ASTRO 2021: Prostate Cancer: Better Living Through Biomarkers

( The 2021 American Society for Radiation Oncology (ASTRO) Hybrid Annual Meeting’s included a session on biomarkers and salvage radiotherapy and a discussant presentation of five presentations in this session by Dr. Brian Baumann. With regards to organizing these five presentations into clinical context, Dr. Baumann provided the following summary outline:

  • In radical prostatectomy patients:
    • Biomarker studies in the salvage setting include AR-V7 (Konieczkowski et al.) and a genomic classifier analysis from the SAKK 09/10 trial (Dal Pra et al.)
    • A phase I trial of neoadjuvant stereotactic body radiotherapy (Hammer et al.)
  • In patients with intact disease:
    • Biomarker studies in high-risk intact disease include the genomic classifier analysis from three high-risk RTOG trials (Nguyen et al.)
    • A PSMA PET-derived risk stratification tool (Ma et al.)

Androgen receptor signaling is central to the biology of castration-sensitive prostate cancer and ADT therapy. Certain splice variants are associated with resistance to ADT in metastatic castration-resistant prostate cancer, including AR-V7, which also seems to be associated with DNA repair. The hypothesis of the Konieczkowski et al. study was: do AR variants correlate with poor response to salvage radiotherapy + ADT in the post-operative setting? The cohort for this study included 46 patients retrospectively analyzed with comprehensive AR-V interrogation by ultra-deep sequencing. Ultimately, AR-V7 was the only AR variant associated with oncologic outcomes:



 Dr. Baumann notes that these are intriguing preliminary results for AR-V7 as a potential predictive biomarker for response to salvage radiotherapy + ADT, but the findings are limited by the small size of the cohort (n = 46). Further, the entire cohort received salvage radiotherapy + ADT, so it is unclear if AR-V7 is impacting on ADT, on salvage radiotherapy, or on both. Ultimately, more studies are needed to elucidate the role of AR-V7 as a biomarker in the salvage radiotherapy + ADT setting.

In the SAKK 09/10 trial, genomic classification was a pre-specific endpoint. This was a multicenter, randomized phase 3 trial performed in 24 centers in Switzerland, Germany, and Belgium. Patients with biochemical progression (PSA >0.1 to 2 ng/mL at randomization) were randomized to 64 Gy versus 70 Gy to the prostate bed; no ADT or pelvic nodal radiotherapy was used. The results of this study were recently published in European Urology [1] and there was no difference in failure from biochemical progression between the two groups after six years of follow-up. The pre-specified statistical analysis was to investigate the impact of genomic classifiers on oncologic outcomes prior to analysis of overall study results. Central review of the pathology was associated with a high rate of passing quality control (>97%), and 226 of 350 patients had genomic classifier data for analysis. The primary endpoint for this particular analysis was freedom from biochemical progression. Dr. Baumann notes that the findings in SAKK 09/10 for biochemical progression were similar to the RTOG 9601 trial [2] with respect to overall survival when patients were stratified by high versus low/intermediate genomic classifier risk:



Additionally, the exploratory analysis on the genomic classifier and pre-salvage PSA with regards to freedom from biochemical progression showed important outcome differentiation when stratified by PSA <= vs > 0.5 ng/mL at randomization:


 Dr. Baumann emphasized that this study is the first phase III trial in prostate cancer with a pre-specified genomic classifier analysis that has reported out, with results demonstrating that a genomic classifier adds independent prognostic data. However, with a median follow-up of only 6.3 years the results are not mature enough for distant metastasis or overall survival analysis. The genomic classifier and pre-salvage PSA interaction exploratory results are intriguing given that stratification in outcomes by genomic classifier score are not subtle (high versus low/intermediate), showing a 15% absolute difference in 5-year failure from biochemical progression for the PSA <= 0.5 cohort and a 51% absolute difference in 5-year failure from biochemical progression for the PSA > 0.5 cohort. Ultimately, among patients with a pre-salvage PSA > 0.5 and high-risk genomic classifier score, only 13% of patients were free of progression at 5-years in a cohort treated without ADT or whole pelvis radiotherapy. These results speak to the potential importance of using a genomic classifier and pre-salvage PSA to counsel patients on treatment intensification strategies, including the use of ADT and/or whole pelvis radiotherapy. The especially poor results for PSA >0.5 ng/mL and high-risk genomic classifier score argues in favor of using ADT +/- whole pelvis radiotherapy, which were the winning arms of the SPPORT and GETUG-AFU 16 trials. Dr. Baumann highlighted that in his clinical practice he obtains a Decipher genomic classifier score on all savage radiotherapy patients. For patients with low pre-salvage PSA (< 0.5 ng/mL) and high-risk Decipher, he routinely recommends short-term ADT and whole pelvis radiotherapy, and genomic classifier testing assists with counseling patients with low pre-salvage PSA on the potential benefits of radiotherapy + short term ADT versus radiotherapy alone. For patients with higher pre-salvage PSA, Decipher is useful for shared decision-making on short term versus long-term ADT, and Dr. Baumann routinely treats the elective nodes when using ADT based on the SPPORT trial data.

The phase I trial of neoadjuvant stereotactic body radiotherapy presented by Hammer et al. had a primary objective of determining the maximum tolerated dose of neoadjuvant stereotactic body radiotherapy that leads to <28% of patients experiencing a dose-limiting toxicity 30 days post-radical prostatectomy. This included a population of high-risk or clinically N+ disease patients who were candidates for surgery. Patients were treated with 25 Gy/5 fractions to the pelvic nodes and either 30 Gy in 5 fractions or 35 Gy in 5 fractions to the prostate and seminal vesicles. Ultimately, this trial was stopped early due to unacceptably high toxicity after 16 patients were enrolled. There were several severe complications including grade 3 urinary tract obstruction in 25% of patients, grade 3 urinary fistula in 25% of patients, two patients (13%) requiring a cystectomy and urinary diversion, and two patients (13%) undergoing incision of bladder neck contracture.

Dr. Baumann notes that the authors should be commended for evaluating a novel use of radiation with pre-op stereotactic body radiotherapy for patients with high-risk disease. Importantly, the target volume is smaller in the pre-op setting and pre-op radiotherapy can improve rates of negative margins. Unfortunately, the high dose of stereotactic body radiotherapy (30-35 Gy) was too toxic, but pre-op radiotherapy is worth exploring further. Future phase I/II trials should focus on toxicity mitigation, such as lowering the dose (ie. 25 Gy in 5 fractions), using a more fractionated scheme, or employing treatment planning techniques to reduce toxicity (ie. urethral sparing, online adaptive radiotherapy, proton therapy, etc).

Use of genomic testing can stratify patients based on risk of distant metastasis and allow for personalization of therapy. Similar to the Feng et al. study investigating Decipher in RTOG 9601 [2], the study by Nguyen et al. presented at ASTRO 2021 was designed to validate the performance of the Decipher genomic classifier in pre-treatment biopsy samples collected in three randomized phase III high-risk definitive radiotherapy trials: NRG/RTOG 9202, 9413, and 9902. On multivariable analysis, the genomic classifier was prognostic for distant metastasis (HR 1.24, 95% CI 1.11-1.39), prostate-cancer specific mortality (HR 1.27, 95% CI 1.13-1.43), and overall survival (HR 1.12, 95% 1.05-1.20). Stratified by low (genomic classifier <0.23), intermediate (genomic classifier 0.23-0.32), and high (genomic classifier > 0.32) risk groups, the cumulative incidences of outcomes by these genomic classifier risk groups was statistically significant for distant metastases (p < 0.001), prostate cancer specific mortality (p < 0.001), and OS (p = 0.002).

Dr. Baumann notes that this is an impactful study, which validates gene expression biomarkers using prospective randomized trial data and found an independent association with the genomic classifier score and risk of distant metastasis, prostate cancer specific mortality, and overall survival. This tool can improve risk stratification and aid in personalized treatment decisions for patients. Importantly, genomic classifier testing is now included in the NCCN guidelines and this study paves the way for using genomic classifier testing as a stratification factor in randomized prostate cancer clinical trials. Dr. Baumann encourages everyone to open NRG-GU009/PREDICT-RT (NCT 04513717) at their center:



 In Dr. Baumann’s practice, he routinely obtains genomic classifier testing for all prostate cancer patients, and his preference is to enroll high-risk patients on the NRG GU009 trial. For high-risk patients who are not eligible or decline the study, he uses the Decipher test for shared decision-making for patients already on treatment who express reservations about continuing ADT and to help patients decide between 1.5 years versus 2 years of ADT.

In the study by Ma et al., they reported results testing a nomogram they developed using 200 patients with high-risk prostate cancer who underwent PSMA PET/CT scan who were cN0M0 on conventional imaging. The objective of this study was to determine the prognostic performance of this PSMA nomogram for predicting clinical outcomes, such as biochemical recurrence, distant metastasis, prostate cancer specific mortality, and overall survival, as well as to compare the prognostic performance of the PSMA nomogram to existing risk-stratification tools. Using a multi-institutional database, NCDB, and SEER, they found that the nomogram was predictive of clinical endpoints and outperformed other nomograms such as STAR-CAP, CAPRA, and the MSKCC nomogram.

Dr. Baumann notes that it is remarkable that a database developed using likelihood of PSMA PET-detected nodal or distant metastasis in 200 patients thought to have localized disease on conventional imaging could outperform other nomograms developed using clinical data from thousands of patients. This provides valuable data that patients predicted risk of upstaging on PSMA PET is closely linked with long-term outcomes such as distant metastasis and prostate cancer specific mortality. Future efforts are needed to further refine the nomogram and externally validate it.


Presented by: Brian Baumann, MD, Chief of Genitourinary Radiation Oncology, Washington University, St. Louis, MO

Written by: Zachary Klaassen, MD, MSc – Urologic Oncologist, Assistant Professor of Urology, Georgia Cancer Center, Augusta University/Medical College of Georgia, @zklaassen_md on Twitter during the 2021 American Society for Radiation Oncology (ASTRO) Hybrid Annual Meeting, Sat, Oct 23 – Wed, Oct 27, 2021.


  1. Ghadar P, Hayoz S, Bernhard J, et al. Dose-intensified versus conventional-dose salvage radiotherapy for biochemically recurrent prostate cancer after prostatectomy: The SAKK 09/10 Randomized Phase 3 Trial. Eur Urol. 2021;80(3):306-315.
  2. Feng FY, Huang HC, Spratt DE, et al. Validation of a 22-gene genomic classifier in patients with recurrent prostate cancer: An ancillary study of the NRG/RTOG 9601 Randomized Clinical Trial. JAMA Oncol. 2021;7(4):544-552.