Study: Traditional prostate cancer risks don’t predict hormone therapy benefit

A large analysis led by researchers at the UCLA Health Jonsson Comprehensive Cancer Center has challenged a widely used assumption in postoperative prostate cancer care: the pathological features that identify patients at high risk of recurrence may not reveal who will benefit most from adding hormone therapy to radiation.

The study, published in European Urology, examined whether conventional findings in prostate tissue removed during surgery could predict the effectiveness of androgen-deprivation therapy when combined with postoperative radiotherapy. The investigators found that these features were strongly associated with a patient’s overall prognosis, but they did not identify a subgroup that gained a greater survival advantage from hormone therapy.

The distinction is clinically important. After prostatectomy, some men have detectable or rising prostate-specific antigen, or PSA, or pathological evidence suggesting that cancer could return. Radiation therapy may then be directed at the surgical bed and nearby tissues. Hormone therapy, also called androgen-deprivation therapy, suppresses testosterone and other androgen signals that can stimulate prostate cancer cells. Although the combination can improve outcomes in selected patients, hormone therapy may also cause hot flashes, fatigue, sexual dysfunction, loss of bone density, weight gain and metabolic changes.

To investigate whether pathology could guide that decision, the research team assembled individual patient data from five phase 3 randomized clinical trials. Together, the studies included 4,781 men who had undergone prostate cancer surgery and subsequently received radiation therapy, either alone or with hormone therapy. Pooling patient-level data allowed the researchers to examine treatment effects across clinically relevant subgroups rather than relying only on broad results from separate trials.

The investigators created an adverse feature count based on four pathological findings associated with aggressive disease. These included Grade Group 4 or 5 cancer, invasion of the seminal vesicles, extension of cancer beyond the prostate and positive surgical margins, in which malignant cells are found at the edge of the removed tissue. Each patient received a score from zero to four according to the number of features present.

The results confirmed that the score was prognostic. In other words, men with more adverse features generally faced a higher risk of cancer spreading, or metastasizing, and a higher risk of death. High-grade tumors reflect more abnormal cellular behavior, while extraprostatic extension and seminal vesicle invasion indicate that the tumor has penetrated beyond the prostate’s main boundaries. Positive margins suggest that microscopic disease may remain after surgery.

Yet the same score was not predictive of treatment benefit. Statistical analyses found no evidence that men with multiple adverse pathological features experienced a larger improvement in overall survival or metastasis-free survival when hormone therapy was added to radiation. Patients with fewer risk factors did not appear to be the only group benefiting, nor did those with the most aggressive pathological profiles show a clearly amplified response.

This finding illustrates a central concept in precision oncology. A prognostic biomarker estimates how a disease is likely to behave regardless of treatment, while a predictive biomarker indicates whether a particular therapy is more or less likely to work. The pathological features examined in this analysis functioned as prognostic markers, but they did not demonstrate the treatment-interaction effect required to guide hormone therapy decisions.

The pattern remained consistent among men whose PSA level was 0.5 nanograms per milliliter or less before radiation. Previous research has suggested that the overall advantage of adding hormone therapy may be limited in this setting, making the question of patient selection especially relevant. The new analysis indicates that traditional postoperative pathology does not solve that selection problem, even when several high-risk features are present.

“These findings highlight an important distinction between predicting risk and predicting treatment benefit,” said lead investigator Dr. Amar Kishan, professor and executive vice chair of radiation oncology at the David Geffen School of Medicine at UCLA. Kishan said the results point to the need for molecular biomarkers capable of revealing tumor biology and identifying which patients are genuinely sensitive to hormone suppression. Such tools could help clinicians intensify treatment for men likely to benefit while sparing others from avoidable toxicity.

The study’s senior author was Dr. Daniel Spratt of University Hospitals Seidman Cancer Center and Case Western Reserve University. The UCLA research team included Tahmineh Romero-Kalbasi, Dr. Michael Steinberg, Dr. Luca Valle, Dr. Kekoa Taparra, Dr. Matthew J. Farrell, Dr. Matthew Rettig, Dr. Adam Singer, Dr. Nikita Baclig, Dr. Robert Reiter, Dr. Scott Eggener, Dr. Wayne Brisbane and Dr. Nicholas Nickols. The investigators’ conclusion is not that pathology has lost its value, but that risk classification alone cannot determine treatment response. Future approaches may combine pathological assessment with genomic signatures, immune markers, imaging and other molecular measurements to make postoperative prostate cancer treatment more precise.

News Publication Date: Not provided.

Web References: UCLA Health Jonsson Comprehensive Cancer Center: https://www.uclahealth.org/cancer ; Dr. Amar Kishan: https://www.uclahealth.org/providers/amar-kishan

References: European Urology, DOI: 10.1016/j.eururo.2026.07.013