Abstract: Introduction Primary central nervous system lymphoma (PCNSL) is a rare subtype of non-Hodgkin lymphoma, of which approximately 95% are diffuse large B-cell lymphomas (DLBCLs). Despite the substantial development of intensive chemotherapy during the past two decades, overall clinical outcome of PCNSL has been poorly improved especially in elderly and so has been our knowledge about the molecular pathogenesis of PCNSL, in terms of driver alterations that are relevant to the development of PCNSL. Method To delineate the genetic basis of PCNSL pathogenesis, we performed a comprehensive genetic study. We first analyzed paired tumor/normal DNA from 35 PCNSL cases by whole-exome sequencing (WES). Significantly mutated genes identified by WES and previously known mutational targets in PCNSL and systemic DLBCL were further screened for mutations using SureSelect-based targeted deep sequencing (Agilent) in an extended cohort of PCNSL cases (N = 90). Copy number alterations (CNAs) have been also investigated using SNP array-karyotyping (N =54). We also analyzed WES and SNP array data of systemic DLBCL cases (N = 49) generated by the Cancer Genome Atlas Network (TCGA) to unravel the genetic difference between PCNSL and systemic DLBCL. Results The mean number of nonsynonymous mutations identified by WES was 183 per sample, which was comparable to the figure in systemic DLBCL and characterized by frequent somatic hypermutations (SHMs) involving non-Ig genes. A higher representation of C 〉 T transition involving CpG dinucleotides and hotspot mutations within the WRCY motif targeted by SHM further suggested the involvement of activation-induced cytidine deaminase (AID) in the pathogenesis of PCNSL. We found 12 genes significantly mutated in PCNSL (q 〈 0.1), including MYD88, PIM1, HLA-A, TMEM30A, B2M, PRDM1, UBE2A, HIST1H1C, as well as several previously unreported mutational targets in systemic DLBCL or PCNSL, such as SETD1B, GRB2, ITPKB, EIF4A2. Copy number analysis identified recurrent genomic segments affected by focal deletions (N = 27) and amplifications (N = 10), most of which included driver genes targeted by recurrent somatic mutations or known targets of focal CNAs such as CDKN2A and FHIT. Subsequent targeted sequencing finally identified a total of 107 significantly mutated genes, of which 43 were thought to be targeted by SHM according to their mutational signature and genomic distribution. Most cases with PCNSL (98%) had mutations and CNAs involving genes that are relevant to constitutive NF-KB/Toll-like receptor (TLR)/BCR activity, including those in MYD88 (80%), CD79B/A (60%), CARD11 (18%), TNFAIP3 (26%), GRB2 (24%) and ITPKB (23%). Genetic alterations implicated in escape from immunosurveillance were also frequently identified in as many as 76% of cases. Mutations of HLA-B (64%), HLA-A (36%), HLA-C (28%), B2M (14%) and CD58 (12%) were commonly detected in addition to CNAs in 6p21.32 (HLA class II), 1p13.1 (CD58) and 15q15.2 (B2M), suggesting the importance of immune escape in the pathogenesis of PCNSL. SHMs were also seen in most cases (98%), which affected not only known targets of AID including PIM1, IGLL5 and BTG2 but also previously unreported genes involved in cell proliferation, apoptosis, or B cell development. The pattern of frequently mutated genes in PNCSL was more uniform compared with that in systemic DLBCL, and similar to that found in the activated B cell subtype of DLBCL (ABC-DLBCL), which was in accordance with the previous report of immunophenotypic analysis of PCNSL. On the other hand, mutations of HLA class I genes (HLA-B, HLA-A) were more frequently mutated in PCNSL compared with ABC-type DLBCL. Conclusion WES, SNP array karyotyping and follow-up targeted sequencing of a large cohort of PCNSL cases revealed the genetic landscape of PCNSL, which were more homogeneous than that of systemic DLBCL, and thought to be involved in activation of constitutive NF-KB/TLR/BCR signaling, escape from immunosurveillance, as well as highly frequent SHMs. Disclosures No relevant conflicts of interest to declare.

Abstract: Background: Allogeneic hematopoietic stem cell transplantation (HSCT) is the only curative therapy for patients with myelodysplastic syndromes (MDS), whose benefit, however, is frequently offset by accompanying mortality and morbidity, underscoring the importance of accurate prognostication before the therapeutic choice. For this purpose, several systems, such as the International Prognostic Scoring System (IPSS), are being successfully applied to clinics, and recent genome profiling studies indicate that molecular diagnostics can further improve the prediction. Nevertheless, existing systems are based on the observation from those patients who were untreated or only supportively treated and therefore, may not successfully be applied to the prognostication of the patients who are actually treated by HSCT. Methods: We analyzed patients with MDS (N = 719) from a cohort of Japan Marrow Donor Program (JMDP) who were treated by unrelated HSCT between 2006 and 2013. Peripheral blood DNA was subjected to targeted deep sequencing in 68 major driver genes for the detection of both somatic mutations and copy number variations (CNVs) with accurate determination of their allelic burdens. Results: The median age at HSCT and observation period were 53 years old (20-66) and 372 days (2-3001), respectively. At the diagnosis, 63, 203, 163 and 65 patients have low, intermediate-1, intermediate-2 and high risk classified on the basis of IPSS, respectively (IPSS data was not available for 250 patients). The median time from diagnosis to HSCT was 274 (9-10900) days. Mutations were observed in 75% of the patients, of which TP53 was most frequently mutated (14.3%), followed by U2AF1 (13.2%), RUNX1 (12.2%), ASXL1 (11.0%) and DNMT3A (9.3%). The mean number of mutations was 2.1 per patient and the mean allelic burden was 23.4%. To evaluate karyotyping we combined metaphase cytogenetics and copy number variations using targeted sequencing data. Complex karyotype, chromosome 7 anomaly, deletion 5q, and deletion 20q were observed in 174 (24.4%), 173 (24.3%), 91 (12.8%), and 50 (7.0%) of the patients, respectively. Combined, 86.6% of the patients had one or more genetic lesions. Patients with one or more mutations or CNVs showed unfavorable overall survival (Hazard Ratio (HR) 2.46, P = 2.12 x 10-5). Univariate analysis for each gene identified mutations in TP53 (HR 2.85, P 〈 2.0 x 10-16), NRAS (HR 1.90, P = 5.4 x 10-4), ETV6 (HR 1.54, P = 0.029), CBL (HR 2.25, P = 5.3 x10-5), EZH2 (HR 1.74, P = 0.014), KRAS (HR 2.01, P = 2.0 x 10-3), U2AF2 (HR 1.97, P = 0.027), JARID2 (HR 2.09, P = 0.039), and RIT1 (HR 2.16, P = 0.023) as the unfavorable factors for the overall survival. Besides, mutations in PRPF8 had a favorable effect on overall survival (HR 0.50, P = 0.029). Then, we performed multivariate analysis with stepwise model selection of these significant mutations and clinical parameters. Mutations in TP53 (HR 2.31, P=0.015), and ETV6 (HR 2.57, P=0.015) remained significant together with complex karyotype (HR 2.15, P = 0.0063), grade of acute graft versus host disease (GVHD) (Grade I or II: HR 1.95, P = 0.011, Grade III or IV: HR 4.18, P = 7.94 x 10-5), and the number of red blood cell transfusion received before HSCT ( 〉 =10 times: HR 2.64, P = 0.027). Next, we analyzed the impact of mutations on relapse in cases who achieved complete response after HSCT (N = 423 (58.8%)). Patients with mutations in one or more genetic lesions showed unfavorable relapse free survival (HR 2.27, P = 1.65 x 10-4). Univariate analysis for each gene revealed mutations in TP53 (HR 3.09, P = 9.5 x 10-16), NRAS (HR 2.21, P = 0.0019), ETV6 (HR 1.90, P = 0.012), PRPF8 (HR 0.40, P = 0.046), and WT1 (HR 2.24, P = 0.013) were significant for the relapse free survival. Multivariate analysis and stepwise model selection identified ETV6 (HR 2.98, P = 0.011), WT1 (HR 4.01, P = 0.014), complex karyotype (HR 2.39, P = 0.0083), IPSS High (HR 6.22, P = 0.0053), and Grade III or IV acute GVHD (HR 2.91, P = 0.0071) as unfavorable factors. Conclusions: This large study of MDS cases treated by unrelated HSCT demonstrated that somatic mutations of several driver genes were novel prognostic factors for overall and relapse free survival. These genetic factors were independent of well-known prognostic makers, and therefore could be used to better guide therapy for MDS patients. Disclosures No relevant conflicts of interest to declare.

Abstract: TP53 and RAS-pathway mutations predict very poor survival, when seen with CK and MDS/MPNs, respectively. For patients with mutated TP53 or CK alone, long-term survival could be obtained with stem cell transplantation.

Abstract: DNA hypermethylation has long been implicated in the pathogenesis of myelodysplastic syndromes (MDS) and also highlighted by the frequent efficacy of demethylating agents to this disease. Meanwhile, recent genetic studies in MDS have revealed high frequency of somatic mutations involving epigenetic regulators, suggesting a causative link between gene mutations and epigenetic alterations in MDS. The accumulation of genetic and epigenetic alterations promotes tumorigenesis, hypomethylating agents such as Azacitidine exert their therapeutic effect through inhibition of DNA methylation. However, the relationship between patterns of epigenetic phenotypes and mutations, as well as their impact on therapy, has not been clarified. To address this issue, we performed genome-wide DNA methylation profiling (Infinium 450K) in combination with targeted-deep sequencing of 104 genes for somatic mutations in 291 patients with MDS. Beta-mixture quantile normalization was performed for correcting probe design bias in Illumina Infinium 450k DNA methylation data. Of the 〉 480,000 probes on the methylation chip, we selected probes using the following steps: (i) probes annotated with "Promotor_Associated" or "Promoter_Associated_Cell_type_specific; (ii) probes designed in "Island", "N_Shore" or "S_Shore"; (iii) removing probes designed on the X and Y chormosomes; (iv) removing probes with 〉 10% of missing value. Consensus clustering was performed utilizing the hierarchical clustering based on Ward and Pearson correlation algorithms with 1000 iterations on the top 0.5% (2,000) of probes showing high variation by median absolute deviation across the dataset using Bioconductor package Consensus cluster plus. The number of cluster was determined by relative change in area under cumulative distribution function curve by consensus clustering. Unsupervised clustering analysis of DNA methylation revealed 3 subtypes of MDS, M1-M3, showing discrete methylation profiles with characteristic gene mutations and cytogenetics. The M1 subtype (n=121) showed a high frequency of SF3B1 mutations, exhibiting the best clinical outcome, whereas the M2 subtype (n=106), characterized by frequent ASXL1, TP53 mutations and high-risk cytogenetics, showed the shortest overall survival with the hazard ratios of 3.4 (95% CI:1.9-6.0) and 2.2 (95% CI:1.2-4.0) compared to M1 and M3, respectively. Finally, the M3 subtype (n=64) was highly enriched (70% of cases) for biallelic alterations of TET2 and showed the highest level of CpG island methylation and showed an intermediate survival. In the current cohort, we had 47 patients who were treated with demethylating agents, including 11 responders and 36 non-responders. When DNA methylation status at diagnosis was evaluated in terms of response to demethylating agents, we identified 54 differentiated methylated genes showing 〉 20% difference in mean methylation levels between responders and non-responders (q 〈 0.1). Twenty-five genes more methylated in responders were enriched in functional pathways such as chemokine receptor and genes with EGF-like domain, whereas 29 less methylated gene in responders were in the gene set related to regulation of cell proliferation. Genetic alterations were also assessed how they affected treatment responses. In responders, TET2 mutated patients tended to more frequently respond (45% vs 34%), whereas patients with IDH1/2 and DNMT3A mutations were less frequently altered (0% vs 14%, 9% vs 14%) in responders, compared in non-responders. In conclusion, our combined genetic and methylation analysis unmasked previously unrecognized associations between gene mutations and DNA methylation, suggesting a causative link in between. We identified correlations between genetic/epigenetic profiles and the response to demethylating agents, which however, needs further investigation to clarify the mechanism of and predict response to demethylation agents in MDS. Disclosures Alpermann: MLL Munich Leukemia Laboratory: Employment. Nadarajah:MLL Munich Leukemia Laboratory: Employment. Haferlach:MLL Munich Leukemia Laboratory: Employment, Equity Ownership. Kiyoi:Taisho Toyama Pharmaceutical Co., Ltd.: Research Funding; Novartis Pharma K.k.: Research Funding; Pfizer Inc.: Research Funding; Takeda Pharmaceutical Co.,Ltd.: Research Funding; MSD K.K.: Research Funding; Sumitomo Dainippon Pharma Co., Ltd.: Research Funding; Alexion Pharmaceuticals.: Research Funding; Teijin Ltd.: Research Funding; Zenyaku Kogyo Company,Ltd.: Research Funding; FUJIFILM RI Pharma Co.,Ltd.: Patents & Royalties, Research Funding; Nippon Shinyaku Co.,Ltd.: Research Funding; Japan Blood Products Organization.: Research Funding; Eisai Co.,Ltd.: Research Funding; Yakult Honsha Co.,Ltd.: Research Funding; Astellas Pharma Inc.: Consultancy, Research Funding; Kyowa-Hakko Kirin Co.,Ltd.: Consultancy, Research Funding; Fujifilm Corporation.: Patents & Royalties, Research Funding; Nippon Boehringer Ingelheim Co., Ltd.: Research Funding; Bristol-Myers Squibb.: Research Funding; Chugai Pharmaceutical Co.,LTD.: Research Funding; Mochida Pharmaceutical Co.,Ltd.: Research Funding. Kobayashi:Gilead Sciences: Research Funding. Naoe:Toyama Chemical CO., LTD.: Research Funding; Otsuka Pharmaceutical Co., Ltd.: Research Funding; Nippon Boehringer Ingelheim Co., Ltd.: Research Funding; Kyowa Hakko Kirin Co., Ltd.: Patents & Royalties, Research Funding; Pfizer Inc.: Research Funding; Astellas Pharma Inc.: Research Funding; FUJIFILM Corporation: Patents & Royalties, Research Funding; Celgene K.K.: Research Funding; Chugai Pharmaceutical Co., Ltd.: Patents & Royalties. Kern:MLL Munich Leukemia Laboratory: Employment, Equity Ownership. Miyazaki:Chugai: Honoraria, Research Funding; Shin-bio: Honoraria; Sumitomo Dainippon: Honoraria; Celgene Japan: Honoraria; Kyowa-Kirin: Honoraria, Research Funding.

Abstract: Adult T-cell leukemia/lymphoma (ATL) is a distinct form of peripheral T-cell lymphoma, which is etiologically associated with human T-cell leukemia virus type 1 (HTLV-1) infection during early infancy. Although HTLV-1 can effectively immortalize human T cells, there is a long latency period of ~50 years before the onset of ATL, suggesting that HTLV-1 infection alone may be insufficient for the development of ATL, but additional acquired genetic events that accumulate during the later life are essential for the development of ATL. However, such somatic alterations underlying the pathogenesis of ATL have not been fully elucidated. To obtain a complete registry of genetic alterations in ATL, we performed an integrated genetic study, in which whole-genome/exome and RNA sequencing (RNA-seq) was performed together with array-based methylation and genomic copy number analysis among a cohort of 50 paired ATL samples, followed by extensive validation using targeted deep sequencing of detected mutations in 〉 400 follow-up samples. Compared with other lymphoid malignancies, ATL cells carried higher numbers of mutations, copy number alterations, and rearrangements than in other lymphoid malignancies, suggesting the presence of global genomic instability in ATL. In addition to previously reported mutational targets in ATL (TP53,TCF8, and FAS) and known targets frequently mutated in other lymphoid malignancies (CARD11, GATA3, IRF4, POT1, and RHOA), we identified a variety of highly recurrent mutations affecting previously unknown mutational targets, many of which are involved in T-cell development, activation and migration, immunosurveillance, and transcriptional regulation. Molecular and functional analysis using human T-cell leukemia cell lines showed that some of these novel mutations actually augment T-cell receptor signaling, validating their biological significance in ATL. A comparison of mutations among disease subtypes revealed that several subtype-specific mutations, including TP53, CD58, IRF4 and TBL1XR1 mutations in acute and lymphoma types, and STAT3mutation in chronic and smoldering types, suggesting that different oncogenic mechanisms underlie different ATL subtypes. Furthermore, ATL cells had a distinct pattern of copy number changes and genomic rearrangements. Interestingly, their gene targets showed a significant overlap to mutational targets. Surprisingly, somatic focal deletions involving the 14q31.1 locus were observed in all the cases examined by whole-genome sequencing and therefore are thought to uniquely characterize ATL genomes, although their gene targets remained to be identified. Like other regions also frequently deleted in ATL, such as 7q31.1 and 1p21.3 loci, these deletions were thought to reflect high levels of genetic instability. Finally and conspicuously, pathway analysis revealed that multiple genes involved in the Tax interactome were systematically altered in ATL, although Tax itself underwent gene silencing in most cases. These data suggested that ATL cells can escape from cytotoxic T-lymphocytes by silencing immunogenic Tax expression, while developing alternative oncogenic mechanisms through acquiring somatic mutations or copy number alterations in the Tax-related pathway. Our findings suggest that deregulated T-cell functionalities caused by genetic alterations, especially those associated with HTLV-1 Tax oncoprotein, are central to ATL pathogenesis, and provide a novel clue to contrive new diagnostics and therapeutics for this intractable disease. Disclosures No relevant conflicts of interest to declare.

Abstract: MDS arises through stepwise acquisitions of multiple mutations. Mutation configuration (bi-allelic vs. mono-allelic), specific distributions (hot spot), timing of acquisitions/ranks within clonal hierarchies, and combinatorial spectra, can all affect the pathogenesis and clinical phenotype of this disease. We performed integrative analyses of these processes to determine which relationships are deterministic vs. random. We analyzed 1,809 clinically annotated MDS patients. Deep targeted NGS was applied to a panel of the 36 genes frequently mutated in myeloid neoplasms. Copy number alterations (CNA) were evaluated by combining karyotyping, microarray, and digital CNA analysis. With a mean coverage of 862x, after removing SNPs and errors, we identified 3,971 somatic mutations. Frequently mutated genes/CNAs were TET2 (27%), SF3B1 (23%), ASXL1 (19%), del(5q) (16%), SRSF2 (14%), DNMT3A (11%) and del(7q) (10%), each present in 〉 10% of cases. DNMT3AMT affected the canonical site (R882) in 17% (35/203) of DNMT3AMT cases, were truncating in 39% (79/203) and were other missense mutations in 44% (89/203). Bi-allelic alterations of EZH2 and TP53 most commonly involved a mutation paired with a copy number deletion or UPD. In contrast, RUNX1 and TET2 commonly involved multiple mutations of the gene. Examining intragenetic relationships, we assumed that the impact of individual mutations depends on their clonal hierarchical position. To discriminate 1st hits ("dominant") from subsequent "secondary mutations", we used a stringent binominal distribution algorithm to compare the expected vs. observed VAFs using read counts (Figure 1). A mutation with the largest VAF in a sample was defined as "dominant". Mutations with overlapping 95% CI of expected VAFs were "co-dominant", and those with non-overlapping 95% CIs were "secondary". These assumptions were validated by Pyclone (concordance rate 〉 95%). Accordingly, 1,474 (36%) and 1,372 (35%) mutations were dominant and secondary, respectively. SF3B1, DNMT3A, U2AF1, and TP53 were more likely to be dominant, ASXL1, CBL, ETV6, and KRAS were more likely to be secondary. For example, SF3B1 mutations were dominant in 17% (303/1809) of patients and secondary in 2% (39/1,809), p 〈 .0001, q 〈 .01. 78 common combinations were identified among these different types of mutations. For instance, compared to U2AF1S34 mutations, U2AF1Q157 mutations were more associated with ASXL1 mutations and del(7q), and less with DNMT3A and BCOR mutations (p 〈 .01). Compared to TET2 mono-allelic mutations, TET2 bi-allelic mutations co-existed more with ZRSR2 and SRSF2 mutations, but less with del(5q). DNMT3AR882 mutations at 77% (27/35) were more likely to be dominant than truncating or other missense DNMT3A mutations at 51% (40/79) and 45% (47/98), p=.0046, q=.07. We also evaluated the impact of different types of mutations and their combinations on disease phenotypes and survival. Many relationships were identified between mutations in different genes and bi-allelic vs. mono allelic hits. Among frequently correlated dominant/secondary pairs, pairs of dominant EZH2 mutations and secondary RUNX1 or ASXL1 mutations were associated with higher-risk subtypes (q=.08) (Figure 2), whereas patients with dominant SF3B1 mutations and secondary JAK2 mutations had myeloproliferative features and lower-risk subtypes (q 〈 .001; q=.09). DNMT3AR882 mutations associated with worse overall survival than did truncating and other missense DNMT3A mutations. Five pairs of dominant and secondary mutations significantly affected survival (p 〈 .01, q 〈 .1). We hypothesized that if a fraction of MDS originates from clonal hematopoiesis of indeterminate potential (CHIP), the compositions of dominant mutations in MDS should be similar to those in CHIP. MDS patients with dominant mutations of TET2, DNMT3A, and ASXL1 were defined as "CHIP-derived MDS" and had more secondary TET2 and ZRSR2 mutations than "De novo MDS". In conclusion, we report a comprehensive analysis of various mutational scenarios and the resultant clinical features. Most significantly, our results demonstrate how invariant patterns of evolution evolve with certain dominant hits dictating high probabilities of specific secondary events. Such patterns coincide with unique combinations of clinical properties that arise with hits along specific pathways. Disclosures Nazha: MEI: Consultancy. Sekeres:Celgene: Membership on an entity's Board of Directors or advisory committees; Celgene: Membership on an entity's Board of Directors or advisory committees; Opsona: Membership on an entity's Board of Directors or advisory committees; Opsona: Membership on an entity's Board of Directors or advisory committees. Haferlach:MLL Munich Leukemia Laboratory: Employment, Equity Ownership. Maciejewski:Alexion Pharmaceuticals, Inc.: Consultancy, Membership on an entity's Board of Directors or advisory committees, Speakers Bureau; Alexion Pharmaceuticals, Inc.: Consultancy, Membership on an entity's Board of Directors or advisory committees, Speakers Bureau; Ra Pharmaceuticals, Inc: Consultancy; Apellis Pharmaceuticals: Consultancy; Apellis Pharmaceuticals: Consultancy; Ra Pharmaceuticals, Inc: Consultancy.

Abstract: RHOA mutations are common in ATLL and show a unique distribution compared with other T-cell lymphomas. Depending on patients, functionally distinct RHOA mutations are clonally selected and involved in the pathogenesis of ATLL.

Abstract: The genetic and clinical characteristics of breast tumors with germline variants, including their association with biallelic inactivation through loss-of-heterozygosity (LOH) and second somatic mutations, remain elusive. We analyzed germline variants of 11 breast cancer susceptibility genes for 1,995 Japanese breast cancer patients, and identified 101 (5.1%) pathogenic variants, including 62 BRCA2 and 15 BRCA1 mutations. Genetic analysis of 64 BRCA1/2 -mutated tumors including TCGA dataset tumors, revealed an association of biallelic inactivation with more extensive deletions, copy neutral LOH, gain with LOH and younger onset. Strikingly, TP53 and RB1 mutations were frequently observed in BRCA1- (94%) and BRCA2- (9.7%) mutated tumors with biallelic inactivation. Inactivation of TP53 and RB1 together with BRCA1 and BRCA2 , respectively, involved LOH of chromosomes 17 and 13. Notably, BRCA1/2 tumors without biallelic inactivation were indistinguishable from those without germline variants. Our study highlights the heterogeneity and unique clonal selection pattern in breast cancers with germline variants.