Abstract IA07: Strategies to eliminate drug tolerant neuroblastoma persister cells

Abstract

High-risk neuroblastoma remains a leading cause of pediatric cancer mortality, and relapse—driven by chemotherapy-resistant "persister" tumor cells that survive induction therapy—accounts for the majority of these deaths. Our recent single-nucleus and whole-genome sequencing analysis of matched diagnosis and definitive-surgery samples from 20 patients with high-risk neuroblastoma identified persister cells characterized by suppressed MYC(N) activity, activation of NF-κB signaling, and cell-cell communication with the tumor microenvironment that reinforces this dormant, non-cycling state. Because these cells evade cytotoxic chemotherapy through cell-cycle arrest rather than classical drug efflux or target mutation, eliminating them requires therapeutic strategies that do not depend on active proliferation. This talk will present an integrated framework spanning four complementary approaches, each addressing a distinct vulnerability of the persister state, with the unifying principle that intervention must occur early—during minimal residual disease—rather than after clinical relapse, when persister clones have already expanded and diversified. First, we showed that lineage-defining surfaceome profiling across adrenergic and mesenchymal neuroblastoma cell states identifies "stable" targets such as B7-H3 that resist downregulation during the persister transition, unlike state-restricted antigens, supporting adoptive cell therapies designed to track tumor cells through phenotypic drift. Second, building on peptide-centric CAR platforms, we present unpublished data demonstrating that despite reduced PHOX2B transcript and peptide-HLA density in persister cells, residual antigen presentation remains sufficient for potent PC-CAR-mediated killing, informing our ongoing PHOX2B PC-CAR T-cell trial (NCT07007117). Third, we will show that a BCL-xL degrader selectively induces apoptosis in dormant, cell-cycle-arrested persister populations by exploiting their heightened dependence on this anti-apoptotic node, representing a senolytic strategy distinct from cytotoxic chemotherapy. Fourth, we discuss a synthetic lethal approach exploiting chromosome 17q/TRIM37 gain, a defining genomic feature of high-risk neuroblastoma, using the PLK4 inhibitor RP-1664 to induce centriole amplification and lethal multipolar mitosis; this agent showed activity in 14 of 15 xenograft models and produced maintained complete responses when combined with GD2-directed chemoimmunotherapy in a MYCN-driven murine model. Together, these data argue that durable cures in high-risk neuroblastoma will require rational sequencing or combination of stable-antigen immunotherapy, senolytic targeting of dormancy programs, and/or genomically-informed synthetic lethal agents, deployed during the minimal residual disease window rather than at relapse, when persister-derived clones have already acquired additional resistance mechanisms.