Introduction ARID1A mutation (ARID1Amut) has a high incidence in OCCC (up to 60%) and EC (up to 40%), with evidence as a negative prognostic marker for treatment resistance and outcomes. EZH2 inhibition in ARID1Amut solid tumors results in tumor growth inhibition (Bitler et al. Nat Med 2015;21:231–238). Preliminary Phase II (NCT04104776) efficacy, safety, and biomarker findings from OCCC and EC cohorts receiving tulmimetostat are reported.

Methods The Phase II study is evaluating tulmimetostat 350 mg once daily in 6 disease-based cohorts, including ARID1Amut OCCC/EC. Per Simon 2-stage design, expansion of enrolment (plus n=19 patients per cohort in Stage 2) requires objective response rate (ORR) ≥1/10 in Stage 1. Primary endpoint is ORR; secondary endpoints include safety. Evaluation of two additional dose levels was implemented for both cohorts, per FDA recommendation of Project Optimus, to inform on optimal tulmimetostat dose.

Results 24 patients were enrolled (OCCC, n=14; EC, n=10); 50% of each cohort have received ≥3 prior treatment lines. Both cohorts are eligible for Stage 2 expansion, with 1 and 2 confirmed partial responses in patients with OCCC and EC, respectively (table 1). The manageable safety profile across all 6 tumor cohorts (n=81) was consistent with known class effects; Grade ≥3 related adverse events (≥10% of patients) included thrombocytopenia, anemia, neutropenia, and diarrhea. Next generation sequencing did not reveal a specific hotspot for ARID1Amut locations impacting clinical outcome in patients with OCCC/EC.

Conclusion/Implications These preliminary findings in heavily pre-treated patients with ARID1Amut OCCC/EC support continued investigation of tulmimetostat monotherapy.