A phase 2, randomized, double-blind, placebo- controlled study of chemo-immunotherapy combination using motolimod with pegylated liposomal doxorubicin in recurrent or persistent ovarian cancer: a Gynecologic Oncology Group partners study

Abstract

Background: A phase 2, randomized, placebo-controlled trial was conducted in women with recurrent epithelial ovarian carcinoma to evaluate the efficacy and safety of motolimod-a Toll-like receptor 8 (TLR8) agonist that stimulates robust innate immune responses-combined with pegylated liposomal doxorubicin (PLD), a chemotherapeutic that induces immunogenic cell death.

Patients and methods: Women with ovarian, fallopian tube, or primary peritoneal carcinoma were randomized 1 : 1 to receive PLD in combination with blinded motolimod or placebo. Randomization was stratified by platinum-free interval (≤6 versus >6-12 months) and Gynecologic Oncology Group (GOG) performance status (0 versus 1). Treatment cycles were repeated every 28 days until disease progression.

Results: The addition of motolimod to PLD did not significantly improve overall survival (OS; log rank one-sided P = 0.923, HR = 1.22) or progression-free survival (PFS; log rank one-sided P = 0.943, HR = 1.21). The combination was well tolerated, with no synergistic or unexpected serious toxicity. Most patients experienced adverse events of fatigue, anemia, nausea, decreased white blood cells, and constipation. In pre-specified subgroup analyses, motolimod-treated patients who experienced injection site reactions (ISR) had a lower risk of death compared with those who did not experience ISR. Additionally, pre-treatment in vitro responses of immune biomarkers to TLR8 stimulation predicted OS outcomes in patients receiving motolimod on study. Immune score (tumor infiltrating lymphocytes; TIL), TLR8 single-nucleotide polymorphisms, mutational status in BRCA and other DNA repair genes, and autoantibody biomarkers did not correlate with OS or PFS.

Conclusions: The addition of motolimod to PLD did not improve clinical outcomes compared with placebo. However, subset analyses identified statistically significant differences in the OS of motolimod-treated patients on the basis of ISR and in vitro immune responses. Collectively, these data may provide important clues for identifying patients for treatment with immunomodulatory agents in novel combinations and/or delivery approaches.

Trial registration: Clinicaltrials.gov, NCT 01666444.

Trial registration: ClinicalTrials.gov NCT01666444.

Keywords: Toll-like receptor 8; biomarkers; immunotherapy; motolimod; oncology; ovarian cancer.

Figure 1.

Kaplan–Meier plots of OS (A) and PFS (B) in the ITT population. No significant increase in either OS or PFS was observed between patients treated with motolimod plus PLD (blue line) and patients treated with placebo plus PLD (red line). Landmark analysis of OS in patients receiving motolimod, comparing those who experienced injection site reaction (ISR+; blue line) to those who did not (ISR−; red line) (C). Among motolimod-treated patients, those who were ISR+ had longer OS compared with those who were ISR−.

Figure 2.

(A) Statistically significant (two-sided P < 0.05) upregulation of plasma analytes following subcutaneous administration of 3.0 mg/m² of motolimod. Analytes that were prospectively identified as being responsive to TLR8 stimulation by motolimod are identified with a dark gray heading. Data shown represent log-fold changes (pre-dose to 8 h post-dose) in each analyte. P-values for each analyte are provided in supplementary Table S3, available at Annals of Oncology online. (B) In vivo upregulation of immunologically important analytes. P-values for each analyte are provided in supplementary Table S4, available at Annals of Oncology online.

Figure 3.

(A) Statistically significant (two-sided P < 0.05) upregulation of analytes in vitro from baseline (pre-treatment) assessment of immune responsiveness to TLR8 stimulation. Data shown represent log-fold changes in each analyte in response to in vitro stimulation with 300 nM motolimod compared with null (no stimulant). Corresponding P-values are provided in supplementary Table S5, available at Annals of Oncology online. Analytes that were prospectively identified as being responsive to in vitro activation of TLR8 by motolimod are identified with a dark gray heading. (B) Analysis of OS by baseline in vitro response in IFN-γ (A), IL-12p40 (B), and TNF-α. (C) In vitro responses for each analyte was discretized as low, medium, or high and their association with survival was tested using a one-sided log-rank test. Within the motolimod plus PLD treatment group only, patients with higher IFN-γ (P = 0.049), TNF-α (P = 0.041), or IL-12p40 (P = 0.024) response had significantly improved survival compared to those motolimod-treated subjects who had a lower level of mediator response.