Introduction/Background*In the randomized phase II NeoPembrOV study (NCT03275506), Pembrolizumab in combination with neoadjuvant chemotherapy (NACT) met its primary endpoint of complete debulking rate (CRR) for the treatment of patients with advanced up-front non-resectable high-grade serous ovarian cancer (HGSOC). However, the CRR in the control group was similar. Identification of potential predictive biomarkers is fundamental to better understand the place of Pembrolizumab in this setting.

Methodology 91 Patients (pts) with HGSOC unable to received complete upfront debulking surgery were included and received Carboplatin (AUC5) Q3W + Paclitaxel (175mg/m²) Q3W +/- Pembrolizumab 200 mg Q3W IV before and after surgery. Pembrolizumab was given until 24 months maximum. After interval debulking surgery, optional bevacizumab was proposed. BRCA1/2 mutation status (BRCA) was assessed using standard algorithms. Immunohistochemical PD-L1 expression was evaluated on both tumour and immune cells (IC) using the Ventana SP263 assay. Associations of progression-free survival (PFS) with BRCA, and PD-L1 expression were evaluated.

Result(s)* BRCA status was available for 81 pts (89%). 3 out of 30 pts (10%) in the control arm harboured a BRCA mutation (mBRCA) versus 13/61 in the experimental arm (21.3%). Median PFS (mPFS) in both arms in the BRCA wild-type (wtBRCA) subgroup were not different (mPFS 20.8 months [95% IC,15.0-25.7] vs 18.2 months [95%IC, 16.8.0-21.4] in control and experimental arms respectively). mPFS in the mBRCA subgroup were not reached in both arms. PD-L1 expression was available for 85/91 patients (93.4%). PD-L1 IC ≥5% was positive in 29/85 patients (34.1%) and correlated to mPFS in the whole population (18.2m in PDL1 IC<5% vs 23.4m PD-L1 IC ≥5% respectively, p=0.02). mPFS was similar HR: 1.39 [0.71-2.74] in patients with PD-L1 IC<5% (mPFS 19.3mo [14.9-25.7] vs mPFS 18.2mo [14.5-19.3] in control and experimental arms respectively). Pembrolizumab improved mPFS (HR: 0.56 [0.19-1.61]) for pts with PD-L1≥5% (mPFS 20.8 mo [9.5-NE] vs mPFS 23.4mo [18.0-NE] in control and experimental arms respectively).

Conclusion*If no benefit in adding Pembrolizumab to CT +/- bevacizumab was found in wtBRCA subgroup, exploratory PFS analyses in the PD-L1 IC ≥5% subgroup showed a trend favouring Pembrolizumab in patients with advanced HGSOC.