(P 212) REPORT ON THE EWING SARCOMA PATIENTS ENROLLED IN ARM D OF THE ACSE-ESMART TRIAL: PHASE I/II STUDY OF THE PARP INHIBITOR OLAPARIB IN COMBINATION WITH IRINOTECAN IN CHILDREN WITH ADVANCED MALIGNANCIES

Objective: Ewing sarcoma (ES) has been identified as a PARP inhibitor (PARPi) sensitive cancer in preclinical studies. The ES-specific fusion protein itself – best investigated for EWSR1-FLI1 – has been shown to be linked to this sensitivity by mimicking a BRCA deficient phenotype; however, the full mechanism of this sensitivity remains unknown. Despite this preclinical sensitivity, PARPi monotherapy in relapsed/refractory ES patients was unsuccessful and several combination therapies including PARPi - mainly in combination with chemotherapy - have been and are being investigated. AcSé-ESMART is a proof-of-concept, phase I/II, platform trial, designed to explore targeted agents in a molecularly enriched relapsed/refractory pediatric population. Arm D was evaluating the PARPi olaparib in combination with the topoisomerase I inhibitor irinotecan. In contrast to other PARPi/chemotherapy combination studies, we opted for a prolonged course of PARPi and low dose irinotecan as sensitizer. The Phase I part established the recommended Phase II dose (RP2D) (Gatz ASCO 2019). The Phase II part of the trial assessed the activity in two separate expansion cohorts (cohort 1: homologous recombination repair defect (HRD) and cohort 2: ES). We here report the results of all ES patients enrolled in this study.

Methods: Comprehensive molecular profiling (WES -/+ RNA sequencing) (ideally from the relapsed tumour) was required as part of the inclusion criteria for all patients. Patients with ES were eligible based on the presence of an ES-specific fusion. Dose escalation followed a continuous reassessment method (CRM) with a dose limiting toxicity (DLT) threshold of ≤0.25. In the escalating dose levels Olaparib was administered orally twice daily at 90-130 mg/m2 on Days 1 to 10 and irinotecan intravenously at 10-30 mg/m2 on Days 4 to 8 of a 21-day cycle. Starting dose level (DL1) was Olaparib 90 mg/m2 on Days 1 to 10 and irinotecan 10 mg/m2 on Days 4 to 8 of a 21-day cycle. Adverse events (AEs) were reported based on National Cancer Institute Common Toxicity Criteria (NCI-CTCAE) version 4.03. Activity assessment in the two expansion cohorts followed a Minimax Simon 2-stage design. Each cohort was to progress to the second stage (additional 9 patients) if 2 or more confirmed responses were observed in the first 16 patients. Patients treated in the Phase I part at the RP2D were counting towards the respective expansion cohorts.

Results: From October 2016 till July 2022 seventy patients (median age: 14 years, range 5-23) in total were included in the study, 36 of them were ES patients; there were four screen failures – all were ES patients: three were not treated due to rapid progressive disease, in one case the patient declined treatment. Four dose levels were explored in the phase I part to identify the RP2D (DL1, DL2, DL3 and DL2a). The RP2D was DL2 (Olaparib 90 mg/m2 and irinotecan 20 mg/m2).
Of the 32 treated ES patients, eight were enrolled in the Phase I part (DL1 n=3; DL2 n=2; DL3 n=2; DL2a n=1) and 24 in the Phase 2 part. Six patients had DLTs, all at DL2 (Phase I part n=1; Phase 2 part, n=5) for an overall DLT rate at DL2 in the ES cohort alone of 0.23. Main AEs including the DLTs were hematological and gastroinstestinal and not different from the non-ES patients. All patients had an ES fusion (ESWR1::FLI1 n=25; EWSR1::ERG n=4; EWSR1 translocation by FISH n=2; data missing n=1). Median number of treatment cycles were 2, range 2;40+. Three patients had objective responses, two of them confirmed: one patient - treated on DL1 - had an unconfirmed partial response (PR) and received 6 cycles; on DL2, one patient had a complete response (10 cycles) and one a PR; this patient is ongoing in cycle 40. Nine patients had stable disease (SD) as best response, three of them prolonged with 6, 10, 16 cycles, respectively (all in DL2). Retrospective correlative analysis of the molecular profiling data and additional tumor tissue expression analysis are ongoing to identify potential constellations linked with these exceptional responses and rapid progression, respectively.

Conclusion: The combination of Olaparib and irinotecan in this protracted schedule is overall well tolerated. Encouraging clinical benefit was observed in a subset of ES patients with one patient remaining on treatment 2.5 years after enrolment. Current molecular hypothesis is insufficient for patient selection and better biomarkers are needed.