Research confirms prognostic performance of genomic classifier

June 10, 2020

The classifier can be a useful adjunct for clinical decision-making, an investigator says.

Transcriptome profiling of tumor specimens collected from patients with biochemically recurrent prostate cancer participating in a phase 3 clinical trial further validates the performance of a commercially available genomic classifier (Decipher) for predicting the risk of distant metastasis.

In addition, the findings indicate that the genomic classifier has value for guiding the selection of patients undergoing salvage radiotherapy for recurrence who are most likely to benefit from treatment intensification with anti-androgen therapy, said Felix Y. Feng, MD.

The study authors analyzed data from NRG/RTOG 9601, a randomized, placebo-controlled phase 3 trial investigating the addition of bicalutamide to salvage radiotherapy. Feng, a professor of radiation oncology, urology, and medicine, director of the Benioff Initiative for Prostate Cancer Research, and vice chair for translational research, department of radiation oncology at the University of California, San Francisco, presented the research at the 2020 Genitourinary Cancers Symposium.

He told Urology Times®, “Our study is the first to investigate the performance of a clinical-grade genomic classifier in the context of a prospective, randomized post-prostatectomy trial, and the findings provide the strongest validation to date of its prognostic value.”

However, another main aim of the current analysis was to see whether the commercially available 22-gene biomarker panel could help guide decisions on implementing hormone therapy.

“Although NRG/RTOG 9601 showed a statistically significant survival benefit for men treated with bicalutamide compared to placebo-treated controls, only a minority of patients who received anti-androgen therapy benefited from this treatment,” Feng explained. “Therefore, the study’s result should not be interpreted to mean that hormone therapy be given to all men with PSA recurrences after prostatectomy.

“Our present analysis indicates that the genomic classifier can be a useful adjunct for clinical decision-making. In that regard, our study makes an important contribution to patient care as the management of prostate cancer moves toward treatment personalization.”

NRG/RTOG 9601 randomized 760 men receiving salvage radiotherapy to treatment with placebo or bicalutamide for 2 years. Patients were eligible for the study if they had pT3 or pT2 disease with a positive margin, no nodal involvement, and a prostate-specific antigen (PSA) of 0.2 to 4.0 ng/mL.

The study investigating the genomic classifier used formalin-fixed, paraffin-embedded prostatectomy specimens. Tissue from 352 of the study’s participants, representing equal numbers of men from the placebo and bicalutamide groups, were included in the analysis and demonstrated to be representative of the overall population.

The performance of the genomic classifier as an outcomes predictor was analyzed considering its generated score (0-1) as a continuous variable. In multivariable analyses adjusting for age, race, Gleason score, T-stage, margin status, entry PSA, and treatment arm, the genomic classifier was a statistically significant, independent predictor of distant metastases, prostate cancer-specific mortality, and overall survival (OS).


Hormone therapy benefit smaller in low-risk group

To analyze its potential benefit for identifying patients likely to benefit from hormone therapy, men were first divided into two groups based on their genomic classifier risk stratification: low risk (n = 183) vs intermediate or high risk (n = 169). The results showed that the benefit from hormone therapy was consistently smaller in the low-risk group vs the intermediate- or high-risk group, whether considering 12-year risk of distant metastases, 12-year risk of prostate cancer-specific mortality, or 12-year OS.

Findings from additional analyses that included only men who received early salvage radiotherapy when the PSA was less than 0.7 ng/mL (n = 175) showed that in this particular subgroup, men categorized as low risk according to the genomic classifier appeared to derive no benefit from hormone therapy when considering any of the three outcomes.

“Although the results were not statistically significant, the data showed that the 12-year overall survival was numerically worse for men in the low-risk group if they received hormone therapy than if they were treated with placebo. This suggests that, if we give hormone therapy to all patients with PSA recurrences after surgery, we may be overtreating those with low Decipher scores and low PSAs at the time of initiation of radiation therapy,” Feng said.

Disclosures: Dr. Feng has consulted for Astellas, Bayer, Blue Earth Diagnostics, Celgene, Clovis Oncology, Janssen, Roivant Sciences, and Sanofi. For full disclosures, see bit.ly/276disclosures.