Search

Introduction: Multiple Myeloma (MM) is an incurable plasma cell malignancy commonly preceded by the asymptomatic stage smoldering multiple myeloma (SMM). MM is characterized with significant genomic heterogeneity of chromosomal gains and losses (CNVs), translocations, and point mutations (SNVs); alterations that are also observed in SMM patients. However, current SMM risk models rely solely on clinical markers and do not accurately capture progression risk. While incorporating some genomic biomarkers improves prediction, using all MM genomic features to comprehensively stratify patients may increase risk stratification precision in SMM. Methods: We obtained a total of 214 patient samples at SMM diagnosis. We performed whole-exome sequencing on 166 tumors; of these, RNA sequencing was performed on 100. Targeted capture was done on 48 additional tumors. Upon binarization of DNA features, we performed consensus non-negative matrix factorization to identify distinct molecular clusters. We then trained a random forest classifier on translocations, SNVs, and CNVs. The predicted clinical outcomes for the molecular subtypes were further validated in an independent SMM cohort of 74 patients. Results: We identified six genomic subtypes, four with hyperdiploidy (>48 chromosomes, HMC, HKR, HNT, HNF) and two with IgH translocations (FMD, CND) (Table 1). In multivariate analysis accounting for IMWG (20-2-20) clinical risk stages, high-risk (HMC, FMD, HKR) and intermediate-risk (HNT, HNF) genetic subtypes were independent predictors of progression (Hazards ratio [HR]: 3.8 and 5.5, P = 0.016 and 0.001, respectively). The low-risk, CND subtype harboring translocation (11;14) was enriched for the previously defined CD-2 MM signature defined by the B cell markers CD20 and CD79A (FDR = 0.003 ), showed upregulation of CCND1, E2F1, and E2F7 (FDR = 0.01, 0.0004, 0.08), and was enriched for G2M checkpoint, heme metabolism, and monocyte cell signature (FDR = 0.003, 0.003, 0.003, respectively). The FMD subtype with IgH translocations (4;14) and (14;16) was enriched for P53, mTORC1, unfolded protein signaling pathways and plasmacytoid dendritic cell signatures (FDR = 0.01, 0.005, 0.008, respectively). The HKR tumors were enriched for inflammatory cytokine signaling, MYC target genes, T regulatory cell signature, and the MM proliferative (PR) signatures (FDR = 0.02, 0.03, 0.007, 0.02, respectively). The APOBEC mutational signature was enriched in HMC and FMD tumors (P = 0.005), while there was no statistical difference across subtypes in the AID signature. The median follow-up for the primary cohort is 7.1 years. Median TTP for patients in HMC, FMD, and HKR was 3.8, 2.6, and 2.2 years, respectively; TTP for HNT and HNF was 4.3 and 5.2, respectively, while it was 11 years in CND patients (P = 0.007). Moreover, by analyzing the changes in MM clinical biomarkers over time, we found that patients from high-risk subgroups had higher odds of developing evolving hemoglobin and monoclonal protein levels over time (P = 0.01 and 0.002, respectively); Moreover, the absolute increase in M-protein was significantly higher in patients from the high-risk genetic subtypes at one, two, and five years from diagnosis (P = 0.001, 0.03, and 0,01, respectively). Applying the classifier to the external cohort replicated our findings where intermediate and high-risk genetic subgroups conferred increased risk of progression to MM in multivariate analysis after accounting for IMWG staging (HR: 5.5 and 9.8, P = 0.04 and 0.005, respectively). Interestingly, within the intermediate-risk clinical group in the primary cohort, patients in the high-risk genetic subgroups had increased risk of progression (HR: 5.2, 95% CI 1.5 - 17.3, P = 0.007). In the validation cohort, these patients also had an increased risk of progression to MM (HR: 6.7, 95% CI 1.2 - 38.3, P = 0.03), indicating that molecular classification improves the clinical risk-stratification models. Conclusion: We identified and validated in an independent dataset six SMM molecular subgroups with distinct DNA alterations, transcriptional profiles, dysregulated pathways, and risks of progression to active MM. Our results underscore the importance of molecular classification in addition to clinical evaluation in better identifying high-risk SMM patients. Moreover, these subgroups may be used to identify tumor vulnerabilities and target them with precision medicine efforts. Figure 1 Figure 1. Disclosures Bustoros: Janssen, Bristol Myers Squibb: Honoraria, Speakers Bureau; Takeda: Consultancy, Honoraria. Casneuf: Janssen: Current Employment. Kastritis: Amgen: Consultancy, Honoraria, Research Funding; Takeda: Honoraria; Pfizer: Consultancy, Honoraria, Research Funding; Genesis Pharma: Honoraria; Janssen: Consultancy, Honoraria, Research Funding. Walker: Bristol Myers Squibb: Research Funding; Sanofi: Speakers Bureau. Davies: Takeda: Consultancy, Honoraria; Amgen: Consultancy, Honoraria; Abbvie: Consultancy, Honoraria; BMS: Consultancy, Honoraria; Roche: Consultancy, Honoraria; Janssen: Consultancy, Honoraria. Dimopoulos: Amgen: Honoraria; BMS: Honoraria; Takeda: Honoraria; Beigene: Honoraria; Janssen: Honoraria. Bergsagel: Genetech: Consultancy, Honoraria; Oncopeptides: Consultancy, Honoraria; Janssen: Consultancy, Honoraria; Pfizer: Consultancy, Honoraria; Novartis: Consultancy, Honoraria, Patents & Royalties: human CRBN mouse; GSK: Consultancy, Honoraria; Celgene: Consultancy, Honoraria. Yong: BMS: Research Funding; Autolus: Research Funding; Takeda: Honoraria; Janssen: Honoraria, Research Funding; Sanofi: Honoraria, Research Funding; GSK: Honoraria; Amgen: Honoraria. Morgan: BMS: Membership on an entity's Board of Directors or advisory committees; Jansen: Membership on an entity's Board of Directors or advisory committees; Karyopharm: Membership on an entity's Board of Directors or advisory committees; Oncopeptides: Membership on an entity's Board of Directors or advisory committees; GSK: Membership on an entity's Board of Directors or advisory committees. Getz: IBM, Pharmacyclics: Research Funding; Scorpion Therapeutics: Consultancy, Current holder of individual stocks in a privately-held company, Membership on an entity's Board of Directors or advisory committees. Ghobrial: AbbVie, Adaptive, Aptitude Health, BMS, Cellectar, Curio Science, Genetch, Janssen, Janssen Central American and Caribbean, Karyopharm, Medscape, Oncopeptides, Sanofi, Takeda, The Binding Site, GNS, GSK: Consultancy.

Background: Recent studies have identified clinical and genomic factors contributing to worse clinical outcomes in patients with multiple myeloma (MM). Clonal hematopoiesis (CH) reflects the presence of somatic driver mutations in the blood or marrow of otherwise asymptomatic individuals. Using a variant allele frequency (VAF) cutoff of 2%, we recently reported CH in 21.6% of MM patients at the time of autologous stem cell transplant (ASCT) and found it was associated with shorter overall survival (OS) and progression-free survival (PFS) in those who did not receive maintenance therapy with an immunomodulatory drug (IMiD). However, this finding was based on a single tertiary center and only included MM patients who received ASCT. Methods: We studied a larger cohort of 986 newly diagnosed MM cases. Whole-exome sequencing (WES) data of peripheral blood and bone marrow samples of 986 MM patients (523 transplanted and 463 non-transplanted) from the Multiple Myeloma Research Foundation (MMRF) Clinical Outcomes in MM to Personal Assessment of Genetic Profile (CoMMpass, NCT0145429) study were analyzed. Both peripheral blood and tumor samples were analyzed to filter out myeloma mutations that could be contaminating the peripheral blood. Given the lower depth of coverage compared to prior targeted sequencing studies, small clones with a VAF below 2% were not detected. Altogether, the WES samples had a total depth of coverage of 117.68X. All data were analyzed using R version 3.5.0 (R Core Team). Results: Among the total cohort, 113 CH mutations were detected in 101/986 (10.24%) patients. CH was detected in 42/523 (8.03%) transplanted patients, compared to 59/463 (12.74%) non-transplanted patients. The most commonly mutated genes were DNMT3A, TET2, ASXL1, PPM1D, and TP53. The median age of the cohort was 63 years (range: 27 - 93), 60% were male, and median follow-up was 3.9 years (95% CI: 3.7 - 4.0). The presence of CH was associated with age (69 vs. 62 years, P < 0.001). As expected, the median age of transplanted patients was lower (60 vs. 67 years) than in the non-transplanted group, which likely explains the higher prevalence of CH detected in the non-transplanted group. CH was associated with recurrent bacterial infections (P = 0.01) and increased cardiovascular disease (P = 0.006), but not with cerebrovascular disease (P = 0.74) or coagulopathies (P = 0.65). There was a trend towards worse PFS in non-ASCT patients with CH who were not treated with IMiDs (1.8 years) compared to non-CH IMiD-treated patients (2.7 years) (P < 0.001). A CH effect on PFS was not detected in ASCT patients. OS was not different in those with or without CH in both ASCT and non-ASCT groups. 8 (0.8%) patients developed a second hematologic malignancy. CH at the time of MM diagnosis was not associated with an increased risk of developing a second hematologic malignancy (P = 0.58). To determine whether CH clones emerged or evolved during treatment, we examined serial samples from 52 patients (36 ASCT patients and 16 non-transplanted patients) with sequential samples. The median time between the first and second time point was 3.1 years (range: 1.0 - 5.4 years). At the first time point, only 3/52 (5.8%) patients had CH, but that number increased to 13/52 (25.0%) at the second time point. Five out of the 13 (38%) were non-transplanted patients. All but 1 patient were exposed to IMiDs. The most common emerging mutated gene was DNMT3A, found in 7 patient samples at the second time point, compared to 2 patients at the first time point. Conclusion: Using WES in a large cohort of newly diagnosed MM patients, we detected CH in 10.2% (VAF ≥ 2%) of patients. CH and non-IMiD treatment confers a shorter PFS in non-transplanted MM patients. However, throughout IMiD-based treatment, MM patients tend to acquire and/or expand previously undetected CH clones, particularly DNMT3A. The clinical significance of this clonal expansion during therapy is yet to be elucidated, and for now, this observation does not yet change clinical management. Figure 1 Figure 1. Disclosures Steensma: Novartis: Current Employment. Ebert: Deerfield: Research Funding; GRAIL: Consultancy; Exo Therapeutics: Membership on an entity's Board of Directors or advisory committees; Celgene: Research Funding; Skyhawk Therapeutics: Membership on an entity's Board of Directors or advisory committees. Soiffer: NMPD - Be the Match, USA: Membership on an entity's Board of Directors or advisory committees; Gilead, USA: Other: Career Development Award Committee; Rheos Therapeutics, USA: Consultancy; Kiadis, Netherlands: Membership on an entity's Board of Directors or advisory committees; Juno Therapeutics, USA: Other: Data Safety Monitoring Board; Precision Biosciences, USA: Consultancy; Jazz Pharmaceuticals, USA: Consultancy; Jasper: Consultancy; Takeda: Consultancy. Sperling: Adaptive: Consultancy. Getz: Scorpion Therapeutics: Consultancy, Current holder of individual stocks in a privately-held company, Membership on an entity's Board of Directors or advisory committees; IBM, Pharmacyclics: Research Funding. Ghobrial: AbbVie, Adaptive, Aptitude Health, BMS, Cellectar, Curio Science, Genetch, Janssen, Janssen Central American and Caribbean, Karyopharm, Medscape, Oncopeptides, Sanofi, Takeda, The Binding Site, GNS, GSK: Consultancy.

Primary central nervous system lymphomas (PCNSLs) and primary testicular lymphomas (PTLs) are extranodal large B-cell lymphomas (LBCLs) with inferior responses to current empiric treatment regimens. To identify targetable genetic features of PCNSL and PTL, we characterized their recurrent somatic mutations, chromosomal rearrangements, copy number alterations (CNAs), and associated driver genes, and compared these comprehensive genetic signatures to those of diffuse LBCL and primary mediastinal large B-cell lymphoma (PMBL). These studies identify unique combinations of genetic alterations in discrete LBCL subtypes and subtype-selective bases for targeted therapy. PCNSLs and PTLs frequently exhibit genomic instability, and near-uniform, often biallelic, CDKN2A loss with rare TP53 mutations. PCNSLs and PTLs also use multiple genetic mechanisms to target key genes and pathways and exhibit near-uniform oncogenic Toll-like receptor signaling as a result of MYD88 mutation and/or NFKBIZ amplification, frequent concurrent B-cell receptor pathway activation, and deregulation of BCL6. Of great interest, PCNSLs and PTLs also have frequent 9p24.1/PD-L1/PD-L2 CNAs and additional translocations of these loci, structural bases of immune evasion that are shared with PMBL.

Classical Hodgkin lymphoma (cHL) and primary mediastinal large B-cell lymphoma (PMBL) are aggressive tumors with distinct cells of origin and pathomorphological features. However, these lymphomas share certain transcriptional signatures and aberrant signaling pathways. CHLs and PMBLs both exhibit constitutive activation of NF-κB and JAK/STAT signaling and genetic bases of PD-1 mediated immune evasion including frequent 9p24.1/PD-L1/PD-L2 copy gains. In both lymphomas, PD-1 blockade is a FDA-approved therapy for relapsed/refractory disease. To characterize genetic bases of response to PD-1 blockade and identify complementary treatment targets in cHL and PMBL, we defined the comprehensive genetic signatures of both diseases. First, we obtained flow cytometry-sorted Hodgkin Reed Sternberg (HRS) cells from 23 biopsies of newly diagnosed cHLs and intact tumor biopsy specimens from 37 newly diagnosed PMBLs. The isolated HRS cells and paired normal DNAs and PMBL biopsy specimens were subjected to whole exome sequencing using an optimized workflow for low input samples and an expanded bait set to capture structural variants (SVs), including translocations. We used newly developed and established analytical pipelines to analyze tumor samples without paired normals (PMBLs) and identify significantly mutated genes (candidate cancer genes [CCGs], MutSig2CV, CLUMPS), SCNAs (GISTIC2.0) and SVs(4 algorithms) in both cHL and PMBL. In cHL, we identified 15 CCGs, 13 recurrent SCNAs, SVs in ETV6 and CIITA, complementary alterations of JAK/STAT, NF-κB and PI3K signaling pathway components and a median number of 11 genetic drivers per tumor. Previously unappreciated aspects of the cHL genetic signature included the increased incidence of driver mutational events in cHLs with ARID1A alterations (p=0.012). Analyses of co-occurring genetic events in EBV+ and EBV- cHLs confirmed that EBV- cHLs were significantly more likely to exhibit alterations of specific NF-κB signaling intermediaries (such as TNFAIP3 mutation and/or focal copy loss, p=0.006) and perturbations of MHC class I antigen presentation pathway components (inactivating B2M mutations, HLA-B mutations or focal copy loss of 6p21.32/HLA-B, p=0.008). The latter findings provide genetic bases for the reported differences in cell surface expression of MHC class I in EBV+ and EBV- cHLs. In PMBL, we defined 15 CCGs and more selective perturbations of specific epigenetic modifiers (ZNF217 and EZH2), transcription factors (PAX5 and IRF2BP2) and TP53, in comparison with cHL. The majority of these alterations were clonal supporting their role as early drivers. We identified 18 SCNAs and additional SVs in CIITA and PD-1 ligands, recurrent alterations of JAK/STAT and NF-κB signaling pathway components and a median of 9 genetic drivers per PMBL. Antigen presentation pathways in PMBL were perturbed by multiple recurrent alterations, including B2M mutations, focal copy losses of B2M and the MHCI/II loci, SVs of CTIIA and EZH2 mutations. There was a significant correlation between genetic perturbations of MHC class I pathway components and absence of MHC class I expression in PMBL, as previously described in cHL. Recurrent cHL alterations including B2M, TNFAIP3, STAT6, GNA13 and XPO1 CCGs and 2p/2p15/2p16.1, 6p21.32, 6q23.2 and 9p/9p24.1 SCNAs were also identified in >20% of PMBLs, highlighting shared pathogenetic mechanisms in these diseases. These tumors of predominantly young adults (median age: cHL 26 yrs; PMBL 34 yrs) both had a high rate of spontaneous deamination of CpGs, a clock-like mutational signature that is typically associated with aging. CHLs and PMBLs both exhibited previously uncharacterized molecular features that may increase sensitivity to PD-1 blockade, including high mutational burdens, in comparison with other lymphoid and solid tumors. In particular, the mutational burden in EBV- cHLs was among the highest reported, similar to that in carcinogen-induced cancers (melanoma and NSCLC). Additionally, both cHLs and PMBLs had an increased incidence of microsatellite instability and APOBEC mutational signatures, features associated with a more favorable response to PD-1 blockade. Taken together, these data define genetic similarities and differences in cHL and PMBL and establish a framework to comprehensively assess molecular bases of response to PD-1 blockade and develop rational combination therapies in these diseases. Disclosures Armand: Merck: Consultancy, Honoraria, Membership on an entity's Board of Directors or advisory committees, Research Funding, Speakers Bureau; Bristol-Myers Squibb: Consultancy, Honoraria, Membership on an entity's Board of Directors or advisory committees, Research Funding; Otsuka: Research Funding; Sigma Tau: Research Funding; Adaptive: Consultancy, Membership on an entity's Board of Directors or advisory committees, Research Funding; Affimed: Consultancy, Membership on an entity's Board of Directors or advisory committees, Research Funding; Roche: Research Funding; Pfizer: Consultancy; ADC Therapeutics: Consultancy; Infinity: Consultancy; Genentech: Research Funding; Tensha: Research Funding. Rodig:Merck: Research Funding; Affirmed: Research Funding; Kite, a Gilead Company: Research Funding; Bristol Myers Squib: Consultancy, Honoraria, Other: Travel Expenses, Speakers Bureau. Fromm:Merck, Inc.: Research Funding. Getz:Pharmacyclics: Research Funding; IBM: Research Funding; MuTect, ABSOLTUE, MutSig and POLYSOLVER: Patents & Royalties: MuTect, ABSOLTUE, MutSig and POLYSOLVER. Shipp:AstraZeneca: Honoraria, Membership on an entity's Board of Directors or advisory committees; Gilead Sciences: Honoraria, Membership on an entity's Board of Directors or advisory committees; Takeda Pharmaceuticals: Honoraria, Membership on an entity's Board of Directors or advisory committees; Bayer: Research Funding; Merck & Co.: Honoraria, Membership on an entity's Board of Directors or advisory committees, Research Funding; BMS: Consultancy, Honoraria, Membership on an entity's Board of Directors or advisory committees, Research Funding.

Diffuse large B-cell lymphoma (DLBCL) is a clinically and molecularly heterogeneous disease with recognized transcriptional subtypes associated with normal cells of origin, activated B-cell (ABC) and germinal center B-cell (GCB) tumors. Emerging data suggested that additional heterogeneity existed, prompting us to comprehensively characterize genomic signatures of 304 newly diagnosed DLBCLs from patients treated with state-of-the-art therapy. We integrated recurrent mutations, somatic copy number alterations (SCNAs) and structural variants (SVs) and identified 5 genetically distinct DLBCL clusters (C1- C5 DLBCLs; Chapuy, Stewart, Dunford, et al. Nat Med 2018). Specifically, we identified two genetically distinct ABC subtypes, including favorable-risk C1 DLBCLs with features of extrafollicular origin and alterations also seen in transformed marginal zone lymphomas (NOTCH2 and NF-κB pathway member mutations and BCL6 SVs). Unfavorable-risk C5 ABC DLBCLs harbored frequent 18q/BCL2 copy gain and co-occurring CD79B and MYD88L265P mutations. We also identified two genetically distinct GCB subtypes, including unfavorable-risk C3 DLBCLs with frequent BCL2 SVs, mutations in chromatin-modifying enzymes (CREBBP, MLL2, EZH2) and BCR/PI3K signaling pathway members (including inactivating PTEN mutations and copy loss). Favorable-risk C4 GCB DLBCLs had frequent mutations in core and linker histones and signaling intermediates (SGK1, BRAF and STAT3). Additionally, we identified an ABC/GCB-independent subtype, C2 DLBCLs, characterized by frequent bi-allelic TP53 inactivation, 9p21.23/CDKN2A copy loss and associated genomic instability reflected in recurrent SCNAs, increased genome doublings and a distinct outcome following induction therapy. A next step in utilizing the characterized genetic substructure was to confirm it in an independent series and develop a molecular classifier that allows prospective identification of C1-C5 DLBCLs. To this end, we accessed whole exome sequencing, copy number and SV data from a recent cohort of newly diagnosed DLBCLs (39 tumor-normal pairs, 462 tumor-only samples; Schmitz et al. NEJM 2018). All samples were re-analyzed using our mutational and SCNA pipelines and our newly generated tumor-only algorithm (Chapuy, Stewart, Dunford, et al. Nat Med 2018) to avoid batch effects and harmonize the datasets. SVs were used as reported. Purity and ploidy were inferred using ABSOLUTE and samples with missing data or low purity were removed. For the combined cohort (579 samples), we assessed our previously characterized 158 genetic drivers (Chapuy, Stewart, Dunford, et al. Nat Med 2018) and confirmed equal distribution of their marginal frequencies (R=0.88, p=1.5e-51), excluding batch effects. Next, we applied non-negative matrix factorization (NNF) consensus clustering to the combined dataset (158 genetic drivers vs. 579 tumors) and confirmed the C1-C5 DLBCL genetic clusters. Notably, tumors from both series contributed at comparable frequencies to the respective C1-C5 DLBCLs. We also noted an enrichment of the alternative genetic labels from Schmitz et al. in 3 of our C1-C5 DLBCL subtypes (B2N in C1 DLBCLs, p Next, we developed a molecular classifier that prospectively identified C1-C5 DLBCLs using a minimum number of easy-to-measure features. The NMF-defined classes of the combined cohort were used as gold-standard training and validation datasets. We tested different models for classification and selected an artificial neural network approach which provides accurate classification of individual samples and well-calibrated confidence metrics. To minimize potential overtraining, we developed a reduced input feature set of the 22 most discriminating features, constructed confidence metrics for each sample and trained an ensemble of Feed-Forward Neural Networks via 10-fold cross validation. With this approach, our classifier had 84% accuracy for the total set and 94% accuracy for the high-confidence samples (70% of all samples). The newly developed parsimonious classifier will allow prospective identification of the independently confirmed C1-C5 DLBCL subtypes in newly diagnosed patients, a necessity for clinical application. Disclosures Getz: Pharmacyclics: Research Funding; IBM: Research Funding; MuTect, ABSOLTUE, MutSig and POLYSOLVER: Patents & Royalties: MuTect, ABSOLTUE, MutSig and POLYSOLVER. Shipp:BMS: Consultancy, Honoraria, Membership on an entity's Board of Directors or advisory committees, Research Funding; Merck & Co.: Honoraria, Membership on an entity's Board of Directors or advisory committees, Research Funding; AstraZeneca: Honoraria, Membership on an entity's Board of Directors or advisory committees; Bayer: Research Funding; Gilead Sciences: Honoraria, Membership on an entity's Board of Directors or advisory committees; Takeda Pharmaceuticals: Honoraria, Membership on an entity's Board of Directors or advisory committees.

Primary mediastinal large B-cell lymphomas (PMBL) typically occur in young women who present with localized, large mediastinal masses. These tumors share certain clinical, pathomorphological and transcriptional features with classical Hodgkin lymphoma (cHL). To date, PMBL genetic analyses focused on limited sets of genes and recurrent somatic copy number alterations (SCNAs). Previously, we identified frequent 9p24.1/PD-L1/PD-L2 copy gains and increased expression of the PD-1 ligands as a genetically-defined immune escape mechanism in PMBL. The demonstrated efficacy of PD-1 blockade in relapsed/refractory PMBL led to recent FDA approval and underscored the importance of characterizing targetable genetic vulnerabilities in this disease. For these reasons, we obtained diagnostic biopsy specimens from 37 patients with PMBL (median age 34; female 70%) and performed whole exome sequencing (WES) with an expanded bait set to capture structural variants (SVs). Somatic alterations (mutations, SCNAs and SVs) were determined using established analytical pipelines including our algorithm for evaluating tumors without paired normal samples. Genes more frequently mutated than by chance, Candidate Cancer Genes (CCGs), were identified with MutSig2CV and recurrent SCNAs were defined with GISTIC2.0. SVs were characterized with a recently described 4-algorithm pipeline (Nature Medicine, 2018;24(5):679-690). First, we identified 15 CCGs (q-value We next analyzed the PMBL mutational signatures and identified 3 cases as hypermutators with MSI signatures, including 2 with MLH1 frameshift mutations and 1 with a nonsense PMS2 mutation. Despite the young age of the PMBL patient cohort, the majority of remaining mutations were caused by spontaneous deamination at CpGs, a genetic signature usually associated with aging. The next most prevalent mutational signatures were APOBEC and, infrequently, AID. We observed a higher median mutational density in PMBL (7.56 mutations/MB), compared to diffuse large B-cell lymphoma (DLBCL) and most solid cancers, providing a potential basis for increased neoantigen production and responsiveness to PD-1 blockade. Next, we identified 18 recurrent SCNAs, including 10 copy gains (2 focal and 8 arm level) and 8 copy losses (7 focal and 1 arm level). Copy gains of 9p24.1/PD-L1/PD-L2 were detected in 70% of cases. SVs were defined at base-pair resolution and included infrequent (2/37) tandem duplications of both PD-1 ligands and inactivating CTIIA SVs (deletions and inversions) in 10% (4/37) of cases. Although PMBL had a higher mutational density than DLBCL, the PMBL alterations involved a smaller number of median genetic drivers (9 [PMBL] vs 17 [DLBCL], respectively). Combined analyses of recurrent CCGs, SCNAs and SVs revealed that certain candidate driver genes were perturbed by multiple mechanisms. Examples include: TNFAIP3 (59% overall, 41% mutations, 24% copy loss, 6% biallelic); and B2M (51% overall, 30% mutations, 27% copy loss, 6% biallelic). Concurrent analyses of the 3 types of genetic alterations also revealed multiple bases of perturbing specific signaling pathways. In this PMBL series, 73% (27/33) of tumors exhibited one or more alterations of JAK/STAT pathway components: IL4R mutations (32%), JAK2 (9p24.1 focal copy gain [70%]) and STAT6 mutations (43%). Additionally, 59% of PMBLs had alterations of antigen presentation pathway components including B2M copy loss or mutations, copy loss of 6q21.33 (which includes the HLA class I/II loci) and SVs of CTIIA. These findings provide a genetic framework for analyzing the precise mechanism of action of PD-1 blockade in PMBL. Taken together, these findings underscore the importance of a comprehensive genomic analysis in PMBL and define additional candidate treatment targets and pathogenetic mechanisms in this disease. ____ BC, CS and AD contributed equally. GG and MAS contributed equally. Disclosures Rodig: Merck: Research Funding; KITE: Research Funding; Affimed: Research Funding; Bristol Myers Squibb: Research Funding. Shipp:Merck: Research Funding; AstraZeneca: Honoraria; Bristol-Myers Squibb: Honoraria, Membership on an entity's Board of Directors or advisory committees, Research Funding; Bayer: Research Funding.

Classical Hodgkin lymphoma (cHL) is composed of rare malignant Hodgkin Reed Sternberg (HRS) cells within an extensive, but ineffective, inflammatory/immune cell infiltrate. Emerging data suggests that cHLs use multiple genetic mechanisms to evade immune recognition. We previously found that HRS cells exhibit near-universal somatic copy number alterations (SCNAs) involving chromosome 9p24.1/PD-1-L1/PD-L2 and rare chromosomal rearrangements of PD-L1 or PD-L2. The 9p24.1 amplicon also includes JAK2, which increases JAK2 copy numbers, augments JAK2/STAT signaling and further induces PD-1 ligand expression. However, HRS cells also have inactivating mutations of B2M and decreased or absent MHC class I expression. In cHL, clinical responses to PD-1 blockade are unrelated to HRS cell expression of MHC class I but closely associated with HRS cell expression of MHC class II, highlighting the potential role of CD4+ T-cell effectors (J Clin Oncol 2018;36:942-50). To define genetic bases of response and resistance to PD-1 blockade and identify complementary treatment targets, we performed whole exome sequencing (WES) of HRS cells. We first used a previously described multi-color flow cytometric sorting protocol (Methods 2012; 57:368-75) to obtain highly purified CD30+ HRS cells and normal B cells from the excisional biopsies of 25 newly diagnosed cHLs. The isolated HRS cells and paired normal B cells were then subjected to WES using an optimized workflow for low input samples and an expanded bait set to capture structural variants (SVs). We used established analytical pipelines to identify significantly mutated genes (candidate cancer genes [CCGs], MutSig2CV), SCNAs (GISTIC2.0) and SVs (4 algorithms). With improved methodology and purity (median of 80%) of the isolated HRS cells, we defined 15 significantly mutated CCGs, 21 recurrent SCNAs, including 6 CN gains (4 focal and 2 arm level) and 15 CN losses (14 focal and 1 arm level), and low frequency SVs. We identified 2 cHLs as hypermutators with MSI signatures due to splice site mutations in MSH2 or missense mutations in POLE. Excluding the 2 hypermutators, the analyzed cHLs had a median mutational density of 6.4 mutations/Mb, that falls within the top quartile of reported cancer mutational frequencies (Nature 2013 499:214). We also identified a previously unappreciated high incidence of ARID1A mutations (24%) in cHL. This is noteworthy because ARID1A deficiency increases mutational load and augments the efficacy of PD-1 blockade in murine models (Nature Med 2018;24:556). Together, the observed MSI signatures, relatively high mutational burden and newly identified ARID1A mutations in cHL represent additional potential genetic bases for the efficacy of PD-1 blockade. Notably, these cHLs also exhibited recurrent 9p24.1 copy gain (80%) and multiple genetic bases of enhanced JAK/STAT signaling including JAK2 copy gain (80%), STAT6 mutations (32%) involving known hotspots (D419 and N421) in the DNA-binding domain and frequent inactivating SOCS1 mutations (68%). We also identified multiple genetic bases for immune evasion, including B2M inactivating mutations (36%), HLA-B mutations (16%) and 6p21.32/HLA-B copy loss (28%), copy loss of the larger 6p21.32 region and inactivating CIITA SVs (8%). Additional signaling pathways were perturbed by multiple genetic mechanisms in these cHLs. For example, NF-κB pathway alterations included: TNFAIP3 mutations (24%) and 6q23.2/TNFAIP3 copy loss (56%), 12% biallelic; NFKBIE mutations (24%) and 6q21.32/NFKBIE copy loss (12%); and NFKBIA mutations (16%). The gene encoding the nuclear export protein, XPO1, was perturbed by E571K mutations (24%) and frequent 2p15/XPO1 copy gain (72%). Additionally, GNA13, an activator of RHOA and modifier of PI3K signaling, was mutated in 24% of cases. Of interest, cHL recurrent alterations including B2M, TNFAIP3, STAT6, and GNA13 mutations and 6q23.2 and 9p24.1 SCNAs were also identified in > 20% of examined primary mediastinal B-cell lymphomas, highlighting shared pathogenetic mechanisms in these diseases. In summary, comprehensive genomic analyses of purified HRS cells reveal new genetic bases of immune evasion, potential mechanisms of response and resistance to PD-1 blockade and additional targetable alterations. KW, BC, CS, AD and DW contributed equally. JF, GG and MS contributed equally. Disclosures Rodig: Affimed: Research Funding; KITE: Research Funding; Merck: Research Funding; Bristol Myers Squibb: Research Funding. Shipp:Merck: Research Funding; Bayer: Research Funding; Bristol-Myers Squibb: Honoraria, Membership on an entity's Board of Directors or advisory committees, Research Funding; AstraZeneca: Honoraria.

Large B-cell lymphomas (LBCL) exhibit multiple genetic alterations, often in specific combinations. However, little is known about the biologic consequences of these concurrent genetic features. We recently characterized the comprehensive molecular signatures of two LBCL subtypes that exclusively involve extranodal disease sites, primary central nervous system lymphoma (PCNSL) and primary testicular lymphoma (PTL). In contrast to systemic diffuse large B-cell lymphomas (DLBCLs) or most activated B-cell (ABC)-type DLBCLs, the majority of PCNSLs and PTLs exhibit oncogenic Toll-like receptor (TLR) signaling (MYD88L265P activating mutations), often with concurrent B-cell receptor (BCR) activation (CD79BY196mut) and/or additional mutations or inactivating rearrangements of TBL1XR1 (Chapuy and Roemer et al, Blood 2016; 127:869-81). TBL1XR1 is a component of the NCoR/SMRT co-repressor complex that modulates TLR/MYD88 signaling by increasing the clearance of NCoR/SMRT transcriptional co-repressors from certain TLR/MYD88 target genes. However, the functional consequences of somatic mutations in TBL1XR1 (TBL1XR1mut) or reduced TBL1XR1wt expression in LBCLs remain to be defined. For these reasons, we performed in silico modeling of the identified TBL1XR1mut and functionally characterized TBL1XR1mut and decreased TBL1XR1wt expression in model LBCL systems. First, we evaluated the frequency and types of TBL1XR1mut in our local series of PCNSLs and 151 primary DLBCLs (102 local and 49 previously published) for which we had whole exome sequenceing data. Thirty-six percent of PCNSLs and 6% of DLBCLs in this series exhibit TBL1XR1mut, over half of which occur in the context of MYD88 and/or CD79B mutations. In these PCNSLs and DLBCLs, we identifed 13 different non-synonymous TBL1XR1 mutations. All identified non-overlapping TBL1XR1 mutations were in the WD40 propellar domains, which are critical for protein-protein interactions. To gain insights into the structural consequences of TBL1XR1mut, we overlayed these alterations onto the TBL1XR1 protein crystal structure, modeled the corresponding mutations in silico and assessed protein stability changes. Eighty percent of the TBL1XR1mut resulted in significant destabilization of the mutant protein (ΔΔGibbsFreeEnergy median: 9.78 kcal/mol, range: 1.91 - 28.31). We found that mutations often perturbed the 3D protein structure by abrogating critical polar interactions (representative example TBL1XR1wt vs. TBL1XR1H390R in Figure 1). These in silico modeling data suggest that genetic perturbations of TBL1XR1mut disrupt the function of the encoded protein. For that reason, we modeled reduced TBL1XR1wt expression in informative DLBCL cell lines with or without concurrent CD79B/MYD88 mutations (TMD8 and OCI-Ly1, respectively) by genetically depleting endogenous TBL1XR1 alleles with CRISPR/Cas9. All of the resulting LBCL cell lines with loss of either one or two TBL1XR1 alleles exhibited significantly increased proliferation in comparison to the parental controls. We next assessed the role of TBL1XR1mut by generating HA-tagged versions of each identified TBL1XR1 mutation with site-directed mutagenesis. Viral particles of TBL1XR1mut constructs and the TBL1XR1wt control were used to reconstitute mutant and wild type protein in the TBL1XR1-/- Ly1 cell line. Reconstitution with each TBL1XR1mut significantly enhanced LBCL proliferation, in comparison to TBL1XR1wt. Taken together, these in silico and experimental data suggest that TBL1XR1mut in PCNSL and DLBCL are inactivating events and establish TBL1XR1 as a tumor suppressor in these diseases. Additional analysis of potential interactions between perturbed TBL1XR1 and MYD88/CD79B are underway. Disclosures Rodig: Bristol-Myers Squibb: Honoraria, Research Funding; Perkin Elmer: Membership on an entity's Board of Directors or advisory committees.

Clonal evolution in response to therapy is a central feature of disease relapse. This raises a fundamental question in cancer biology: what enables the relapse clone to replace the pre-treatment clone? In other words, is the increased fitness of the relapse clone due to a lower death rate during therapy (less sensitivity to therapy) or a higher growth rate following therapy (superior ability to compete during repopulation)? We sought to address this question in chronic lymphocytic leukemia (CLL), as its relatively indolent disease kinetics enable the study of serially collected samples from the same patient over time. We recently reported the genetic characterization of 278 samples from patients enrolled in the German CLL Study Group CLL8 trial (Nature, in press). These samples were collected prior to first therapy with FC or FCR, and studied using whole-exome sequencing (WES). From this cohort, we further analyzed by WES 59 patients (FC [n = 28] or FCR [n = 31]) at time of relapse. We found that clonal evolution is the rule rather than the exception (57 / 59 CLLs), with TP53 alterations found in relapse in 15 cases. This series constitutes a unique opportunity to dissect the clonal dynamics of treated CLL. We therefore quantified clone-specific death and growth rates by targeted deep sequencing of serial peripheral blood samples, beginning at pre-treatment and ending at relapse. Given the expected minimal mutation detection sensitivity (0.1-1%) by targeted deep sequencing, we only selected samples with >1% CLL cells by flow cytometry. Such samples were available for 23 of 59 patients, with a median of 6 samples/patient (range 3-10). Based on the mutations identified by WES in the pre-treatment and relapse samples, we designed patient-specific multiplexed assays for targeted deep sequencing (median sequencing depth - 6561). A series of normal samples were sequenced together with patient samples to account for sequencing errors. The measurements of the CLL cell fraction in the sample, by sequencing and by flow cytometry, were highly correlated (r=0.89, p Clone-specific growth rates following therapy were calculated based on the measurements taken after therapy end, following exponential growth rate calculation. To calculate the clone-specific death rate during therapy, we applied two complementary approaches. First, measurements were taken after 3 cycles of therapy and the death rate per cycle was calculated. Second, clone-specific growth rates were back extrapolated to estimate the size of the population at the end of therapy, a method we have validated with an ultrasensitive emulsion droplet sequencing approach for targeted mutation detection. We discerned different mechanisms of relapse based on whether the relapse clone harbored mutated TP53 (TP53mut) or other mutations. In CLLs where the relapse clone contained TP53mut(n=10), the TP53mut clone showed lower death rate during therapy compared with the pre-treatment TP53 wildtype (TP53wt) clone (2.4 and 3.8 median log10 reduction, respectively; P = 0.02). On the other hand, the TP53mut clone showed only modestly higher growth rates during repopulation compared with the TP53wt clone (median growth rate of 0.8%/day vs. 0.56%/day, P = 0.13). Thus, differential sensitivityto therapy plays a primary role in TP53mut clonal evolution. In contrast, in the remaining cases whose relapse clone harbored mutations other than in TP53 (e.g., NOTCH1, ATM, SF3B1), we did not find differential sensitivity (median log10 clone reduction of 3.9 for the pre-treatment clone vs. 3.8 for the relapse clone, P=0.9). The primary engine leading to takeover by the relapse clone was a median of 1.5-fold higher growth rate during repopulation compared with the pretreatment clone. These data uncover evolutionary mechanisms in a personalized fashion directly from patient samples. Complementary efforts to apply these methods to define evolutionary mechanisms with targeted therapy are well under way. Thus, precise quantitation of clone-specific fitness in the context of therapy provides the required knowledge infrastructure to design the next generation of therapeutic algorithms, to maximize overall tumor elimination, instead of merely selecting one clone over another. Disclosures Tausch: Gilead: Other: Travel support. Fink:Roche: Honoraria, Other: travel grant. Hallek:Mundipharma: Honoraria, Other: Speakers Bureau and/or Advisory Boards, Research Funding; Boehringher Ingelheim: Honoraria, Other: Speakers Bureau and/or Advisory Boards; Celgene: Honoraria, Other: Speakers Bureau and/or Advisory Boards, Research Funding; Janssen: Honoraria, Other: Speakers Bureau and/or Advisory Boards, Research Funding; Roche: Honoraria, Other: Speakers Bureau and/or Advisory Boards, Research Funding; Gilead: Honoraria, Other: Speakers Bureau and/or Advisory Boards, Research Funding; AbbVie: Honoraria, Other: Speakers Bureau and/or Advisory Boards, Research Funding; Pharmacyclics: Honoraria, Other: Speakers Bureau and/or Advisory Boards, Research Funding. Stilgenbauer:AbbVie, Amgen, Boehringer-Ingelheim, Celgene, Genentech, Genzyme, Gilead, GSK, Janssen, Mundipharma, Novartis, Pharmacyclics, Roche: Consultancy, Honoraria, Research Funding.

Diffuse large B-cell lymphoma (DLBCL) is a genetically heterogeneous disease characterized by multiple low-frequency alterations including somatic mutations, copy number alterations (CNAs) and chromosomal rearrangements. We sought to identify previously unrecognized low-frequency genetic events, integrate recurrent alterations into comprehensive signatures and associate these signatures with clinical parameters. For these reasons, our multi-institutional international group assembled a cohort of 304 primary DLBCLs from newly diagnosed patients, 87% of whom were uniformly treated with state-of-the-art therapy (rituximab-containing CHOP regimen) and had long term followup. Tumors were subjected to whole exome sequencing with an extended bait set that included custom probes designed to capture recurrent chromosomal rearrangements. In this cohort, 47% of samples had available transcriptional profiling and assignment to associated disease subtypes. Analytical pipelines developed at the Broad Institute were used to detect mutations (MuTect), CNAs (Recapseq+Allelic Capseq) and chromosomal rearrangements (dRanger+Breakpointer) and assess clonality (Absolute). To analyze formalin-fixed paraffin-embedded tumors without paired normals we developed a method which utilized 8334 unrelated normal samples to stringently filter recurrent germline events and artifacts. Significant mutational drivers were identified using the MutSig2CV algorithm and recurrent CNAs were assessed with GISTIC2.0. In addition, we utilized a recently developed algorithm, CLUMPS2, to prioritize somatic mutations which cluster in 3-dimensional protein structure. With this approach, we identified > 90 recurrently mutated genes, 34 focal amplifications and 41 focal deletions, 20 arm-level events and > 200 chromosomal rearrangements in the DLBCL series. Of note, 33% of the mutational drivers were also perturbed by chromosomal rearrangements or CNAs, underscoring the importance of a comprehensive genetic analysis. In the large DLBCL series, we identified several previously unrecognized but potentially targetable alterations including mutations in NOTCH2 (8%) and TET2 (5%). The majority of identified chromosomal rearrangements involved translocations of potent regulatory regions to intact gene coding sequences. The most frequently rearrangements involved Ig regulatory elements which were translocated to BCL2, MYC, BCL6 and several additional genes with known roles in germinal center B-cell biology. After identifying recurrent somatic mutations, CNAs and chromosomal rearrangements, we performed hierarchical clustering and identified subsets of DLBCLs with comprehensive signatures comprised of specific alterations. A large subset of tumors shared recurrent alterations previously associated with follicular lymphoma including mutations of chromatin modifiers such as CREBBP, MLL2, and EZH2 in association with alterations of TNFRSF14 and GNA13 and translocations of BCL2. This cluster was enriched in GCB-type DLBCLs and contained a subset with select genetic alterations associated with an unfavorable outcome. An additional cohort of tumors was characterized by alterations perturbing B-cell differentiation including recurrent BCL6 translocations or alterations of PRDM1. A subset of these DLBCLs had alterations of NOTCH2 and additional pathway components or mutations of MYD88 in association with TNFAIP3, CD70 and EBF1, a master regulator of B-cell differentiation. An additional group of DLBCLs exhibited frequent MYD88 mutations in association with alterations of CD79B, PIM1, TBL1XR1 and ETV6 and BCL2 copy gain; these tumors were highly enriched for ABC-type DLBCLs. This coordinate signature and additional alterations of p53 pathway components were associated with outcome. We explored bases for the identified genetic alterations in DLBCL by performing an in silico mutational signature analysis. The most frequent mutational signatures were those of spontaneous deamination (aging) and AID with rare cases of microsatellite instability. We also assessed the clonality of identified genetic features to define cancer cell fraction and establish the timing of specific genetic events. The comprehensive genetic signatures of clinically annotated DLBCLs provide new insights regarding approaches to targeted therapy. Disclosures Link: Kite Pharma: Research Funding; Genentech: Consultancy, Research Funding. Rodig:Perkin Elmer: Membership on an entity's Board of Directors or advisory committees; BMS: Research Funding. Pfreundschuh:Boehringer Ingelheim, Celegene, Roche, Spectrum: Other: Advisory board; Roche: Honoraria; Amgen, Roche, Spectrum: Research Funding. Shipp:Gilead: Consultancy; Sanofi: Research Funding; BMS: Membership on an entity's Board of Directors or advisory committees, Research Funding; Merck: Membership on an entity's Board of Directors or advisory committees; Bayer: Membership on an entity's Board of Directors or advisory committees, Research Funding.

Intra-tumoral genetic heterogeneity and clonal evolution play a central role in disease relapse, and constitute one of the foremost obstacles to cancer cure. To measure intra-tumoral genetic heterogeneity, we recently developed an analytic framework which utilizes whole-exome sequencing (WES) to calculate the proportion of cancer cells within a sample harboring each mutation. Applying this approach to a clinically heterogeneous cohort of chronic lymphocytic leukemia (CLL), we previously discovered that the intra-tumoral genetic heterogeneity in primary patient samples prior to treatment initiation anticipates clonal evolution and is linked to adverse clinical outcome with therapy (Landau et al., Cell 2013). To definitively evaluate the value of intra-tumoral heterogeneity in predicting clinical outcome in CLL, we performed an analogous analysis of pre-treatment samples collected from study subjects enrolled on the CLL8 trial of the GCLLSG (Hallek et al., Lancet 2010). In this multicenter phase III study, 817 patients were randomized to receive as first-line therapy fludarabine and cyclophosphamide either alone or in combination with rituximab. Paired pre-treatment tumor DNA from CD19+ selected peripheral blood mononuclear cells (PBMC), and non-tumor DNA from CD19- PBMC were available from 309 subjects (61% IGHV unmutated, 26% del(11q), 4.5% del(17p)). From these 309 sample pairs, we performed SNP array analysis and WES. Somatic mutations were identified as single nucleotide variants (SNVs), indels and copy number alterations (CNAs), detected in the tumor DNA but not in matched germline DNA. Subsequently, we utilized the algorithm ABSOLUTE to calculate the purity and ploidy of each sample, and classify every mutation as either clonal (i.e., present in all leukemic cells in the sample) vs. subclonal (i.e., present in a subset of leukemic cells). Progression free survival (PFS) was defined as the primary outcome measure. Preliminary sequencing results for 275 of 309 CLL samples are currently available. Across these samples, WES identified 7290 somatic SNVs and indels. Overall mutation rate for these samples was 0.8/Mb, similar to prior large-scale reports of DNA sequencing of CLL. By applying the algorithm MutSig, which identifies putative driver genes recurrently mutated more than expected by chance, we detected 16 of 24 previously described CLL drivers (e.g., TP53, ATM, SF3B1 and NOTCH1) to significantly affect this cohort. In addition, due to the large sample size we have identified 9 novel putative drivers in CLL including MGA, IKZF3 and ARID1A. CNA analysis identified the previously reported recurrent CNAs including (del(13q), del(11q), del(17p), trisomy 12 and del(8p)). At this time, subclonal analysis by ABSOLUTE is available for 216 of 309 CLLs. Across the 216 samples, we identified 2041 clonal and 2106 subclonal mutations. A putative driver gene mutation was identified as clonal in 102 (47.2%) patients, and as subclonal in 81 (37.5%) patients. The presence of a clonal driver was not significantly associated with differences inPFS (logrank P= 0.236). In contrast, the presence of a subclonal driver was associated with a significantly shortened PFS (logrank P= 0.006, Figure 1A-B). A Cox regression model that included the presence of a subclonal driver as well as the treatment arm assignment (FC vs. FCR), retained the presence of a subclonal driver as statistically significant in association with shorter PFS (HR 1.56 [95%CI 1.12-2.19], P=0.009). The presence of a subclonal driver was furthermore associated with shorter overall survival (logrank P=0.02, Fig 1C-D). This preliminary analysis demonstrates that pre-treatment characterization of intra-tumoral genetic heterogeneity can identify patients at risk for inferior outcome with first-line fludarabine-based therapy in CLL. With completion of this analysis for the entire cohort of 309 patients, we plan to perform a multivariable analysis to evaluate the clinical impact of clonal and subclonal driver mutations in CLL. Further ongoing studies include a longitudinal genetic evaluation of the subset of patients with subsequent disease relapse in order to unravel the evolutionary dynamics underlying disease progression. Figure 1 Figure 1. Disclosures Fischer: Roche: Travel grants Other. Hallek:Janssen, Pharmacyclics: Consultancy, Research Funding.

Introduction. Primary testicular lymphoma (PTL) and primary central nervous system lymphoma (PCNSL) are large B-cell lymphomas (LBCL) that occur in immune privileged (IP) sites and share certain clinical and molecular features. To date, the treatment of these IP lymphomas is largely empiric and more effective targeted therapies are needed. Methods. To define actionable genetic features of IP lymphomas, we performed comprehensive genomic analyses of 21 PCNSLs and 7 PTLs and validated specific alterations in an independent cohort of 43 additional PTLs. Recurrent copy number alterations (CNAs) were detected using high-density single nucleotide polymorphism (SNP) arrays and the GISTIC algorithm and integrated with transcriptional profiles to identify candidate driver genes. Recurrent somatic mutations were identified using a combination of whole exome sequencing (WES) of paired tumor/normal samples and whole transcriptome sequencing (RNA-Seq) of the additional tumors without paired normal samples. Results. In systemic diffuse large B-cell lymphomas (DLBCLs), multiple low-frequency CNAs and associated target genes decrease p53 activity and perturb cell cycle regulation; infrequent somatic mutations of TP53 also deregulate these pathways (Cancer Cell, 2012; 22:359-372). In contrast, PCNSLs and PTLs primarily exhibit bi-allelic deletion of the upstream regulator of p53 activity and cell cycle, CDKN2A (~70% PCNSLs and ~80% of PTLs) and rarely have copy loss or somatic mutations of TP53 or CNAs of additional pathway components. The most commonly mutated genes in PCNSL and PTL, CD79B and MYD88, are also perturbed in a subset of systemic DLBCLs. However, mutations of these two genes are much more frequent in IP lymphomas (70% MYD88 and 61% CD79B of analyzed PCNSLs and PTLs) and these alterations are commonly found in the same cases (57% of cases in this series). These data indicate that concurrent oncogenic activation of the B-cell receptor (BCR) and the Toll-like receptor (TLR) signaling pathways is a characteristic feature of IP lymphomas with implications for targeted therapies. Among the IP lymphomas, ~20% of PCNSLs and ~40% PTLs exhibit 3q12.3/NFKBIZ copy gain and increased expression of the NFKBIZ protein product, IκB-ζ, an atypical IκB family member induced by TLR signaling. In our PTL series, MYD88 wild-type tumors had the highest 3q12.3/NFKBIZ copy gains, and ~90% of all analyzed PTLs had structural bases for NFκB activation via the TLR pathway. Lentiviral-mediated IκB-ζ knockdown decreased expression of the IκB-ζ target genes, IκB-α and BCL-xL, and induced apoptosis of LBCL cell lines with MYD88 L265P mutations, NFKBIZ gain or both alterations. In addition, enforced expression of NFKBIZ enhanced the growth of LBCLs with normal NFKBIZ copy numbers. Taken together, these data suggest that many IP lymphomas depend upon oncogenic MYD88/NFKBIZ signaling. Although the majority of CNAs and somatic mutations were shared by PCNSLs and PTLs, certain alterations were primarily observed in PTL. In both the initial and independent validation series, > 40% of PTLs exhibited copy gain of chromosome 9p24.1/CD274 (PD-L1) / PDCD1LG2 (PD-L2) and associated overexpression of the PD-1 ligands. These observations were of particular interest because 9p24.1 copy gain is a characteristic abnormality in two additional lymphoid malignancies, primary mediastinal LBCL and classical Hodgkin lymphoma, PD-1 signaling promotes tumor immune evasion and the PD-1 pathway is targetable. We also identified one PTL in which a novel translocation juxtaposed the regulatory elements of TBL1XR1 (chromosome 3) to the start codon-bearing exon 2 of PDCD1LG2 (PD-L2) (chromosome 9). This translocation, which was detected by RNA-Seq and confirmed by 5’ RACE and a newly developed split-apart FISH assay, resulted in dramatic overexpression of the PD-L2 protein. These data suggest that PTLs utilize several genetic mechanisms to deregulate the PD-1 ligands and limit anti-tumor immunity. Conclusions. Integrative and comparative genomic studies define PCNSL and PTL as related but unique lymphoid malignancies with targetable genetic alterations, and associated p53 deficiency and cell cycle deregulation, concurrent oncogenic BCR and TLR signaling and PD-1 dependent immune evasion that warrant further clinical investigation. Note: B.C. and M.G.M.R have made equal contributions to this research. Disclosures Feuerhake: Roche Pharma Research and Early Development (pRED) from 2008-2012: Employment. Freeman:Merck: on the PD-1 pathway Patents & Royalties; EMD-Serrono: on the PD-1 pathway Patents & Royalties; Boehringer-Ingelheim: on the PD-1 pathway Patents & Royalties; Amplimmune: on the PD-1 pathway Patents & Royalties; Roche: on the PD-1 pathway Patents & Royalties; Bristol-Myers-Squibb: on the PD-1 pathway Patents & Royalties; Novatis: on the PD-1 pathway, on the PD-1 pathway Patents & Royalties. Shipp:Sanofi: Research Funding; Bayer: Membership on an entity's Board of Directors or advisory committees; Gilead: Consultancy, Membership on an entity's Board of Directors or advisory committees; Bristol-Myers-Squibb: Membership on an entity's Board of Directors or advisory committees, Research Funding; Pharmacyclics: Membership on an entity's Board of Directors or advisory committees; Merck: Membership on an entity's Board of Directors or advisory committees; Janssen R&D: Membership on an entity's Board of Directors or advisory committees.

Unbiased high-throughput massively parallel sequencing methods have transformed the process of discovery of novel putative driver gene mutations in cancer. In chronic lymphocytic leukemia (CLL), these methods have yielded several unexpected findings, including the driver genes SF3B1, NOTCH1 and POT1. Recent analysis, utilizing down-sampling of existing datasets, has shown that the discovery process of putative drivers is far from complete across cancer. In CLL, while driver gene mutations affecting >10% of patients were efficiently discovered with previously published CLL cohorts of up to 160 samples subjected to whole exome sequencing (WES), this sample size has only 0.78 power to detect drivers affecting 5% of patients, and only 0.12 power for drivers affecting 2% of patients. These calculations emphasize the need to apply unbiased WES to larger patient cohorts. To this end, we performed a combined analysis of CLL WES data joining together our previously published cohort of 159 CLLs with data from 103 CLLs collected by the International Cancer Genome Consortium (ICGC). The raw sequencing reads from these 262 primary tumor samples (102 CLL with unmutated IGHV, 147 with mutated IGHV, 13 with unknown IGHV status) were processed together and aligned to the hg19 reference genome. Somatic single nucleotide variations (sSNVs) and indels were detected using MuTect. Subsequently, inference of recurrently mutated genes was performed using the MutSig algorithm. This method combined several characteristics such as the overall mutation rate per sample, the gene specific background mutation rate, non-synonymous/synonymous ratio and mutation clustering to detect genes that are affected by mutations more than expected by chance. This analysis identified 40 recurrently mutated genes in this cohort. This included 22 of 25 previously identified recurrently mutated genes in CLL. In addition, 18 novel candidate CLL drivers were identified, mostly affecting 1-2% of patients. The novel candidates included two histone proteins HIST1H1D and HIST1H1C, in addition to the previously identified HIST1H1E. Another was IKZF3, affected by a recurrent sSNV resulting in a p.L162R change in its DNA binding domain, in close proximity to a region recently identified as critical for lenalidomide resistance in multiple myeloma (MM). An additional recurrently mutated gene was nuclear RNA export factor 1 (NXF1), which along with previously known recurrently mutated genes (SF3B1, XPO1, DDX3X), highlights the importance of RNA processing to CLL biology. Finally, this search for putative CLL driver genes also identified ASXL1 and TRAF3, already characterized as drivers in acute myeloid leukemia and MM, respectively. Of the 59 of 262 samples for which RNA-seq data were available, 76% of the identified driver mutations were detected and thereby validated. Validation using RNAseq detection of driver mutations and targeted sequencing within the entire cohort are ongoing. The larger size of our cohort enabled the separate application of the somatic mutation discovery process to samples with mutated or unmutated IGHV. Among the 147 samples with mutated IGHV, only 5 driver genes (TP53, SF3B1, MYD88, CHD2, RANBP2) retained significance. In contrast, analysis of the 102 IGHV unmutated samples revealed a distinct and more diverse pattern of recurrently mutated genes (lacking MYD88 and CHD2, and including NOTCH1, RPS15, POT1, NRAS, EGR2, BRAF, MED12, XPO1, BCOR, IKZF3, MAP2K1, FBXW7 and KRAS). This extended cohort also allowed for better resolution of the clinical impact of those genetic variants with greater than 4% prevalence in the cohort. For example, samples with POT1 mutations were found to be associated with shorter time from sample to therapy compared with those with wild-type POT1 (P= 0.02). Our study demonstrates that with larger cohort size, we can effectively detect putative driver genes with lower prevalence, but which may nonetheless have important biological and clinical impact. Moreover, our interrogation shows that subset analysis can reveal distinct driver patterns in different disease subsets. In particular, the marked clinical difference between CLLs with mutated and unmutated IGHV may reflect the higher likelihood of the latter group to harbor a broader spectrum of driver mutations with a more complex pattern of co-occurrence. Disclosures Brown: Sanofi, Onyx, Vertex, Novartis, Boehringer, GSK, Roche/Genentech, Emergent, Morphosys, Celgene, Janssen, Pharmacyclics, Gilead: Consultancy.