Sarangi, Vivekananda

Background: Follicular Lymphoma (FL) is the second most common non-Hodgkin lymphoma and presents with significant clinical, cellular, molecular, and genetic heterogeneity. Despite the name and defining microanatomic location, the genetic and molecular identity and pathogenesis of the FL tumor cell is largely uncharacterized. Prior clinical and molecular classifications of FL have been primarily driven by pathologic classification (Grade 1-3b), genetic classification (M7-FLIPI), or gene expression profiling (IR-1/2). Using a unique cohort of 93 FL tumors, we have explored the transcriptomic signature of purified FL B cells, along with their matched whole tumor, and identified unique molecular subsets which are defined by distinct pathway activation, immune content, and genomic signatures identified through whole exome sequencing (WES). Methods: Frozen tumor biopsies from 93 untreated FL (Grade 1-3b) patients enrolled in the University of Iowa/Mayo Clinic Lymphoma SPORE were used for the study. DNA was isolated from whole tumor cell suspensions and RNA was isolated from both whole tumor and B cell enriched cell suspensions. RNA sequencing (RNAseq) and WES was performed in the Mayo Clinic Genome Analysis Core. RNAseq and WES data were processed using the Mayo Clinic standard pipeline and novel driver genes were identified using 20/20+ driver analysis. Copy number variants were identified using GISTIC 2.0. NMF clustering and single sample gene set testing, for B cell lineage and tumor microenvironment (TME) signatures, was performed in R using the NMF and SingScore packages. Results: NMF consensus clustering of FL B cell RNAseq data identified two distinct subsets, C1 (n=32) and C2 (n=57). Clinically, C1 was associated with being FL grade 3 (p<0.001) and more aggressive therapy (e.g. immunochemotherapy) at diagnosis while other clinical factors were similar between the groups. To determine the biologic underpinnings that drove the NMF analysis, the top NMF metagenes were incorporated (n=389 for C1 and n=293 for C2) for GO analysis. Among those genes were AICDA, IRF4, and EZH2 in C1, and BCL2 in C2. Additionally, we identified significant gene set enrichment with NMF metagenes and calculated differential gene set expression comparing C1, C2, and normal B cells (n=5). Together, these analyses suggest C1 displays significantly (p<0.05) enhanced DNA replication and metabolic activity, whereas C2 displays significantly enhanced cytoskeleton rearrangement, repressed inflammatory cytokine signaling, and NF-kB activity. Based on recent scRNAseq data suggesting that FL derives from distinct germinal center B cells, the B lineage difference between C1 and C2 was explored. Both subgroups were heterogeneous for dark (DZ), intermediate (INT), light (LZ) zones, and plasmablast (PBL) gene signatures. When compared to benign B cell samples, C1 was enriched for a pre-PBL (p<0.01), and C2 for an INT phenotype related to a LZ signature (p<0.01). To define if a genomic signature was associated with C1 and C2, mutations and copy number variants (>5% frequency) were examined. TNFAIP3, TP53, and BCL6 alterations were enriched in C1 samples, whereas C2 associated with alterations in BCL2, KMT2D, CREBBP, REL, and MYC. Finally, B cell clusters were analyzed for TME signatures. C1 samples displayed significant enrichment of macrophage, cytotoxic cell, gamma-delta-Tcell, and endothelial cell TME elements (p<0.05) consistent with an inflamed/hot tumor. In contrast, C2 samples exhibited significantly lower signatures (p<0.05) in all entities except for the fibroblast signature, compared to C1 and normal, consistent with an immune desert phenotype. Conclusion: Our results suggest that B cells from FL patients display two distinct transcriptomic signatures. C1 identifies an immunologically active tumor, driven by TNFAIP3 alterations, with pre-PBL characteristics, DNA replication and repair, inflammatory cytokine secretion/signaling, and hyper-metabolic characteristics. C2 identifies an immunologically quiet tumor, driven by alterations in BCL2 and chromatin modifiers, with an intermediate GC phenotype, repressed cytokine signaling, and active cell cycle progression and cytoskeleton rearrangement. This study improves our understanding of the mechanisms driving FL tumors and motivates further investigation into the relationship between tumor intrinsic factors that may influence the TME. Disclosures Maurer: Morphosys: Membership on an entity's Board of Directors or advisory committees; Kite: Membership on an entity's Board of Directors or advisory committees; Pfizer: Membership on an entity's Board of Directors or advisory committees; Nanostring: Research Funding; Celgene / BMS: Research Funding. Ansell:ADC Therapeutics: Research Funding; Trillium: Research Funding; Affimed: Research Funding; Regeneron: Research Funding; AI Therapeutics: Research Funding; Takeda: Research Funding; Seattle Genetics: Research Funding; Bristol Myers Squibb: Research Funding. Cerhan:NanoString: Research Funding; BMS/Celgene: Research Funding. Novak:Celgene/BMS: Research Funding.

Abstract Background: Genetic studies have shifted to sequencing-based rare variants discovery after decades of success in identifying common disease variants by Genome-Wide Association Studies using Single Nucleotide Polymorphism chips. Sequencing-based studies require large sample sizes for statistical power but often inadvertently introduce batch effects because samples are typically collected, processed, and sequenced at multiple centers. Conventionally, batch effects are first detected and visualized using Principal Components Analysis and then controlled by including batch covariates in the disease association models. For sequencing-based genetic studies, because all variants included in the association analyses have passed quality control measures, this conventional approach treats every variant as equal and ignores the substantial differences still remaining in variant qualities and characteristics such as genotype quality scores, alternative allele fractions (fraction of reads supporting alternative allele at a variant position) and sequencing depths. Results: In the Alzheimer’s Disease Sequencing Project (ADSP) exome dataset of 9,904 cases and controls, we discovered hidden variant-level differences between sample batches of three sequencing centers and two exome capture kits. Although sequencing centers were included as a covariate in our association models, we observed differences at the variant level in genotype quality and alternative allele fraction between samples processed by different exome capture kits that significantly impacted both the confidence of variant detection and the identification of disease-associated variants. Furthermore, we found that the association signals of a subset of top disease risk variants came exclusively from samples processed by one exome capture kit that was more effective at capturing the alternative alleles compared to the other kit. Conclusions: Our findings highlight the importance of additional variant-level quality control for large sequencing-based genetic studies. More importantly, we demonstrate that automatically filtering out variants with batch differences may lead to false negatives if the batch discordance came largely from quality differences and if the variants from one batch had better quality scores.

Single nucleotide variants in the open reading frames (ORFs) of pharmacogenes are important causes of interindividual variability in drug response. The functional characterization of variants of unknown significance within ORFs remains a major challenge for pharmacogenomics. Deep mutational scanning (DMS) is a high‐throughput technique that makes it possible to analyze the functional effect of hundreds of variants in a parallel and scalable fashion. We adapted a “landing pad” DMS system to study the function of missense variants in the ORFs of cytochrome P450 family 2 subfamily C member 9 ( CYP2C9 ) and cytochrome P450 family 2 subfamily C member 19 ( CYP2C19 ). We studied 230 observed missense variants in the CYP2C9 and CYP2C19 ORFs and found that 19 of 109 CYP2C9 and 36 of 121 CYP2C19 variants displayed less than ~ 25% of the wild‐type protein expression, a level that may have clinical relevance. Our results support DMS as an efficient method for the identification of damaging ORF variants that might have potential clinical pharmacogenomic application.

Introduction: Current classifications of low-grade B-cell lymphomas (LGL), including splenic marginal zone lymphomas (SMZL), nodal marginal zone lymphomas (NMZL), and Lymphoplasmacytic lymphomas (LPL) are based on a mixture of clinical features and morphologic, immunophenotypic, and genetic findings from the tumor biopsy specimen. While this approach to classification makes pathologic diagnosis more precise, the corresponding clinical impact for the timing and choice of treatment is limited, and differentiating between cases can be challenging. Although LGL are considered indolent and the 10-year overall survival (OS) is about 80%, 70% of cases will eventually require treatment and approximately 30% of patients display a more aggressive phenotype and have a poor prognosis. Consequently, further investigations into the driving genetic, biological, and immune mechanisms of LGL are essential for early identification of high-risk patients and design of personalized treatments. Materials and Methods: RNA-seq was performed on 63 newly diagnosed LGL patient samples from the Mayo Clinic/University of Iowa Lymphoma SPORE: SMZL (N=48), NMZL (N=6), LPL (N=5), MALT (N=2), and BCL (N=2) as well as 5 normal memory B cell controls (CD19+CD27+). For identification of biologic clusters, filtered RNA-seq data was analyzed using the Non-negative Matrix Factorization (NMF) clustering tool from the Broad Institute against the normal samples. Each cluster was analyzed for differential gene expression. This analysis generated cluster-specific T values for each gene. Genes that were significantly associated with a cluster (FDR<0.05) were analyzed for genetic ontology. Cibersort was used to deconvolute the tumor microenvironment (TME). We performed whole exome sequencing (WES) analyses on 60/63 matched LGL to characterize mutations and copy number alterations. Results: WES revealed the presence of previously reported mutations in genes such as MYD88, SPEN, NOTCH, and KLF2 as well as copy number alterations 7q31.2, BCL6, TNFAIP3, and BCL2. NMF clustering of RNA-seq data resulted in a best fit of 5 clusters, named LGL1 through LGL4 and "Normals". Pathologic subtypes were not exclusive to a specific cluster. Differential RNA expression analysis resulted in gene sets that were differentially expressed in each cluster. A cluster signature of the top 1% of associated genes (N=94) was created for each. LGL1 genes were associated with higher BCR signaling. LGL2 was characterized by regulatory dysfunction, particularly concerning mitochondrial integrity. LGL3 genes were associated with high TME and NOTCH and STAT signaling components. Specifically, LGL3 was associated with high CD8+ T-cell and M2 macrophage TME presence. LGL4 was distinguished by the presence of aggressive genetic programs related to B-cell activation such as NF-kB and IRF4. Consequently, LGL4 patients had significantly poorer rates of event-free survival (EFS) (P=0.029) and OS (P=0.006) when compared against the other groups. The top 1% gene signature (N=94) was next validated against an independent cohort of 84 LGL samples with gene expression data (SMZL [N=34], NZML [N=24], and LPL [N=24]). A similar cluster pattern emerged, with pathological subtypes distributing across LGL classifications based on respective transcriptomic signatures. Association of DNA variants within each cluster was also analyzed, but due to the low frequency of genomic variants detected in LGL, nothing was significantly associated. However, we did see an enrichment of BCL2, BCL6, and TNFAIP3 alterations in LGL4 and an overall lower driver gene mutation frequency in LGL2. Conclusions: Gaining a greater understanding of LGL based on their genetic, biologic, and immune profiles will establish a valuable platform for clinicians to identify high-risk cases and make better therapeutic decisions. Using a large cohort of well-annotated cases we identify novel clusters of LGL that present unique genomic and clinical profiles. Our study lays the groundwork for a precision therapy approach in LGL in which DNA or RNA profiles can be used to identify patients early in treatment who may not benefit from the current standard of care and who would benefit from closer monitoring and targeted agents. Disclosures Anagnostou: American Society of Hematology, Mayo Clinic/Iowa Lymphoma SPORE, Mayo Clinic Immune Monitoring Core, Mayo Clinic Hematology Small Grant: Research Funding. Ansell:Bristol-Myers Squibb: Research Funding; Regeneron: Research Funding; Regeneron: Research Funding; Seattle Genetics: Research Funding; Trillium: Research Funding; Seattle Genetics: Research Funding; Affimed: Research Funding; Bristol-Myers Squibb: Research Funding; Regeneron: Research Funding; LAM Therapeutics: Research Funding; LAM Therapeutics: Research Funding; Seattle Genetics: Research Funding; Affimed: Research Funding; Regeneron: Research Funding; Bristol-Myers Squibb: Research Funding; Bristol-Myers Squibb: Research Funding; Seattle Genetics: Research Funding; Trillium: Research Funding; Seattle Genetics: Research Funding; Trillium: Research Funding; LAM Therapeutics: Research Funding; Affimed: Research Funding; Mayo Clinic Rochester: Employment; Mayo Clinic Rochester: Employment; Mayo Clinic Rochester: Employment; Regeneron: Research Funding; Affimed: Research Funding; Seattle Genetics: Research Funding; Affimed: Research Funding; Trillium: Research Funding; Bristol-Myers Squibb: Research Funding; Seattle Genetics: Research Funding; Regeneron: Research Funding; Bristol-Myers Squibb: Research Funding; LAM Therapeutics: Research Funding; LAM Therapeutics: Research Funding; Mayo Clinic Rochester: Employment; Mayo Clinic Rochester: Employment; Regeneron: Research Funding; Regeneron: Research Funding; Trillium: Research Funding; Mayo Clinic Rochester: Employment; Trillium: Research Funding; Affimed: Research Funding; Bristol-Myers Squibb: Research Funding; Mayo Clinic Rochester: Employment; Affimed: Research Funding; Bristol-Myers Squibb: Research Funding; Affimed: Research Funding; Seattle Genetics: Research Funding; Trillium: Research Funding; LAM Therapeutics: Research Funding; Affimed: Research Funding; Seattle Genetics: Research Funding; Regeneron: Research Funding; Trillium: Research Funding; Trillium: Research Funding; Bristol-Myers Squibb: Research Funding; LAM Therapeutics: Research Funding; Mayo Clinic Rochester: Employment; LAM Therapeutics: Research Funding; LAM Therapeutics: Research Funding; Mayo Clinic Rochester: Employment. Cerhan:Janssen: Membership on an entity's Board of Directors or advisory committees; Celgene: Research Funding; NanoString: Research Funding. Novak:Celgene Coorperation: Research Funding.

Introduction: New Diffuse Large B-cell (DLBCL) treatments remain a clinical need despite the success of rituximab combined with CHOP chemotherapy (RCHOP), which results in durable responses in 60-70% of patients. Those refractory to, or who relapse following, first-line therapy have a very poor outcome, with only 20% surviving beyond 5 years. Rationally-targeted frontline strategies are needed, especially for those with high-risk disease. Significant advances have been made in the genomic classification of DLBCL, but none have impacted the design of phase III trials for untreated DLBCL patients. Building on new genetic profiling studies to personalize clinical treatment, an NCI initiative, could allow clinicians to add targeted therapies to the RCHOP backbone based on individual tumor signatures. The phase II ECOG-ACRIN1412 trial, which compared RCHOP combined with lenalidomide (R2CHOP) versus RCHOP showed significantly superior event free survival (EFS) and overall survival benefits for those treated with R2CHOP. Herein, we report the profile of a high-risk ABC/non-GCB subset of DLBCL driven by genomic alterations in inflammatory genes that are susceptible to front-line R2CHOP, but continue to experience poor outcome with RCHOP alone. Methods: We studied a total of 196 DLBCL patients. 47 were treated with R2CHOP from an investigator-initiated, open-label, single-arm phase II study (NCT00670358). 149 were newly diagnosed DLBCL cases treated with RCHOP, or R-immunochemotherapy (herein called RCHOP), and followed prospectively through the Molecular Epidemiology Resource (MER) of the University of Iowa/Mayo Clinic SPORE that served as a contemporary cohort. Patients from each treatment group were divided based on their event free survival at 24 months (EFS24). DNA alterations within these populations were identified through whole exome sequencing (WES). Variants were analyzed for their presence in EFS24 achievement or failure groups in both RCHOP and R2CHOP. Both tumor and germline sequencing was performed for 47/47 R2CHOP cases and 49/149 RCHOP cases. Gene expression data from the PanCan panel of 730 B-cell-related genes was analyzed to determine gene expression profiles characteristic of the high-risk/R2CHOP-profile on 59 available non-GCB DLBCL cases (45 RCHOP; 14 R2CHOP). A two-sided comparative marker analysis T statistic test was applied to assess what genes displayed differential expression based on achieving EFS24 in both R2CHOP and RCHOP populations. Positive values represented associations with achieving EFS24 and negative values were associated with cases of EFS24 failure. Results: Three genes were enriched in the RCHOP cases that failed EFS24 but achieved EFS24 with R2CHOP among non-GCB patients: PIM1, SPEN, and MYD88 (L265P), herein referred to as Responder Alterations. In R2CHOP cases, patients with a Responder Alteration had a better overall EFS (P = 0.051) compared to wild type patients. In contrast, RCHOP treated patients with a Responder Alteration in their tumor had a significantly worse overall EFS (P = 0.0004) compared to patients without a mutation. Together, PIM1, SPEN, or MYD88 (L265P) mutations were present in 38.0% (30/79) of all non-GCB cases. Collective R2CHOP and RCHOP EFS24 differential gene expression T values were significantly different for 18 previously-defined signatures. The R2CHOP cases that achieved EFS24 were enriched for genes involved in cell cycle, JAK-STAT, cytokine signaling, and NF-κB pathways based on these signatures and ontology analyses. Lastly, cases with WES and PanCan data were analyzed together to observe differential gene expression patterns between cases with and without signature mutations. These data suggest that R2CHOP disrupts tumors reliant on IRF4, NF-κB, and STAT transcription factors, leading to a loss of proliferative feedback systems. Conclusions: Our combined analysis of DNA and RNA across R2CHOP and RCHOP treatment cohorts identifies a high-risk non-GCB phenotype that encompasses approximately 38% of non-GCB patients, is capable of sustaining JAK-STAT and NF-κB signaling, and is sensitive to R2CHOP. Our study lays the groundwork for a precision therapy approach in DLBCL in which DNA or RNA profiles can be used to identify patients early in treatment who may not benefit from the current standard of care, RCHOP, and who would benefit from the addition of lenalidomide or other targeted agents. Disclosures Gandhi: Celgene Corporation: Employment, Equity Ownership, Patents & Royalties. Ansell:LAM Therapeutics: Research Funding; Seattle Genetics: Research Funding; LAM Therapeutics: Research Funding; Seattle Genetics: Research Funding; Regeneron: Research Funding; Bristol-Myers Squibb: Research Funding; Affimed: Research Funding; Trillium: Research Funding; Trillium: Research Funding; Regeneron: Research Funding; Bristol-Myers Squibb: Research Funding; LAM Therapeutics: Research Funding; Trillium: Research Funding; Affimed: Research Funding; LAM Therapeutics: Research Funding; Bristol-Myers Squibb: Research Funding; Trillium: Research Funding; Affimed: Research Funding; Bristol-Myers Squibb: Research Funding; Trillium: Research Funding; LAM Therapeutics: Research Funding; Bristol-Myers Squibb: Research Funding; Regeneron: Research Funding; Affimed: Research Funding; Seattle Genetics: Research Funding; Mayo Clinic Rochester: Employment; Regeneron: Research Funding; Seattle Genetics: Research Funding; Mayo Clinic Rochester: Employment; Seattle Genetics: Research Funding; LAM Therapeutics: Research Funding; LAM Therapeutics: Research Funding; LAM Therapeutics: Research Funding; Regeneron: Research Funding; Regeneron: Research Funding; Seattle Genetics: Research Funding; Seattle Genetics: Research Funding; Trillium: Research Funding; Trillium: Research Funding; Affimed: Research Funding; Affimed: Research Funding; Mayo Clinic Rochester: Employment; Mayo Clinic Rochester: Employment; Mayo Clinic Rochester: Employment; Seattle Genetics: Research Funding; Mayo Clinic Rochester: Employment; Bristol-Myers Squibb: Research Funding; Bristol-Myers Squibb: Research Funding; Affimed: Research Funding; Regeneron: Research Funding; Regeneron: Research Funding; Bristol-Myers Squibb: Research Funding; Regeneron: Research Funding; Seattle Genetics: Research Funding; Trillium: Research Funding; Mayo Clinic Rochester: Employment; Mayo Clinic Rochester: Employment; Trillium: Research Funding; Affimed: Research Funding; Mayo Clinic Rochester: Employment; Bristol-Myers Squibb: Research Funding; Affimed: Research Funding; LAM Therapeutics: Research Funding. Cerhan:Celgene: Research Funding; NanoString: Research Funding; Janssen: Membership on an entity's Board of Directors or advisory committees. Nowakowski:Curis: Research Funding; Bayer: Consultancy, Research Funding; Celgene: Consultancy, Research Funding; F. Hoffmann-La Roche Ltd: Research Funding; Genentech, Inc.: Research Funding; MorphoSys: Consultancy, Research Funding; NanoString: Research Funding; Selvita: Membership on an entity's Board of Directors or advisory committees. Novak:Celgene Coorperation: Research Funding.

Abstract Background Next Generation sequencing (NGS) is a powerful tool to identify somatic mutations associated with tumor onset and drug response. While it is well suited for high quality fresh/frozen samples, NGS is not proven for FFPE tissue which is the most common type of clinical specimen. Since the nucleic acids can be readily extracted from FFPE samples for a variety of genomic analyses, a comparative mutational analysis of paired frozen and FFPE tissues is urgently needed. Our long term goal is to establish a lab protocol to detect mutations in FFPE tumors using a targeted capture and sequencing approach for genes of interest. This pilot study focuses on the comparison of FFPE and frozen samples to test the validity of using FFPE tissues in such application. Methods Gene Selection: 128 genes associated with known pathogenic mutations in lymphoma Sample Selection: 9 diffuse large B-cell lymphoma (DLBCL) cases with FFPE, frozen and germline samples, as well as 10 frozen normal lymphatic tissues as references for CNV detections Capture Probe Design: We targeted coding exons and UTR, as well as the evolutionarily conserved intronic regions. The capture probes were designed using the Agilent eArray tool. The titling density of the probes was set to 3 probes overlapping with every base in the target region to improve the capture efficiency in FFPE samples. The least stringent masking of the repeat regions was allowed to include regions with small repeats that are shorter than the length of the sequencing reads (100-bp). In addition, boosting parameters were picked to set various levels of probe replication in different regions in order to minimize the local coverage differences (e.g. between regions of different GC contents) Sequencing and Bioinformatics: The target capture and sequencing were performed by the Mayo Clinic Medical Genome Facility. The reads were mapped to Human Reference Genome Build 37 using Novalign, and SNVs were called using GATK. The CNVs were identified using an in-house developed algorithm, patternCNV. Results The designed probes covered 99.65937% of the target regions. We generated 2.2-6.7 Gbp of reads per sample, 57.4-71.5% of which were on target. This equalled an average coverage of 2100-6700 folds which is 10-30 times higher than the minimal coverage recommended by Agilent. Due to this high coverage, we observed duplicate reads that accounted for 7.7-73.5% of the total reads. When we analysed the data with and without the duplicated reads, the concordance of the called SNVs was between 84-93% out of 207-249 mutated positions per trio-sample. There were 7.8-8.9% and 1.1-2.2% unique SNVs per sample by excluding or including duplicate reads, respectively. The dis-concordances were mostly missed calls, where a SNV was observed in only 1 or 2 of the trio samples. The missed calls from frozen samples ranged from 0-10.4% compared to 1.4-10.4% from the FFPE tissues, with 0.88-2.4% more SNVs missed in FFPE. Further analyses showed that all of the missing calls came from the lack of or low coverage of the corresponding positions. There were also differences of the called SNVs between the trio samples. However, this was extremely rare. Only 2 out of the 9 trio samples at a total of 3 positions had disagreements in called SNVs between FFPE and frozen tissues, all due to the allelic imbalance where the percentage of reads supporting the alternative alleles were below 20%. Therefore, this dis-concordance can be removed by back-filling of the read-level information for each position. Unfortunately only 11.9-47.4% of the CNVs called in frozen tissues were identified in FFPE samples, due to the widely various coverage in FFPE samples. The consequent large noises of the log ratio values between the FFPEs and normal references significantly reduced the sensitivity for CNV calling. Conclusions This pilot study compared the performance of SNV and CNV detection in FFPE and paired frozen tissues using a target capture and sequencing approach. With a capture probe design strategized to benefit FFPE samples, we observed SNV detection rates in FFPE that were only slightly lower (0.88-2.4%) than those of frozen tissues due to poor coverage of some positions in FFPE samples. With a proper back-filling step, there was no dis-concordance of the called SNVs between FFPE and frozen samples. However, CNV detections in FFPE were more problematic due to the un-predictable regional coverage in FFPE samples. Disclosures: No relevant conflicts of interest to declare.

Abstract Tumors arise and evolve by iterative steps of mutation, subclonal selection, and clonal expansion due to growth advantage of the fittest subclones and external mutation induction and selection pressure from radiation and drug therapies. Various studies have shown that greater clonal complexities in primary tumors are correlated with poor clinical outcome of tumor progression and/or drug resistance. Recently, we have reported that very few of the driver mutations, all of which were clonal, detected in primary diffuse large B-cell lymphoma (DLBCL) were associated with DLBCL outcome as measured by 24-month event-free survival (EFS24). In the current project, we identified and studied subclonal mutations in primary DLBCL tumors and assessed their associations with EFS24, as well as the role of activation-induced cytidine deaminase (AID) in genomic mutations occurred in DLBCL tumors. The detection of subclonal mutations is still a significant bioinformatics challenge. In addition, identification of clonal mutations in samples with lower tumor purities faces similar challenges due to the low concentration of reads supporting the mutant alleles. The current methods of somatic mutation calling are not sufficiently sensitive to identify the low concentration mutations in tumor DNA sequencing data. We implemented a bioinformatics workflow for low-concentration and subclonal mutation detection which is based on the positional read pile-up data and a back-fill approach (PUB). The pile-up files were first generated using the re-aligned and re-calibrated BAM files after read alignment using Burrows-Wheeler Aligner (BWA). The variant positions with alternative bases were annotated by attributes defined in the Variant Quality Score Recalibration (VQSR). PUB then used a boosting method and a generalized linear model (GLM) to train a model of 'good quality' variants using common variants from HapMap, and prioritized and called clonal and subclonal variants based on the trained model. The VQSR attributes related to alternative allele depth were less-weighed in order to call subclonal mutations. The somatic mutations were then identified by Fisher's Exact Test using sequencing depths of the reference and alternative alleles from paired tumor and germline sequencing data. The exome sequencing data of paired tumor and peripheral blood from 48 newly diagnosed DLBCL patients were analyzed using PUB. Thirty-six of the 48 patients achieved EFS24 with the other 12 patients experienced primary treatment failure. Most of the tumors studied had tumor purities between 40-70%. PUB identified substantially higher number of somatic mutations, both clonal and subclonal, compared to those detected using existing somatic callers. We observed that the prevalence of mutations in previously reported driver genes were higher using thresholds of mutation concentration ≥ 5% and Fisher's Exact Test p ≤ 0.05, including EZH2 (mutated in 26% of DLBCL tumors analyzed by PUB, compared to 12.7% as previously reported), MLL2 (72% vs. 31%), CD79B (22% vs. 14%), TNFRSF14 (36% vs. 20%), MEF2B (22% vs. 16.4%), CARD11 (46% vs. 21.8%), and MYD88 (18% vs. 11%). In addition, other genes involved tumorigenesis that have not been previously linked to DLBCL also harbored both clonal and subclonal mutations with substantial prevalence, including FGFR3 (12%), KIT (24%), and ATM (28%). Furthermore, association of genes displaying clonal and subclonal mutations with EFS24 identified potential biomarkers for DLBCL outcome. Among these, two genes EPGN and ASTE are involved in epithelial growth factor receptor signaling and had association p values of 0.0001 and 0.0016, respectively. The oncogene MAFB was also associated with EFS24 (p= 0.0016). We searched for AID induced mutations among all identified variant positions and concluded that there was no evidence of AID site enrichments compared to a simulated data set. In summary, we developed and applied a sensitive bioinformatics pipeline for the identification of both clonal and low concentration somatic mutations in primary DLBCL exome sequencing data which further revealed the clonal complexity of the primary DLBCL tumors. Several of the genes were identified as potential biomarkers for DLBCL outcome. Disclosures Maurer: Kite Pharma: Research Funding. Ansell:Bristol-Myers Squibb: Research Funding; Celldex: Research Funding. Link:Genentech: Consultancy, Research Funding; Kite Pharma: Research Funding. Cerhan:Kite Pharma: Research Funding.

Abstract To date, whole genome and exome sequencing studies of follicular lymphoma (FL) have primarily focused on identification of small site mutations that are recurrent in FL tumorigenesis or involved in tumor clonal evolution. A comprehensive genomic and transcriptomic survey of various mutation types including large structural variants (SVs) in FL cases with detailed clinical annotations and long-term follow-up has not been accomplished. To gain insight into genetic biomarkers that may predict clinical features, we performed exome and whole genome mate-pair sequencing of fresh frozen tumor and paired peripheral blood DNAs, and transcriptome sequencing of tumor RNAs, from 8 FL patients. The patients we selected were either below the median age of FL onset (n=7, median 54.5 yrs) or had a family history of lymphoma (n=1). These patients were clinically diverse, and included patients who had Grade 1 or 2 disease (n=4), classified as “indolent”; and patients with Grade 3a disease (n=2) or who subsequently had pathologic transformation (n=2), classified as “aggressive”. The coding regions of the genome (exome) were captured using Agilent SureSelect Target Capture Kit V2.0, and sequenced at 100-bp paired-end. The single nucleotide variants (SNVs) and small insertions and deletions (INDELs) were called using The Genome Analysis Toolkit (GATK), and the exon level copy number variants (CNVs) we identified using patternCNV. The whole-genome mate-pair libraries of both normal and tumor DNAs with 3kb insert sizes were sequenced paired end at 50-bp. The large SVs of CNVs, large INDELs, translocations, and inversions were identified using the SnowShoes-SV algorithm. The RNA sequencing libraries of 8 FL tumors were constructed using Illumina TruSeq protocol, and sequenced at 50-bp paired end. The RNA-Seq data were analyzed using TopHat and the fusion transcripts were identified using SnowShoes-FTD. Our analysis of SNVs and INDELs revealed mutations in previously reported genes including MLL2, CREBBP, TNFRSF14, and histone cluster genes (HIST1H2AM, HIST1H2BD). In addition, we identified novel recurrent mutations in cysteine-rich PAK1 inhibitor (CRIPAK) in 25% of the tumors. In a secondary analysis performed by Sanger sequencing or re-analysis of publically available RNA sequencing data, we identified CRIPAK mutations in 44% of FL (n=32) and 28% of DLBCL tumors (n=102). Bioinformatics analysis shows that the coding region of CRIPAK is highly enriched with the protein functional domain, post-SET, which is usually found in histone lysine methyltransferases (HMTase) genes including MLL2 and EZH2 that are known to be important in lymphomagenesis. Interestingly, CRIPAK is part of the same regulatory network consisted of previously identified lymphoma genes including MLL2, EZH2, CREBBP, and EP300, according to the shortest path algorithm by MetaCore (Philadelphia, Pennsylvania). Recurrent SVs identified in the FL tumors included the well-known IGH-BCL2 translocation, or t(14;18), in 6 out of 8 cases (75%) and the chr1q amplifications in 4 out of 8 (50%) tumors. Other non-recurrent large CNVs involving entire chromosomes or chromosome arms, as well as other inter- and intra-chromosomal structural variants were detected in individual tumors. In addition, we identified and validated 6 fusion transcripts from the transcriptome sequencing data in 3 out of 8 cases (38%). While our sample size is small, we found that SNVs and INDELs in MLL2, CREBBP, TNFRSF14, CRIPAK, and histone cluster genes, as well as t(14;18), did not distinguish indolent or aggressive tumors. However, all aggressive (4/4, 100%) and none of the indolent tumors had gains in chr1q; and the presence of RNA fusion transcripts were observed in aggressive tumors only (3/4, 75%). Additionally, we found that aggressive tumors had higher numbers of genes with point mutations (SNVs and short INDELs) (40 ± 7.6 vs. 26 ± 2.6; aggressive vs. indolent), higher numbers of genes impacted by copy number aberrations (1060 ± 263.7 vs. 233 ± 133.9), and higher numbers of large SVs (24 ± 10 vs. 6 ± 1.6). Taken together, our comprehensive analysis of 8 FL tumors reveals genetic diversity among newly diagnosed FL patients, identifies novel and recurrent mutations in CRIPAK, and finds that high tumor complexity and DNA instability may be indicators of aggressive disease. Disclosures No relevant conflicts of interest to declare.