Objective: The natural history, treatment options, and clinical outcomes of pancreatic metastases to the brain remain largely unknown. Here, we seek to investigate characteristics that influence OS in pancreatic metastases to the brain.
Methods: This is a population-based retrospective study of OS in 508 patients with pancreatic metastases to the brain using the SEER database. Univariate and multivariate Cox regression analyses were utilized, and a predictive nomogram was developed.
Results: There were 508 patients identified for this study, with a median OS of 2 months. In the univariate analysis, patients older than 65 years had significantly reduced OS (P < 0.001). Patients with liver metastases (P < 0.001) and liver and lung metastases (P < 0.001) exhibited significantly reduced OS. Treatment of the primary tumor with chemotherapy only (P < 0.001), radiation only (P = 0.01), radiation and chemotherapy (P < 0.001), and surgery only (P = 0.01) were associated with increased OS. Resection of a distant metastasis site (P = 0.009) and of a brain metastasis (P = 0.03) were associated with increased OS. In the multivariable analysis, factors that remained significant included patient age (P = 0.01), liver metastases (P < 0.001), liver and lung metastases (P < 0.001), treatment with chemotherapy (P < 0.001), treatment with radiation and chemotherapy (P < 0.001), and treatment with surgery and chemotherapy (P < 0.001). The nomogram had a C-index of 0.766, suggesting congruence between the findings on the nomogram and the results in the internal verification.
Conclusions: Median OS is influenced by age, multiorgan metastases, and treatment of the primary tumor. These data highlight the marginal benefit of treatment, yet improved quality of life (QOL) remains to be elucidated.
Keywords: Brain metastases; Nomogram; Pancreatic cancer; Prognosis.
Copyright © 2023 Elsevier Inc. All rights reserved.