Traditional Prostate Cancer Risk Assessment Scales Do Not Predict Outcomes from Brain Metastases: A Population-Based Predictive Nomogram

TL;DR

This study finds that traditional prostate cancer risk tools don't predict brain metastasis outcomes, and develops a new survival prediction model based on factors like race, tumor size, and treatment.

Contribution

A novel nomogram is developed to predict survival in prostate cancer brain metastases, independent of traditional prostate cancer risk metrics.

Findings

Median overall survival for brain metastases from prostate cancer is 15 months.

Hispanic patients and those with smaller tumors or no additional metastases had better survival.

Traditional prostate cancer metrics like Gleason and ISUP grading are not predictive of survival in brain metastases.

Abstract

Brain metastases from systemic cancer are the most common tumors of the central nervous system. For prostate metastases to the brain, the clinical progression is poorly understood. This retrospective study aims to elucidate clinical risk factors associated with overall survival (OS; months post-diagnosis) in prostate metastases to the brain, and then develop a nomogram to aid in clinical decision making for this vulnerable population. We identified several factors associated with survival, including race, tumor size, and the presence of additional metastases. This study should serve as a clinical framework for prognostication in metastatic prostate cancer to the brain.

Brain metastases are an uncommon yet life-limiting manifestation of prostate cancer. However, there is limited insight into the natural progression, therapeutics, and patient outcomes for prostate cancer once metastasized to the brain. This is a retrospective study of 461 patients with metastatic prostate cancer to the brain with a primary outcome of median overall survival (OS). The Surveillance, Epidemiology, and End Results (SEER) database was examined using Cox regression univariate and multivariable analyses, and a corresponding nomogram was developed. The median overall survival was 15 months. In the multivariable analysis, Hispanic patients had significantly increased OS (median OS 17 months, p = 0.005). Patients with tumor sizes greater than three centimeters exhibited significantly reduced OS (median OS 19 months, p = 0.014). Patients with additional metastases to the liver exhibited significantly reduced OS (median OS 3.5 months, p < 0.001). Increased survival was demonstrated in patients treated with chemotherapy or systemic treatment (median OS 19 months, p = 0.039), in addition to radiation and chemotherapy (median OS 25 months, p = 0.002). The nomogram had a C-index of 0.641. For patients with prostate metastases to the brain, median OS is influenced by race, tumor size, presence of additional metastases, and treatment. The lack of an association between traditional prostate cancer prognosis metrics, including Gleason and ISUP grading, and mortality highlights the need for individualized, metastasis-specific prognosis metrics. This prognostic nomogram for prostate metastases to the brain can be used to guide the management of affected patients.

Figures