < 0.05 and performed using Stata/SE version 14.2 (StataCorp, College Station, TX, USA). A waiver was obtained from the Dana-Farber/Harvard Cancer Center Institutional Review Board before undertaking this study.
2.3. Tumor grading convention
The 2014 International Society of Urological Pathology consensus conference noted the difference in outcomes between Gleason 8 (GG
4) and Gleason 9–10 (GG 5) disease and recommended the use of the GG system to describe PCa grading . We agree with the importance of the GG system and have used it alongside the Gleason score when appropriate in this manuscript. Since a significant portion of our analyses concerns the difference in the survival benefit of ADT for Gleason 8 versus 9 versus 10 disease and]FID$T2[the GG system does not capture potential differences between Gleason 9 and 10 disease, we chose to also use the Gleason score system to describe tumor grading.
3.1. Baseline patient characteristics
On multivariable analysis, receipt of ADT was associated with a significant improvement in OS for patients with
Table 1 – Baseline patient characteristics
No ADT ADT p value No ADT ADT p value
Year of diagnosis
Clinical tumor stage
Zip code-level income
Percentages may not add to 100 due to rounding.
ADT = androgen deprivation therapy; PSA = prostate-specific antigen.
On sensitivity analysis of the 16 830 men who were treated to 66.6–86.4 Gy and had information on the use of whole-pelvis versus prostate-only radiation, the AHR for the association between use of ADT and OS was 0.75 (95% CI:
Fig. 1 – Adjusted Kaplan-Meier plots of overall survival: (A) Gleason 8,
(B) Gleason 9–10. The p values were derived from the multivariable Cox regression model.
ADT = androgen deprivation therapy.
We analyzed the association between receipt of ADT and OS in a contemporary, national cohort of over 20 000 patients with localized or locally advanced, Gleason 8–10 PCa treated with EBRT. Our results demonstrate that while ADT was associated with an approximately 20% decrease in all-cause mortality for patients with Gleason 8 disease, it was not associated with a significant survival advantage for patients with Gleason 9–10 disease. In addition, the AHRs for the association between use of ADT and OS increased with a higher Gleason score. Overall, these results suggest that
Fig. 2 – Forest plot of the overall survival benefit of ADT for Gleason 8 versus Gleason 9–10 disease.
ADT = androgen deprivation therapy; CI = confidence interval; OS = overall survival.
Fig. 3 – Forest plot of the overall survival benefit of ADT for Gleason 8 versus Gleason 9 versus Gleason 10 disease. ADT = androgen deprivation therapy; CI = confidence interval; OS = overall survival.
compared with Gleason 8 PCa, Gleason 9–10 disease derives less survival benefit from ADT and that a higher Gleason score predicts lesser benefit. Our findings potentially have major clinical implications. The current standard of care for patients with high-risk localized or locally advanced PCa managed with radiation therapy, including men with Gleason 8–10 disease, is the addition of long-course ADT, typically in the form of a luteinizing hormone-releasing hormone (LHRH) agonist with or without a first-generation antiandrogen [1–3]. Our study challenges the current paradigm of treating all patients with Gleason 8–10 disease similarly, as our results indicate that any survival advantage from the use of ADT is significantly weaker for Gleason 9–10 disease than for Gleason 8 disease.
Notably, our results do not contradict prospective evidence regarding the efficacy of ADT. The EORTC 22863, RTOG 8531, RTOG 8610, RTOG 9408, Dana-Farber Cancer Institute 95-096, and TROG 96.01 randomized controlled trials examined the benefit of adding ADT to radiation
therapy [1,3,19–22]. Additionally, EORTC 22961, RTOG 9202, GICOR DART 01/05, TROG 03.04, and PCS IV evaluated longer-course versus shorter-course ADT with radiation therapy [2,4,23–25]. To our knowledge, publications or presentations of these trials have not presented subgroup analyses of patients with Gleason 9–10 disease as patients with Gleason 8–10 disease have historically been grouped together. Furthermore, while it is known that 20% of the PCS IV cohort had Gleason 9–10 disease, the proportion of the other trial cohorts that harbored Gleason 9–10 disease is unknown. It may be that the survival advantage observed from the use of ADT is driven largely by patients with Gleason 8 or less PCa. A post hoc analysis of these trials may be helpful in further elucidating the survival benefit of ADT for Gleason 8–10 patients.