ALBERT

Description: Background Historically diagnosis and prognosis of myeloid disorders including acute myeloid leukemia (AML) have been determined using a combination of morphology, immunophenotype, cytogenetic and more recently single gene, if not single mutation, analysis. The introduction of NGS technology has resulted in an explosion in the quantity of mutation data available. However, the feasibility and utility of NGS technology with regards to decision-making in routine clinical practice of myeloid disorders is currently unknown. We therefore developed an advanced NGS tool for simultaneous assessment of multiple myeloid candidate genes from low amounts of input DNA and present clinical utility analysis below. Methods We designed a targeted resequencing assay using a TruSeq Custom Amplicon panel with the MiSeq platform (both Illumina) consisting of 341 amplicons (~56 kb) designed around exons of genes frequently mutated in myeloid malignancies (ASXL1, ATRX, CBL, CBLB, CBLC, CEBPA, CSF3R, DNMT3a, ETV6, EZH2, FLT3, HRAS, IDH1, IDH2, JAK2, KIT, KRAS, MPL, NPM1, NRAS, PDGFRA, PHF6, PTEN, RUNX1, SETBP1, SF3B1, SRSF2, TET2, TP53, U2AF1, WT1 & ZRSR2). Filtering, variant calling and annotation were performed using Basespace and Variant Studio (Illumina) with additional indel detection achieved using Pindel. A cohort of samples previously characterised with conventional techniques was used for validation and the lower limit of detection established using qPCR. Post-validation, DNA from 152 diagnostic blood or bone marrow samples from patients with confirmed or suspected myeloid disorders; both AML (n=46) and disorders with the potential to transform to AML i.e. myelodysplasia (confirmed n=54, suspected n=10) and myeloproliferative neoplasms (n=42), were analysed using this assay. To gather clinical utility data we developed a reporting algorithm to feed back information to clinicians; only those variants with a variant allele frequency (VAF) of 〉10% and described as acquired in publically available databases were reported with the exception of novel mutations predicted to result in a truncated protein. Further utility data was obtained using published mutation algorithms to determine the proportion of patients in whom mutation data altered prognosis. Results In the validation cohort, initial concordance for detection of clinically significant mutations was 88% rising to 100% once Pindel was used to identify FLT3 ITDs. The lower limit of detection was 3% VAF, and mean amplicon coverage was 390 reads. Using our reporting algorithm 66% of patients in the post-validation cohort had a suspected pathogenic mutation relevant to a myeloid disorder, rising to 74% in patients with confirmed diagnoses. The median number of reported variants per sample for all diagnoses was one (range 0-6). When mutation data for patients with AML with intermediate risk cytogenetics was analysed using the algorithm of Patel et al (N Engl J Med. 2012;366:1079-1089), 4/22 (18%) moved into another risk category. A further two patients had double CEBPA mutations, improving their prognosis. Identification of complex mutations in KIT exon 8 in 2/6 patients with core binding factor AML resulted in more intensive MRD monitoring due to the increased risk of relapse. Interpretation of mutation data for patients with confirmed myelodysplasia using the work of Bejar et al (N Engl J Med. 2011;364:2496-2506) revealed 13/54 (24%) had a high risk mutation independently associated with poor overall survival. 2/8 (25%) patients with chronic myelomonocytic leukemia and 1/12 (8.3%) patients with primary myelofibrosis had high risk ASXL1 exon 12 mutations, independently associated with a poor prognosis. Among suspected diagnoses confirmatory mutations were found in 2/19 (11%), while the absence of mutations reduced the probability of myeloid disease in 11/19 (58%), in some cases sparing elderly patients invasive bone marrow sampling. A further 20 patients had clinically relevant mutations. Conclusions The NGS Myeloid Gene Panel provided extra information to clinicians in 57/152 patients (38%) helping inform diagnosis, individualize disease monitoring schedules and support treatment decisions. The targeted panel approach requires rigorous validation and standardisation in particular of bio-informatics pipelines, but can be adapted to incorporate new genes as their relevance is described and will become central to treatment decisions. Disclosures No relevant conflicts of interest to declare.

Description: Abstract 1383 Explorative genome-wide next-generation sequencing of leukaemias and lymphomas has revealed a wide spectrum of acquired mutations and considerable tumour heterogeneity that might be responsible for disease initiation, resistance to treatments and relapse. There is, therefore, a clinical need to identify these genetic abnormalities in a diagnostic setting. Here, we present the development and validation of a targeted next generation mutation analysis tool. To compare the distribution pattern of genetic abnormalities in chronic lymphocytic leukemia (CLL), we performed targeted deep sequencing on CLL samples using a TruSeq custom designed targeted amplicon assay (TSCA, Illumina). We reveal differential mutation distribution patterns depending on clinical CLL subgroups. The TSCA panel was designed to amplify 21 genes (table 1) with known or suspected links to either the development of CLL or as response predictors, including TP53, SF3B1 (Puente, Nature, 2011; Quesada et al, 2012) and NOTCH1 (Rossi, Blood, 2012). Where genes have known mutational hotspots in CLL, only these regions were included in our panel, for example exons 5–8 of TP53. For genes such as MAP2K1, where mutations are distributed throughout the coding region, every exon was targeted. In total, we were able to design an amplicon panel able to cover 99% of our desired 36,035bp target region. Table 1. List of genes included in CLL custom amplicon panel ASXL1 ATM CHD2 DDX3X FBXW7 HMCN1 IRF4 KLHL6 LRP1B MAP2K1 MAPK1 MED12 NOTCH1 PCLO POT1 SAMHD1 SF3B1 TP53 XPO1 ZFPM2 ZMYM3 In order to validate our approach, we used samples previously subjected to whole genome sequencing as controls. Of the 13 individual mutations in the control cohort, we were successfully able to detect 10 (77%) with our custom assay to an average depth of 1380x. A 19bp deletion in TP53 failed to be picked up by the variant calling software, and 2 point mutations in ATM were not detected due to the targeted nature of the assay. There was a single false positive mutation across all samples in ZFPM2, caused by a sequencing error in a homopolymer region. The sample group consisted of 45 representative CLL cases, split into two cohorts. The first cohort consisted of 11 cases that have yet to receive any treatment, whilst the second cohort comprised 34 relapsed/refractory cases. Analysis of further samples is in progress. We performed library preparation according to the manufacturers instructions. Each sample was dual indexed with two 8bp “barcodes” prior to equimolar pooling, and the final pooled library was processed on an Illumina MiSeq instrument using the TruSeq 2×150bp paired end sequencing protocol. The run produced 1.6Gb of passed filter sequence data, with 92.8% of above the quality threshold of Q30. The average depth of coverage across all samples was 849x. Primary analysis of the sequencing data was performed using the cloud based data analysis package from Illumina, which carried out the alignment and variant calling. A conservative quality score threshold of 〉99 was set, with all variants above this carried forward for further analysis. Our custom amplicon panel detected mutations in 35 of the samples, comprising 8 indels and 45 point mutations. Of the 54 mutations, 40 were missense, 8 were frame-shifts, 1 was a nonsense mutation and 5 are predicted to have functional effects on splicing domains. The most frequently mutated gene was TP53, followed by SF3B1, PCLO and NOTCH1 (figure 1). Fig 1 Frequency of genes with somatic mutations in our CLL cohort. Fig 1. Frequency of genes with somatic mutations in our CLL cohort. Importantly, there was good correlation between mutation allele frequencies from whole genome sequencing, targeted deep sequencing and TSCA, demonstrating that the high sensitivity of large-scale genome sequencers can be reliably applied in a diagnostic setting. We describe mutation hotspots and mutation distribution patterns and link them to clinical behaviour. For example: SF3B1 mutations occurred in 15% of patients and were linked to reduced progression free survival. In conclusion, our technique allows for rapid mutation detection of the most frequently mutated genes in CLL. Further refinements in amplicon design and variant calling will lead to added precision. TSCA design and validation for other haematological diseases is in progress. Disclosures: No relevant conflicts of interest to declare.

Description: Key Points Acquired pathogenic mutations in SAMHD1 are found in up to 11% of relapsed/refractory patients with CLL. SAMHD1 is mobilized to sites of DNA damage.

Description: Key Points Germline JAK2V617I mutation as a sole genetic event does not suppress hematopoietic stem cells. JAK2V617I induces weaker constitutive activation than JAK2V617F but considerable cytokine hyperresponsiveness.

Description: Using sickle cell disease as a model, Cutts et al describe a highly sensitive method for prenatal diagnosis of known single-gene defects using next-generation sequencing of maternal plasma cell-free DNA.

Description: Chronic lymphocytic leukemia is characterized by relapse after treatment and chemotherapy resistance. Similarly, in other malignancies leukemia cells accumulate mutations during growth, forming heterogeneous cell populations that are subject to Darwinian selection and may respond differentially to treatment. There is therefore a clinical need to monitor changes in the subclonal composition of cancers during disease progression. Here, we use whole-genome sequencing to track subclonal heterogeneity in 3 chronic lymphocytic leukemia patients subjected to repeated cycles of therapy. We reveal different somatic mutation profiles in each patient and use these to establish probable hierarchical patterns of subclonal evolution, to identify subclones that decline or expand over time, and to detect founder mutations. We show that clonal evolution patterns are heterogeneous in individual patients. We conclude that genome sequencing is a powerful and sensitive approach to monitor disease progression repeatedly at the molecular level. If applied to future clinical trials, this approach might eventually influence treatment strategies as a tool to individualize and direct cancer treatment.

Description: Abstract 713 Chronic lymphocytic leukaemia (CLL) has a highly variable clinical course. Recent whole genome sequencing (WGS) data reflects this heterogeneity revealing low level recurrent somatic mutations (Puente Nature 2011, Quesada Nat Gen 2011, Wang NEJM 2011, Schuh Blood 2012). Our WGS study of sequential samples from 3 patients revealed candidate founder mutations that were present in all cells at all time-points. We focus on one of the genes affected by founder mutations; SAMHD1. SAMHD1 has been identified as an anti-viral restriction factor that targets HIV-1 by blocking reverse transcription of viral RNA (Laguette Nature 2011, Hrecka Nature 2011). The recently elucidated triphosphohydrolase activity of SAMHD1 leads to depletion of deoxynucleotide triphosphates (dNTP) during reverse transcription, thus interrupting the viral replication cycle before integration into the genome (Goldstone Nature 2011). In constitutional disease, recessive mutations in SAMHD1 have been implicated in deregulation of the innate immune response and development of a congenital autoimmune encephalopathy, Aicardi–Gouti ères syndrome (AGS) (Crow Hum Mol Gen 2009). In our single institutional cohort of 100 1st line and relapsed CLL patients we identified 8 patients with acquired mutations in SAMHD1, of which 6 were chemorefractory (Table 1). This is much higher than the expected frequency of 25% chemorefractory patients in this cohort (Knight Leukemia 2012), implying a correlation between SAMHD1 mutations and poor outcome. In order to precisely establish the incidence of SAMHD1 mutations in patients requiring 1st line treatment, we sequenced 200 samples from patients recruited into first line UK clinical trials. We determined a mutation frequency of 3% (Table 1). Whole genome SNP arrays on our SAMHD1 mutated patients reveals monoalleleic deletions or copy neutral loss of heterozygosity at the SAMHD1 locus in 14 of 15 samples. TABLE 1: SAMHD1 mutated patients Sample Age IgHV mutation TP53 disruption Clinical Status Mutation 1 72 U Neg Pre Treatment c.-166G〉T 2 72 M Neg Refractory M1K K523X 3 NK U Neg Refractory Y155C 4 65 U Neg Refractory R145X 5 NK NK NK Pre Treatment R145Q 6 63 U Neg Pre Treatment I136T 7 77 M Neg Refractory E355K 8 68 U Neg Refractory L431S 9 72 U Neg Refractory F545L 10 NK NK NK Pre Treatment W572X 11 69 NK NK Pre Treatment N/A 12 66 M Pos Refractory T365P 13 NK NK Neg Pre Treatment R371H 14 77 M Neg Refractory N/A 15 25 NK Neg AGS N/A U=Unmutated, M=Mutated. Next, we questioned whether patients with congenital SAMHD1 mutations are more susceptible to developing B-cell malignancies. We reviewed 20 patients with AGS and homozygous SAMHD1 mutations. Intriguingly, 2 of these patients have developed a B-cell malignancy. One of these patients presented at 1 month of age with features typical of AGS and has been subsequently diagnosed with CLL at the age of 25. Sequencing the SAMHD1 locus of both germline and CLL cells from this patient confirmed the homozygous 1609-1G4C mutation. We screened the patient's CLL cells for acquired mutations recently found to be recurrent in CLL. None of these genes were mutated suggesting the SAMHD1 germline mutation was sufficient to cause CLL. In addition, whole exome analysis is in progress for a more complete view of acquired mutations potentially contributing to CLL pathogenesis. To evaluate the interplay of recurrent somatic mutations in CLL in the context of our SAMHD1 mutations, we used a custom designed targeted sequencing panel (TruSeq Custom Amplicon, Illumina). SAMHD1 mutations were found exclusively in SF3B1 negative patients. Only one SAMHD1 patient had a TP53 mutation. To begin to functionally define the role of SAMHD1 mutations in CLL, we examined the impact of SAMHD1 mutations on SAMHD1 mRNA gene expression by quantitative PCR analysis of purified CLL cells and normal B cell controls. Expression in the mutated CLL cells was significantly lower compared to normal B cells. From this, we hypothesise that CLL cells with SAMHD1 mutations might show an increase in intracellular dNTP levels. We are currently evaluating the levels of dNTPs using a custom designed qPCR to measure dNTP incorporation onto template DNA. In conclusion, we provide the first evidence that the lentiviral restriction factor and dNTP triphosphohydrolase SAMHD1 acts as a tumour suppressor in human B cells. We propose that deregulation of the dNTP pool in B cells caused by mutations in SAMHD1 might contribute to lymphomagenesis. Disclosures: Ross: Illumina: Employment. Bentley:Illumina: Employment. Hillmen:Alexion Pharmaceuticals, Inc: Consultancy, Honoraria, Membership on an entity's Board of Directors or advisory committees.

Description: Abstract 1738 Since the initial description of V617F somatic mutation in patients with Philadelphia chromosome negative myeloproliferative neoplasms (MPNs), a remarkable association between alterations in the JAK2 gene and MPNs has emerged. In addition to V617F, a number of other mutations have been detected in exons 12–15 of the JAK2 gene. Furthermore, a specific JAK2 haplotype predisposes to somatic V617F mutation and MPN. However, the link between JAK mutation and MPNs is not straightforward. For example, in familial cases of MPNs, occurrence of V617F is heterogeneous and occurs as a somatic rather than germline mutation. The underlying inherited genetic abnormality in many of these cases remains unknown. Similarly, occurrence of JAK2 V617F in sporadic MPNs is also heterogeneous and is associated with variable disease characteristics. Thus, understanding of the relationship between JAK2 mutation and MPN disease phenotype remains far from complete. We herein report a family with germline V617I mutation (Figure 1), associated with mild/moderate thrombocythaemia and thrombosis. All patients had normal haematocrit, WBC and peripheral blood (PB) morphology without splenomegaly or other abnormality on physical examination. P1: Presented in 2006 at the age of 53 with a significant ischaemic cerebrovascular event and a platelet count of 750 × 109/l, with a history of longstanding thrombocythaemia (〉10 years; 700–970 × 109/l). Bone marrow examination showed normal architecture with increased numbers of morphologically normal megakaryocytes and no fibrosis. She was commenced on aspirin and hydroxycarbamide resulting in good control of the platelet count. Aspirin was not tolerated due to recurrent epistaxis. Subsequent JAK2 mutation screening by pyrosequencing demonstrated an abnormal pyrogram pattern subsequently identified to be V617I by Sanger sequencing. Quantification of allelic level by a pyrosequencing assay designed to detect V617I confirmed heterozygous (≈50%) V617I in PB mononuclear cells (MNCs), CD3+ cells, CD66+ myeloid cells, buccal swab DNA and hair follicle DNA. P2: Daughter of P1. 34 years. Asymptomatic. Persistent thrombocythemia (470–604 × 109/l). Heterozygous (≈50%) V617I in PB MNCs, CD3+ T cells, CD66+ myeloid cells, buccal swab DNA and hair follicle DNA. P3: Son of P1. 36 years. Asymptomatic. Persistent thrombocythemia (606–648 × 109/l). Heterozygous (≈50%) V617I in PB MNCs, CD3+ T cells, CD66+ myeloid cells and buccal swab DNA. P4: Son of P1. 38 years. Asymptomatic. Persistent thrombocythemia (456–526 × 109/l). Heterozygous (≈50%) V617I in PB MNCs, CD3+ T cells, CD66+ myeloid cells and buccal swab DNA. P5: Daughter of P1. 40 years. Platelet count 294 × 109/l. V617I negative in all tissues. V617I has been previously reported to occur rarely in MPN (PMID: 19074595) and to be constitutively-activating in cell line models (PMID: 18326042) and molecular dynamic simulations (PMID: 19744331). Single cell intracellular quantitative pSTAT3 FACS analysis of PB cells from P1-4 demonstrated GCSF hyper-responsiveness of V617I positive PB CD33+ and CD34+ cells. For example, in comparison with normal controls (NC; n=6) V617I CD33+ myeloid cells (n=4) showed a 14-fold increased pSTAT3 mean fluorescent intensity relative to unstimulated cells in response to 15 minutes stimulation with 0.8 ng/ml GCSF (11% vs 156%; P