Therapy of PARP inhibitor-associated myeloid neoplasms in ovarian cancer patients

ABSTRACT Therapy-related myeloid neoplasms (tMNs), including myelodysplastic syndrome (MDS) and acute myeloid leuk- emia (AML), are severe, often rapidly fatal bone marrow disorders that occur in a small fraction of solid tumor patients who have received cytotoxic treatment. Ovarian cancer (OC) survivors have one of the highest inci- dences of tMNs among all cancer patients, reflecting both an increased risk due to platinum exposure and a recently recognized further increase in risk with PARP inhibitor (PARPi) therapy. Collectively, these PARPi-as- sociated myeloid neoplasms are approximately as frequent as acute promyelocytic leukemia, a rare AML sub- type that previously had a high fatality rate but has become among the most curable myeloid neoplasms through the concerted efforts of the research community. In this context, our team has contributed to studies showing that a disproportionate number of PARPi-associated tMNs in OC patients have biallelic inactivation of the TP53 tumor suppressor gene. Like other TP53-mutant myeloid neoplasms, tMNs in OC patients receiving PARPis respond poorly to conventional chemotherapy and relapse quickly, with a median survival of 4-8 months, high- lighting the need for improved therapies. In preliminary studies, we have shown that replication checkpoint mod- ulators, including inhibitors of the kinases CHK1, WEE1 and ATR, exhibit antileukemic activity in AML cell lines, xenografts, and primary AML samples, including TP53-mutant AMLs. This activity reflects TP53-independent induction of apoptosis at therapeutically achievable concentrations through at least two distinct apoptotic path- ways, one of which involves leukemia cell-intrinsic transactivation of the TNF gene followed by TNF receptor- dependent caspase activation and the other of which involves the mitochondrial apoptotic pathway but is less completely defined. These observations lead to the hypothesis that replication checkpoint modulators, admin- istered as monotherapy or in combination, will have antileukemic effects in the tMNs that develop in OC patients. To further advance the study of these agents, we now propose to define the second pathway by which replication checkpoint modulators induce apoptosis in TP53-mutant AML, thereby providing deeper insight into determi- nants of tMN sensitivity to this emerging class of drugs (Aim 1); determine the antineoplastic effects of replication checkpoint modulators in vitro, alone and in combination with hypomethylating agents and BH3 mimetics (two classes of current standard-of-care AML agents), in samples of PARPi-emergent tMNs from OC patients (Aim 2); and assess the efficacy of replication checkpoint modulators in vivo, alone and in combination with hypomethyl- ating agents or BH3 mimetics, in patient-derived xenografts generated from PARPi-emergent AMLs of OC pa- tients (Aim 3). Collectively, these studies are designed to provide new insight into the antineoplastic action of replication checkpoint modulators and simultaneously position our multidisciplinary team to perform a future early phase clinical trial of replication checkpoint modulator-based therapy in OC patients with PARPi-emergent tMNs.