Your AI-Trained Oncology Knowledge Connection!
The predictive value of PD-L1 expression on response to chemoimmunotherapy was greater on metastatic lesions vs primary tumors in HGSOC.
The predictive value of PD-L1 expression was found to be greater when it was assessed on metastatic lesions compared with primary tumor cells in high-grade serous ovarian cancer (HGSOC), according to data from an analysis of the phase 2 NeoPembrOV/GINECO trial (NCT03275506).1 The abstract was selected as the best fellowship paper during the
Findings from the analysis showed that the progression-free survival (PFS) HR interaction in metastatic samples per PD-L1 immune cell (IC) score was 0.46 (95% CI, 0.22-0.96; P = .05) in favor of the pembrolizumab arm compared with the control arm. The PFS HR interaction in metastatic samples per tumor proportion score (TPS) was 0.71 (95% CI, 0.31-1.6; P = .41) in favor of the investigational arm. Comparatively, the PFS HR interactions by IC score and TPS in primary tumor samples were 1.9 (95% CI, 0.48-7.4; P = .36) and 0.77 (95% CI, 0.24-2.4; P = 66).
“PD-L1 [analysis] was successfully performed on 56 metastatic samples and on 29 primary tumors,” Laetitia Collet, MD, MSc, a medical oncologist and PhD student at Institut Jules Bordet and in Brussels, Belgium, said during the presentation. “[PD-L1–expressing tumor cells] were more frequently observed in tubo-ovarian primary tumors in 66% of cases vs 34% of metastases.”
NeoPembrOV/GINECO was an open-label, non-comparative study that enrolled women with newly diagnosed FIGO stage IIIC/IV HGSOC who were not suitable for primary debulking surgery.1,2 Other key inclusion criteria included being at least 18 and a maximum of 75 years old, having an ECOG performance status of 2 or less, and having adequate hepatic, renal, and bone marrow function.2
Eligible patients were randomly assigned 2:1 to receive pembrolizumab at 200 mg every 3 weeks plus standard-of-care (SOC) carboplatin plus paclitaxel or SOC alone prior to surgery.1,2 Patients received 2 to 5 cycles of the same chemotherapy doublet after surgery; those in the investigational arm also received maintenance pembrolizumab with or without bevacizumab for up to 15 months after surgery.
The primary end point was independently assessed complete resection rate at interval debulking surgery.1 Secondary end points included overall response rate, progression-free survival, and overall survival.
To conduct their analysis, Collet and her coauthors examined post-treatment samples following debulking surgery. They performed bulk RNA sequencing, multiplex immunofluorescence, PD-L1 expression analysis via the Ventana SP263 assay, and spatial transcriptomic analysis.
The objectives of the analysis were to characterize the tumor microenvironment in tubo-ovarian primary tumors and metastases and to assess predictive biomarkers of response to chemo-immunotherapy according to sample origin (primary tumor vs metastasis).
Additional findings demonstrated that metastases were enriched in immune cells and immune-associated pathways. PD-L1 expression was also found to be associated with immune enrichment in metastatic lesions and was also associated with immune evasion and pathways related to proliferation and metabolism in tubo-ovarian primary tumors.
Moreover, tumor vessel signature was found to be predictive to response to pembrolizumab. Samples with tumor lysis syndrome (P = .014), high CD3 (P < .001), and high CD20 (P = .004) were all found to be associated with extended OS. The colocalization of tumor vessels impaired the tumor microenvironment.
Disclosures: Collet received the ESMO Translational Research Fellowship. She also received travel expenses from AstraZeneca, MSD, Pharmamar, and GSK, and was an invited speaker for AstraZeneca and GSK.
