Randomized Phase II Study of Trabectedin/olaparib Compared to Physician’s Choice in Subjects with Previously Treated Advanced or Recurrent Solid Tumors Harboring Dna Repair Deficiencies

Genomic aberrations affecting the repair of DNA double-strand breaks (DSB) by homologous recombination (HR) are found in various cancers and result in sensitivity to inhibitors of the DNA repair enzyme PARP1. Trabectedin induces DNA DSB and PARP1 activation and may thus increase the effect of PARP inhibitors in HR-deficient cancers. Whole-exome/genome sequencing (WES/WGS) can identify mutations in DNA repair pathways and mutational signatures generated by these aberrations. We present a randomized phase II trial comparing trabectedin and olaparib (TrO) with treatment of physician’s choice (PC) in adults with advanced/metastatic cancers with defective HR DNA repair (“BRCAness”), as determined by WES/WGS. Molecular eligibility is determined based on a newly developed BRCAness score. Main exclusion criteria are hematologic/primary brain cancers, ECOG PS >1, platinum-refractory disease, and prior PARP inhibitor treatment. Patients are randomized 1:1 to TrO (days 1/1-21) vs. PC, until progressive disease (PD). Cross-over upon PD is allowed. The primary endpoint is the disease control rate (DCR [RECIST 1.1]) at 16 weeks. As of 04/2022, 85 patients have been randomized (TrO, n=42; PC, n=43). Various entities were included (sarcoma, n=39; cholangiocarcinoma, n=6; gynecologic cancer, n=7; uveal melanoma, n=5; other, n=28); prior treatment was balanced between arms. Main AE associated with TrO were cytopenias, infections, and gastrointestinal side effects, requiring dose reductions of trabectedin in more than 50%. In both arms, a median of 3 treatment cycles were applied; PD was the main cause of treatment termination. Cross-over to TrO occurred in 58%. The planned interim analysis after assessment of the primary endpoint after 30 patients showed a comparable DCR at week 16 in the TrO (38.5%) and PC (36.8%) arms while the DCR after cross-over was 55.6%. Based on a 51% conditional power to achieve superiority of TrO as initially planned, the study was continued. The trial successfully passed interim analysis and is expected to be completed in 2022.