ALBERT

Description: Abstract 3519 Pediatric Acute Myeloid Leukemia (AML) offers a unique window into the genesis and progression of hematologic cancers because of its rapid progression and limited confounding factors. As part of the NCI TARGET AML initiative (target.cancer.gov), we report the initial integrative analysis of 58 pediatric AML whole-genome sequences [WGS, performed by Complete Genomics Inc. (CGI)] along with 225 Affymetrix microarrays, clinical cytogenetic and biomarker tests. WGS was performed on matched samples collected at diagnosis and remission from each patient. Candidate variants were identified by CGI and defined as being present only in tumor samples. To focus on the highest confidence variant calls, we developed stringent filters based on a set of 281 true-positive and 44 false-positive verified variants. We identified a total of 629 high-confidence variants in 58 cases (∼ 11 per patient). 99 of these variants occurred in 39 genes (2 to 8 per gene, up to 4 per sample). Analysis of the potential functional consequence of these variants identified 28 as deleterious impacting 17 genes in 21 of the 58 relapsed AMLs, including multiple mutations in AML-associated genes such as KIT, NRAS, PTPN11, and WT1, and several key hematopoietic genes (e.g. IKZF1, GATA2). Using microarray expression data, we identified 1,992 (1,051) genes that were differentially expressed in pediatric AML cases compared to 4 normal bone marrow samples [FDR-adjusted p-value = 0.05 (0.01)]. 249 differentially expressed genes were more than 6 standard deviations away from the control average in more than half of the samples (e.g. WT1, MYCN, miR155), suggesting the existence of a shared set of dysregulated processes across most pediatric AMLs. Network-oriented enrichment analysis using the Bioconductor (http://bioconductor.org) package DEGraph revealed differentially regulated interactions among 3,496 genes, including 1,437 cancer genes and highly enriched interactions involving growth factor/RTK signaling, down-regulated immune processes, and up-regulated transcription and translation. To pinpoint processes specific to AML subtypes, we used the Bioconductor package WGCNA to cluster the 225 microarray expression datasets and align them with clinically-identified cytogenetic and mutation data. We identified five distinct expression clusters. Three of these clusters corresponded to known cytogenetic abnormalities: MLL fusions, t(8;21), and Inv16. The other two clusters were cytogenetically normal, but all members of one cluster carry CEBPA mutations. The expression patterns of cases with t(8;21) and Inv16, while distinct, shared a number of features that distinguished them from samples with MLL abnormalities. For example, in both t(8;21) and Inv16 cases (but not in MLL cases) the fibroblast growth factor (FGF) receptor FGFR1 and the FGF ligand FGF11 were over-expressed compared to control samples. These differences may explain the higher rates of remission associated with Core Binding Factor (CBF) abnormalities. To identify potential relationships between sequence variants and differentially expressed genes, we used the Graphite Bioconductor package to search four publicly available databases (Reactome, NCI PID, KEGG, Biocarta) for known interactions of our candidate deleterious genes. Surprisingly, all but two of our candidate deleterious genes shared many interactors, suggesting they impacted shared processes. The core connected components of the mutation interaction network were all members of the set of dysregulated interactions that we identified by gene expression analysis and included multiple members of well-known pathways implicated in AML and other cancers (e.g. RTK/growth factor signaling, JAK/STAT signaling). Candidate mutations impacted both shared and distinct pathways. Furthermore, within shared pathways, the candidate mutations impacted shared and distinct targets. These findings have important implications for pathway-specific drug targeting. Comparing the interactions of candidate WGS mutations with those of clinically-identified chromosomal abnormalities, a further pattern emerges: CBF and MLL associated gene fusions appeared to impact a different set of genes and processes compared to WGS (presumably second-hit) variants, suggesting complementary roles. This finding has important implications for ongoing research and testing of targeted treatments. Disclosures: No relevant conflicts of interest to declare.

Description: Although survival of children with B-cell acute lymphoblastic leukemia (B-ALL) has improved substantially over time, 15% to 20% of patients will relapse, and most of those who experience a bone marrow relapse will die. A better understanding of genetic and epigenetic aberrations in relapsed ALL will facilitate new strategies for risk stratification and targeted therapy. In this collaborative study with the Therapeutically Applicable Research to Generate Effective Treatments (TARGET) project, we performed high resolution genome-wide DNA methylation profiling using the HELP (HpaII tiny fragment Enrichment by Ligation-mediated PCR) array on a total of 178 (110 diagnosis, 68 relapse) leukemia samples from 111 patients with childhood B-ALL enrolled on the Children’s Oncology Group (COG) clinical trials who experienced relapsed, and 12 normal preB samples isolated from the bone marrows of 12 healthy individuals. The HELP array covers 117,521 CpG sites, annotated to ∼22,000 gene promoters. For eight diagnosis/relapse pairs, base-pair resolution DNA methylation using the eRRBS (enhanced Reduced Representation Bisulfite Sequencing) method was also performed on Illumina HiSeq2000. The median relapse time for the 111 patients was 21.8 months (range 2.1 to 56.2). Unsupervised clustering analysis using the HELP data revealed seven clusters: one cluster contained only the 12 normal preB samples; four clusters were enriched with MLLr, ETV6/RUNX1, Trisomy 4+10, and TCF3/PBX1 samples, respectively. The sixth cluster was not enriched for specific cytogenetic cases, but interestingly, all cases in this cluster were NCI High Risk (age〉10 years or WBC〉=50,000; p25%, FDR

Description: Abstract 123 Comprehensive cataloguing of somatic changes in acute myeloid leukemia (AML) has revealed novel molecular pathways that contribute to the development and progression of the disease, however many lack prognostic significance. The challenge that lies ahead is to understand how these molecular changes relate to transcription and protein deregulation to influence treatment responses and outcomes. Also, some mutations found in adult AML are detected at lower frequency or not at all in pediatric AML. To better define the genomic profile of somatic mutations in childhood AML and evaluate the extent of clonal evolution from diagnosis to relapse, we performed whole exome sequencing in matched trios of specimens (diagnostic, germline and relapse) from 20 children with AML who lacked known karyotypic high-risk features. Candidate variants were verified by 454 sequencing. In total, 364 somatic variants were successfully verified, of which 263 were non-silent mutations (127 at diagnosis and 136 at relapse). Of the 195 unique nonsynonymous mutations present in 180 genes, mutations in 6 genes were detected in more than one patient, including ETV6, KIT, KRAS, NRAS in 2 patients and TET2 and WT1 in 3 patients. A polyphen prediction estimated 72% of the missense and nonsense mutations as possibly and/or probably damaging. Mutations identified in single patients included known genes implicated in leukemias (e.g., GATA2, IKZF1), as well novel mutations in genes not previously implicated in AML. Prevalence and clinical significance of ETV6 mutations were assessed in a large cohort of pediatric AML. The entire coding sequence of ETV6 gene was sequenced in 180 cases of pediatric AML. ETV6 mutations, including missense (N=5), Indels (N=4) or splice site mutations (N=1) were detected in 6% of the patients without karyotypic or molecular risk factors. Those with and without ETV6 mutations had an overall survival of 33% and 71%, respectively (p=0.006) with a corresponding relapse rate of 80% and 29% (p=0.004). We examined the clonal evolution of mutations from diagnosis to relapse by comparing the mutation profile at diagnosis to that in relapse. In the discovery phase, at diagnosis, 234 somatic mutations in 202 genes were detected, of which 174 were non-silent events, including 36 (21%) indels, 126 (72%) missense, 11 (6%) nonsense mutations and 1 (1%) splice site mutation. At relapse, 350 mutations in 309 genes were detected, of which only 112 mutations were detectable at diagnosis (29%). In verification phase, 71 mutations in 70 genes were identified only in diagnostic specimens, and 105 mutations in 104 genes were detected in relapse only, with an overlap of 192 variants, suggesting significant molecular evolution and loss or gain of genomic events from diagnosis to relapse. With secondary targeted deep sequencing the overlap of mutations present at diagnosis and relapse increased to 52%. Comparison of allele fractions between diagnosis and relapse revealed a significant enrichment of the variant from diagnosis to relapse with evolution of a minor clone at diagnosis to a more dominant presence at relapse. It also highlights that in a substantial subset of patients, novel genomic events may emerge, leading to disease relapse. In addition, analysis of the allele fractions of SNPs in the normal and tumors of these patients allowed us to detect regions of recurrent loss of heterozygosity (LOH). Six of 20 patients exhibited LOH in one or more regions. The most common region of LOH, found in 4 patients (20%) was the q-arm of chromosome 13 (3 with FLT3/ITD), including 10 genes with germ line variation across 4 patients, and in two patients the q-arm of chromosome 16 was involved. Two of the 4 patients had LOH only in the relapse specimen. This study highlights the potential impact of applying novel sequencing approaches to pediatric AML and not only identifies clinically significant somatic mutations, but defines genomic complexity in childhood AML by demonstrating significant clonal evolution from diagnosis to relapse. Mutations emerging at relapse may represent initially rare, chemotherapy-resistant events that cooperate with other mutations and lead to adverse outcome. Identifying mutations that may be therapeutically targeted early in the course of therapy may provide improved outcomes for patients. Disclosures: No relevant conflicts of interest to declare.

Description: Abstract 1287 The NCI-directed TARGET AML initiative has provided an opportunity to evaluate the genomic, transcriptomic and epigenomic patterns of large numbers of pediatric patients with newly diagnosed and relapsed AML with the goal of identifying distinctive changes in pathways that are prognostic or can be therapeutically targeted. Genomic sequencing has revealed an important, but limited number of gene mutations, suggesting that epigenetic alterations may be required to explain the development and biological behavior of AML. We have analyzed genome-wide promoter methylation patterns using Illumina Infinium Methylation 27 arrays on 200 paired diagnostic (Dx) and remission (Rm) cases plus 47 case-matched relapsed (Rl) samples. Marked differences in DNA methylation were observed in the AML compared to remission bone marrows. DNA methylation patterns could discriminate specific types of cytogenetically characterized AML, but no subset of specific methylation changes were identified that predicted outcome across the entire cohort irrespective of AML subtype. Due to the strong correlations of methylation pattern with cytogenetically defined subtypes, such a global signature for outcome would be expected to require either a very strong signal or very large sample size to detect a prognostic, methylation signature that was able to validated in heterogeneous cohorts. We therefore evaluated prognostic methylation features within each cytogenetic group. Using a linear model and empirical Bayes statistic, selective groups of high quality methylation probe signals revealed an association with survival that independently predicted outcome within each cytogenetic subtype. Using a top-scoring pair (TSP) algorithm for each karyotype subtype, five different TSPs each for Inv16, MLL, t(8;21) and normal karyotype subgroups were identified that predicted outcome with an accuracy of 100% except for the t(8;21) subtype in which accuracy of predicting outcome was 87%. We used an alternative analytical approach to determine whether specific epigenetic promoter patterns were associated with diagnosis and/or relapse AML compared to remission bone marrows across all subtypes. These patterns could potentially be considered to be essential for AML development and/or progression. We divided patient data into balanced training and test sets. Using a novel pathway analysis based on outlier statistics, we identified methylation changes independent of cytogenetic subtype in gene promoters associated with specific cellular pathways. Cellular pathways of potential interest characterized by preferentially hypermethylated promoters of genes in the AML diagnostic samples compared to remission samples included Hedgehog, extracellular matrix receptor interactions, cell adhesion molecules, (NCAM/CAM) and phospholipase pathways. Those gene sets that showed hypomethylation in the AML diagnostic samples were selectively associated with cytochrome P450, tyrosine, and arachidonic acid metabolism. We generated a 10 gene signature from genes for two of the most significant pathways, i.e., the Hedgehog and cytochrome P450 metabolism pathways, using TSP and applied this TSP signature using a Fisher's exact test for both Dx vs Rm and Rl vs Rm samples (Table). Test Set Dx v Rm Call Dx Call Rm Test Rl v Rm Call Rl Call Rm True Dx 93 3 True Rl 24 1 True Rm 19 77 True Rm 7 18 OR 120, p 〈 2.2 × 10−16 OR 55, p 〈 8.1 × 10−7 The confirmation of the signature in the independent test set validated the significance of the Hedgehog and P450 pathways in newly diagnosed and relapsed pediatric AML. These data demonstrate the unique contribution of epigenetic changes as a prognostic indicator in subgroups of AML. Importantly, outlier analysis followed by pathway-based interrogation identified several key gene sets that are selectively epigenetically altered in an otherwise uncharacterized subset of pediatric AML. Of note, the hypermethylation and hypomethylation of gene promoters characterizing the Hedgehog and of promoters of cytochrome P450 pathways respectively was highly associated with AML diagnosis and relapse samples compared to remission marrows, suggesting a specific role for these pathways in the physiology and maintenance of pediatric AML. Work in progress is examining samples for mutations, gene copy number, expression and functional consequences of the pathways that are selectively altered by epigenetic mechanisms. Disclosures: No relevant conflicts of interest to declare.

Description: Abstract 125 Despite identification of genomic alterations in AML, clinically actionable variants are limited. Further, mutations identified in adult AML are less frequent or not found in childhood AML. TARGET AML initiative, a collaboration between National Cancer Institute Office of Cancer Genomics (NCI/OCG) and Children's Oncology Group (COG) provided resources to perform multi-platform analysis of a large number of pediatric AML cases in order to identify therapeutic targets in childhood AML. As part of this effort, diagnostic, remission and relapse specimens from a large cohort of children with AML were subjected to whole genome sequencing (WGS). In addition, all specimens underwent characterization by expression profiling, SNP genotyping, methylation profiling and miRNA sequencing. Specimens from a subset of patients also underwent whole exome sequencing (WES) and whole transcript sequencing (RNA Seq). Here we present data from an interim analysis of the data that identified clinically significant, novel somatic mutations, deletions and translocations involving the ETV6 gene in pediatric patients with AML. Non-silent somatic mutations of ETV6, including missense, splice site mutations and indels were identified by WGS and WES. Large deletions involving ETV6 gene were identified by WGS and SNP/CGH arrays, and novel fusion transcripts (cryptic translocations) involving ETV6 were identified by WGS and RNA Seq. We present the prevalence and clinical significance of genomic alterations of ETV6 that were identified as part of TARGET AML study. WGS and WES data were available from 54 and 22 pediatric AML patients, respectively. In these 76 selected cases, somatic mutations in ETV6 were identified and verified in 3 patients. Frequency validation was performed by sequencing of the entire coding region of the ETV6 gene in diagnostic specimens from 180 children with AML. Somatic mutations of ETV6 gene were identified in 6% of patients, of which none had cytogenetic or molecular risk features. Those with somatic ETV6 mutations had an overall survival (OS) of 20% vs. 76% for those without mutations (p=0.002) with a corresponding relapse risk (RR) of 80% vs. 29% (p=0.004), demonstrating the potential impact of ETV6 mutations in pediatric AML. Deletions involving ETV6 gene (del ETV6) were detected in 3 patients by WGS and confirmed by SNP genotyping array and by fluorescent in situ hybridization (FISH). Subsequent interrogation of the SNP genotyping data identified deletions of 12p13 involving ETV6 in 19/242 patients (8%) ranging from 0.25 Mb to 〉20 Mb. Those with del ETV6 lacked cytogenetic (−7, or −5/del5q) or molecular (FLT3/ITD) high-risk features, but 50% of those with del ETV6 had core binding factor (CBF) AML. All patients with del ETV6 achieved a morphologic complete remission (CR) and had no evidence of minimal residual disease at the end of induction chemotherapy. Relapse risk for patients with and without del ETV6 was 63% vs. 45% (p=0.3) with a corresponding disease-free survival (DFS) of 32% vs. 53% (p=0.2). Due to the high prevalence of CBF AML in patients with del ETV6, we inquired whether this alteration might impact outcome in patients with CBF AML. In patients with CBF AML, del ETV6 was associated with adverse outcome, where CBF patients with and without del ETV6 had an event-free survival of 0% vs. 63%, respectively (p=0.002). Of the CBF AML patients who achieved an initial CR, those with compared to those without a del ETV6 had a RR of 88% (vs. 38%, p=0.08) with a DFS of 0% (vs. 61%, p=0.009) respectively. Cryptic ETV6 translocations involving at least two distinct translocation partners represented a third class of ETV6 alterations identified by WGS and RNA Seq. FISH evaluation of select patients was available and identified 16 patients with positive FISH for ETV6 variants. Those who had ETV6 alterations by FISH had a median age at diagnosis of 3.7 years. Evaluation of remission induction rate as well as post remission outcome demonstrated that all but 3 of these 16 patients either failed to achieve a CR (N=4) or relapsed after an initial remission (82% failure rate). These data identified varying genomic alterations of ETV6 by whole genome sequencing that identify a significant proportion of pediatric AML cases with adverse outcome who may be targeted for altered therapy. Disclosures: No relevant conflicts of interest to declare.

Description: The PIM family of serine/threonine kinases (PIM1, PIM2, PIM3) are important regulators of signal transduction that phosphorylate proteins essential for cell proliferation, survival, and apoptosis. The PIM kinases are constitutively active and broadly expressed in multiple tissues and up-regulated in various malignancies. We report the discovery of a novel fusion transcript encoding the kinase domain of PIM3 fused to SCO2, a cytochrome c oxidase assembly protein. In transcript sequencing (RNA Seq) of 68 pediatric AML cases, PIM3-SCO2 fusion transcript was computationally identified and experimentally verified in index cases and studied in an independent cohort of pediatric patients with AML. RNA-Seq performed on the Illumina HiSeq in 68 diagnostic specimens from children with AML treated on COG clinical trials. Sequence reads were mapped to human genome using Novoalign. Four computational methods including Defuse, TophatFusion, FusionMap, and Snowshoes-FTD, were utilized to identify fusion transcripts and after filtering to eliminate false positives, fusions were selected based on observation in 2 or more fusion methods and presence in chimerDB. PIM3-SCO2 was identified as an in-frame fusion transcript in 3 cases with Inv(16) and subsequently verified by RT-PCR and Sanger sequencing. Frequency validation was performed by semi-quantitative expression analysis of PIM3-SCO2 expression levels in 235 AML diagnostic specimens as well as 6 normal bone marrow (NBM) controls. PIM3-SCO2 fusion protein was assessed by Western blot on whole cell lysates from cases with the fusion transcript. After verification of the fusion, available whole genome sequencing data in matching cases was interrogated and failed to demonstrate genomic counterpart to this fusion transcript, suggesting that this fusion may be the result of transcriptional read-through; also called transcription-induced chimera (TIC). Such fusion transcripts are generated when genes in proximity on a genome strand are spliced together to generate a chimeric product. Frequency validation studies in 235 diagnostic specimens from COG AAML0531 demonstrated that PIM3-SCO2 fusion transcript was highly prevalent in AML and expressed in 187 of the 235 cases of AML (80%) with variable prevalence across different cytogenetic cohorts, with prevalence of 87% in CBF, 56% in MLL, 79% in normal karyotype, and 70% in those with “other” karyotypes (p