br Patient summary In this study we examined
Patient summary: In this study, we examined the effect of androgen deprivation therapy (ADT) for Gleason 8 (Grade Group 4) versus Gleason 9–10 (Grade Group 5) prostate cancer. We found that Gleason 9–10 disease may derive a smaller survival benefit from ADT than Gleason 8 disease.
© 2018 European Association of Urology. Published by Elsevier B.V. All rights reserved.
The addition of androgen deprivation therapy (ADT) to external beam 129-46-4 therapy (EBRT) is known to improve overall survival (OS) in men with high-risk prostate cancer (PCa) [1–3]. A Gleason score of 8–10 or Grade Groups (GGs) 4 and 5 is an established high-risk feature, and post hoc analysis of a randomized controlled trial suggests that the benefit of long-course ADT may be greatest for patients with Gleason 8–10 disease . However, it has been hypothesized that Gleason pattern 5 disease, that is, predominantly tumors with a Gleason score of 9–10, may be less sensitive to ADT .
Determining whether Gleason 9–10 PCa has decreased sensitivity to ADT is critically important. It is generally assumed that ADT will substantially enhance survival for patients with Gleason 9–10 disease. If such is not the case, then there will be a dire need to utilize additional agents to optimize care for these patients. Phase 3 trials have shown an OS improvement from the use of novel antiandrogens [6– 8] and docetaxel [9–11] in metastatic castration-resistant and castration-sensitive PCa, as well as a metastasis-free survival benefit from the use of novel antiandrogens in nonmetastatic castration-resistant disease [12,13]. Addi-tional trials have suggested a possible benefit for these agents in the initial management of high-risk localized and locally advanced disease [14,15]. Should Gleason 9–10 PCa prove to be less sensitive to ADT, the use of novel antiandrogens and docetaxel may prove to be critical to its management.
Therefore, we investigated the association between ADT and OS for patients with Gleason 8 versus 9–10 PCa treated with EBRT.
2. Materials and methods
The National Cancer Database (NCDB) is a hospital-based tumor database created as a joint effort between the Commission on Cancer and the American Cancer Society . It captures information regarding socio-demographic factors, tumor characteristics, first-course treatment details, including type of radiation therapy and receipt of hormone therapy, and OS from an estimated 70% of incident cancers in the USA.
We identified 20 139 men in the NCDB diagnosed with localized or locally advanced, Gleason 8–10 (GGs 4 and 5) PCa from 2004 to 2011 who were treated with EBRT as their only form of local therapy (Supplemen-tary Fig. 1). Patients with nodal or metastatic disease were excluded. Additionally, patients with missing sociodemographic (age, race, insurance status, treatment facility type, and zip code-level income) or clinicopathologic (prostate-specific antigen [PSA], Gleason score, or clinical tumor stage) data of interest were also excluded. The NCDB captures OS but not disease-specific survival. To more accurately use OS to approximate disease-specific survival, we excluded patients with more than one documented malignancy or a Charlson-Deyo comorbidity score >0 as these patients have the highest risk of mortality from causes not related to PCa. Furthermore, patients with missing data on the use of ADT or those who began EBRT >365 d after PCa diagnosis or ADT >180 d before or after EBRT initiation were excluded as these patients may have received EBRT or ADT for palliative purposes. Follow-up was obtained from 2004 through 2012.
2.2. Statistical analysis
Baseline patient characteristics were compared using Pearson’s x2]FID$T1[test for categorical variables and the Wilcoxon rank-sum test for ordinal or continuous variables. Cox proportional hazards regression was utilized to examine the association between ADT and OS. Included in models were natural log-transformed PSA (in ng/ml), clinical tumor stage (T1-T2a, T2b-T2c, T3a, T3x, T3b, T4), age (by year), race (white, black, other), treatment facility type (academic, comprehensive, community), and zip code-level income (by quartile). To investigate whether a significant interaction existed between receipt of ADT and Gleason score (8 vs 9–10), models included the use of ADT (yes, no), Gleason score (8, 9–10), and an ADT (yes, no) Gleason score (8, 9–10) interaction term.
Furthermore, to investigate whether a higher Gleason score (8, 9, 10) correlated with an increased adjusted hazard ratio (AHR) for the association between ADT and OS, which would indicate that a higher Gleason score predicts lesser OS benefit from ADT, additional Cox regression models were constructed that included the use of ADT (yes, no), Gleason score (8, 9, 10) and ADT (yes, no) Gleason score (8, 9, 10) interaction terms, in addition to the other variables listed above.