Using Patient-Derived Xenograft (PDX) Models as a ‘Black Box’ to Identify More Applicable Patients for ADP-Ribose Polymerase Inhibitor (PARPi) Treatment in Ovarian Cancer: Searching for Novel Molecular and Clinical Biomarkers and Performing a Prospective Preclinical Trial

Abstract

(1) The accuracy of patient-derived xenografts (PDXs) in predicting ADP-ribose polymerase inhibitor (PARPi) efficacy in ovarian cancer was tested, novel biomarkers were investigated, and whether PARPis could replace platinum-based chemotherapy as a first-line therapy was explored. (2) PDXs were reconstructed for 40 patients with ovarian cancer, and niraparib, olaparib and paclitaxel, and carboplatin (TC) sensitivity tests were conducted. Whole exon sequencing and homologous recombination deficiency (HRD) scores were performed, and patient clinical information was collected. The molecular biomarkers were identified by reverse-transcription quantitative PCR and immunoblotting. (3) Niraparib and olaparib sensitivity were tested in 26 patients and showed high consistency. Approximately half of BRCA wild-type, HRD-negative, and platinum-resistant patients may benefit from PARPis. AKT1 enrichment indicated PARPi resistance; high KRAS expression indicated PARPi sensitivity. CA125 below 10 U/mL during chemotherapy has a sensitivity and specificity similar to platinum sensitivity in predicting PARPi efficacy. Niraparib and TC sensitivity tests were performed on 23 patients, and TC showed a better response in this preclinical trial. (4) PDX can indicate individualized PARPi efficacy. Decreased CA125 levels and KRAS and ATK1 expression levels may be novel biomarkers. The preclinical evidence does not support the implementation of PARPis as the first-line treatment in an unselected population.