Search

Distinct patterns of DNA methylation characterize the epigenetic landscape of promyelocytic leukemia/retinoic acid receptor-α (PML-RARα)-associated acute promyelocytic leukemia (APL). We previously reported that the microRNAs (miRNAs) clustered on chromosome 14q32 are overexpressed only in APL. Here, using high-throughput bisulfite sequencing, we identified an APL-associated hypermethylation at the upstream differentially methylated region (DMR), which also included the site motifs for the enhancer blocking protein CCCTC-binding factor (CTCF). Comparing the profiles of diagnostic/remission paired patient samples, we show that hypermethylation was acquired in APL in a monoallelic manner. The cytosine guanine dinucleotide status of the DMR correlated with expression of the miRNAs following a characteristic position-dependent pattern. Moreover, a signature of hypermethylation was also detected in leukemic cells from an established transgenic PML-RARA APL mouse model at the orthologous region on chromosome 12, including the CTCF binding site located upstream from the mouse miRNA cluster. These results, together with the demonstration that the region does not show DNA methylation changes during myeloid differentiation, provide evidence that 14q32 hypermethylation is implicated in the pathogenesis of APL. We propose a model in which loss of imprinting at the 14q32 domain leads to overexpression of the miRNAs in APL.

BACKGROUND: The incidence of human papillomavirus-associated vulval neoplasia is increasing worldwide; yet the associated genetic changes remain poorly understood. METHODS: We have used single-nucleotide polymorphism microarray analysis to perform the first high-resolution investigation of genome-wide allelic imbalance in vulval neoplasia. Our sample series comprised 21 high-grade vulval intraepithelial neoplasia and 6 vulval squamous cell carcinomas, with paired non-lesional samples used to adjust for normal copy number variation. RESULTS: Overall the most common recurrent aberrations were gains at 1p and 20, with the most frequent deletions observed at 2q, 3p and 10. Copy-neutral loss of heterozygosity at 6p was a recurrent event in vulval intraepithelial neoplasia. The pattern of genetic alterations differed from the characteristic changes we previously identified in cutaneous squamous cell carcinomas. Vulval neoplasia samples did not exhibit gain at 5p, a frequent recurrent aberration in a series of cervical tumours analysed elsewhere using an identical protocol. CONCLUSION: This series of 27 vulval samples comprises the largest systematic genome-wide analysis of vulval neoplasia performed to date. Despite shared papillomavirus status and regional proximity, our data suggest that the frequency of certain genetic alterations may differ in vulval and cervical tumours.

Historically, genes targeted by recurrent chromosomal deletions have been identified within the smallest genomic region shared in all patients, the minimally deleted region (MDR). However, deletions this small do not occur in all patients and are a simplification of the impact larger heterogeneous deletions have during carcinogenesis. We use the example of 13q14 deletions in chronic lymphocytic leukemia to show that genes outside MDRs are associated with disease progression. Genomic profiling of 224 patients identified 205 copy number alterations on chromosome 13 in 132 cases. Deletions including DLEU2 were heterogeneous (845 Kb-96.2 Mb) and identified two breakpoint cluster regions within short interspersed nuclear elements proximal to DLEU2 and within long interspersed nuclear elements/L1 repeats distal to GUCY1B2. After defining a deletion class on the basis of size and location, we show that (a) at diagnosis, larger deletions (class II) were associated with a significantly increased risk of disease progression (odds ratio=12.3; P=0.005), (b) in progressive patients, class II deletions were enriched (P=0.02) and (c) this association was independent of IgVH mutational status, ZAP70 expression and ATM/TP53 deletion. Deletion of a 1 Mb gene cluster (48.2-49.2 Mb), including SETDB2, PHF11 and RCBTB1, was significantly associated (P

Early genetic events in the development of high-grade serous ovarian cancer (HGSOC) may define the molecular basis of the profound structural and numerical instability of chromosomes in this disease. To discover candidate genetic changes we sequentially passaged cells from a karyotypically normal hTERT immortalised human ovarian surface epithelial line (IOSE25) resulting in the spontaneous formation of colonies in soft agar. Cell lines transformed ovarian surface epithelium 1 and 4 (TOSE 1 and 4) established from these colonies had an abnormal karyotype and altered morphology, but were not tumourigenic in immunodeficient mice. TOSE cells showed loss of heterozygosity (LOH) at TP53, increased nuclear p53 immunoreactivity and altered expression profile of p53 target genes. The parental IOSE25 cells contained a missense, heterozygous R175H mutation in TP53, whereas TOSE cells had LOH at the TP53 locus with a new R273H mutation at the previous wild-type TP53 allele. Cytogenetic and array CGH analysis of TOSE cells also revealed a focal genomic amplification of CXCR4, a chemokine receptor commonly expressed by HGSOC cells. TOSE cells had increased functional CXCR4 protein and its abrogation reduced epidermal growth factor receptor (EGFR) expression, as well as colony size and number. The CXCR4 ligand, CXCL12, was epigenetically silenced in TOSE cells and its forced expression increased TOSE colony size. TOSE cells had other cytogenetic changes typical of those seen in HGSOC ovarian cancer cell lines and biopsies. In addition, enrichment of CXCR4 pathway in expression profiles from HGSOC correlated with enrichment of a mutated TP53 gene expression signature and of EGFR pathway genes. Our data suggest that mutations in TP53 and amplification of the CXCR4 gene locus may be early events in the development of HGSOC, and associated with chromosomal instability.

Histone methyltransferases (HMTs) are important epigenetic regulators of gene transcription and have been found disrupted at the genomic level in a spectrum of human tumors including hematological malignancies. In CLL, recurring genomic lesions targeting chromatin modifiers are emerging in the literature (Puente 2015, Nature) but their biological and clinical significance remain uncertain. We studied 802 CLL patients, divided into a discovery [n=261, sampled pre-treatment] and two validation cohorts [n=541, 431 pre-treatment and 110 ultra-high risk], using high-resolution SNP-arrays [n=572, Affymetrix SNP6.0 and HumanOmniS-8] and high-throughput re-sequencing [n=320, Haloplex and TruSeq] to identify genomic lesions targeting HMT genes. In our discovery cohort, we identified nine novel regions of copy-number changes; the prime finding was a recurrent deletion of chromosome 3p in 4% of patients between genomic positions 47.12-47.36Mb. This region included SETD2, KIF9 and KLHL18 (Fig 1A), with SETD2 being the most significantly under-expressed gene (p=0.001) in 3p-deleted [n=6] versus non-deleted cases [n=8]. Further validation in two independent cohorts showed that SETD2 deletions were enriched in ultra high-risk CLL and associated with loss of TP53 (p=0.003), genomic complexity (in TP53 wild-type cases, p=0.01) and chromothripsis. Next, we screened for somatic mutations in SETD2 using targeted re-sequencing and identified non-synonymous mutations in four (4.5%) discovery cases (p.D99G, p.W1306X, p.Q1545K, p.E1955Q), and 7/231 (3%) in the pre-treatment validation cases (p.A50T, p.P167L, p.E670K, p.M1742L, p.M1889T (x2), p.I2295M) (Fig 1B). Mutations were somatic in all samples tested [n=5]. To study the clonal nature of the SETD2 deletions, we assigned each genomic CNA with a relative copy-number by normalizing CNA intensity values from array features, and could infer that the 3p deletion was in the dominant clonal population in 14/21 cases with data available for analysis. Employing either the ABSOLUTE algorithm for our discovery cohort or manually correcting for tumor sample purity and local copy-number changes in our validation cohort, we observed that 10/11 SETD2 mutations exhibited a clonal cell fraction. These data strongly imply that SETD2 aberrations represent early clonal events in the pathobiology of CLL. Next, we extended our gene expression analysis to include additional wild-type, deleted, and mutated patients, showing reduced expression both in deleted and mutant cases (p=0.035), thus confirming haplo-insufficiency also for the SETD2-mutant cases. W e reviewed the DNA methylation status of the SETD2 gene body and promoter regions (15 & 9 CpG probes, respectively) from published data, and found no correlation between SETD2 methylation and RNA expression, suggesting that DNA methylation does not play a substantial role in regulating SETD2 expression in CLL. Finally, we analyzed the impact of SETD2 lesions on treatment-free survival (TFS) and overall survival (OS). For TFS and OS, we observed a significantly worse outcome (TFS: 44 vs. 105 months; p=0.004, OS: 85 vs. 199 months; p=0.002) in SETD2 deleted cases that were wild-type for TP53/ATM, compared to cases wild type for TP53/ATM/SETD2 (Fig 1D). Mutant SETD2 cases (wild-type for TP53/ATM) and those wild-type for TP53/ATM/SETD2 exhibited median TFS of 74 and 106 months respectively, differences that did not reach significance (p=0.1). Whilst these data suggest that SETD2 lesions may be clinically relevant, further investigations in larger materials are warranted to understand their full impact on survival. In conclusion, we report for the first time somatic deletions and mutations in SETD2, a gene found disrupted invarious human solid and hematological tumors, in ~7% of CLL patients requiring treatment. These are likely to be early clonal events and associate with TP53 dysfunction, genomic complexity and chromothripsis, with deletions enriched in ultra high-risk CLL. Figure 1. SETD2 lesions. A. SNP6.0 data for the del(3p) cases. B. Schematic diagram highlighting the prevalence and positioning of SETD2 mutations. C. Clonal cell fraction data for SETD2 and other gene mutations. D. Treatment-free survival (TFS). Figure 1. SETD2 lesions. A. SNP6.0 data for the del(3p) cases. B. Schematic diagram highlighting the prevalence and positioning of SETD2 mutations. C. Clonal cell fraction data for SETD2 and other gene mutations. D. Treatment-free survival (TFS). Disclosures Tausch: Gilead: Other: Travel support. Steele:Portola Pharmaceuticals: Other: Travel bursary to ASH 2015; Janssen: Other: Travel bursary to EHA 2015. Hillmen:Janssen: Consultancy, Honoraria, Research Funding; GSK: Consultancy, Honoraria, Research Funding; Novartis: Honoraria, Research Funding; Roche: Consultancy, Honoraria, Research Funding; Celgene: Research Funding; Gilead: Consultancy, Honoraria, Research Funding; AbbVie: Consultancy, Honoraria, Research Funding. Dyer:Gilead: Research Funding; Roche Pharmaceuticals: Speakers Bureau; ONO Pharmaceuticals: Research Funding. Stilgenbauer:AbbVie, Amgen, Boehringer-Ingelheim, Celgene, Genentech, Genzyme, Gilead, GSK, Janssen, Mundipharma, Novartis, Pharmacyclics, Roche: Consultancy, Honoraria, Research Funding. Schuh:Acerta Pharma BV: Research Funding. Strefford:Roche: Research Funding.

Prostate cancer is significantly more common in Western men than in Asian men, but the basis for this difference remains unknown. Because genomic studies of Asian prostate cancer are very limited, we used a genome-wide approach to reveal the genomic alterations in Chinese prostate cancers. We found a significant reduction in the frequency of certain somatic genomic changes that are commonly found in Western prostate cancers, including the 21q22.2-22.3 deletion, which involves the TMPRSS2:ERG fusion gene, and 10q deletion, which causes PTEN inactivation. Array results were confirmed by PCR-based molecular copy-number counting in selected samples. The different frequencies of these genomic changes were further evaluated by fluorescent in situ hybridization and immunohistochemistry analyses of tissue microarray samples. These alterations might be key genetic changes underlying the regional/ethnic difference in clinical incidence and might be induced by specific environmental and/or genetic risk factors that Western men are exposed to. Our findings suggest that tumors arise in Western and Chinese populations by alternative pathogenetic mechanisms. Cancer Res; 70(13); 5207–12. ©2010 AACR.

A novel exon in AML1-ETO negatively influences the clonogenic potential of the t(8;21) in acute myeloid leukemia

The acquisition of uniparental disomy (aUPD) in acute myeloid leukemia (AML) results in homozygosity for known gene mutations. Uncovering novel regions of aUPD has the potential to identify previously unknown mutational targets. We therefore aimed to develop a map of the regions of aUPD in AML. Here, we have analyzed a large set of diagnostic AML samples (n = 454) from young adults (age: 15-55 years) using genotype arrays. Acquired UPD was found in 17% of the samples with a nonrandom distribution particularly affecting chromosome arms 13q, 11p, and 11q. Novel recurrent regions of aUPD were uncovered at 2p, 17p, 2q, 17q, 1p, and Xq. Overall, aUPDs were observed across all cytogenetic risk groups, although samples with aUPD13q (5.4% of samples) belonged exclusively to the intermediate-risk group as defined by cytogenetics. All cases with a high FLT3-ITD level, measured previously, had aUPD13q covering the FLT3 gene. Significantly, none of the samples with FLT3-ITD(-)/FLT3-TKD(+) mutation exhibited aUPD13q. Of the 119 aUPDs observed, the majority (87%) were due to mitotic recombination while only 13% were due to nondisjunction. This study demonstrates aUPD is a frequent and significant finding in AML and pinpoints regions that may contain novel mutational targets.

BACKGROUND: Oral submucous fibrosis (OSF) is a high-risk pre-cancerous condition where 7‐13% of these patients develop head and neck squamous cell carcinoma (HNSCC). To date there is no cancer predictive markers for OSF patients. Genomic instability hallmarks early genetic events during malignant transformation causing loss of heterozygosity (LOH) and chromosomal copy number abnormality. However, to date there is no study on genomic instability in OSF. Although this condition is known as a high-risk pre-cancerous condition, there is no data regarding the genomic status of this disease in terms of genetic susceptibility to malignant transformation. METHODS: In this study, we investigated the existence of genetic signatures for carcinogenesis in OSF. We employed the high-resolution genome-wide Affymetrix Mapping single nucleotide polymorphism microarray

In a small fraction ( approximately 2%) of cases of childhood acute lymphoblastic leukemia (ALL) clinical presentation of leukemia is preceded, some 2-9 months earlier, by a transient, remitting phase of nonclassical aplastic anemia, usually in connection with infection. The potential "preleukemic" nature of this prodromal phase has not been fully explored. We have retrospectively analyzed the blood and bone marrow of a child who presented with aplastic anemia 9 months before the development of ETV6-RUNX1 fusion gene positive ALL. High resolution SNP genotyping arrays identified 11 regions of loss of heterozygosity, with and without concurrent copy number changes, at the presentation of ALL. In all cases of copy number change, the deletion or gain identified by single nucleotide polymorphism (SNP) analysis was confirmed in the ALL blasts by FISH. Retrospective analysis of aplastic phase bone marrow showed that the ETV6-RUNX1 fusion was present along with all of the additional genetic changes assessed, albeit subclonal to ETV6-RUNX1. These data identify for the first time the leukemic genotype of an aplasia preceding clinical ALL and indicate that multiple secondary genetic abnormalities can contribute to a dominant subclone several months before a diagnosis of ALL. These data have implications for the biology of ALL and for management of similar patients.

Understanding the dynamics of evolution of Follicular Lymphoma (FL) clones during disease progression is important for monitoring and targeting this tumor effectively. Genetic profiling of serial FL biopsies and examples of FL transmission following bone marrow transplant suggest that this disease may evolve by divergent evolution from a common ancestor cell. However where this ancestor cell resides and how it evolves is still unclear. The analysis of the pattern of somatic hypermutation of the immunoglobulin gene (Ig) is traditionally used for tracking the physiological clonal evolution of B cells within the germinal center and allows to discriminate those cells that have just entered the germinal center and display features of ancestor cells from those B cells that keep re-circulating across different lymphoid organs. Here we investigated the pattern of somatic hypermutation of the heavy chain of the immunoglobulin gene (IgH-VH) in 4 flow-sorted B cells subpopulations belonging to different stages of differentiation, from sequential lymph node biopsies of cases displaying diverse patterns of evolution, using the GS-FLX Titanium sequencing platform. We observed an unexpectedly high level of clonality, with hundreds of distinct tumor subclones in the different subpopulations from the same sample, the majority detected at a frequency

Gene expression profiles were determined from presentation peripheral blood and bone marrow samples of 28 patients with acute myeloid leukemia (AML). Hierarchical clustering sorted the profiles into separate groups, each representing one of the major cytogenetic classes in AML [i.e., t(8;21), t(15;17), inv(16), 11q23, and normal karyotype]. Statistical group comparison identified genes whose expression was strongly correlated with these chromosomal classes. Moreover, the normal karyotype AMLs were characterized by distinctive up-regulation of certain members of the class I homeobox A and B gene families, implying a common underlying genetic lesion. These data reveal novel diagnostic and therapeutic targets and demonstrate the potential of microarray-based dissection of AML.

Genomic changes affecting tumour suppressor genes are fundamental to cancer. We applied SNP array analysis to a panel of testicular germ cell tumours to search for novel tumour suppressor genes and identified a frequent small deletion on 6q25.3 affecting just one gene, ZDHHC14. The expression of ZDHHC14, a putative protein palmitoyltransferase with unknown cellular function, was decreased at both RNA and protein levels in testicular germ cell tumours. ZDHHC14 expression was also significantly decreased in a panel of prostate cancer samples and cell lines. In addition to our findings of genetic and protein expression changes in clinical samples, inducible overexpression of ZDHHC14 led to reduced cell viability and increased apoptosis through the classic caspase-dependent apoptotic pathway and heterozygous knockout of ZDHHC14 increased [CORRECTED] cell colony formation ability. Finally, we confirmed our in vitro findings of the tumour suppressor role of ZDHHC14 in a mouse xenograft model, showing that overexpression of ZDHHC14 inhibits tumourigenesis. Thus, we have identified a novel tumour suppressor gene that is commonly down-regulated in testicular germ cell tumours and prostate cancer, as well as given insight into the cellular functional role of ZDHHC14, a potential protein palmitoyltransferase that may play a key protective role in cancer.

This is an open-access paper.-- et al., ATM mutation and BIRC3 deletion and/or mutation have independently been shown to have prognostic significance in chronic lymphocytic leukemia. However, the relative clinical importance of these abnormalities in patients with a deletion of 11q encompassing the ATM gene has not been established. We screened a cohort of 166 patients enriched for 11q-deletions for ATM mutations and BIRC3 deletion and mutation and determined the overall and progression-free survival among the 133 of these cases treated within the UK LRF CLL4 trial. SNP6.0 profiling demonstrated that BIRC3 deletion occurred in 83% of 11q-deleted cases and always co-existed with ATM deletion. For the first time we have demonstrated that 40% of BIRC3-deleted cases have concomitant deletion and mutation of ATM. While BIRC3 mutations were rare, they exclusively occurred with BIRC3 deletion and a wildtype residual ATM allele. In 11q-deleted cases, we confirmed that ATM mutation was associated with a reduced overall and progression-free survival comparable to that seen with TP53 abnormalities, whereas BIRC3 deletion and/or mutation had no impact on overall and progression-free survival. In conclusion, in 11q-deleted patients treated with first-line chemotherapy, ATM mutation rather than BIRC3 deletion and/or mutation identifies a subgroup with a poorer outcome., The authors would like to thank Leukaemia and Lymphoma Research, the Kay Kendall Leukaemia Fund and Wessex Medical Research for funding this study. The UK LRF CLL4 trial was funded by a core grant from Leukaemia and Lymphoma Research, with associated research work supported by the MRC (G8223452) and Cancer Research UK, and laboratory studies by the Arbib Foundation, Schering Healthcare UK, Schering AG, Germany and Leukaemia and Lymphoma Research.

Many human cancers present as multifocal lesions. Understanding the clonal origin of multifocal cancers is of both etiological and clinical importance. The molecular basis of multifocal prostate cancer has previously been explored using a limited number of isolated markers and, although independent origin is widely believed, the clonal origin of multifocal prostate cancer is still debatable. We attempted to address clonal origin using a genome-wide copy-number analysis of individual cancer and high-grade prostatic intraepithelial neoplasia (HGPIN) lesions. Using Affymetrix array 6.0 copy-number analysis, we compared the genomic changes detected in 48 individual cancer and HGPIN lesions, isolated from 18 clinically localized prostate cancer cases. Identical genomic copy-number changes, shared by all same-case cancer foci, were detected in all 13 informative cases displaying multiple tumor foci. In addition, individual HGPIN lesions in the two multifocal-HGPIN cases available shared identical genomic changes. Commonly known genomic alterations, including losses at 6q15, 8p21.3-8p21.2, 10q23.2-10q23.31, 16q22.3, 16q23.2-16q23.3 and 21q22.2-21q22.3 regions and gain of 8q24.3 were the most frequently detected changes in this study and each was detected in all same-case foci in at least one case. Microarray data were confirmed by fluorescence in situ hybridization in selected foci. Our high-resolution genome-wide copy-number data suggest that many multifocal cases derive from a single prostate cancer precursor clone and that this precursor may give rise to separate HGPIN foci and may further progress to multifocal invasive prostate cancer. These findings, which demonstrate the monoclonal origin of multifocal prostate cancer, should significantly enhance our understanding of prostate carcinogenesis. © 2012 Wiley Periodicals, Inc.

Background Inherited metabolic diseases (IMDs) comprise a diverse group of generally progressive genetic metabolic disorders of variable clinical presentations and severity. We have undertaken a study using microarray gene expression profiling of cultured fibroblasts to investigate 68 patients with a broad range of suspected metabolic disorders, including defects of lysosomal, mitochondrial, peroxisomal, fatty acid, carbohydrate, amino acid, molybdenum cofactor, and purine and pyrimidine metabolism. We aimed to define gene expression signatures characteristic of defective metabolic pathways. Methods Total mRNA extracted from cultured fibroblast cell lines was hybridized to Affymetrix U133 Plus 2.0 arrays. Expression data was analyzed for the presence of a gene expression signature characteristic of an inherited metabolic disorder and for genes expressing significantly decreased levels of mRNA. Results No characteristic signatures were found. However, in 16% of cases, disease-associated nonsense and frameshift mutations generating premature termination codons resulted in significantly decreased mRNA expression of the defective gene. The microarray assay detected these changes with high sensitivity and specificity. Conclusion In patients with a suspected familial metabolic disorder where initial screening tests have proven uninformative, microarray gene expression profiling may contribute significantly to the identification of the genetic defect, shortcutting the diagnostic cascade.

Development of chemoresistance limits the clinical efficiency of platinum-based therapy. Although many resistance mechanisms have been demonstrated, genetic/molecular alterations responsible for drug resistance in the majority of clinical cases have not been identified. We analyzed three pairs of testicular germ cell tumor cell lines using Affymetrix expression microarrays and revealed a limited number of differentially expressed genes across the cell lines when comparing the parental and resistant cells. Among them, CCND1 was the most significantly differentially expressed gene. Analysis of testicular germ cell tumor clinical samples by quantitative reverse transcription PCR analysis revealed that overall expression of CCND1 was significantly higher in resistant cases compared with sensitive samples (P < 0.0001). We also found that CCND1 was dramatically overexpressed both in induced and intrinsically resistant samples of ovarian and prostate cancer. Finally combined CCND1 knockdown using small-interfering RNA and cisplatin treatment inhibited cell growth in vitro significantly more effectively than any of these single treatments. Therefore, deregulation of CCND1 may be a major cause of cisplatin resistance in testicular germ cell tumors and may also be implicated in ovarian and prostate cancers. CCND1 could be potentially used as a marker for treatment stratification and as a molecular target to improve the treatment of platinum-resistant tumors.

Chimeric fusion genes are highly prevalent in childhood acute lymphoblastic leukemia (ALL) and are mostly prenatal, early genetic events in the evolutionary trajectory of this cancer. ETV6-RUNX1–positive ALL also has multiple (∼ 6 per case) copy number alterations (CNAs) as revealed by genome-wide single-nucleotide polymorphism arrays. Recurrent CNAs are probably “driver” events contributing critically to clonal diversification and selection, but at diagnosis, their developmental timing is “buried” in the leukemia's covert natural history. This conundrum can be resolved with twin pairs. We identified and compared CNAs in 5 pairs of monozygotic twins with concordant ETV6-RUNX1–positive ALL and 1 pair discordant for ETV6-RUNX1 positive ALL. We compared, within each pair, CNAs classified as potential “driver” or “passenger” mutations based upon recurrency and, where known, gene function. An average of 5.1 (range 3-11) CNAs (excluding immunoglobulin/T-cell receptor alterations) were identified per case. All “driver” CNAs (total of 32) were distinct within each of the 5 twin pairs with concordant ALL. “Driver” CNAs in another twin with ALL were all absent in the shared ETV6-RUNX1–positive preleukemic clone of her healthy co-twin. These data place all “driver” CNAs secondary to the prenatal gene fusion event and most probably postnatal in the sequential, molecular pathogenesis of ALL.

We report genetic aberrations that activate the ERK/MAP kinase pathway in 100% of posterior fossa pilocytic astrocytomas, with a high frequency of gene fusions between KIAA1549 and BRAF among these tumours. These fusions were identified from analysis of focal copy number gains at 7q34, detected using Affymetrix 250K and 6.0 SNP arrays. PCR and sequencing confirmed the presence of five KIAA1549-BRAF fusion variants, along with a single fusion between SRGAP3 and RAF1. The resulting fusion genes lack the auto-inhibitory domains of BRAF and RAF1, which are replaced in-frame by the beginning of KIAA1549 and SRGAP3, respectively, conferring constitutive kinase activity. An activating mutation of KRAS was identified in the single pilocytic astrocytoma without a BRAF or RAF1 fusion. Further fusions and activating mutations in BRAF were identified in 28% of grade II astrocytomas, highlighting the importance of the ERK/MAP kinase pathway in the development of paediatric low-grade gliomas.

Cutaneous squamous cell carcinomas (SCC) are the second most commonly diagnosed cancers in fair-skinned people; yet the genetic mechanisms involved in SCC tumorigenesis remain poorly understood. We have used single nucleotide polymorphism (SNP) microarray analysis to examine genome-wide allelic imbalance in 16 primary and 2 lymph node metastatic SCC using paired non-tumour samples to counteract normal copy number variation. The most common genetic change was loss of heterozygosity (LOH) on 9p, observed in 13 of 16 primary SCC. Other recurrent events included LOH on 3p (9 tumors), 2q, 8p, and 13 (each in 8 SCC) and allelic gain on 3q and 8q (each in 6 tumors). Copy number-neutral LOH was observed in a proportion of samples, implying that somatic recombination had led to acquired uniparental disomy, an event not previously demonstrated in SCC. As well as recurrent patterns of gross chromosomal changes, SNP microarray analysis revealed, in 2 primary SCC, a homozygous microdeletion on 9p23 within the protein tyrosine phosphatase receptor type D (PTPRD) locus, an emerging frequent target of homozygous deletion in lung cancer and neuroblastoma. A third sample was heterozygously deleted within this locus and PTPRD expression was aberrant. Two of the 3 primary SCC with PTPRD deletion had demonstrated metastatic potential. Our data identify PTPRD as a candidate tumor suppressor gene in cutaneous SCC with a possible association with metastasis.

Basal cell carcinoma is the most common human cancer with increasing incidence reported worldwide. Despite the aberrant signaling role of the Hedgehog pathway, little is known about the genetic mechanisms underlying basal cell carcinomas. Towards a better understanding of global genetic events, we have employed the Affymetrix Mapping 10K single nucleotide polymorphism (SNP) microarray technique for “fingerprinting” genomewide allelic imbalance in 14 basal cell carcinoma–blood pair samples. This rapid high-resolution SNP genotyping technique has revealed a somatic recombination event–uniparental disomy, leading to a loss of heterozygosity (LOH), as a key alternative genetic mechanism to allelic imbalances in basal cell carcinomas. A highly conserved LOH region at 9q21-q31 was found in 13 of 14 (93%) basal cell carcinomas. Further statistical and fluorescence in situ hybridization analyses confirmed that the 9q LOH was a result of uniparental disomy in 5 of 13 (38%) basal cell carcinomas. De novo mutations in the Patched 1 gene (PTCH) were found in 9 of 13 (69%) basal cell carcinomas with 9q LOH. A second important locus, containing LOH at 6q23-q27 was found in 5 of 14 (36%) basal cell carcinomas, suggesting that the presence of an additional putative tumor suppressor gene may be contributing to basal cell carcinoma development. This study shows that the rate of 9q LOH in basal cell carcinomas has been previously underestimated. Furthermore, we provide the first evidence that uniparental disomy due to somatic recombination constitutes one of the mechanisms of LOH in basal cell carcinoma tumorigenesis.

We have prospectively analysed and correlated the gene expression profiles of children presenting with acute leukaemia to the Royal London and Great Ormond Street Hospitals with morphological diagnosis, immunophenotype and karyotype. Total RNA extracted from freshly sorted blast cells was obtained from 84 lymphoblastic [acute lymphoblastic leukaemia (ALL)], 20 myeloid [acute myeloid leukaemia (AML)] and three unclassified acute leukaemias and hybridised to the high density Affymetrix U133A oligonucleotide array. Analysis of variance and significance analysis of microarrays was used to identify discriminatory genes. A novel 50-gene set accurately identified all patients with ALL and AML and predicted for a diagnosis of AML in three patients with unclassified acute leukaemia. A unique gene set was derived for each of eight subtypes of acute leukaemia within our data set. A common profile for children with ALL with an ETV6-RUNX1 fusion, amplification or deletion of ETV6, amplification of RUNX1 or hyperdiploidy with an additional chromosome 21 was identified. This suggests that these rearrangements share a commonality in biological pathways that maintains the leukaemic state. The gene TERF2 was most highly expressed in this group of patients. Our analyses demonstrate that not only is microarray analysis the single most effective tool for the diagnosis of acute leukaemias of childhood but it has the ability to identify unique biological pathways. To further evaluate its prognostic value it needs to be incorporated into the routine diagnostic analysis for large-scale clinical trials in childhood acute leukaemias.

Chronic lymphocytic leukemia (CLL) is associated with profound defects in immune function, resulting in failure of anti-tumor immunity and increased susceptibility to infection. We have previously demonstrated alterations in gene expression profiles of T cells from CLL patients, which translate into functional defects in T-cell immune synapse formation, motility and cytotoxicity (Gorgun et al. JCI 2005; Ramsay et al. JCI 2008, Blood 2013). However a comparison of the transcriptome of natural killer (NK) cells from CLL patients and controls has not been investigated. NK cells were isolated from the peripheral blood of patients with CLL and healthy donors, followed by gene expression profiling using the Affymetrix U133Plus2.0 platform. 117 probes showed a >2-fold decrease in expression while only 18 probes showed a >2-fold increase in expression (adjusted p-value < 0.05) in CLL NK cells compared to healthy donor NK cells. Strikingly, 52 out of the 117 significantly down-regulated probes (44.4%) were for interferon-inducible genes including STAT1 (Signal Transducers and Activator 1), SOCS1 (Suppressor of cytokine signaling 1), interferon regulatory factor genes IRF7 and IRF9, and oligoadenylate synthetase genes OAS1, OAS2, and OAS3. The majority of these genes were inducible by both type 1 and type 2 interferons. Many of these genes have been implicated in host immunity to viral infections, and so it is possible that decreased NK-cell responsiveness to interferon contributes to the increased susceptibility of CLL patients to viruses. Notably, there was also altered expression of signaling pathways in common with T cells from CLL patients, with dysregulation of the cytoskeleton genes RAB3GAP1, RAB38, and EPHA1 and down-regulation of JUN mirroring the dysregulated JNK-signaling and the altered actin cytoskeleton pathways we have found in T cells from CLL patients. These changes were not due to differences in the relative frequencies of CD56DIM and CD56BRIGHTNK cells. Lenalidomide has significant clinical activity in CLL. It is not directly toxic to tumor cells in vitro, but instead is thought to activate anti-tumor immunity and block pro-tumor micro-environmental factors. NK-cell proliferation has been shown to correlate with clinical response to lenalidomide in CLL (Chanan Khan et al. BJ Haem 2011). Therefore, we investigated the effect of lenalidomide treatment on the gene expression profiles of NK cells from CLL patients in comparison to healthy controls. PBMCs from CLL patients or healthy controls were cultured in the presence of 1μM lenalidomide or vehicle control for 48 hours, followed by NK-cell isolation, RNA extraction and gene expression profiling. There were striking differences in the effect of lenalidomide on NK cells from CLL patients compared with healthy NK cells. In CLL NK cells, lenalidomide repaired the down-regulation of interferon-inducible genes, by increasing the expression of genes such as OAS3, IFIT1, IFI44L, IFIT3, OAS1, PDK4, and ACTN1. Pathway analysis highlighted the effect of lenalidomide on inducing interferon signaling, showing significant activation of interferon α, γ, and λ as upstream regulators. While many of the interferon-inducible genes were up-regulated >3-fold in CLL NK cells, only OAS3 was significantly up-regulated in healthy NK cells with lenalidomide. Furthermore, the gene for IFNγ, IFNG, was actually significantly down-regulated in healthy NK cells by this agent. Lenalidomide also significantly down-regulated the expression of 5 genes encoding killer-cell immunoglobulin-like receptors (KIRs): KIR2DL1, KIR2DL2, KIR2DS3, KIR2DS4, and KIR3DL2, in healthy NK cells, but did not significantly down-regulate KIR genes in the CLL NK-cell dataset. Lenalidomide treatment did have some overlapping effects on CLL and healthy NK cells, including up-regulation of genes ARL11, CYFIP, and CORO1B that regulate the actin cytoskeleton pathway. In conclusion, NK cells from CLL patients have down-regulation of interferon response genes and pathways known to regulate normal immune function in response to bacteria and viruses. Lenalidomide has a differential effect on CLL and healthy NK cells: in CLL NK cells it repairs defective interferon responses, whereas in healthy NK cells it down-regulates inhibitory pathways. Disclosures: Riches: Celgene: Research Funding. Gribben:Celgene: Research Funding; Pharmacyclics: Honoraria; Roche: Honoraria.

Abstract 790 Several investigators have defined gene expression profiling (GEP) signatures in Diffuse Large B-cell Lymphoma (DLBCL) enriched for macrophage and stromal genes, suggesting an active innate immune response against lymphoma. We hypothesize that the malignant B-cells drive tumor-associated macrophage (TAM) dysfunction in a subset of patients with DLBCL which is relevant for their biology and prognosis. We performed GEP on TAM from diagnostic DLBCL in order to recognise key genes and pathways open to functional validation. Single cell suspensions from 8 DLBCL and 8 reactive lymph nodes were used in this study. TAMs were flow sorted using CD36 expression. cDNA synthesis and amplification was performed using the Nugen Ovation Pico WTA system and the Affymetrix GeneChip human gene 1.0 ST platform was used. We used bioinformatics analysis of GEP data from DLBCL whole tumors, DLBCL purified B-cells, in vitro manipulated macrophages and other immune cells in order to define macrophage-enriched genes as well as specific M1/M2 signatures. We identified a 221 gene signature that significantly distinguished DLBCL TAMs from control macrophages, with 165 genes upregulated and 56 genes downregulated. Moreover, a comparative transcriptome analysis of 22 diverse immune cell phenotypes/activation states (IRIS: GSE22886) revealed that 26% of these genes are highly macrophage-enriched. Gene Ontology analysis revealed an over-representation for transcripts involved in inflammatory response (p 6.8×10−23), wound healing (p 1.9×10−20), chemotaxis (p 3.2 ×10−9), and cell motility (p 1.7×10−7). Upregulated genes in TAMs included well known M1 (complement components, CXCL9 or CXCL10) as well as M2 genes (MSR, CD163 or MARCO) (Table 1). In our signature, there was enrichment for M1 compared to M2 genes as defined by bioinformatics analysis. TAMs showed overexpression of the CSF1R gene as well as the chemokines CCL2 and CCL5, suggesting an autocrine feed-back loop of macrophage chemotaxis and survival in DLBCL. Moreover, TAMs showed upregulation of the lymphocyte attractants CCL20, CXCL9 and CXCL10, together with T-cell immunosupressants indoleamine 2,3-dioxygenase 1 and PD-L1, which would support a role for macrophages in T-cell recruitment and dysfunction in DLBCL. We also saw strong upregulation of 7 metallothionein isoforms in TAMs. These are proteins known to be expressed in macrophages and linked to response to oxidative damage, modulation of inflammation and cell proliferation. However their role in cancer microenvironment is unclear. We describe for the first time the GEP from DLBCL TAMs. The TAM transcriptome has partial overlapping genes with both M1 and M2 gene signatures, but also has a characteristic GEP potentially driven by their presence in the DLBCL microenvironment. Although further molecular and functional validation is required, this data provides a platform of genes which serve as excellent candidates for future exploration to understand DLBCL pathogenesis and to define new therapeutic targets. Table 1: Genes of particular interest represented in our TAM signature. Probe collapse was done using the highest expression. The Benjamini-Hochberg multiple hypothesis test was used to determine significance (FDR 0.05). Gene IDs Gene description Log2 Fold Change Adjusted p value MT1E-H, L, M, X, MT2A metallothionein 1E-H, 1M, 1X, 2A 2.3 – 4.4 0.00145 - 0.00017 C1QA, B and C complement component 1, q subcomponent, A, B and C chains 3.1 – 3.6 0.00136 - 0.00255 C2 complement component 2 3.2 0.00136 CCL2, 4, 5, 8, 20 chemokine (C-C motif) ligand 2, 4, 5, 8 and 20 2.4 - 3.7 0.046 - 0.00130 CD163 CD163 molecule 4.6 0.00025 CD14 CD14 molecule 3.4 0.00091 CD274 CD274 molecule 3.7 0.00447 CSF1R colony stimulating factor 1 receptor 2.1 0.00886 CXCL9-11 chemokine (C-X-C motif) ligand 9, 10 and 11 4.2 – 4.4 0.012 - 0.00158 FCGR1B Fc fragment of IgG, high affinity Ib, receptor (CD64) 4.9 0.00091 FCGR2A Fc fragment of IgG, low affinity IIa, receptor (CD32) 3.3 0.00139 FCGR3A Fc fragment of IgG, low affinity IIIa, receptor (CD16a) 5.2 0.00011 IDO1 indoleamine 2,3-dioxygenase 1 4.8 0.00232 MARCO macrophage receptor with collagenous structure 2.6 0.02029 MSR1 macrophage scavenger receptor 1 3.7 0.01467 Disclosures: Gribben: Celgene: Honoraria; Roche: Honoraria; Pharmacyclics: Honoraria; GSK: Honoraria; Mundipharma: Honoraria; Gilead: Honoraria.

Tumor suppressor genes (TSGs) play critical roles in preventing tumorigenesis and they are frequently inactivated in tumours. Recently developed high-density microarrays can detect subchromosomal deletions, recurrence of which usually indicates the location of TSGs within the deleted region. We analyzed testicular germ cell tumour (TGCT) clinical samples using SNP arrays and found a frequent small deletion on the region 6q25.3 containing only one known gene, ZDHHC14. While its cellular function is unknown, ZDHHC14 belongs to the recently discovered DHHC family, which are predicted to be involved in protein palmitoylation, a reversible lipid modification that regulates membrane tethering for key proteins in cell signaling, cancer, neuronal transmission, and membrane trafficking. Consistently, we found a dramatic under-expression of ZDHHC14 mRNA and protein in TGCTs, and this associated with chemoresistance. Oncomine database mining showed that ZDHHC14 is also under-expressed in lymphoma, liposarcoma, brain, kidney, lung and colorectal cancers. Thus, it appears that ZDHHC14 downregulation may be involved in other cancers. We studied ZDHHC14 expression in prostate cancer (PCa), detecting a decrease at mRNA and protein level. We also detected that ZDHHC14 mRNA was downregulated in a pilot study on breast cancer samples. As genomic loss of the ZDHHC14 region was only detected in a small number of PCa samples, we checked whether promoter hypermethylation was the cause for ZDHHC14 downregulation. However, no changes in methylation status were found. We then sequenced the whole genomic region surrounding ZDHHC14 by next generation sequencing in TGCTs and PCa and found several mutations in the promoter, the coding region, as well as in intronic regions. Finally, we tested the function of ZDHHC14 in cell-based studies. We generated a 293 T-REx tetracycline inducible ZDHHC14 overexpressing stable cell line, which showed that ZDHHC14 overexpression decreased cell viability. The induction of apoptosis by ZDHHC14 overexpression was detected both by FACS and caspase 7 and PARP cleavage analyses. This was confirmed by transient ZDHHC14 overexpression in the PCa cell line 22RV1. In vivo we xenografted mice using both tetracycline inducible ZDHHC14 overexpressing 293 T-REx cells and control cells transfected with the empty vector. ZDHHC14 expression was induced by tetracycline at the beginning of inoculation and we detected that ZDHHC14 overexpression blocked tumour initiation completely. In conclusion, these results implicate ZDHHC14 as a tumour suppressor gene commonly inactivated in human cancers, indicating that it might exert its tumor suppressor role through the induction of programmed cell death. This is the first study showing the involvement of ZDHHC14 in a specific pathway, the classic caspase-dependent apoptosis. Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 103rd Annual Meeting of the American Association for Cancer Research; 2012 Mar 31-Apr 4; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2012;72(8 Suppl):Abstract nr 4858. doi:1538-7445.AM2012-4858

Abstract 976 Lenalidomide has recently been demonstrated to have significant activity in chronic lymphocytic leukemia (CLL). Its mechanism of action in this disease is not well understood, but it is thought to act primarily by enhancing anti-tumor immunity and reducing production of pro-tumoral factors in the CLL microenvironment. We have previously demonstrated alterations in the expression of cytoskeletal genes in T-cells from patients with CLL and have subsequently shown that these changes translate into a deficit in T-cell function, due to impaired actin polymerization resulting in defective immunological synapse formation. Treatment of both autologous T-cells and CLL cells with lenalidomide was necessary to repair this defect, suggesting that this may be a key component of this agent's activity in CLL. Therefore we examined the effect of lenalidomide on the global gene expression profiles of isolated B-cells and T-cell subsets from CLL patients and healthy donors. Peripheral blood mononuclear cells from patients with untreated CLL or healthy donors were cultured in the presence of 1 μM lenalidomide or vehicle control for 48 hours. The lymphocyte subsets were isolated, followed by RNA extraction and gene expression profiling using the Affymetrix HGU133Plus2.0 platform. Lenalidomide treatment had similar effects on gene expression in T-cells from both patients with CLL and healthy donors. The most prominent changes in expression were of genes involved in cytoskeletal signaling including a 20-fold increase in WASF1 (Wiskott Aldrich Syndrome protein family, member 1), and greater than 2-fold increases in the expression of Rac-family member RHOC, (Ras homolog gene family, member C), actin binding proteins CORO1B (Coronin 1B), PARVA (Parvin alpha), and the Rho guanine nucleotide exchange factors (GEFs), ARHGEF5 and ARHGEF7. We also observed changes in genes regulating integrin signaling including PXN (Paxilin) and FAK (Focal adhesion kinase), and a shift towards Th1 differentiation with upregulation of TNF, IL-12R, and IL-18R. In addition, we noted increased expression of the transcription factors IKZF1, IKZF4 and IRF4, genes involved in the Ikaros pathways that are essential for hematopoiesis and control of lymphoid proliferation. These changes in gene expression provide further evidence that an important mechanism of action of lenalidomide is the upregulation of the actin cytoskeletal network including Rho-GTPases and integrin activation signaling, and are consistent with our previous observations concerning the functional repair of T-cells in CLL. Initial analysis of the effect of lenalidomide on the gene expression profiles of the CLL B-cells showed similar changes to those previously described in vivo from CLL patients receiving single agent lenalidomide in a clinical trial (Chen et al. JCO 2010). In our system, lenalidomide treatment resulted in a greater than 2-fold upregulation of 189 genes, and a greater than 2-fold downregulation of 85 genes in CLL B-cells. We observed increased expression of several genes belonging to the TNF superfamily including TNF-α, OX40L, and APRIL, and the receptors DR5, DCR2, and OX40. Many of these are known to mediate apoptosis signaling, and we also observed increased expression of pro-apoptotic genes such as FAS, BID (BH3 interacting domain death agonist), HRK (Harakiri), and CFLAR (CASP8 and FADD-like apoptosis regulator), and cell cycle regulators CDKN1A and CDKN1C (Cyclin-dependent kinase inhibitors 1A and 1C). Lenalidomide also upregulated expression of several genes of known importance in the CLL microenvironment, including the chemokines CCL3 and CCL4, CD40, CD274 (PD-L1), CD279 (PD-1), and adhesion molecules LFA3 and ICAM1. The effect of lenalidomide on the gene expression profiles of normal B-cells was less marked, with greater than 2-fold upregulation of 51 genes and downregulation of 12 genes. However, we did observe that lenalidomide treatment induced upregulation of genes involved in cytoskeletal pathways such as RND1 (Rho family GTPase 1), RHOQ (Ras homolog gene family, member Q), and MYO1B (myosin 1B). In conclusion, investigation of the effect of lenalidomide on gene expression profiling in CLL suggests that the drug acts both to enhance T-cell function, and to render the CLL cells more susceptible to immune cell mediated killing. Disclosures: Gribben: Roche: Honoraria; Celgene: Honoraria; GSK: Honoraria; Mundipharma: Honoraria; Gilead: Honoraria; Pharmacyclics: Honoraria.

Prostate cancer is the most common male cancer in Western countries, but much less frequent in Asian countries. We systematically investigated genomic changes in prostate cancers from UK and China to determine genetic similarity and difference and potential alternative mechanisms of prostate carcinogenesis in these high and low cancer incidence populations. Using Affymetrix array 6.0 microarrays, we analyzed genome-wide genomic alterations in 32 UK and 39 Chinese prostate cancer samples. Distinct difference in genomic alterations between Western and Chinese prostate cancers were found, including loss of 21q22 and PTEN deletion, which are the most common genomic changes in Western prostate cancer but rarely detected in the Chinese samples. To further evaluate the difference between Western and Chinese samples, FISH analysis for PTEN deletion and ERG rearrangements and immunohistochemistry analysis for PTEN and ERG expressions were applied to UK (n=160) and Chinese (n=143) prostate cancer tissue microarrays (TMAs). PTEN deletion and ERG rearrangements were found at a significantly higher frequency in samples from UK (42.3% and 37.4%) than China (14.3% and 7.5% respectively) (P In conclusion, we compared genetic alterations in prostate cancer samples from UK and China and found there are significant differences between the two groups, commencing at pre-invasive, HGPIN, stage. Our findings suggest that tumours arise in Western and Chinese populations by alternative pathogenetic mechanisms. These alterations, differentially presented in these two populations may be key genetic changes underlying the regional/ethnic difference in clinical incidence and may be induced by specific environmental and/or genetic risk factors that Western men are exposed to. Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 102nd Annual Meeting of the American Association for Cancer Research; 2011 Apr 2-6; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2011;71(8 Suppl):Abstract nr 3878. doi:10.1158/1538-7445.AM2011-3878

Prostate cancer is commonly detected as multiple foci in a single prostate gland and it is of etiological and clinical importance to identify the origin of these multifocal cancer lesions. Due to the limited number of markers analyzed, previous studies to address this issue are inconclusive. A high-resolution genome-wide study should address this issue better. Therefore, we have performed a genome-wide analysis using Affymetrix Array 6.0 of individually laser-capture microdissected foci to compare the genetic similarities and/or differences of these foci within the same prostate gland. Fluorescence in situ hybridization was used to confirm SNP array results. A total of 43 cancer and high-grade prostatic intraepithelial neoplasia (PIN) foci from 16 prostate cancer cases were analyzed for genomic copy number changes. Copy number changes from same-case foci were compared to determine the genetic relationship between the foci. In many cases, we found extensive genetic differences between different foci isolated from the same case. However, omniclonal copy number gains and losses were detected in 11/11 informative case-matched multifocal tumor cases and 8/8 informative case-matched PIN and tumor cases, suggesting that multiple cancer foci arise from a common prostate cancer precursor clone. In addition, a higher degree of similarity was observed between PIN and adjacent cancer foci, as compared to distant lesions. The strong genetic similarities between PIN and prostate cancer foci from the same region suggest clonal expansion of separate PIN lesions to invasive cancer. Based on the monoclonal model, we examined the step-wise accumulation of these genomic copy number changes. Loss of 10q23.2-10q23.31 and 16q were commonly shared in same-case tumor foci and, in many cases, also matched PIN foci, suggesting that they are early events in prostate carcinogenesis. Other common genomic changes, such as loss of 6q, 8p and 21q22.2-21q22.3, were found less frequently in all same-case foci and rarely in PIN lesions, suggesting these genetic events occur at a relatively later stage. In conclusion, using a high-density SNP array, we reveal a monoclonal origin for the majority of multifocal prostate cancer samples analyzed. From these SNP array data, we have also begun to elucidate the step-wise genetic events in prostate carcinogenesis, including early events such as loss of 10q23.2-10q23.31 and 16q Taken together, these findings should significantly enhance our understanding of prostate carcinogenesis and potentially affect the clinical management of this disease. Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 102nd Annual Meeting of the American Association for Cancer Research; 2011 Apr 2-6; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2011;71(8 Suppl):Abstract nr 4831. doi:10.1158/1538-7445.AM2011-4831

Abstract 1696 Sequential karyotypic and/or FISH analysis in patients with CLL show evidence of genomic evolution in 10 – 20% of cases. Common secondary abnormalities include loss of the second allele of 13q14, additional trisomies or translocations involving the immunoglobulin gene loci in patients with trisomy 12 and complex karyotypic abnormalities in patients with loss and/or mutation of the p53 and ATM genes. However the clinical significance of genomic evolution in patients presenting with early stage disease remains uncertain. Nor is it clear whether the use of high resolution genomic arrays can detect additional clinically relevant abnormalities in this cohort of patients. To address these issues we performed sequential genomic studies using the affymetrix SNP 6.0 platform (Affymetrix, Santa Clara, CA), G-banding and FISH in 29 patients presenting with either stage A0 CLL or MBL in whom IGVH gene mutational status and CD38 and ZAP 70 expression were known. In each case, Partek Genomic Suite (Partek Incorporated, St Louis, MS) was used to identify acquired copy number alterations (CNAs) and copy number neutral LOH events (CNNLOH) by comparing the disease presentation (PT) and subsequent (FU) SNP tumor profiles to each other and to matched germ-line DNA extracted from buccal cells. Patients were subdivided into those that subsequently exhibited stable disease for a minimum of 5 years (n=9) with a median time between sampling of 8 years (range 2 – 15) or those that developed progressive disease (n=20) with a median time between sampling of 4 years (range 1 – 10). 8 patients with progressive disease were treated prior to the follow up sample. All stable cases and 13/20 with progressive disease had mutated IGHV genes. The concordance between G-banding, FISH and SNP array data was >99%. In each case, a more complex genome was present at FU. The acquisition of additional CNAs was not associated with IgVH mutation status, ZAP-70, CD38 positivity or whether a patient had been treated. When these data were stratified for whether a patient subsequently exhibited stable or progressive disease, the following observations were made. Firstly, there was no difference in the acquisition of small CNAs (1Mb) per patient and the total genomic size of these additional changes were greater in progressive cases; stable and progressive cases exhibited a mean of 0.67 and 1.1 additional CNAs at FU and a mean total imbalance of 2.33Mb and 16.9Mb, respectively. In this limited series these differences did not reach statistical significance. Interestingly, the frequency of copy number variations (CNV, annotated in the Database of Genomic Variation; http://projects.tcag.ca/variation/) was higher in cases that went on to develop progressive disease (P=0.002). The size of established recurrent CNAs also changed between time-points. Most notably, several cases exhibited larger 13q14 deletions at FU. Two patients with deletion or CNNLOH of 17p displayed a more complex genome at FU, characterized by multiple CNAs larger than 1Mb. One patient acquired CNNLOH on 17p at FU in addition to five additional CNAs (>1Mb). Conversely, the degree of complexity did not increase in a patient with an ATM deletion. Six cases without an apparent abnormality of ATM or TP53 showed complex genomes at PT and additional CNAs (mean 2.8) at FU. A series of novel, recurrent CNAs were observed exclusively in FU samples, including duplications of 13q and 15q observed in 38% and 53% of patients, respectively. In conclusion, using high resolution microarray analysis we show genomic evolution in early stage patients with predominantly good risk biomarkers who pursued either an indolent or progressive course. However, our data suggests that larger acquired CNAs may be more prognostically relevant than smaller regions. We show preliminary data implicating CNVs in the pathogenesis of CLL and we have also identified a series of novel acquired regions of CNAs that may harbor cancer genes which it is hoped will contribute to our understanding of the pathophysiology of CLL and/or have clinical relevance. Disclosures: No relevant conflicts of interest to declare.

Abstract 2699 Transformation from chronic myelomonocytic leukemia (CMML) to acute myeloid leukemia (AML) is associated with a poor prognosis. Clinical features have been used to score those at high risk, e.g. increased marrow blast percentage. Recently, several gene mutations have been identified including TET2, KRAS, NRAS, CBL and RUNX1, of which TET2 is the most frequent (Kohlmann et al 2010, Kosmider et al 2009). Initial descriptions of TET2 mutations in myeloid malignancy suggested that they were early mutations (Delhommeau et al 2009), however subsequent studies have shown in some patients with myeloproliferative neoplasia (MPN) they can be preceded by mutations in JAK2 (Schaub et al 2010). We report a male patient who was diagnosed with CMML aged 43 years. His marrow blast count was 6% and within 2 months he had transformed to AML (FAB M4) with marrow blasts of 50%. He was treated with daunorubicin and AraC (Barts 12 schedule) and obtained a complete morphological remission (CR). However he relapsed 88 days after CR in his skin, central nervous system and marrow. He died with conservative treatment after 6 months. Using this case as a model for transforming CMML, we aimed to identify clonal changes characteristic of the three stages of his disease, chronic leukaemia, transformation to acute leukemia and subsequent relapse. We used cytogenetic analysis, single nucleotide polymorphism (SNP) arrays (Affymetrix SNP 6.0 and 10K; 10K data previously reported by Raghavan et al 2008) to identify the clonal changes. Gene mutation analysis of genomic and cDNA was performed using PCR, cloning and Sanger and amplicon deep-sequencing for the mutational burden of TET2, CBL and KRAS (454 Life Sciences, Branford, CT). At all stages of his disease, he had a normal karyotype, a NPM1 mutation, but no mutations of FLT3, JAK2, KRAS or CBL (Table). Uniparental disomy (UPD) was demonstrated on chromosome 4q at relapse but not present at diagnosis, suggesting an associated TET2 mutation that would be heterozygous at diagnosis and homozygous at relapse. The heterozygous mutation was indentified in the CMML and at transformation, in a region adjacent to the splice site of the end of exon 7. The transcript resulted in a mis-splicing event leading to a stop codon. A second truncated mutant due to a deletion in exon 6 was found expressed at low level. Deep-sequencing of genomic DNA confirmed a deletion in exon 6 consistent with this product at 2.4% (918-fold coverage) in the CMML, and increased to 5.7% (865-fold coverage) burden at transformation. At relapse, the genomic mutation was homozygous. Expression of only the mis-spliced product was observed, with no wild type TET2, and no evidence of the second truncated product by cloning or deep-sequencing (832 reads). Homozygous TET2 mutations are associated with 4qUPD in MPN and myelodysplastic syndromes (Mohamedali et al 2009, Janowska et al 2010), but this is the first description of clonal evolution of a hetero- to homozygous TET2 mutation at disease progression. The case is consistent with the hypothesis that TET2 mutations evolve early in the pathogenesis of myeloid malignancies. However, it also shows there is considerable instability of TET2 mutations, with multiple mutations at various levels, and with mitotic recombination being a late event leading to relapse. CMML Transformed AML Relapsed AML CGN 46,XY 46,XY 46,XY SNP array Not done No somatic copy number abnormalities detected 4qUPD 88201900-qter TET2 c4340+5G>T+10_+21del het result: Cys1298X c4340+5G>T+10_+21del het result: Cys1298X c4340+5G>T+10_+21del hom result: Cys1298X c3732_3733delCT 2.4% result: Tyr1245LeufsX22 c3732_3733delCT 5.7% result: Tyr1245LeufsX22 Absent NPM1 Dup956_959TCTG het Dup956_959TCTG het Dup956_959TCTG het Disclosure: No relevant conflicts of interest to declare.

Prostate cancer is a highly heterogeneous disease both in terms of its pathology and clinical presentation. Pathologically, multiple cancer foci are commonly detected in a single prostate gland. Previously, microsatellite analysis has been used to analyse the origin of multifocal prostate cancer cases. However, it is still debatable whether these separate cancer foci originate from a single clone, or arise independently. To date no extensive high-resolution genome-wide study has been performed to address this. For this reason, we have performed a genome-wide analysis using Affymetrix Array 6.0 with 18 million gene probes to compare the genetic similarities and/or differences of multiple cancer foci within the same prostate gland. We considered two lesions to be separate if they were located O1 cm apart throughout the depth of the tissue. Cancer foci, high-grade prostatic intraepithelial neoplasia (PIN), and matched normal prostate tissue were identified and microdissected separately from H+E stained fresh-frozen tissue sections. SNP 6.0 array data were analysed to determine the genetic relationship between the cancer foci. In many cases we found extensive genetic differences between different cancer foci isolated from the same case. However, all same-case cancer foci and/or PIN shared at least one common genetic alteration, suggesting that multiple cancer foci arise from a common prostate cancer precursor clone at stage of PIN or earlier. In addition, a higher degree of similarity was observed between PIN and adjacent cancer foci, as compared to distant lesions. The strong genetic similarities between PIN and prostate cancer foci from the same region suggest that clonal expansion of separate PIN lesions may account for the multifocal nature of prostate cancer. BREAKPOINT PATTERN ANALYSIS IN SOLID TUMORS

Background The human cell cycle transcription factor FOXM1 is known to play a key role in regulating timely mitotic progression and accurate chromosomal segregation during cell division. Deregulation of FOXM1 has been linked to a majority of human cancers. We previously showed that FOXM1 was upregulated in basal cell carcinoma and recently reported that upregulation of FOXM1 precedes malignancy in a number of solid human cancer types including oral, oesophagus, lung, breast, kidney, bladder and uterus. This indicates that upregulation of FOXM1 may be an early molecular signal required for aberrant cell cycle and cancer initiation. Results The present study investigated the putative early mechanism of UVB and FOXM1 in skin cancer initiation. We have demonstrated that UVB dose-dependently increased FOXM1 protein levels through protein stabilisation and accumulation rather than de novo mRNA expression in human epidermal keratinocytes. FOXM1 upregulation in primary human keratinocytes triggered pro-apoptotic/DNA-damage checkpoint response genes such as p21, p38 MAPK, p53 and PARP, however, without causing significant cell cycle arrest or cell death. Using a high-resolution Affymetrix genome-wide single nucleotide polymorphism (SNP) mapping technique, we provided the evidence that FOXM1 upregulation in epidermal keratinocytes is sufficient to induce genomic instability, in the form of loss of heterozygosity (LOH) and copy number variations (CNV). FOXM1-induced genomic instability was significantly enhanced and accumulated with increasing cell passage and this instability was increased even further upon exposure to UVB resulting in whole chromosomal gain (7p21.3-7q36.3) and segmental LOH (6q25.1-6q25.3). Conclusion We hypothesise that prolonged and repeated UVB exposure selects for skin cells bearing stable FOXM1 protein causes aberrant cell cycle checkpoint thereby allowing ectopic cell cycle entry and subsequent genomic instability. The aberrant upregulation of FOXM1 serves as a 'first hit' where cells acquire genomic instability which in turn predisposes cells to a 'second hit' whereby DNA-damage checkpoint response (eg. p53 or p16) is abolished to allow damaged cells to proliferate and accumulate genetic aberrations/mutations required for cancer initiation.

Abstract 671 Chronic lymphocytic leukaemia (CLL) is a heterogeneous disease characterised by a variable clinical course. Chromosomal abnormalities involving 13q14 are the most common genetic alterations in CLL, occurring in ∼60% of cases and when identified as a sole abnormality, are associated with a favourable clinical outcome. A minimally deleted region (MDR), found in most cases, encompasses DLEU2, DLEU1, miR15a and miR16-1. However, recent SNP profiling data showed that 13q deletions extends beyond the MDR and exhibit considerable heterogeneity both in size and location (Ouillette et al, Cancer Res 2008 68(4);1012). Furthermore, this study showed that larger genomic deletions were associated with a higher Rai stage and were enriched in treated patients. To explore this association further, we performed an initial analysis on samples taken at diagnosis from 100 Stage A patients, of whom 50% had stable disease for over 10 years, and on pre-treatment samples from 100 patients enrolled on the LRF CLL4 trial who had either a complete or partial response to fludarabine and cyclophosphamide [n=200] using the Affymetrix SNP6.0 microarray platform. Deletions of 13q were identified in 99/200 patients (49.5%), the majority (94%) of which included a mono- or biallelic deletion of the MDR. The smallest deletion in this series was 130Kb, while the largest was an 82Mb terminal deletion (mean deletion size 7.25Mb). Furthermore, we identified 12 cases with acquired copy number neutral LOH affecting 13q, exclusive in the presence of a bi-allelic deletion of the MDR. In agreement with the previous study we demonstrated that the size and genomic location of the deletions were highly heterogeneous. Common deletion breakpoints within a 1.9Mb region on 13q between genomic locations 48.7 and 50.6Mb defined 42 cases with small deletions. The remaining 57 cases (58%) showed deletions extending further towards the telomere or centromere and defined those cases with a large deletion. The presence of a large 13q deletion at diagnosis was associated with disease progression (p Disclosures: No relevant conflicts of interest to declare.

166 Recurrence of acute myeloid leukemia AML has a poor prognosis with only 20% of adults surviving to 5 years. Therefore it is of importance to identify molecular changes that explain the pathogenesis of relapsed AML. Previous studies had not identified consistently acquired cytogenetic changes at relapse. Recently, acquired uniparental disomy due to mitotic recombination was described in 40% of relapsed AML (Raghavan et al 2008). Most of the events lead to homozygosity for FLT3 mutations. This study aimed to discover if there are further genetic abnormalities acquired at disease recurrence that cannot be identified by conventional cytogenetics, i.e. microdeletions or gains. Twenty-one presentation and relapse paired AML patient blood and marrow samples were stored with consent at St Bartholomew's Hospital, London. Eleven patient samples had a normal karyotype at diagnosis, two had favourable prognosis cytogenetics (inv(16) and t(8;21)) and others had varying numerical cytogenetic abnormalities and rearrangements associated with an intermediate prognosis. DNA from the samples was analysed by array based high-resolution single nucleotide polymorphism (SNP) genotyping (Affymetrix Human SNP array 6.0). Data was analysed using Partek Genome Browser (Partek, MO). In all cases, the leukemia infiltrate of the marrow or blood was greater than 60% and most cases were greater than 90% allowing accurate identification of DNA copy number changes. Abnormalities of a size that would be identified by cytogenetics were disregarded. Using segmentation analysis using a p-value less than 0.001, over 400 microdeletions and gains were detected that were acquired at relapse in the 21 pairs. Each of the copy number changes was less than 2 megabases in size. One AML sample with a normal karyotype at diagnosis and trisomy 8 and add(9)(q34) at relapse had not acquired any microdeletions or gains. In contrast, in other samples as many as 69 microdeletions/gains were detected. There was no correlation between increased complexity of the karyotype of the leukemia and the number of microdeletions/gains. Several of the acquired microdeletions/gains were in regions containing genes known to be involved in AML, including a deletion of 234Kb at 13q12.2 involving FLT3 and CDX2 , and an acquired deletion at 21p11.2 of 150Kb involving exons encoding the runt domain of RUNX1 . Another copy number gain was detected at the MLL locus, suggestive of partial tandem duplication. Other detected locations are in [Table 1][1]. | Location by cytoband | Copy number change | Size / Kb | P value | Gene | |:--------------------:| ------------------ | --------- | ------- | ---------- | | 13q12.2 | Deletion | 234 | 10−33 | FLT3, CDX2 | | 21q22.12 | Deletion | 150 | 10−13 | RUNX1 | | 11q23.3 | Gain | 5.1 | 0.0099 | MLL | | 11p15.4 | Gain | 83 | 0.00001 | NUP98 | | 17q21.31 | Deletion | 8.0 | 0.0007 | BRCA1 | * The results indicate that recurrent AML may be associated with the deletion or gain of several genes involved in leukaemogenesis. Many other locations are involved throughout the genome, suggesting at least some of these are also involved in the clonal evolution of the leukaemia at recurrence. Further studies should identify novel genes from these regions involved in the pathogenesis of AML. Table 1 Disclosures: No relevant conflicts of interest to declare. [1]: #T1

Background Cancer associated with smoking and drinking remains a serious health problem worldwide. The survival of patients is very poor due to the lack of effective early biomarkers. FOXM1 overexpression is linked to the majority of human cancers but its mechanism remains unclear in head and neck squamous cell carcinoma (HNSCC). Methodology/Principal Findings FOXM1 mRNA and protein expressions were investigated in four independent cohorts (total 75 patients) consisting of normal, premalignant and HNSCC tissues and cells using quantitative PCR (qPCR), expression microarray, immunohistochemistry and immunocytochemistry. Effect of putative oral carcinogens on FOXM1 transcriptional activity was dose-dependently assayed and confirmed using a FOXM1-specific luciferase reporter system, qPCR, immunoblotting and short-hairpin RNA interference. Genome-wide single nucleotide polymorphism (SNP) array was used to ‘trace’ the genomic instability signature pattern in 8 clonal lines of FOXM1-induced malignant human oral keratinocytes. Furthermore, acute FOXM1 upregulation in primary oral keratinocytes directly induced genomic instability. We have shown for the first time that overexpression of FOXM1 precedes HNSCC malignancy. Screening putative carcinogens in human oral keratinocytes surprisingly showed that nicotine, which is not perceived to be a human carcinogen, directly induced FOXM1 mRNA, protein stabilisation and transcriptional activity at concentrations relevant to tobacco chewers. Importantly, nicotine also augmented FOXM1-induced transformation of human oral keratinocytes. A centrosomal protein CEP55 and a DNA helicase/putative stem cell marker HELLS, both located within a consensus loci (10q23), were found to be novel targets of FOXM1 and their expression correlated tightly with HNSCC progression. Conclusions/Significance This study cautions the potential co-carcinogenic effect of nicotine in tobacco replacement therapies. We hypothesise that aberrant upregulation of FOXM1 may be inducing genomic instability through a program of malignant transformation involving the activation of CEP55 and HELLS which may facilitate aberrant mitosis and epigenetic modifications. Our finding that FOXM1 is upregulated early during oral cancer progression renders FOXM1 an attractive diagnostic biomarker for early cancer detection and its candidate mechanistic targets, CEP55 and HELLS, as indicators of malignant conversion and progression.

The long-term disease-free survival rate of adult acute lymphoblastic leukemia (ALL) patients remains less than 40%, in contrast to pediatric cases where it approaches 80%. Furthermore, whereas genetic abnormalities are widely used in childhood ALL for assigning patients to prognostic risk groups, their use in adult ALL is generally restricted to the presence or absence of the t(9;22)/[BCR-ABL1]. Novel prognostic markers, allowing better treatment stratification, and new treatment targets are therefore much needed. We have investigated diagnostic samples from 45 adult ALL cases using genome-wide, high-resolution single nucleotide polymorphism (SNP) array analysis with the Affymetrix 10K, 250K Nsp, and 250K Sty chips, in total including more than 500,000 SNPs with a median inter-marker distance of

Children with Down syndrome (DS) have a 30 fold increased risk of developing leukaemia compared to their non-DS counterparts. While it is now known that cooperating mutations in the haemopoietic transcription factor GATA1 occur in all cases of acute megakaryoblastic leukaemia in Down syndrome patients, the additional genetic events which confer an increased risk of acute lymphoblastic leukaemia in Down syndrome (ALL-DS) are unknown. We initiated a search for mutations in the coding region of candidate genes including RAS, B-RAF, FLT3, KIT and JAK2 in a series of ALL-DS cases. No mutations were identified. We then carried out high resolution (Affymetrix 250K Nsp and 250K Sty) SNP array analysis of leukaemic blast cell DNA from 9 cases of ALL-DS, the majority of which had matched remission DNA. Overall, the whole and partial chromosome gains and losses were in agreement with the karyotype, but in all cases these were updated and refined. There were between 1 and 12 additional copy number alterations per case, with small focal deletions comprising 1 or 2 genes being more frequent than gains. The most common of these was a focal deletion of the CDKN2A gene (4 cases), all of which had either a partial deletion or copy number neutral LOH of the whole of 9p. The other most common focal deletions were of 12p13.3 (ETV6 gene) and 9p13.2 (PAX5), found in 2 cases each. Other regions affected included 3q13.2 (BTLA), 5q33.3 (EBF1), 13q14.2 (RB1) and 20p12.2 (including c20orf94), identified in 1 case each. These results indicate that the secondary genetic events in ALL-DS are similar to those for ALL overall (Mullighan et al., Nature446: 758–764, 2007). The number and pattern of submicroscopic genetic abnormalities more closely resembles that of ETV6-RUNX1 positive ALL than high hyperdiploid ALL (which includes acquired trisomy 21).

MicroRNAs (miRNAs) are single stranded non-coding RNAs of ∼ 22 nucleotides involved in gene regulation. Several examples of an association between disrupted expression of miRNAs and cancer have been shown. Using a real-time quantitative PCR, designed to amplify only from the mature miRNA (TaqMan® -MicroRNA Looped-PCR assay, Applied Biosystems), we measured the expression levels of 157 miRNAs in 100 acute myeloid leukaemia (AML) patients representing the spectrum of known karyotypes common in AML, 2 leukaemic cell lines (KG1 and NB4), and the bone marrow from 2 healthy donors. ANOVA analysis using a 5% false discovery rate threshold was performed to identify differentially expressed miRNAs between leukaemia samples and normal bone marrow. MiRNAs specific to karyotype groupings were also identified in the same way. A method was developed to demonstrate the spatial localisation, in situ, of specific miRNA identified in the quantification and confirm the expression of miRNA with relation to karyotype. Commercial locked nucleic acid (LNA) oligonucleotides (Exiqon) were obtained for two miRNAs (miRNA-127 and miRNA-154), as well as for positive and negative control probes. LNAs were labelled with digoxigenin and probes applied to both cytospins and/or trephines of a sample set representative of the different AML subtypes. Detection of hybridisation signals was either by colorimetric or fluorescent reaction and visualised by confocal microscopy. Unsupervised cluster analysis revealed an association of miRNA expression with the karyotype of the samples. Promyelocytic leukaemias (APML) bearing the t(15;17) translocation, show a distinct pattern characterised by the high expression of a subset of 10 miRNAs located in the human 14q32 imprinted domain, including miR-127 and miR-154. ANOVA data analysis revealed the de-regulation of 33 miRNAs across the leukaemic set in respect to bone marrow from healthy donors. Seventeen miRNAs were up-regulated and 16 down-regulated. MiRNAs miR-155, miR-181a, miR-181b, miR-181c, miR-142-5p, miR-221, and miR-222, were among those commonly highly expressed. Down-regulated were miR-26a, miR-34c, and miR-199a. MiRNAs miR-10a and miR-125b showed the highest variability throughout the samples, being associated with specific subgroups. In situ hybridisation analysis of miR-127 and miR-154 confirmed the results obtained by real-time PCR of their expression associated with APML. This study, conducted on about a third of the miRNAs reported in the Sanger database (microrna.sanger.ac.uk), demonstrates the potential for using miRNA expression to sub-classify cancer. The expression analysis of larger number of miRNAs coupled with the in situ hybridisation of leukaemic cells will allow the investigation of miRNA expression on stored samples during the disease course and provide valuable insights into the leukaemogenic process.

The recent discovery of acquired uniparental disomy (aUPD) in acute myeloid leukemia (AML) has been linked to homozygosity for mutations in certain genes (Raghavan et al, Cancer Res. 2005, Fitzgibbon et al, Cancer Res. 2005). Although this phenomenon, which is undetectable by conventional cytogenetics, has been confirmed in subsequent small-scale studies, its extent and frequency remains uncertain. To determine the frequency and distribution of aUPD, DNA samples from 455 young adult AML patients entered in the UK Medical Research Council AML10 trial were analyzed using Mapping 10K 2.0 single nucleotide polymorphism (SNP) arrays (Affymetrix Inc.). Genomic DNA from blood samples of ten non-leukemic individuals was used as control to estimate the copy number values (control set I). We defined aUPD as 50 consecutive homozygous markers but allowed 2 heterozygous calls to accommodate contaminating normal tissue. Using this criterion a false positive rate of 3.3% was calculated from an available data of 90 independent controls (control set I). Overall, 120 regions of UPD were observed in 79 AML cases (17%), 87% of which involved at least one breakpoint, i.e. resulted from mitotic recombination, and 13% were whole chromosome aUPDs arising from chromosomal non-disjunction. They were the sole aberration, as detected by SNP arrays, in 61 samples (13%), and 84% of these had only a single region of aUPD. There was a non-random distribution across chromosomes; 13q (n=18 cases), 11p (n=8) and 11q (n=9) were the most frequently affected. Other chromosomes with regions of recurrent aUPD were 2p (n=7), 2q (n=6), 1p (n=5), 19q (n=4), 17q12–q21.2 (n=4), 21q (n=4), 9p (n=3), Xq (n=3), 6p (n=2), and 17p (n=2). Acquired UPDs were observed across all cytogenetic risk groups: in 25% of adverse risk patients, 11% of favorable risk, 19% of normal karyotype and 10% of the remaining intermediate risk patients. Samples with aUPD13q (5% of samples) belonged exclusively to the intermediate risk group. Chromosome 13 was the only chromosome to show recurrent whole chromosome aUPD. Fifteen samples with aUPD13q covered the region of the FLT3 gene at 13q12.2; all 15 had a FLT3-internal tandem duplication (ITD) and all cases with a high FLT3-ITD mutant level > 50% of total had 13q aUPD. Gains and losses were observed in 12% and 14% of the samples respectively. As expected, gains on chromosome 8 and losses on chromosomes 5 and 7 were common, confirming the general utility of this approach. No homozygous losses were observed. Comparison of arrays with cytogenetic analysis showed that additional information (aUPDs and/or copy number changes) was obtained in 23% of cases with a normal karyotype and 38% of cases without available cytogenetics. This study highlights the importance of aUPD in the development of AML and pinpoints regions that may contain novel mutational targets.

The use of Single Nucleotide Polymorphism (SNP) array profiling has uncovered extensive regions of acquired uniparental disomy (aUPD) in the cancer genome that go undetected using cytogenetics or array-CGH platforms. These regions usually arise by mitotic recombination and can render a cell homozygous for a pre-existing abnormality. SNP array analysis was performed using the Affymetrix 10K Gene-chip mapping array on DNA extracted from a series of Follicular Lymphoma (FL) lymph nodes biopsies from 42 patients, taken at time of diagnosis (n=20), progression (n=16) and transformation (t-FL) (n=32), and the t(14;18) positive lymphoma cell lines DoHH2 and RL 2261. Analysis was performed using the genome oriented laboratory file system, a software package designed to interpret SNP data. The criteria of > 96% homozygosity in at least 50 contiguous SNPs was found to detect no abnormalities in 24 normal remission bone marrows and was therefore adopted for the detection of abnormal runs of homozygosity. Abnormalities were detected in 53/68 primary specimens; these were non-random with recurring sites of aUPD on several chromosomes including 6p, 9p, 12q and 17p. This panel included 26 paired FL and t-FL samples; in 9 FL cases regions of homozygosity were identified which were not present in the subsequent t-FL sample, suggesting that t-FL may arise in a proportion of patients by a mechanism other than a process of direct clonal evolution. Homozygosity of 9p and 17p was seen primarily in the transformation samples and in three cases rendered the cell homozygous for a pre-existing mutation of either CDKN2A or TP53. Thus mutation precedes mitotic recombination which leads to the removal of the remaining wild-type allele. Nine out of 42 patients studied (21%) have aUPD of chromosome 6p. This appears to be an early event in lymphomagenesis as it was present in both FL and t-FL samples obtained from the same patient. The sites of mitotic recombination cluster in a region immediately proximal to the MHC complex at 6p21-12 in 8/9 cases. Mutations in CCND3, CDKN1A, and two translocation partners of BCL6: SRP20 and HIST1H4I, which are all located within or just distal of this cluster, were excluded by direct sequence analysis. Loss of MHC class II expression has been frequently observed in Diffuse Large B-cell lymphoma and a selective advantage gained by the lymphoma cell here would have implications in tumour surveillance. This study highlights the frequency of aUPD in follicular lymphoma and ongoing studies will elucidate the selective basis of a UPD at this location in the pathogenesis of lymphoma.

SNP array technology permits the simultaneous analysis of copy number and allelotype data. This approach has revealed the somatic acquisition of uniparental disomy (UPD) in approximately 20% of acute myeloid leukemias (AMLs) (Raghavan et al. Cancer Res2005;65:375–378). UPD, which mostly appears to be the consequence of mitotic recombination, cannot be detected by conventional analysis. We have conducted a pilot investigation of samples from the UK MRC AML 10 trial in order to confirm these findings in a larger data set from a large clinical trial. The Affymetrix 10K GeneChip Mapping Array was used to type DNA from 100 AML blast samples of which 87 produced arrays with call rates in excess of 90%. Analysis was performed using the genome orientated laboratory file (GOLF) system, an in-house software package designed to interpret SNP array data. Control germline DNA was not available for each AML and GOLF was used to create a control copy number experiment from the mean of 52 array data sets from normal tissue (blood and remission marrow). The copy number ratio was calculated for each SNP for each sample. In 44 samples the karyotype concurred with the SNP array results (excluding balanced translocations which cannot be detected by SNP array). In the other samples, 54 abnormalities were detected that were not seen in the karyotype (6 samples had no karyotype information). Nine were amplifications, 12 were deletions and 31 were UPDs (35%). Of the UPDs, 12 were either whole chromosome or extended to the most telomeric SNP, with the others therefore being interstitial changes. The mean size of an interstitial LOH was 13.3 Mb, with the smallest detected being 3.2 Mb. Recurrent chromosomal abnormalities not detected by giemsa banding are listed in table 1. Twenty samples that were regarded as normal karyotype by gene banding had abnormalities by SNP array. In three examples numerical karyotypic abnormalities were not seen on SNP array analysis (add4q, −8 and a hypodiploid AML). This may be because a minority clone of the AML cells had this karyotype. UPD is known to be associated with homozygous mutations in AML (Fitzgibbon et al. Cancer Res 2005; in press). In this series, two patients had UPD13, both with biallelic FLT3 internal tandem duplication mutations. However, this study has identified several new areas to look for potential homozygous mutations. Given that in this study germ-line comparison has not been made one should not rule out the possibility of some of the smaller abnormalities being copy number polymorphisms (Bailey et al. Science2005;297;1003–1007). However, consanguinity can probably be ruled out given the lack of widespread homozygosity. This study confirms the frequency of UPD and illustrates the potential of SNP arrays for highlighting novel genetic events in AML. Table 1: Recurrent chromosomal abnormalities not detected by gene banding Previously Undetected Abnormality Number UPD1p 2 Amplified 1p 2 UPD2p 2 Del7q 3 UPD8p 2 UPD11p 2 UPD11q 3 Amplified 12 3 Amplified 13 2 UPD16p 4 UPD16q 2 Del20q 2

Many cases of AML have either a normal karyotype or non-recurrent chromosomal abnormalities and hence their pathogenesis remains obscure. The introduction of array-based analysis of single nucleotide polymorphisms (SNPs) allows the rapid determination of genome-wide allelic information at a high density for a DNA sample. High-resolution SNP genotype analysis was performed on 64 presentation AML samples with full karyotype information as follows: normal karyotype [40], t(8;21) [5], t(15;17) [4], inv16 [3], 11q23 [2],−7 [3],+8 [2] and other structural abnormalities [7]. Using the 10K SNP array (Affymetrix, Inc., Santa Clara) 9, a mean call rate of 93.3% yielded more than 10,000 SNP genotype calls per sample. Large unexpected regions of homozygosity were observed in 12 AMLs (18.75%). These regions ranged in size from 16 million base pairs to 113 million base pairs and would have been visible in the karyotypes if due to deletion. Remission bone marrow samples from 5 of those patients were subjected to SNP genotype analysis. The SNP call data demonstrated clearly that the homozygosity seen in the leukemic DNA was not present in the respective remission bone marrow DNA. Fluorescence in situ hybridisation (FISH) demonstrated 2 signals for probes within regions of homozygosity. Furthermore, hybridisation signal values on the SNP arrays demonstrated that regions of homozygosity did not differ from the rest of the chromosome. It was therefore concluded that such homozygous regions corresponded to uniparental disomy (UPD) due to somatic recombination events occurring during development of the leukemias. There appears to be a non-random distribution of UPD with 5 events on chromosome 11, 2 on chromosome 6, 2 on chromosome 9 and 1 on chromosomes 13, 19 and 21. As expected for somatic recombination, homozygosity continued to the telomere in most cases. Any parental bias in UPD could be evidence of a role for imprinted genes. This issue was investigated using the H19 gene, which is located at 11p15 and is normally methylated only on the paternal allele. Two leukemias exhibited UPD including 11p15 and the methylation status of the H19 gene was therefore determined by bisulfite sequencing. One leukemia with UPD11p exhibited a homozygous methylated paternal pattern, while the other example of UPD11p showed a homozygous non-methylated maternal pattern. These data show that the UPD seen on 11p is not restricted to a single parental origin. In a previous analysis, the leukemia with UPD19q was shown to be homozygous for a CEBPA mutation and FISH demonstrated 2 copies of the CEBPA gene. This gene is located at 19q13.1, within the area of UPD and we conclude that the mutation occurred prior to the UPD. We can therefore speculate that an important consequence of UPD could be to unmask pre-existing mutations. A total of 8 different chromosomal regions have been shown to be affected by UPD in this study and this may suggest that there are at least this number of mutational targets. The discovery of widespread, somatically acquired, UPD in leukemias has potentially important clinical implications. 20% of the normal karyotype AMLs was found to have UPD, and this could offer a valuable new approach to the classification of this important subgroup of AML. The prognostic consequences of such cryptic abnormalities for the patient are uncertain, and larger studies will be required to assess the clinical significance of this phenomenon.

Abstract 2591 Poster Board II-567 Although childhood high hyperdiploid acute lymphoblastic leukemia is associated with a favorable outcome, 20% relapse. This makes it important to identify these patients already at diagnosis to ensure proper risk-stratification. To identify changes associated with relapse and ascertain the genetic evolution patterns, SNP array and mutation analyses of FLT3, KRAS, NRAS, and PTPN11 were performed on 11 paired diagnostic/relapse samples. The “triples trisomies” +4, +10, and +17 were detected in 64%, a frequency similar to the one generally observed at diagnosis, thus questioning their favorable prognostic impact. Structural changes, mainly cryptic hemizygous deletions, were significantly more common at relapse (P Disclosures: No relevant conflicts of interest to declare.

Abstract 757 Chromosomal deletions involving 13q14 are the commonest genetic abnormality in chronic lymphocytic leukemia, occurring in ~60% of cases and are associated with a favourable clinical outcome when identified as a sole abnormality. A minimally deleted region (MDR), found in most cases, encompasses DLEU2, DLEU1, and miR15a/16-1. However, deletions this small do not occur in all patients and are a simplification of the impact larger heterogeneous deletions have during carcinogenesis. We have previously shown that large 13q14 deletions are associated with disease progression (Strefford et al, ASH 2009, 114(22), 671). To further characterize this abnormality and identify putative cancer genes, we report an updated analysis of our Affymetrix SNP 6.0 genomic profiling study of 224 patients. These patients were subdivided into three cohorts: Cohort I samples were taken at disease presentation, from patients with either stable disease for >5 years (n=38) or progressive disease within 3 years (n=25). Cohort II comprised 64 unselected patient samples taken at disease progression. Cohort III consisted of samples taken at enrollment to the UK CLL4 treatment trial from patients treated with fludarabine and cyclophosphamide, sub-divided based on complete (n=49), partial (n=40) or no response (n=8) to treatment. In total 205 copy number alterations targeted 13q14 in 132 cases, facilitating the identification of a MDR encompassing DLEU2 and the miR15a/16-1 cluster in 119 cases. The remaining cases had partial loss of the MDR, including (n=11) or excluding (n=2) miR15a/16-1. Although deletion size and location was heterogeneous (0.13-96.2Mb), a 210kb proximal breakpoint cluster region (p-BCR) was identified, targeting TRIM13, KCNRG and exons 7–11 of the DLEU2 gene (n=31), whilst a 210kb distal BCR (d-BCR) targeted RNASEH2B and GUCY1B2 (n=46). Breakpoints in these regions frequently targeted DNA repeats, specifically short (SINE) and long interspersed nuclear elements (LINE) in the p- and d-BCR, respectively. Based on size and location, we defined two deletion classes; both encompassed the MDR, whilst class I deletions were Disclosures: No relevant conflicts of interest to declare.