Clinical outcomes of prexasertib monotherapy in recurrent wild-type high-grade serous ovarian cancer involve innate and adaptive immune responses.

BACKGROUND

Preclinical data suggest cell cycle checkpoint blockade may induce an immunostimulatory tumor microenvironment. However, it remains elusive whether immunomodulation occurs in the clinical setting. To test this, we used blood and fresh tissue samples collected at baseline and post therapy from a phase II trial of the cell cycle checkpoint 1 inhibitor (CHK1i) prexasertib in recurrent ovarian cancer.

METHODS

Paired blood samples and fresh core biopsies, taken before treatment was started at baseline (cycle 1 day 1 (C1D1)) and post second dose on day 15 of cycle 1 (C1D15), were collected. To evaluate changes in the immune responses after treatment, multiparametric flow cytometry for DNA damage markers and immune cell subsets was performed on paired blood samples. RNA sequencing (RNAseq) of paired core biopsies was also analyzed. Archival tissue immune microenvironment was evaluated with immunohistochemistry. All correlative study statistical analyses used two-sided significance with a cut-off of p=0.05.

RESULTS

Flow cytometric analysis showed significantly increased γ-H2AX staining after CHK1i treatment, accompanied by increased monocyte populations, suggestive of an activated innate immune response (median 31.6% vs 45.6%, p=0.005). Increased expressions of immunocompetence marker HLA-DR (Human Leukocyte Antigen DR antigen) on monocytes and of a marker of STING (stimulator of interferon genes) pathway activation, in biopsies were associated with improved progression-free survival (PFS) (9.25 vs 3.5 months, p=0.019; 9 vs 3 months, p=0.003, respectively). Computational analysis of RNAseq data indicated increased infiltration of tumor niches by naïve B-cells and resting memory T-cells, suggestive of a possibly activated adaptive immune response, and greater T-reg infiltration after treatment correlated with worse PFS (9.25 vs 3.5 months, p=0.007). An immunosuppressive adaptive immune response, perhaps compensatory, was also observed on flow cytometry, including lymphodepletion of total peripheral CD4+ and CD8+T cells after CHK1i and an increase in the proportion of T-regs among these T-cells. Additionally, there was a trend of improved PFS with greater tumor-infiltrating lymphocytes (TILs) in archival tissues (13.7 months >30% TILs vs 5.5 months ≤30% TILs, p=0.05).

CONCLUSION

Our study demonstrates that a favorable clinical response in high-grade serous ovarian carcinoma patients treated with CHK1i is possibly associated with enhanced innate and adaptive immunity, requiring further mechanistic studies. It is supportive of current efforts for a clinical development strategy for therapeutic combinations with immunotherapy in ovarian cancer.