Predictive genomic biomarkers in non-metastatic castration resistant prostate cancer (nmCRPC) treated with androgen receptor pathway inhibitors (ARPi)

Predictive genomic biomarkers in non-metastatic castration resistant prostate cancer (nmCRPC) treated with androgen receptor pathway inhibitors (ARPi)


Predictive genomic biomarkers in non-metastatic castration resistant prostate cancer (nmCRPC) treated with androgen receptor pathway inhibitors (ARPi)

, , , , , , , , , , , , , ,


Although recent trials have demonstrated overall survival (OS) and metastasis free survival (MFS) benefit with the use of ARPi in high-risk nmCRPC, individual treatment outcomes may vary. This retrospective multicentre analysis explores the association between genomic mutations with ARPi treatment outcomes in nmCRPC patients to identify predictive biomarkers.


In high-risk nmCRPC patients treated with apalutamide, enzalutamide or darolutamide at APCaRI affiliated cancer centres, median MFS, OS, PSA decline ≥ 50% from baseline (PSA50), and second progression free survival (PFS2) were calculated. Next generation gene sequencing (NGS) was performed on archival tumour tissue examining for genomic alterations in 500 genes, including homologous recombination repair (HRR), mismatch repair, tumour suppressor (TS), and PI3K/AKT oncogene (OG) groups. Analysis was conducted using Cox proportional hazards regression using wildtype cases as the reference group, while adjusting for PSA doubling time and presence of pelvic lymph nodes.


Of 32 nmCRPC patients, 25 (78%) were treated with apalutamide, 5 (16%) with darolutamide and 2 (6%) with enzalutamide. 10 patients (31%) had TS/OG mutations (5 PTEN, 8 TP53, 2 PIK3CA), 3 (9%) had HRR gene mutations (2 ATM, 1 BRCA2) and 1 (3%) had 2 MLH1 mutations (microsatellite instability). All 5 patients who received subsequent therapy received abiraterone. Patients with TS/OG mutations had significantly shorter MFS (16.4 mo; HR 5.2; 95% CI 1.4 – 25.7; p = 0.018), PFS2 (22.1 mo; HR 15.4; 95% CI 1.9 – 126.3; p = 0.011) and OS (24.1 mo; HR 8.3; 95% CI 1.2 – 58.8; p = 0.035). Those with HRR mutations had significantly reduced PFS2 (median not reach [NR]; HR 40.4; 95% CI 1.6 – 1034.2; p = 0.025) and OS (NR; HR 21.7; 95% CI 1.1 – 446.1; p = 0.045). 3 (9%) patients did not achieve PSA50, including a patient with 2 MLH1 mutations.


This analysis demonstrates that nmCRPC patient with TS/OG and HRR gene mutations have poor clinical outcomes and may benefit from close follow-up. Our results underline the need for ongoing trials which examine novel targeted therapies (e.g. PARP Inhibitors, AKT inhibitors) in these molecularly defined nmCRPC subgroups.

goes to…APCaRI member Russ Greiner

Image of DREAM challenge winners, Russ Greiner pictured on far left.

Dr. Russ Greiner, Canada CIFAR AI Chair, Fellow-in-Residence at Amii, University of Alberta Professor, and APCaRI member, received the CAIAC Lifetime Achievement Award announced at the Canadian AI Conference on May 27, 2021. This the highest honour bestowed by CAIAC, given in recognition to researchers who have distinguished themselves through outstanding research excellence in AI during the course of their academic career. APCaRI congratulates Russ Greiner for his well-deserved CAIAC Lifetime Achievement Award!

“Using machine learning techniques to produce effective, evidence-based personalized treatment”

The main foci of Russ Greiner’s current work are (1) bioinformatics and medical informatics; (2) learning and using effective probabilistic models and (3) formal foundations of learnability. He has published over 200 refereed papers and patents, most in the areas of machine learning and knowledge representation, including 4 that have been awarded Best Paper prizes.

One of these four papers was an entry into an international machine learning competition hosted by DREAM, an open-science effort dedicated to improving health and health care through crowdsourcing problem-solving. DREAM’s challenge was to develop an algorithm to predict which prostate cancer patients would respond to certain treatments and which would follow the medication regimen. The algorithm could be used by clinicians to help chose the best treatment plans for the patient.

Greiner and a team of students tied for the top place in the competition against over 50 teams from around the world. Then the winners collaborated to create an even better solution to the problem!







- Perrin Beatty