Comprehensive Transcriptome Profiling of Cryptic CBFA2T3-GLIS2 Fusion-Positive AML Defines Novel Therapeutic Options — a COG and Target Pediatric AML Study

Abstract The cryptic fusion CBFA2T3-GLIS2 was initially described in infants with acute megakaryoblastic leukemia (AMKL) and is associated with adverse outcome. This fusion was subsequently shown to be present in other AML subsets with similar adverse associations. In order to define genes and pathways that are dysregulated and define therapeutic targets in these high-risk patients, diagnostic specimens from 1,049 children and young adults with AML underwent whole transcriptome sequencing (mRNA and miRNA Seq). Initial interrogation of the mRNA transcriptome identified 25 patients (2.4%) with the cryptic CBFA2T3-GLIS2 fusion. Patients with CBFA2T3-GLIS2 were significantly younger (median age 1.51 vs 10.2 years, p < 0.001), with 92% of fusion-positive patients < 3 years old, and an overall incidence of 9% in all patients < 3 years (23/255, p < 0.001). Half of all CBFA2T3-GLIS2 cases were classified as AMKL (13/25, 52%). Patients with CBFA2T3-GLIS2 lacked other risk features and the overwhelming majority (88%, p < 0.001) were classified as standard risk by conventional cytogenetic and molecular classifications. However, fusion-positive patients had an overall survival of 10% compared to 67.3% for the fusion-negative cohort (hazard ratio (HR) = 3.31, p < 0.001, Fig. 1A), with similar adverse event-free survival (HR = 2.48, p < 0.001). Genomic interrogation of fusion-positive patients by whole genome or targeted exome capture sequencing showed a lack of recurrent somatic mutations or other structural events in this patient population (Bolouri et al. 2017). Transcriptional profiling contrasting CBFA2T3-GLIS2 positive (N = 25) and negative cohorts (N = 1,024) identified several differentially enriched pathways, including TGFB/BMP (FDR < 0.001), WNT (FDR < 0.001), and Hedgehog signaling (HH, FDR = 0.002). Fusion-positive cases highly expressed GLIS2, the HH pathway genes, HHIP, PTCH1, and GLI1, as well as BMP2 (Fig. 1B). In addition, CASP1 and TRAIL-R2, which are involved in apoptosis, as well as the tumor suppressor GLIPR1 were significantly down-regulated. Integration of the miRNA with mRNA transcriptome demonstrated that expression of both CASP1 and TRAIL-R2 were significantly anti-correlated with and both are putative targets of the overexpressed miR-181b-5p; while downregulation of GLIPR1 was associated with an increase in miR-130a-3p, as well as being a predicted target (Fig. 1C). Gene-set enrichment analysis revealed highly up-regulated cell-adhesion and cell-surface markers, including extracellular matrix binding (FDR < 0.001), cell-adhesion molecule binding (FDR < 0.001), and integrin binding genes (FDR < 0.001). Investigating the most highly differentially expressed genes (90th percentile log2 fold-change (LFC), FDR < 0.001) with demonstrated localization to the plasma membrane and extracellular matrix, showed that CBFA2T3-GLIS2 exhibits a unique expression profile with a distinct set of dysregulated genes with plasma membrane localization (Fig. 1D). One of the most highly expressed cell-surface genes included NCAM1 (aka CD56; LFC = 6.97, FDR < 0.001). Given the extreme level of CD56 transcript expression in fusion-positive cases, we studied CD56 protein level using multidimensional flow cytometry (Hematologics, Seattle, WA). Surface CD56 in fusion-positive cases was massively over-expressed, with a median mean fluorescent intensity (MFI) of 1,960.0 versus 9.0 MFI for those without the fusion (p < 0.001, Fig. 1E). We evaluated CD56 as a potential therapeutic target using an anti-CD56 antibody-drug conjugate (m906-PBD-ADC) developed by NCI. Leukemic blasts or control lymphocytes from a patient with relapsed fusion-positive AML were incubated with varying doses of m906 and cytotoxicity was assessed after 72 hours (Notable Labs, Foster City, CA). The CD56-ADC exhibited a dose-dependent and CD56-specific toxicity on leukemic blasts (LIN-CD33+CD14-CD38-, p < 0.001, Fig. 1F), suggesting that CD56 might prove to be a therapeutic target in this high-risk cohort of patients. This study provides comprehensive transcriptome profiling of CBFA2T3-GLIS2 that defines a distinct and highly lethal subset of AML predominantly seen in infants. We also identify pathways and networks that further provide potential therapeutic options. Most importantly, CD56 may present the most immediately actionable target in this cohort of high risk patients. Figure 1. Figure 1. Disclosures Gekas: Notable Labs: Employment. Kolb:Roche- Genentech: Membership on an entity's Board of Directors or advisory committees; Servier: Membership on an entity's Board of Directors or advisory committees. Eidenschink Brodersen:Hematologics, Inc: Employment. Loken:Hematologics, Inc: Employment, Equity Ownership.