IntroductionMixed phenotype acute leukemia (MPAL) is a rare subgroup of acute leukemia characterized by blasts presenting both myeloid and lymphoid lineages. Standard of care for MPAL is not defined and prognosis of MPAL is poor. Genetic basis for mixed immunophynotype and distinct molecular characteristics from other leukemia is not well understood.MethodsWe studied 31 bone marrow samples from previously untreated MPAL patients (median age 53) meeting WHO criteria. Genomic DNA/RNA from bone marrow samples were analyzed by comprehensive genomic analysis including targeted capture sequencing of 295 genes (median 393x coverage), RNA sequencing, and Infinium methylation EPIC array (Illumina). Copy number alterations (CNA) were analyzed from methylation array data using ChAMP algorithm. Mutational landscape of MPAL was compared with194 AML, 71 B-ALL, and 14 T-ALL cases sequenced with the same platform. CpG methylation pattern of MPAL was also compared with 194 AML (TCGA), 505 B-ALL and 101 T-ALL cases (data shared by Nordlund et al. Genome Biology. 2013).ResultsOf the 31 MPAL cases, 18 (58%) had myeloid-T and 13 (42%) had myeloid-B phenotype. Four (13%) cases had Philadelphia chromosome, one (3%) had 11q23 rearrangement, and 8 (26%) had complex karyotypes. Eighteen patients (62%) were treated with ALL regimen, 8 (28%) received AML regimen, and 3 (10%) received other therapy.Targeted sequencing revealed 86 driver mutations (55 SNVs and 31 indels) in 35 genes in 29 samples (94%). MPAL carried median 2 (range 0-6) mutations per case that was similar to AML (P = 0.79) and T-ALL (P = 0.92) but was significantly higher than B-ALL(P < 0.001). The most frequently mutated genes were DNMT3A , IDH2 , NOTCH1 , and RUNX1 in 7 (23%) cases each (Figure 1). Myeloid-T and myeloid-B showed distinct patterns of mutations, in which mutations in NOTCH pathway genes had strong trend of enrichment in myeloid-T (P = 0.06), whereas RUNX1 mutations were significantly enriched in myeloid-B (P = 0.02). When mutation landscape was compared among AML, B-ALL, T-ALL and MPAL, TP53 , NRAS , and DNMT3A mutations were shared across all leukemia. In contrast, NPM1 mutation was specific to AML and mutations in NOTCH1 , FBXW7 , and IL7R were specific to myeloid-T MPAL and T-ALL. CNA analysis showed that CDKN2B as a significantly deleted gene in MPAL.Unsupervised hierarchical clustering of promoter methylation revelaed two distinct clusters. This methylation cluster had significant correlation with immunophenotype (73% of cluster 1 was myeloid-B and 88% of cluster 2 was myeloid-T, P < 0.001). Overall, myeloid-T had more hypermethylated CpG loci than myeloid-B (mean beta value 0.28 vs. 0.52 in myeloid-B vs. myeloid-T, P = 0.002). This difference in methylation was independent from IDH mutation. Molecular pathway analysis of differentially methylated probes (DMP) showed that T-cell receptor pathway, T cell differentiation, and activation pathway genes (CD3D , CD247 , LCK , PRF1 , RHOH , and PRKCQ ) were differentially methylated and were expressed between myeloid-B and myeloid-T. Furthermore, motif enrichment analysis of promoter DMP revealed that IRF8 and IRF4 binding motifs were significantly hyper-methylated in myeloid-T compared to myeloid-B. Consistent with this, RAG1 , RAG2 , CXCR4 and PAX5 , downstream targets of IRF4/IRF8 andessential genes in B-cell development, were differentially expressed between two phenotypes. RNA sequencing revealed two known translocations (NSD1-NUP98 , KMT2A-MLLT4 ) in addition to the novel translocations such as FOXP1-DNAJC15 , TNKS - LYZ , and NOP14 - PLEC in MPAL. We did not detect ZN384 fusions that have been reported in pediatric MPAL (Alexander et al. ASH 2016). Unsupervised hierarchical clustering of methylation pattern among MPAL, AML, B-ALL and T-ALL revealed that 41% and 59% of MPAL had AML-like and ALL-like methylation pattern, respectively (Figure 2). Patients who received therapy that matched with their methylation pattern had significantly better response to treatment than patients who received unmatched therapy (CR rate: 77% vs. 31%, P = 0.024).ConclusionMyeloid-T and myeloid-B MPAL showed distinct patterns of mutation, methylation and gene expressions. Methylation data could identify AML-like and ALL-like MPAL which corresponded to better response to therapy when matched therapy was given. These data suggest that molecularly guided therapy may be beneficial in the treatment of MPAL. [Display omitted] DisclosuresTakahashi:Symbio Pharmaceuticals: Consultancy. DiNardo:Agios: Honoraria, Research Funding; Novartis: Honoraria, Research Funding; Celgene: Honoraria, Research Funding; Daiichi-Sankyo: Honoraria, Research Funding; AbbVie: Honoraria, Research Funding. Jabbour:Bristol-Myers Squibb: Consultancy. Andreeff:Daiichi Sankyo: Consultancy. Kantarjian:Bristol-Meyers Squibb: Research Funding; Amgen: Research Funding; ARIAD: Research Funding; Pfizer: Research Funding; Delta-Fly Pharma: Research Funding; Novartis: Research Funding.