Risk factors and treatment outcomes in metastatic prostate cancer with brain metastases.

Abstract

62 Background: Prostate cancer with brain metastases (PCBM) are rare and carry poor prognosis. While prior studies suggest neuroendocrine (NE) histology and prolonged hormone therapy (HT) may contribute to PCBM, risk factors and optimal treatment strategies for PCBM are poorly understood. Methods: The National Cancer Database was queried for patients diagnosed with metastatic prostate cancer (mPCa) from 2010-2021. Demographics and tumor characteristics were compared between patients with and without PCBM. PCBM incidence and risk factors were assessed with temporal analysis and multivariable logistic regression. Overall survival (OS) was analyzed using Kaplan-Meier and Cox proportional hazards model. Subgroup analyses examined combinations of systemic (HT, chemotherapy, and immunotherapy) and local treatments including surgery, whole brain radiation (WBRT), and stereotactic radiosurgery (SRS). Results: Of 231,702 patients with mPCa, 2,562 (1.1%) had PCBM. No significant temporal increase in PCBM incidence was observed. PCBM more often had NE histology, other visceral metastases (particularly liver), and high Gleason grade. NE histology (aOR: 5.77, 95% CI: 4.90-6.79), liver metastases (aOR: 2.26, 95% CI: 2.06-2.49), and other visceral metastases (aOR: 2.47, 95% CI: 1.10-5.54) were independently associated with PCBM. Patients with PCBM had worse OS (15.1 vs 32.6 months, p<0.001). HT was associated with improved OS compared to no systemic therapy (17.2 vs 4.2 months, p<0.001). Immunotherapy and HT were associated with improved OS compared to HT alone (42.6 vs 17.2 months, p=0.002). In multivariable analysis, immunotherapy had a positive OS benefit (aHR: 0.20, 95% CI: 0.05-0.75) while WBRT was associated with worse OS (aHR=3.38, 95% CI: 1.69-6.76). Chemotherapy, surgery, and SRS had no significant OS benefit after controlling for confounding. Conclusions: PCBM is associated with aggressive disease features such as NE histology. However, we observed no significant temporal increase in incidence despite increasing NHA adoption. Liver metastases were associated with PCBM risk, warranting consideration of further brain imaging in these patients. Durable survival appears primarily dependent on systemic disease control, with local therapies potentially offering benefit in select cases. These findings underscore the need for molecular stratification and advocate for the inclusion of PCBM patients in future clinical trials. Multivariable Cox proportional hazards model for overall survival in PCBM. Variable aHR p-value Age (Cont) 1.035 0.002 CCI (Ref = 0) 1 ≥2 1.6871.557 0.0650.229 cT Stage (Ref <cT4) cT4 1.360 0.179 Histo (Ref = Adeno) NE 3.447 0.003 Other Mets (Ref = None) ≥1 Organ 1.696 0.119 HT (Ref=No) Yes 1.071 0.830 Chemo (Ref=No) Yes 0.780 0.327 Immuno (Ref = No) Yes 0.197 0.017 Surgery of Distant Site (Ref=No) Yes 1.588 0.262 SRS (Ref=No) Yes 0.386 0.198 WBRT (Ref = No) Yes 3.380 <0.001