FcγR3A polymorphism influences natural killer cell activation and response to anti-PD-L1 (avelumab) in gestational trophoblastic neoplasia

Abstract

Background: Low-risk gestational trophoblastic neoplasia are currently receiving monochemotherapy as first-line therapy. In the case of a resistance, a second-line monochemotherapy or polychemotherapy is proposed. As an alternative to these toxic and historic chemotherapy agents, the efficacy of the anti-PD-L1 monoclonal antibody (avelumab) was assessed in the TROPHIMMUN phase II trial Cohort A. Avelumab yielded a 53% cure rate with an acceptable tolerance profile, including normal further pregnancy and delivery. Beyond the blockade of PD-1/PD-L1 interactions, avelumab effect could rely on the induction of antibody-dependent cell-mediated cytotoxicity mediated by FcγR3A-expressing natural killer cells.

Objective: This translational study aimed at testing whether antibody-dependent cell-mediated cytotoxicity is involved in avelumab efficacy on gestational trophoblastic neoplasia and if FcγR3A affinity polymorphism could help predicting the response to avelumab in gestational trophoblastic neoplasia.

Study design: The expression of PD-L1 by the tumor and the phenotype of natural killer cells infiltrating gestational trophoblastic neoplasia were verified by performing transcriptomic and proteomic analyses. Then, JEG-3 choriocarcinoma cells were cocultured with human natural killer cells in the presence and absence of avelumab. The impact of FcγR3A functional polymorphism was assessed on the activation status of natural killer cells and the viability of JEG-3 choriocarcinoma cells. Finally, the data from TROPHIMMUN trial were re-analyzed to determine the impact of the FcγR3A polymorphism of patients on their response to avelumab.

Results: We confirmed that FcγR3A+ natural killer cells infiltrated PD-L1-expressing gestational trophoblastic neoplasia. In vitro, avelumab-coated JEG-3 choriocarcinoma cells induced natural killer cell activation, which promoted the destruction of JEG-3 cells. Natural killer cell activation was abolished when the Fc portion of avelumab was removed, demonstrating the importance of Fcγ receptor in this process. Using this model of antibody-dependent cell-mediated cytotoxicity, we demonstrated that high-affinity FcγR3A polymorphism on natural killer cells was associated with better in vitro response to avelumab. In line with this result, patients from the TROPHIMMUN trial homozygous for the high-affinity FcγR3A polymorphism had better clinical response to avelumab.

Conclusion: Our work demonstrates that antibody-dependent cell-mediated cytotoxicity contributes to the therapeutic effect of avelumab in gestational trophoblastic neoplasia and that the individual patient response is impacted by the FcγR3A polymorphism. The FcγR3A polymorphism could be used as a biomarker to identify patients diagnosed with monochemoresistant gestational trophoblastic neoplasia who are most likely to respond to avelumab.

Keywords: FcγR3A polymorphism; NK cells; anti-PD-L1; antibody-dependent cell-mediated cytotoxicity; avelumab; gestational trophoblastic neoplasia; immunotherapy.