SAN FRANCISCO, September 28, 2025 — A new randomized study finds that a lab test that reads tumor genes can identify which patients with recurrent prostate cancer will benefit from adding hormone therapy to radiation after surgery — the first predictive biomarker in this setting.
In the first prospective, randomized trial to validate a predictive gene expression test for hormone therapy in prostate cancer, patients with a prostate tumor subtype known as luminal B had much lower risks of recurrence and metastasis when radiation was complemented by apalutamide, a type of hormone therapy. Patients without this tumor subtype, however, saw no improvement. Results of the phase II BALANCE trial (NRG Oncology GU006) will be presented today at the American Society for Radiation Oncology (ASTRO) Annual Meeting .
"We've been searching for decades for a way to determine which patients are most likely to respond to hormone therapy after prostatectomy," said Daniel Spratt, MD, principal investigator of the trial and professor and chair of radiation oncology at University Hospitals Seidman Cancer Center and Case Western Reserve University School of Medicine. "We now have a tool that lets us tailor treatment based on a tumor's biology and recommend hormone therapy only for those patients who we think can expect to see a benefit."
Prostate cancer is the second most common cancer worldwide, with more than 314,000 new diagnoses expected in the U.S. this year. Patients whose tumors are confined to the prostate gland have several treatment options including definitive radiation therapy or surgical removal of the prostate, known as radical prostatectomy. But for up to 30% of patients, the disease will recur or persist, often signaled by rising prostate-specific antigen (PSA) levels. For patients with a rising PSA level after prostatectomy, radiation therapy is the standard treatment and has been shown to improve survival.
Hormone therapy is commonly added to radiation in this setting to block or reduce testosterone, a hormone that fuels prostate cancer growth. While it can enhance the effects of radiation and improve cancer control for some patients, it also carries a wide range of side effects, including fatigue, bone loss, hot flashes, metabolic changes and cardiovascular risk.
"Testosterone is important for maintaining bone, muscle, cognitive and cardiac health, but it's also the key fuel driving prostate tumors," said Dr. Spratt. "Until now, we haven't had a reliable way to tell who really needs hormone therapy and who does not."
The answer came from PAM50, a gene expression test that was originally developed to guide breast cancer treatment and adapted to capture prostate cancer biology by Dr. Spratt and his colleagues Felix Feng, MD, FASTRO, and Shuang Zhao, MD.
Prostate tumors, like breast tumors, can be grouped into molecular subtypes; in this study, tumors were classified as either luminal B or non-luminal B. Luminal B tumors grow more quickly and are highly responsive to hormone therapy. Non-luminal B tumors, including luminal A and basal-like subtypes, are generally less dependent on testosterone and may not respond to hormone-based treatment. Dr. Spratt compared the advance to breast cancer, where estrogen receptor status helps guide endocrine therapy decisions.
In the BALANCE trial, 295 patients with recurrent prostate cancer and no signs of metastasis were enrolled at cancer centers across the U.S. All patients had undergone prostatectomy and were experiencing a rising PSA (86% with entry PSA of <0.5 ng/mL). They were randomized to receive a standard course of radiation therapy with six months of either apalutamide (a second-generation anti-androgen therapy) or placebo.
Tumors were tested for their PAM50 subtype and grouped as luminal B (n=127) or non-luminal B (n=168). The primary endpoint was biochemical progression-free survival, a measure of whether the cancer returned based on PSA levels, recurrence, metastasis or death. Median follow-up was 5 years.
Patients with luminal B tumors saw a significant benefit from hormone therapy. Their five-year biochemical progression-free survival was 72.4% with apalutamide, compared to 53.9% with placebo (p=0.0062). In contrast, those with non-luminal B tumors saw no benefit, with nearly identical five-year biochemical control between the hormone and placebo arms (70.2% vs. 71.1%, p=0.44).
A similar pattern emerged for metastasis-free survival. Among luminal B patients, five-year survival without metastasis was 94.7% with hormone therapy and 81.8% with placebo (p=0.029). No difference was seen in the non-luminal B group (89.9% vs. 89.3%, p=0.90).
"The patients with luminal B tumors saw a large benefit, both in reducing recurrence and lowering the risk of metastatic disease," said Dr. Spratt. "But for patients without this subtype, hormone therapy didn't change the outcome. That's incredibly valuable information when we're trying to personalize care."
Dr. Spratt said he believes the findings will be practice changing.
"This is the first prospectively validated predictive biomarker in prostate cancer," he said. "It gives us a promising way to personalize care, recommending hormone therapy for those who respond, and avoiding unnecessary treatment when it is unlikely to help."
While the BALANCE trial was designed as a phase II biomarker-stratified study, Dr. Spratt said the results were so definitive that a phase III trial is unlikely. "The magnitude of benefit — and the complete lack of benefit in some patients — makes it unlikely that we could ethically enroll patients in a follow-up trial," he said. "PAM50 could now be used in recurrent prostate cancer to support shared decision-making."
