(P 303) PRECLINICAL EVALUATION OF PARP1 SELECTIVE INHIBITOR AZD5305 IN COMBINATION WITH DNA DAMAGING AGENTS IN EWING SARCOMA ORTHOTOPIC XENOGRAFT MODELS

Objective: Ewing sarcoma (EWS) is an aggressive malignancy of childhood with a dismal prognosis in patients with metastatic or recurrent disease. We previously demonstrated that EWS orthotopic patient-derived xenografts (O-PDXs) with the EWS-FLI1 translocation were sensitive to Poly (ADP ribose) polymerase inhibitors (PARPi) combined with DNA damaging agents which was translated to clinical trials in pediatric and adult patients. While responses were seen in patients with EWS, the combination of the PARP1/2 inhibitor talazoparib with irinotecan and/or temozolomide did have a narrow therapeutic window with concern for hematologic toxicity. AZD5305 is a next generation PARPi with 500-fold selectivity for PARP1 over PARP2. Previous studies have shown that PARP2 inhibition may play a role in the hematologic toxicity seen in animal models prompting the study of selective PARP1 inhibitor combinations as an attractive strategy to widen the therapeutic window. We present here the preclinical evaluation comparing the PARP1 selective inhibitor AZD5305 combined with DNA damaging agents nano-liposomal irinotecan (nal-IRI) and temozolomide (TMZ) to similar combinations with talazoparib.

Methods: EWS O-PDXs were created by injecting luciferase labeled EWS cells into the femur of athymic nude mice. Pharmacokinetic studies on tumor-bearing mice were performed to determine matched human AUC-guided dosing and schedule of chemotherapy agents. Preclinical phase I tolerability studies were performed for all drug combinations. EWS O-PDXs were randomized to the following five treatment groups: placebo, AZD5305, AZD5305 + nal-IRI, AZD5305 + nal-IRI + TMZ, and talazoparib + nal-IRI + TMZ. Enrollment was randomized for each of the O-PDXs tested. Four courses of therapy were given on a clinically relevant schedule with each course consisting of 21 days. Bioluminescent imaging was performed weekly to monitor tumor burden and mice were removed from the study and classified as progressive disease at any point when tumor approached 20% of body weight.

Results: Toxicity studies showed that AZD5305 was well tolerated in mice when combined with nal-IRI. Dose reduction was required for tolerability when combined with TMZ; however, this was less when compared to TMZ dose reductions required in talazoparib combinations. Preclinical efficacy studies showed that single agent AZD5305 had no effect on tumor inhibition when compared to placebo. In contrast, AZD5305 + nal-IRI and AZD5305 + nal-IRI + TMZ combinations had significant efficacy in EWS O-PDXs and mice survived all 4 courses of therapy, similar to that of talazoparib + nal-IRI + TMZ. Tumor burden in these treatment groups was not detected by bioluminescence at the end of the study and mice are currently undergoing an observation period to assess for durable response.

Conclusion: DNA damaging agents combined with PARPi have shown effect both preclinically and clinically in EWS; however, the narrow therapeutic window represents a significant challenge in treating patients. The combination of nal-IRI with talazoparib is currently being tested in early phase pediatric trials as one strategy to improve tolerability as nal-IRI combinations were shown in adult studies to have less grade 3/4 dose-limiting toxicities. The development of PARP1 selective agents such as AZD5305 may further mitigate toxicity and allow for dose escalation and potentially improved therapeutic benefit in this patient population. Further preclinical testing of AZD5305 in additional models is warranted along with combination testing with other novel agents such as ATR inhibitors to help prioritize treatments to move forward in future clinical trials.