Heather L. Schuback

The genomic and clinical information used to develop and implement therapeutic approaches for acute myelogenous leukemia (AML) originated primarily from adult patients and has been generalized to patients with pediatric AML. However, age-specific molecular alterations are becoming more evident and may signify the need to age-stratify treatment regimens. The NCI/COG TARGET-AML initiative used whole exome capture sequencing (WXS) to interrogate the genomic landscape of matched trios representing specimens collected upon diagnosis, remission, and relapse from 20 cases of de novo childhood AML. One hundred forty-five somatic variants at diagnosis (median 6 mutations/patient) and 149 variants at relapse (median 6.5 mutations) were identified and verified by orthogonal methodologies. Recurrent somatic variants [in (greater than or equal to) 2 patients] were identified for 10 genes (FLT3, NRAS, PTPN11, WT1, TET2, DHX15, DHX30, KIT, ETV6, KRAS), with variable persistence at relapse. The variant allele fraction (VAF), used to measure the prevalence of somatic mutations, varied widely at diagnosis. Mutations that persisted from diagnosis to relapse had a significantly higher diagnostic VAF compared with those that resolved at relapse (median VAF 0.43 vs. 0.24, P < 0.001). Further analysis revealed that 90% of the diagnostic variants with VAF >0.4 persisted to relapse compared with 28% with VAF <0.2 (P < 0.001). This study demonstrates significant variability in the mutational profile and clonal evolution of pediatric AML from diagnosis to relapse. Furthermore, mutations with high VAF at diagnosis, representing variants shared across a leukemic clonal structure, may constrain the genomic landscape at relapse and help to define key pathways for therapeutic targeting. Cancer Res; 76(8); 2197-205. ©2016 AACR.

Abstract The cryptic CBA2T3-GLIS2 fusion generated by the inv(16)(p13.3q24.3) was initially identified in megakaryocytic leukemia and later implicated in other acute myeloid leukemia (AML) subtypes. Presence of this fusion may lead to altered expression of the potentially targetable sonic hedgehog and bone morphogenic protein pathways. We determined the prevalence of CBA2T3-GLIS2 in children treated on COG AAML03P1 and AAML0531 protocols, which collectively enrolled 1361 eligible children, adolescents, and young adults with de novo AML, and correlated the presence of this fusion with patient demographics, laboratory features, and clinical outcomes. We also determined the prevalence and clinical implications of CBA2T3-GLIS2 in 71 children with FAB M7 AML treated on 4 consecutive COG AML trials. Of the 1042 diagnostic samples available and tested for CBFA2T3-GLIS2, 45 (4.3%) were positive for the fusion. Fusion-positive patients were significantly younger than fusion-negative patients (2.1 vs. 10.3 years; P&lt;0.001). CBFA2T3-GLIS2 was most prevalent in the youngest patients (10.6% for 0 to &lt;2 year olds and 8.6% for 2 to &lt;5 year olds) [Figure A]. In contrast, no fusion transcripts were identified in 299 unselected adult patients. All FAB subtypes were represented in fusion-positive patients. Overall, FAB M5 and M7 were equally prevalent in fusion-positive patients, and each subtype accounted for 20% of cases [Figure B]. There was a preponderance of MLL rearrangements (N=7) in fusion-positive patients, and 3 more patients had either t(8;21) or inv(16), and 25.6% (N=11) without karyotypic alterations (CN-AML). None of the fusion-positive patients had the t(7;12) or 12p abnormality. There were few common AML-associated mutations: 1 patient had FLT3ITD and 1 had the WT1 mutation (no NPM1 or biallelic CEBPA mutations were identified). Figure 1 Figure 1. Rates of morphologic complete remission (CR) at the end of induction course 1 were similar for fusion-positive and -negative patients (68.9% vs. 77.7%; P=0.17). However, fusion-positive patients were more likely to have minimal residual disease (MRD) by flow cytometry at this time point (50% vs. 28.9%; P=0.006) with a correspondingly higher relapse rate (RR) from remission of 58% vs. 35% (p=0.005). Disease-free survival for those with and without fusion was 42% vs. 58% (p=0.060) In a subset analysis of 193 patients with CN-AML, the prevalence of CBFA2T3-GLIS2 was 4.7%. Fusion-positive patients were younger than fusion-negative patients (1.6 vs. 13.1 years; P&lt;0.001) and more likely to have MRD at the end of induction (85.7% vs. 40.8%; P=0.043). CN-AML fusion-positive patients had significantly worse 5-year OS (36% vs. 67% P=0.025) and EFS (18% vs. 51%, P=0.017) than fusion-negative patients. All fusion-positive patients in CR had a higher 5-year RR than fusion-negative patients (88% vs. 33%; P&lt;0.001), with a corresponding disease-free survival (DFS) of 13% vs. 59% (P&lt;0.001). Implications of CBFA2T3-GLIS2 were evaluated in 71 cases of children with FAB M7 AML where the fusion was identified in 12 patients (17%). Fusion-positive FAB M7 patients had significantly lower CR rate than fusion-negative patients (33.3% vs. 77.6%, P=0.005) and all were MRD positive at the end of induction (100% vs. 30%, P=0.001). All FAB M7 fusion-positive patients relapsed (100% vs. 36%, P=0.007), with a DFS of 0% vs. 60% (p=0.013). As CBFA2T3-GLIS2 was most common in younger patients, we compared clinical implications in children &lt;2 years of age. Although CR rates in fusion-positive and -negative patients were similar (68.2% vs. 74.9%; P=0.50), fusion-positive patients were more likely than fusion-negative patients to have MRD at the end of course 1 (63.2% vs. 24.8%, P=0.006). Five-year RR rates in fusion-positive and -negative patients were 71% vs. 39% (P=0.012), with corresponding DFS of 29% vs. 56% (P=0.030). This study provides a comprehensive evaluation of incidence and prognostic implications of the CBFA2T3-GLIS2 fusion in pediatric de novo AML. Along with MLL rearrangements, 12p abnormalities, and the recently described NUP98-JARID1A fusion, this cryptic inversion 16, as defined by presence of the CBFA2T3-GLIS2 fusion, represents a distinct, recurrent chromosomal abnormality associated with poor prognosis in infant AML and is a potential therapeutic target. Disclosures No relevant conflicts of interest to declare.

&lt;p&gt;Clonal cluster analysis. Mutations were clustered using DBSCAN based on the VAF and graphed according to timepoint. Panels A-T illustrate possible scenarios of how individual mutations may originate, evolve, and resolve based on VAF.&lt;/p&gt;

&lt;div&gt;Abstract&lt;p&gt;The genomic and clinical information used to develop and implement therapeutic approaches for acute myelogenous leukemia (AML) originated primarily from adult patients and has been generalized to patients with pediatric AML. However, age-specific molecular alterations are becoming more evident and may signify the need to age-stratify treatment regimens. The NCI/COG TARGET-AML initiative used whole exome capture sequencing (WXS) to interrogate the genomic landscape of matched trios representing specimens collected upon diagnosis, remission, and relapse from 20 cases of &lt;i&gt;de novo&lt;/i&gt; childhood AML. One hundred forty-five somatic variants at diagnosis (median 6 mutations/patient) and 149 variants at relapse (median 6.5 mutations) were identified and verified by orthogonal methodologies. Recurrent somatic variants [in (greater than or equal to) 2 patients] were identified for 10 genes (&lt;i&gt;FLT3, NRAS, PTPN11, WT1, TET2, DHX15, DHX30, KIT, ETV6, KRAS&lt;/i&gt;), with variable persistence at relapse. The variant allele fraction (VAF), used to measure the prevalence of somatic mutations, varied widely at diagnosis. Mutations that persisted from diagnosis to relapse had a significantly higher diagnostic VAF compared with those that resolved at relapse (median VAF 0.43 vs. 0.24, &lt;i&gt;P&lt;/i&gt; &lt; 0.001). Further analysis revealed that 90% of the diagnostic variants with VAF &gt;0.4 persisted to relapse compared with 28% with VAF &lt;0.2 (&lt;i&gt;P&lt;/i&gt; &lt; 0.001). This study demonstrates significant variability in the mutational profile and clonal evolution of pediatric AML from diagnosis to relapse. Furthermore, mutations with high VAF at diagnosis, representing variants shared across a leukemic clonal structure, may constrain the genomic landscape at relapse and help to define key pathways for therapeutic targeting. &lt;i&gt;Cancer Res; 76(8); 2197–205. ©2016 AACR&lt;/i&gt;.&lt;/p&gt;&lt;/div&gt;