EP239/#800 Prognostic analysis of splenic metastasis in advanced ovarian cancer: does parenchymal metastasis matter?

Abstract

group, whereas 131 proteins were in the poor response group. Proteins significantly upregulated in the good response group included ribosomal-and infection-related proteins. Proteins significantly upregulated in the poor response group included extracellular matrix receptor-and coagulation-related proteins. To identify a protein signature that stratifies good and poor responders to PARP inhibitors, we performed four feature selection algorithms with leave-one-out cross-validation to improve the accuracy. High expression of Proteins A and B were associated with worse and better progression-free survival, respectively. Conclusions We successfully identified protein signatures associated with response to PARP inhibitors. This study was the most extensive proteomic analysis to predict PARP inhibitor response in ovarian cancer.