Background

Clinical outcomes for rare subtypes of relapsed gynaecological cancers (GC) including ovarian clear cell carcinomas (CCC) and carcinosarcomas (CS) are poor with limited treatment options. Preclinical data demonstrate GC with deleterious ARID1A mutations (and thus ARID1A-protein loss) display increased sensitivity to ATR inhibition. In the absence of ARID1A mutations, ATR inhibitor sensitivity can be enhanced by combination with PARP inhibitors. ATARI is an academic, international, parallel cohort platform phase II trial assessing ceralasertib (CERA) in ARID1A stratified GC.

Methods

Relapsed CCC (ovarian OCCC/endometrial ECCC) patients (pts) with ARID1A loss received CERA alone (160 mg BD day (D) 1-14 of 28D cycle) (Cohort 1A); CCC with ARID1A no loss (Cohort 2) and other non-CCC histological subtypes (endometrioid, CS, cervical) CERA (160 mg OD D1-7) and olaparib (300 mg BD, D1-28) (Cohort 3). Primary endpoint is best overall objective response rate (ORR, RECIST v1.1); key secondary endpoints include disease control rate (DCR) and progression free survival (PFS). Each cohort recruited 29 pts under a Simon 2-stage design (p0=0.1, p1=0.3, alpha=5%, power=80%, 6+ responses to be observed).

Results

We report efficacy results of the first 29 evaluable pts per cohort (Table 1). Grade 3+ toxicities (safety population) were 47% (1A n=36), 41% (2 n=32), 39% (3 n=33); most common anaemia (1A 36%, 2 19%,3 27%) & any other <10%. Discontinuation rates due to AEs were ≤10% in all cohorts. Preliminary translational results will be presented.

Table 1: Antitumour activity results

Conclusions

Celarasertib alone or in combination with olaparib has relevant clinical activity in rare gynae cancers. Demonstrable activity for combination in the non-clear cell ‘basket’ cohort, including carcinosarcoma, provides hypothesis-generating evidence for further investigation.

Clinical trial identification

ClinicalTrials.gov NCT04065269

Release date August 22 2019