Post ASH speaker abstracts 2018


Abstracts to support presentations from Anna Schuh and Cathy Burton

MYC Translocations Identified By Sequencing Panel in Smoldering Multiple Myeloma Strongly Predict for Rapid Progression to Multiple Myeloma

Result Type: Paper
Number: 393
Presenter: Niamh Keane
Program: Oral and Poster Abstracts
Session: 651. Myeloma: Biology and Pathophysiology, excluding Therapy: Genomics of The Pathogenesis and Progression of Multiple Myeloma

Niamh Keane, MB, MRCP1,2*, Caleb K Stein, MS3*, Daniel Angelov, MSc, MB3*, Shulan Tian4*, David Viswanatha, MD5, Shaji K. Kumar, MD5, Angela Dispenzieri, MD5, Veronica Gonzalez De La Calle, MD3*, Kristine Misund, PhD3,6*, Robert A Kyle, M.D5, Michael E O’Dwyer, MD2, Rafael Fonseca, MD3, A. Keith Stewart, MBChB, MBA7, Esteban Braggio, PhD8, Yan Asmann, PhD4, S. Vincent Rajkumar, MD5 and P. Leif Bergsagel, MD8

1Mayo Clinic, Scottsdale, AZ
2National University of Ireland Galway, Galway City, Ireland
3Division of Hematology and Oncology, Mayo Clinic, Scottsdale, AZ
4Mayo Clinic, Jacksonville, FL
5Division of Hematology, Mayo Clinic, Rochester, MN
6KG Jebsen Center for Myeloma Research, Trondheim, AZ, Norway
7Center for Individualized Medicine, Mayo Clinic, Scottsdale, AZ
8Mayo Clinic Arizona, Scottsdale, AZ


Introduction: Smoldering Myeloma (SMM) is a heterogeneous asymptomatic stage between Monoclonal Gammopathy of Unknown Significance (MGUS) and Multiple Myeloma (MM). Risk of progression to MM is 10% in the first 5 years following diagnosis and greatly diminishes thereafter (Kyle et al NEJM 2007). Early intervention in SMM patients at high risk of progression extends survival (Mateos et al NEJM 2013). The IMWG 2014 Updated Criteria establish a subset of SMM patients who are at ultra-high risk for progression to MM within 2 years, and therefore merit treatment (Rajkumar et al Lancet Oncol 2014). Analyses of the genetic and molecular landscape of SMM to date report a near identical picture to MM, however most patients in these studies progress rapidly and are therefore not representative of the entire group. We performed a comprehensive analysis of SMM and MGUS patients to determine markers of high risk of progression that could identify patients to benefit from early treatment

Methods: MGUS not progressing after at least 10 years of follow up and SMM in which follow up data were available were extracted from the Plasma Cell Dyscrasia Biobank. A clinically applicable Custom Capture MM-specific sequencing platform was developed for detection of the most frequently mutated pathways in MM based on analysis of CoMMpass dataset. Coding exons of actionable genes, clinically relevant copy number abnormalities, and regions surrounding IgH (0.5Mb), IgK (0.1Mb), IgL (0.1Mb) and MYC loci (1.6Mb) to identify relevant structural variants (SVs) were included, with combined design 2.2Mb. 12 samples were pre-pooled before capture, and 2 captures were sequenced per lane of Illumina HiSeq4000. Paired-end 150bp reads were mapped to hg19 using BWA-MEM. Single nucleotide variants (SNVs) and small INDELs in capture regions were identified using the GenomeGPS analytic pipeline following Broad GATK variant discovery practices. Copy number variants (CNVs) were identified by patternCNV. SVs of translocations, inversions, large INDELs, and segmental duplications were called by the SnowShoes-SV algorithm developed in-house. False positive SVs, polymorphic SVs, and other artefacts were filtered out using in-house normal SV database.

Results: We identified and sequenced 128 patients including 32 MGUS patients not progressing after 10 years. Of 96 SMM patients included 36 had not progressed to SMM after minimum follow up of 5 years, while 37 and 23 progressed to MM in less than 2 years and between 2-5 years, respectively. The genetic subtype of each patient was determined and verified by clinical FISH. Proportions in each genetic subgroup in MM and SMM/MGUS were similar, indicating that these are primary genetic lesions occurring early in MM pathogenesis. Median SMM time to progression (TTP) was 46 months. As in other series, HRD with IGH translocation, and t(4;14) predicted shorter TTP.

Analysis of CoMMpass dataset found frequent MYC SV (38%) in untreated MM with higher frequency in HRD versus NHRD MM: 53% versus 28% (Misund, ASH 2016). No MYC SV were detected in MGUS cohort, SMM non-progressors at >5 years or SMM progressing between 2-5 years. By contrast, MYC SV were detected in 49% SMM that progressed within 2 years, 55% in HRD and 41% NHRD. SMM with MYC SV had a significantly shorter median TTP compared to patients without MYC SV (11.5 months vs 61 month; p<0.0001). Multivariate analysis with high risk genetic groups and biomarkers for progression (BMPCs >/=60% and FLC ratio 100) confirm MYC SV as an independent variable for progression to MM (hazard ratio=7, 95% confidence interval 3.6-13.7, p=0.00001).

RAS and NFKB pathway mutations were observed with similar frequencies in MM and SMM progressing within 5 years of diagnosis, but with lower frequency in those not progressing by 5 years follow up, and were not observed in the MGUS cohort. A trend toward shorter TTP was observed in patients with RAS pathway mutations but did not reach statistical significance.

Conclusion: In conclusion, we describe MYC translocations as a genetic marker of and likely cause of progression to MM that are absent in MGUS and SMM with TTP >2 years. In contrast MM and SMM early progressors (TTP <2 years) share a similar genetic landscape. Identification of MYC translocations at diagnosis of SMM predicts short TTP to MM, defining a novel ultra-high risk category that merits validation in prospective clinical trials.[/et_pb_text][et_pb_image src="" _builder_version="3.0.92"][/et_pb_image][et_pb_text _builder_version="3.0.92"]

Utility of Clinical-Grade Sequencing of Relapsed Multiple Myeloma Patients; Interim Analysis of the Multiple Myeloma Research Foundation (MMRF) Molecular Profiling Protocol

Result Type: Paper
Number: 395
Presenter: Daniel Auclair
Program: Oral and Poster Abstracts
Session: 651. Myeloma: Biology and Pathophysiology, excluding Therapy: Genomics of The Pathogenesis and Progression of Multiple Myeloma

Daniel Auclair, PhD1*, Sikander Ailawadhi2, Jesus G. Berdeja, MD3, Xuhong Cao4*, Craig E. Cole, MD5, Craig C Hofmeister, MD6, Sundar Jagannath, MD7, Andrzej J. Jakubowiak, MD8, Amrita Krishnan, MD9, Shaji K. Kumar, MD10, Moshe Levy, MD11, Sagar Lonial, MD12, Gregory Orloff, MD13*, Dan Robinson, PhD4*, David Siegel, MD, PhD14, Suzanne Trudel, MD, FRCP(C)15, Saad Z Usmani, MD16, Ravi Vij, MD MBA17, Jeffrey Lee Wolf, MD18, Jennifer Yesil, MS19*, Jeffrey A Zonder, MD 20, Arul Chinnaiyan, MD21* and P. Leif Bergsagel, MD22

1Multiple Myeloma Research Foundation, Norwalk, CT
2Division of Hematology-Oncology, Mayo Clinic, Jacksonville, FL
3Hematology, Sarah Cannon Research Institute, Nashville, TN
4U. of Michigan, Michigan Center for Translational Pathology, Ann Arbor, MI
5Division of Hematology/Oncology, University of Michigan School of Medicine, Ann Arbor, MI
6Comprehensive Cancer Center, The Ohio State University, Columbus, OH
7Department of Hematology and Medical Oncology, Tisch Cancer Institute, Icahn School of Medicine at Mount Sinai, New York, NY
8Department of Medicine, Section of Hematology/Oncology, University of Chicago Medical Center, Chicago, IL
9City of Hope Medical Center, Duarte, CA
10Division of Hematology, Mayo Clinic, Rochester, MN
11Baylor University Medical Center at Dallas, Dallas, TX
12Department of Hematology and Medical Oncology, Emory University – Winship Cancer Institute, Atlanta, GA
13Virginia Cancer Specialists, Fairfax, VA
14John Theurer Cancer Center at Hackensack University Medical Center, Hackensack, NJ
15University of Toronto, Princess Margaret Cancer Centre, Toronto, ON, Canada
16Levine Cancer Institute/Carolinas HealthCare System, Charlotte, NC
17Washington University School of Medicine, St Louis
18University of Cal. SF, San Francisco, CA
19Multiple Mu, Norwalk, CT
20Karmanos Cancer Institute, Detroit, MI
21University of Michigan, Ann Arbor, MI
22Division of Hematology/Oncology, Mayo Clinic, Scottsdale, AZ


Introduction: Multiple myeloma (MM) is the second most prevalent blood cancer, representing approximately 1% of all cancers. Although overall survival has improved in recent years due to new approved agents, the vast majority of MM patients ultimately stop responding to treatment. Whereas current therapeutic approaches have focused mostly on the plasma cell biology of the disease, seminal genomic sequencing research efforts, such as the MMRF CoMMpass study, have highlighted that a large number of MM cases harbor potentially actionable oncogenic molecular alterations. Published reports on small numbers of cases suggest that Precision Medicine (PM) interventions clinically targeting such actionable alterations might be of benefit to MM patients who are running out of options. In order to study on a larger scale the potential and clinical utility of PM approaches in myeloma, the MMRF Molecular Profiling Protocol (NCT02884102) was opened in 2016 across the entire Multiple Myeloma Research Consortium (MMRC) with the goal of enrolling and following 500 relapsed patients that would be molecularly profiled using clinical-grade sequencing performed on the Michigan Oncology Sequencing (MI-ONCOSEQ) platform.

Methods: Bone marrow aspirates (BMAs) and matched normal peripheral blood (PB) are shipped overnight to the Michigan Center for Translational Pathology (MCTP) Clinical Sequencing Lab where CD138 enrichment is carried out. The MCTP sequencing lab is CLIA-CAP certified. DNA and RNA are isolated from MM cells and matched normal. Libraries are generated and subjected to the Oncoseq1500 gene exome capture. Deep targeted re-sequencing (>600x) is carried out on HiSeq2500 run in rapid mode. Data is computationally analyzed for mutation status. A molecular report highlighting actionable findings is produced, reviewed internally by a genomic Tumor Board and returned within 10 days.

Results: We are reporting on 228 consecutive cases analyzed with 84% of the sequenced samples (192) showing a very good tumor content. Importantly, 76% of cases were found to harbor at least one potentially actionable alteration. Of those cases, 53% had alterations in the MAPK pathway, 14% in the CCND1 and cyclin-dependent kinase (CDK) pathways, 6% had activating FGFR3 mutations followed by a group of events at 3% or less. In this cohort (n>2 priors on average), 16% of cases presented with TP53 mutations of which 1/4 could also be detected in blood. A search for other genes where a significant percentage of mutations were also detected in PB identified a small number of those including, among others, SF3B1, TET2, ASLX1, ASLX2 and DNMT3A with such mutations (typically subclonal) often co-occurring in the same specimen. In all, 10% of all cases presented with this mutational signature in both BMAs and PB of genes generally associated with MDS, AML and other myeloid disorders. With regards to actionability, in 10% of cases the treating clinician acted upon the information with the indicated targeted agent. Examples of the responses obtained will be presented. Analysis of progression-free survival and overall benefit for this cohort is ongoing.

Conclusion: Actionable alterations were identified in over three quarter of cases analyzed. Deep sequencing of both BMAs and normal blood could also identify events that would have been missed had sequencing been only performed on the marrow. Although clinical applicability has been limited so far by the lack of availability of targeted agents for myeloma patients, the results suggest that Precision Medicine approaches in MM are possible and should be further studied clinically. To that end, we are launching MyDRUG, a master protocol aimed at developing new myeloma regimens based on individual patient’s genomics.

Plasma Cell Disease Genotyping with the Liquid Biopsy

Result Type: Paper
Number: 327
Presenter: Bernhard Gerber
Program: Oral and Poster Abstracts
Session: 651. Myeloma: Biology and Pathophysiology, excluding Therapy: Genomics

Bernhard Gerber1*, Martina Manzoni2*, Valeria Spina, PhD3*, Alessio Bruscaggin, PhD3*, Sonia Fabris, BSc4*, Marzia Barbieri4*, Gabriella Ciceri5*, Marta Lionetti2*, Alessandra Pompa4*, Gabriela Forestieri3*, Erika Lerch, MD6*, Paolo Servida6*, Francesco Bertoni, MD7, Emanuele Zucca, MD8, Michele Ghielmini, MD6, Agostino Cortelezzi, MD4,9*, Franco Cavalli, MD7,8, Georg Stussi, MD1, Luca Baldini4,5*, Davide Rossi, MD, PhD1,3 and Antonino Neri4,5*

1Division of Hematology, Oncology Institute of Southern Switzerland, Bellinzona, Switzerland
2Department of Oncology and Hemato-Oncology, University of Milan, Milan, Italy
3Institute of Oncology Research, Bellinzona, Switzerland
4Hematology Unit, Foundation IRCCS Ca’ Granda Ospedale Maggiore Policlinico, Milan, Italy
5Department of Oncology and Hemato-oncology, University of Milan, Milan, Italy
6Division of Oncology, Oncology Institute of Southern Switzerland, Bellinzona, Switzerland
7Institute of Oncology Research, Oncology Institute of Southern Switzerland, Bellinzona, Switzerland
8Oncology Institute of Southern Switzerland, Bellinzona, Switzerland
9Foundation IRCCS Ca’ Granda Ospedale Maggiore Policlinico di Milano, Milano, Italy

Introduction: Easily accessible, real-time genotyping is desirable for patients suffering from plasma cell (PC) disorders for diagnostic, prognostic and therapeutic purpose. However, PC disorders usually lack a leukemic phase, and up to now genotyping of tumor cells required purified material from bone marrow (BM) or tissue biopsies. Circulating cell-free DNA (cfDNA) might be an accessible source of tumor material in patients with PC diseases to identify cancer-gene somatic mutations. Accessing the peripheral blood (PB) has clear advantages with respect to the sampling procedure itself, and has the potential to better reflect tumor heterogeneity. Here we aimed at tracking the genetic profile of different PC dyscrasias using plasma cfDNA.

Methods: Twenty-eight consecutive patients (median age= 64; male:female ratio=15:13) with PC disorders [two with monoclonal gammopathy of undetermined significance (MGUS), five with smoldering MM (SMM), and 21 with symptomatic MM] were included between September 2016 and May 2017. The following material was collected: (1) cfDNA isolated from plasma, (2) tumor genomic DNA (gDNA) from CD138+ purified BM PCs for comparative purposes, and (3) normal germ line gDNA extracted from PB granulocytes after Ficoll separation, to filter out polymorphisms. The sampling was done at diagnosis in 25 patients, and during the course of the disease in three MM cases. Median BM PC infiltration was 40% (range: 7-90%). A targeted resequencing gene panel, including coding exons and splice sites of 14 genes (target region: 31 kb) was specifically designed and optimized to allow a priorithe recovery of at least one clonal mutation in 68% (95% confidence interval: 58-76) of MM patients. Ultra-deep next-generation sequencing (NGS) of the gene panel was performed on MiSeq (Illumina) using the CAPP-seq library preparation strategy (NimbleGen). The somatic function of VarScan2 was used to call non-synonymous somatic mutations, and a stringent bioinformatic pipeline was developed and applied to filter out sequencing errors (detection limit 3x10-3). The sensitivity and specificity of plasma cfDNA genotyping were calculated in comparison with tumor gDNA genotyping as the gold standard.

Results: In this prospectively collected cohort of 28 unselected PC dyscrasia patients, circulating tumor cfDNA was detectable in plasma samples with a median of 12011.60 haploid genome-equivalents per mL. The application of our targeted ultra-deep NGS approach for plasma cfDNA genotyping resulted in ≥90% of the target region covered ˃1000X in all plasma samples, and ≥90% of the target region covered ˃2000X in 23/28. Overall, within the interrogated genes, 18/28 (64%) patients harbored somatic mutations (total number: 28; range: 1-4 mutations per patient) that were detectable in plasma cfDNA. Quite consistent with the typical spectrum of mutated genes in MM, plasma cfDNA genotyping revealed somatic variants of NRAS in 25%; KRAS in 14%; TP53TRAF3 and FAM46C in 11%, respectively; CYLD and DIS3 in 7%, respectively; and BRAF and IRF4 in 4% of cases, respectively (Figure 1A). By comparing the genotype of circulating tumor cfDNA with that of gDNA from purified malignant BM PCs of the diagnostic biopsy (gold standard) (Figure 1B), we found that cfDNA genotyping correctly identified 72% of the mutations confirmed in the tumor biopsy (Figure 1C). Notably, the remaining mutations not discovered in cfDNA had a low representation in the purified BM PCs (median allelic abundance=2.5%; range 1.1-4.96%). ROC analysis showed that circulating tumor cfDNA genotyping had the highest sensitivity (92.9%) if mutations were represented in >5% of the alleles of the purified BM PCs (Figures 1D-E). In none of the cases, cfDNA genotyping identified additional somatic mutations not detected in the purified BM PCs.

Conclusions: Overall, our results provide the proof of principle that circulating tumor cfDNA genotyping is a feasible, non-invasive real-time approach that reliably detects clonal and subclonal somatic mutations represented in at least 5% of alleles in tumor PCs.

Progression to AML Is Predictable and Distinct from Age Related Clonal Hematopoiesis

Result Type: Paper
Number: 471
Presenter: Sagi Abelson
Program: Oral and Poster Abstracts
Session: 617. Acute Myeloid Leukemia: Biology, Cytogenetics, and Molecular Markers in Diagnosis and Prognosis I

Sagi Abelson, PhD1*, Stanley W.K. Ng2*, Omer Wiessbrod3*, Philip Zuzarte4*, Lawrence Heisler4*, Yogi Sundaravadanam4*, Ting Ting Wang5*, Trevor J. Pugh, PhD6,7*, David Soave4*, Paul Brennan8*, Mattias Johansson8*, Phillip Awadalla4*, Scott V. Bratman5,7,9*, Jean Wang, MD1,10,11, Mark D. Minden, MD, PhD12,13,14, John E. Dick, PhD1,15 and Liran I. Shlush, MD, PhD16,17*

1Princess Margaret Cancer Centre, University Health Network, Toronto, ON, Canada
2Institute of Biomaterials and Biomedical Engineering, University of Toronto, Toronto, ON, Canada
3Department of Computer Science and Applied Mathematics, Weizmann Institute of Science, Israel, Rehovot, Israel
4Ontario Institute for Cancer Research, Toronto, Canada
5Princess Margaret Cancer Centre, University Health Network, Toronto, Canada
6University of Toronto, Princess Margaret Cancer Centre, Toronto, ON, Canada
7Department of Medical Biophysics, University of Toronto, Toronto, Canada
8International Agency for Research on Cancer (WHO-IARC), Lyon, France
9Department of Radiation Oncology, University of Toronto, Toronto, Canada
10Department of Medicine, University of Toronto, Toronto, Canada
11Division of Medical Oncology and Hematology, UHN, Toronto, Canada
12Department of Medical Biophysics, University of Toronto, Toronto, ON, Canada
13Princess Margaret Cancer Centre, Division of Medical Oncology and Hematology, University Health Network, University of Toronto, Toronto, ON, Canada
14Department of Medicine, University of Toronto, Toronto, ON, Canada
15Department of Molecular Genetics, University of Toronto, Toronto, ON, Canada
16Department of Immunology, Weizmann Institute of Science, Rehovot, Israel
17Division of Hematology Rambam Healthcare Campus, Haifa, Israel

Introduction: Despite marked improvements in our understanding of the biology and genetic basis of acute myeloid leukemia (AML) gained in the past 4 decades, the overall survival of AML has only improved slightly. The strategy for targeting AML once the disease is diagnosed remains grim due to subclonal genetic diversity and the stemness properties of relapse originating cell types. As studies in solid tumors have shown, strategies focused on targeting the cancer at earlier stages hold considerable promise, however no predictive tests are available for AML. The existence of clonally expanded pre-leukemic hematopoietic stem and progenitor cells (preL-HSPC) within the diagnostic blood sample of AML patients provides strong evidence that the disease must evolve over long time periods prior to diagnosis. However, with the discovery that age related clonal hematopoiesis (ARCH) is exceedingly common during normal aging; a refined search for the factors that distinguish ARCH from those rare individuals who will actually develop AML had to be taken.

Methods: We hypothesized that the basis for clonal expansion in ARCH was different than in the clonal expansion leading to AML. To investigate genetic differences, we developed sensitive and accurate error correction sequencing method to study blood samples of healthy individuals from the EPIC cohort before AML diagnosis. These 96 pre-AML cases donated their blood on average 6 years before eventually being diagnosed with AML; a set of 420 matched controls were sequenced in the same way. Furthermore, 3.2 million electronic health records (EHR) from the Clalit database were inspected to identify non-genetic AML-risk prediction parameters.

Results: The overall variant allele frequency (VAF) distribution of the identified mutations was significantly different between cases and controls. Specifically, a greater proportion of cases carried somatic mutations with intermediate-high VAF (70.8% versus 36.6% in the controls). The median number of all somatic mutations was significantly higher in the cases as compared with the controls (7 and 5 respectively) and was correlated with aging in both groups. Interestingly, the pre-AML group had a higher rate of mutation accumulation compared to controls, resulting in significantly greater numbers of mutations occurring after the age of 55. Recurrent AML mutations were more prevalent in cases compared to controls (Age>=55: OR, 7.571; 95% CI, 4.055-14.387). Specifically, highly recurrent splicing factors mutations (e.g. SRSF2 P95H, U2AF1 Q84P, and SF3B1 K700E) were found solely in the pre-AML group suggesting that specific mutations that rewire the splicing machinery inevitably leads to AML when acquired in HSPC. The VAF of recurrent AML mutations was significantly higher in the cases compared to controls. Furthermore we observed a higher proportion of individuals with more than one recurrent AML mutation in the pre-AML group compared to controls (OR, 14.316; 95% CI, 6.318-34.933). Incorporating these factors, we construct a computational model for predicting progression toward AML. The AML risk prediction model predicted progression on average 7 years before actual diagnosis (HR, 20.1, 95% CI 8.59-46.8). EHR data was found to be highly valuable for the identification of high-risk population. Compared to controls, AML patients (N=982) showed aberrant blood counts up to 6 months prior diagnosis (relative risk, 22.2).

Conclusions: These results reveal that the clonal expansion occurring during normal aging is highly distinct from pre-AML clonal expansion, which has a highly predictable evolutionary trajectory. Our study provides clear proof of concept for the feasibility of early AML prediction based on somatic mutation patterns, yet, since ARCH is common, the real-world feasibility of applying our AML risk prediction model will require identifying higher-risk individuals through routine clinical testing. We have shown that information obtained from EHR is useful for that purpose. Future studies of independent population cohorts with access to serial viable blood and leukemia samples will allow for incorporation of information such as the specific identity of the mutated cells and clone expansion kinetics. Such studies will aid in developing better understanding of which individuals should be selected for future clinical trials to study the potential benefits of early interventions in this deadly disease.

Prognostic Impact of Clonal Hematopoiesis in Acute Myeloid Leukemia Patients Receiving Allogeneic Hematopoietic Stem Cell Transplantation in Complete Remission

Result Type: Paper
Number: 406
Presenter: Juliane Grimm
Program: Oral and Poster Abstracts
Session: 617. Acute Myeloid Leukemia: Biology, Cytogenetics, and Molecular Markers in Diagnosis and Prognosis II

Juliane Grimm1*, Marius Bill, PhD1*, Madlen Jentzsch, MD1*, Stefanie Beinicke, MSc1*, Janine Häntschel, MSc1*, Karoline Goldmann1*, Julia Schulz1*, Michael Cross, PhD1*, Wolfram Pönisch, MD1*, Vladan Vucinic1*, Gerhard Behre, MD1*, Thoralf Lange, MD1*, Georg-Nikolaus Franke, MD2*, Dietger Niederwieser, MD1 and Sebastian Schwind, MD1*

1Department of Hematology and Oncology, University Leipzig, Leipzig, Germany
2Department of Hematology and Oncology, University of Leipzig, Leipzig, Germany


Introduction: Clonal hematopoiesis of indeterminate potential (CHIP) is defined as presence of hematologic malignancy-associated mutations, e.g. in the genes DNMT3ATET2ASXL1, but absence of hematologic neoplasms & is a frequent phenomenon in healthy individuals with increasing age. Some individuals with CHIP eventually develop acute myeloid leukemia (AML) & persistent clonal hematopoiesis-associated mutations (CH‑mutations) in complete remission (CR) may give rise to relapse. To date in AML little is known about the biological & clinical implications of the presence of CH-mutations in CR, especially in the context of allogeneic hematopoietic stem cell transplantation (HSCT).

Patients & Methods: We analyzed peripheral blood samples of 113 AML patients (median age 63.6 years, range 31.9-75.8 years) in CR (CR1 61.9%, CR2 14.2%) or CR with incomplete peripheral recovery (CRi; 23.9%) for the presence of CH‑mutations by targeted amplicon sequencing (mean amplicon coverage per sample 7205x) prior to HSCT. Genes evaluated for somatic CH‑mutations based on database entries (dbSNP & COSMIC) & with a variant allele frequency (VAF) of ≥3% were DNMT3ATET2ASXL1SRSF2SF3B1IDH1IDH2JAK2PPM1D, & IKZF1. Codon 646 ASXL1 mutations were validated using proof‑reading polymerase Sanger sequencing. For 76 patients (67.3%) diagnostic bone marrow samples were available for applying a targeted amplicon panel comprising 54 recurrently mutated genes in myeloid malignancies (mean amplicon coverage per sample 8624x). At diagnosis, mutations in the genes NPM1CEBPA, presence of FLT3-ITD, & cytogenetics were determined using standard techniques. All patients received HSCT after non-myeloablative conditioning at our institution between 2001 & 2015. Samples were collected within 30 days before HSCT. Median follow‑up for patients alive was 4.4 years after HSCT.

Results: We identified 70 CH-mutations present in CR/CRi in 48 AML patients (42.5%) with a mean VAF of 19.1% (58.6% of mutations with VAF>10%). The genes DNMT3A (31.4%), TET2 (28.6%) & ASXL1 (14.3%) were found mutated most frequently. We observed no associations between present CH-mutations in CR/CRi & clinical characteristics, except that patients with ≥1 CH-mutation were more often transplanted in CR2 compared to patients without CH-mutation by trend (20.8% vs 9.2%, P=0.10). The presence of ≥1 CH-mutation in CR/CRi did not impact leukemia-free survival (LFS; P=0.95, Figure 1A) or OS (P=0.37, Figure 1B), whereas patients with ≥2 CH‑mutations had longer LFS (P=0.02, Figure 1C) & OS (P=0.007, Figure 1D) compared to patients with no or 1 CH‑mutation. In CR/CRi DNMT3A mutations did not impact LFS (P=0.73) or OS (P=0.71), the presence of TET2mutations did not influence LFS (P=0.25) but associated with longer OS (P=0.06) & ASXL1 mutations associated with longer LFS (P=0.11) & OS (P=0.13) by trend. At diagnosis, we identified 83 CH-mutations in 56/76 AML patients (73.7%) of which 29 (34.9%) persisted in 19/76 patients (25.0%) in CR/CRi. Mutations in the genes JAK2 (80.0%), SRSF2 (57.1%) & DNMT3A (50.0%) were most likely to persist. We identified 9 CH‑mutations (3 DNMT3A, 3 TET2, 2 ASXL1, 1 SF3B1 mutation) in CR/CRi in 8 patients (10.5%), which were not detectable at diagnosis, possibly following the expansion of subclones under chemotherapy selection pressure. Whether patients had persistent, newly detected or lost CH-mutation in CR/CRi did not impact LFS (P=0.22) or OS (P=0.83).

Conclusion: In AML patients in CR/CRi CH-mutations are frequently present (42.5%) & show high VAFs (mean 19.1%) indicating presence of clonal hematopoiesis. Compared to diagnosis 34.9% of CH‑mutations persisted in CR/CRi. 10.5% of patients had CH‑mutations only detectable in CR/CRi prior to HSCT. While we did not observe a negative prognostic impact of presence of ≥1 CH-mutation in CR/CRi, the presence of distinct mutations (TET2ASXL1) may positively influence outcome. Patients with ≥2 CH‑mutations had longer LFS & OS. This finding may be explained by an increased immunogenic potential with higher number of CH‑mutations leading to potent graft vsleukemia effects in the HSCT context.

What’s Next in CML – a Prospective Study Evaluating Sanger Sequencing and Next Generation Sequencing (NGS) for BCR-ABL1 Kinase Domain (KD) Mutation Screening

Result Type: Paper
Number: 248
Presenter: Simona Soverini
Program: Oral and Poster Abstracts
Session: 632. Chronic Myeloid Leukemia: Therapy: NGS-based and Mechanistic Studies

Simona Soverini, PhD1, Luana Bavaro2*, Caterina De Benedittis, PhD1*, Margherita Martelli3*, Stefania Stella, PhD4*, Paolo Vigneri, MD, PhD5*, Alessandra Iurlo, MD, PhD6*, Nicola Orofino, MD6*, Simona Sica, MD7,8*, Federica Sorà, MD7*, Sara Galimberti, PhD, MD9*, Claudia Baratè, MD, PhD10*, Francesco Albano, MD11*, Antonella Russo Rossi, MD12*, Fabio Ciceri, MD13*, Francesca Lunghi, MD14*, Fausto Castagnetti, MD15, Gabriele Gugliotta, MD, PhD2, Gianantonio Rosti, MD16*, Elena Tenti, PhD15*, Fabio Stagno, MD, PhD17, Elisabetta Novella, PhD18*, Eros Di Bona, MD18*, Massimiliano Bonifacio, MD19*, Elisabetta Abruzzese, MD, PhD20*, Marzia Salvucci, MD21*, Mariella D’Adda, MD22*, Rosaria Sancetta, MD23*, Elisabetta Calistri, MD24*, Monica Bocchia, MD25*, Isabella Capodanno, MD26*, Giuseppina Spinosa, MD27*, Maria Antonella Laginestra, PhD28*, Santa Errichiello, PhD29*, Anna Serra, Dr30*, Francesca Carnuccio, Dr30*, Sabrina Coluzzi, PhD31*, Mario Annunziata, MD32*, Caterina Musolino, MD33*, Fabrizio Pane, MD34, Giuseppe Saglio, MD, PhD30, Michele Baccarani, MD35*, Michele Cavo, MD36* and Giovanni Martinelli, MD37

1Hematology/Oncology “L. e A. Seragnoli”, University of Bologna, Bologna, ITA
2Department of Experimental Diagnostic and Specialty Medicine – DIMES Institute of Hematology “L. e A. Seràgnoli”, University of Bologna, Bologna, Italy

3Department of Experimental, Diagnostic and Specialty Medicine – DIMES, Institute of Hematology “L. and A. Seràgnoli”, University of Bologna, Bologna, Italy
4Department of Clinical and Experimental Medicine, University of Catania, Catania, Italy
5Dep. Clinical and Experimental Medicine, University of Catania Medical School, Catania, Italy
6Hematology Division, IRCCS Cà Granda – Maggiore Policlinico Hospital Foundation, Milano, Italy
7Institute of Hematology, Catholic University of Rome, Rome, Italy
8Institute of Hematology, Università Cattolica del Sacro Cuore – Policlinico A. Gemelli, Roma, Italy

9Hematology Division, AOUP, Pisa, Italy
10Department of Clinical and Experimental Medicine, Section of hematology, University of Pisa, Pisa, Italy
11University of Bari, Department of Emergency and Organ Transplantation (D.E.T.O.), Hematology Section, Bari, Italy
12Hematology, University of Bari, Bari, ITA
13Hematology and BMT Unit IRCCS San Raffaele Scientific Institute, Milan, Italy
14U.O. Ematologia e Trapianto di Midollo Osseo, Ospedale San Raffaele IRCCS, Milan, Italy
15Hematology/Oncology “Seragnoli”, University of Bologna, Bologna, Italy
16Institute of Hematology ” L. & A. Seràgnoli”,, St. Orsola University Hospital, Bologna, Italy

17Chair and Hematology Section, Ferrarotto Hospital, Catania, Italy
18Ospedale San Bortolo, UO Hematology, Vicenza, Italy
19Department of Medicine, Section of Hematology, University of Verona, Verona, Italy
20Haematology Unit, S. Eugenio Hospital, Rome, Italy
21Hematology Unit, S. Maria delle Croci Hospital, Ravenna, Italy
22Department of Hematology, Spedali Civili, Brescia, Italy
23Hematology Unit, Ospedale dell’Angelo Mestre, Venezia, Italy
24UO Hematology, Ca Foncello Hospital, Treviso, Italy
25Azienda Ospedaliera Universitaria, University of Siena, Siena, Italy
26UO Hematology, Arcispedale Santa Maria Nuova IRCCS, Reggio Emilia, Italy
27UO Hematology, Foggia, Italy
28Department of Experimental, Diagnostic, and Specialty Medicine, Hematopathology Section, Bologna, Italy
29Department of Clinical Medicine and Surgery- University of Naples “Federico II” – CEINGE-Biotecnologie Avanzate Napoli-Italy, Naples, Italy

30Department of Clinical and Biological Sciences, University of Turin, Orbassano, Turin, Italy
31UO Hematology, San Carlo Hospital, Potenza, ITA
32Hematology, Ospedale Cardarelli, NAPOLI, Italy
33Section of Hematology, A.O.U. Policlinico, Messina, Italy
34Hematology – Departments of Clinical Medicine and Surgery, University of Naples Federico II, Naples, Italy
35Institute of Hematology ” L. & A. Seràgnoli”, St. Orsola University Hospital, Bologna, Italy

36Department of Physics and Astronomy, University of Bologna, Bologna, Italy
37Institute of Hematology ‘Seragnoli’, Department of Experimental, Diagnostic and Specialty Medicine, University of Bologna, Bologna,, Bologna, Italy

Small retrospective studies have suggested that NGS may have several advantages over Sanger sequencing (seq) for BCR-ABL1 KD mutation screening in Philadelphia chromosome-positive leukemia patients (pts). However, how frequently low burden mutations can be detected in pts with Failure or Warning response to tyrosine kinase inhibitors (TKI) and which role low burden mutations play in these pts remain to be addressed prospectively in large series of unselected pts. Moreover, the implementation of routine, NGS-based BCR-ABL1 mutation screening in a molecular diagnostic lab network has never been attempted. To assess the feasibility, performance, cost, turnaround times and clinical utility of NGS, we have conducted a multicenter prospective study: ‘Next in CML’.

The first phase of the study was aimed to 1) create a network of 4 reference labs sharing a standardized protocol, an optimized pipeline of data analysis and a joint database; 2) assess the accuracy and inter-lab reproducibility of results. NGS of amplicons generated by nested RT-PCR was done on Roche GS Junior instruments. A dedicated software (Amplicon Suite; SmartSeq srl) was implemented for read alignment and variant calling. Identical batches of 32 blinded cDNA samples (23 pt samples with known mutation status and 9 serial dilutions of BaF3 T315I-positive cells in BaF3 unmutated cells, simulating mutation loads of 20% to 1%) were sequenced in parallel to check individual lab performances. Of 64 expected mutations, 52 were detected (and accurately quantitated) by 4 labs, 7 by 3 labs, 4 by 2 labs and 1 by 1 lab only. Given that i) mutations that some labs failed to detect ranged between 1 and 3%, and ii) in mutation-negative samples, mutations likely to be PCR/sequencing errors were all <3%, 3% was set as lower detection limit.

The second phase of the study was aimed to apply NGS in parallel with Sanger seq in a consecutive series of prospectively collected samples from 211 chronic myeloid leukemia (CML) pts with Failure or Warning treated at 39 Hematology Centers. Pts positive for any point mutation (the 35bp insertion between exon 8 and 9 was excluded) were 42 (20%) by Sanger seq and 94 (45%) by NGS. Low burden mutations detectable by NGS were found in 52 pts negative for mutations by Sanger seq; in addition, 24 pts positive for mutations by Sanger seq were found to carry additional low burden mutations by NGS, for a total of 76 (36%) pts harboring low burden mutations. Median mutation burden was 11.1% (3%-18.8%). Eight pts had a low burden T315I and 43 pts had other IM/DAS/NIL/BOS resistant mutations: overall, 51 (24%) pts had relevant TKI-resistant low burden mutations missed by Sanger seq. The remaining 25 pts had only mutations with an unknown resistance profile. Pts positive by Sanger seq showing additional low burden mutations detectable by NGS (n=24) mainly had high/intermediate Sokal risk (n=19) and were receiving ≥2nd line therapy (n=17). Pts negative by Sanger seq but positive for low burden mutations by NGS (n=52) mainly had high/intermediate Sokal risk (n=29), Warning response (n=29) and were receiving a reduced TKI dose (n=37). Among pts negative for mutations both by Sanger seq and NGS (n=117), 39 (33%) had loss of MMR reported as a Failure after a single RQ-PCR test only and/or a BCR-ABL1 transcript increase <1 log not confirmed by subsequent follow-up evaluations, suggesting that in several cases failure to detect mutations results from improper triggers for mutation screening.

Longitudinal follow-up, available at the time of writing for 27 pts with low burden mutations, showed that TKI-resistant mutations remain consistently detectable and tend to increase in burden at subsequent timepoints unless treatment is changed. Low burden mutations of unknown resistance profile, in contrast, do not necessarily require treatment intervention to be eliminated.

Our study demonstrates that a robust and reproducible NGS-based BCR-ABL1 KD mutation screening can successfully be implemented in national diagnostic lab networks and is feasible with turnaround times and costs comparable to those of Sanger seq. In a large, prospective series of CML pts with Failure or Warning, known IM/DAS/NIL/BOS resistant mutations were missed by Sanger seq in 24% of the pts. Low burden (≥3%) TKI-resistant mutations were found to be sufficient to drive clonal expansion, whereas more data are needed to understand the clinical significance of those with an unknown resistance profile.

Prospective Molecular MRD Detection By NGS: A Powerful Independent Predictor for Relapse and Survival in Adults with Newly Diagnosed AML

Result Type: Paper
Number: LBA-5
Presenter: Tim Grob
Program: General Sessions
Session: Late-Breaking Abstracts Session

Mojca Jongen-Lavrencic, MD, PhD1Tim Grob, MD1*, Francois G. Kavelaars1*, Adil S.A. Al Hinai1*, Annelieke Zeilemaker1*, Claudia A.J. Erpelinck-Verschueren1*, Yvette Norden2*, Rosa Meijer, PhD2*, Bart J. Biemond, MD, PhD3*, Carlos Graux4*, Marinus van Marwijk Kooij, MD, PhD5*, Markus G. Manz, MD6, Thomas Pabst, MD, PhD7, Violaine Havelange, MD, PhD8*, Jakob R. Passweg, MD, PhD9*, Gert J. Ossenkoppele, MD, PhD10, Gerrit Jan Schuurhuis, PhD10*, Mathijs A. Sanders, PhD1*, Bob Löwenberg, MD, PhD1 and Peter J.M. Valk, PhD1

1Department of Hematology, Erasmus University Medical Center, Rotterdam, Netherlands
2HOVON Data Center, Erasmus University, Rotterdam, Netherlands
3Hematology, Academic Medical Center, Amsterdam, Netherlands
4Université catholique de Louvain, CHU UCL Namur (site Godinne), Godinne, Belgium
5Isala Hospital, Zwolle, Netherlands
6Department of Hematology and Oncology, University and University Hospital Zurich, Zurich, Switzerland
7Department of Medical Oncology, Inselspital, University Hospital and University of Bern, Bern, Switzerland
8St. Luc, Bruxelles, Belgium
9Division of hematology, University Hospital Basel, Basel, Switzerland
10Cancer Center Amsterdam, Department of Hematology, VU University Medical Center, Amsterdam, Netherlands

Introduction: Although the majority of Acute Myeloid Leukemia (AML) patients achieve a complete morphological remission (CR) after induction therapy, relapse rates remain high. Molecular Minimal Residual Disease (MRD) detection by PCR-based technologies has been shown to improve relapse prediction but has been restricted to specific genetically-defined subsets of AML only. Next-Generation Sequencing (NGS) has the advantage that it allows for the assessment of a broad range of disease-related gene mutations in a single assay. Residual leukemia-specific mutations in bone marrow in morphological CR after induction therapy are supposed to represent the source of relapse. However, persistent mutations may also represent clonal hematopoiesis, analogous to age-related clonal hematopoiesis of indeterminate potential (CHIP) present in healthy individuals. It is currently unknown which and to what extent the persisting somatic mutations after induction therapy contribute to AML relapse. Here, we present a comprehensive study, detailing the value of molecular MRD detection by NGS, in a large prospective cohort of newly diagnosed AML.

Methods: 482 AML patients (<65 years) were treated with 2 cycles of standard induction chemotherapy followed by consolidation in HOVON-SAKK clinical trials ( NGS was performed to detect mutations in a panel of 54 genes frequently mutated in myeloid malignancies (Illumina) at diagnosis and in bone marrow in morphological CR after completion of induction therapy. Thompson-Tau outlier testing was performed to reliably detect persisting mutations above background error rates. The primary and secondary endpoints of the study were relapse and overall survival, respectively. To establish and subsequently test our definition of NGS MRD, the cohort was split into a representative training (n=283) and validation cohort (n=147). The Cumulative Incidence of Relapse (CIR) was estimated with competing-risks regression analyses according to the method of Fine & Gray. The Cox proportional hazard model was used to calculate overall survival estimates. P-values <0.05 were considered significant.

Results: In 430 out of 482 (89.2%) AML patients somatic driver mutations were present at diagnosis. In 51.4% of subjects persisting mutations were detected in bone marrow in morphological CR at highly variable variant allele frequencies (VAF 0.0002-0.47), predominantly persisting in DNMT3A (78.7%), TET2 (54.2%) and ASXL1 (51.6%). These persistent mutations in DNMT3A, TET2 and ASXL1 (DTA) in the training cohort did not associate with the incidence of relapse at any VAF cut-off, indicating a stage of clonal hematopoiesis rather than a condition of impending relapse. In contrast, in the subset of AML patients with persisting DTA mutations, a significant correlation with relapse was observed when any other persistent non-DTA mutation was considered (training cohort: 5-years CIR 76.4% vs. 39.4%; p=0.002). In the training cohort NGS MRD, as defined by persistent non-DTA mutations, was found to be highly associated with the risk of relapse (SHR:1.85 [95%CI 1.27-2.70]; p=0.001), which was confirmed in the validation set (SHR:2.81 [95%CI 1.64-4.79]; p<0.001) (Fig. 1). In fact, NGS MRD was significantly associated with CIR when the training and validation series were combined (5-years CIR 58.3% versus 33.9% (p<0.001)) (Fig. 2). In addition, NGS MRD predicted for reduced survival in both cohorts (training: HR:1.64 [95%CI 1.12-2.42]; p=0.012 and validation: HR:3.08 [95%CI 1.87-5.08]; p<0.001). Finally, multivariable analysis including the data of all 430 AML patients, with adjustment for age, WBC, ELN2017 risk and number of induction cycles needed to achieve CR, revealed that NGS MRD expresses profound independent prognostic significance for relapse (SHR:1.89 [95%CI:1.34-2.65]; p<0.001) and overall survival (HR:1.64 [95%CI:1.18-2.27]; p=0.003). In sensitivity analysis with time-dependent correction for allogeneic stem cell transplantation NGS MRD remained highly prognostic for relapse and survival.

Conclusions: In an unprecedentedly large prospective study including training and validation cohorts, targeted NGS MRD detection is established as a powerful and independent predictor for relapse and survival. NGS MRD is applicable in virtually all newly diagnosed adults with AML while persistent CHIP-related mutations lack prognostic value.

Mutations in SRP54 Gene Cause Severe Primary Neutropenia As Well As Shwachman-Diamond-like Syndrome

Result Type: Paper
Number: LBA-3
Presenter: Christine Bellanné-Chantelot
Program: General Sessions
Session: Late-Breaking Abstracts Session

Christine Bellanné-Chantelot, PharmD, PhD1,2*, Caroline Marty, PhD1*, Barbara Schmaltz-Panneau, PhD1*, Odile Fenneteau, MD3*, Isabelle Callebaut, PhD4*, Séverine Clauin5*, Aurélie Docet5*, Gandhi Damaj, MD6, Thierry Leblanc, MD7*, Isabelle Pellier, MD8*, Cécile Stoven, MD9*, Sylvie Souquere, PhD10*, Iléana Antony-Debre, PhD11*, Blandine Beaupain, MsC12*, Nathalie Aladjidi, MD13*, Vincent Barlogis, MD14*, Frederic Bauduer, MD, PhD15*, Philippe Bensaid, MD16*, Odile Boesflug-Tanguy, MD-PhD17*, Claire Berger, MD18*, Yves Bertrand, MD19*, Liana Carausu, MD20*, Claire Fieschi, MD, PhD21*, Claire Galambrun, MD22*, Aline Schmidt, MD23*, Hubert Journel, MD24*, Francoise Mazingue, MD25, Brigitte Nelken, MD26*, Thuan Chong Quah, MD27*, Eric Oksenhendler, MD28*, Marie Ouachee, MD29*, Marlène Pasquet, MD30*, Felipe Suarez, MD, PhD31, Gérard Pierron, PhD32*, William Vainchenker, MD, PhD1*, Isabelle Plo, PhD1* and Jean Donadieu, MD33

1INSERM UMR1170, Gustave Roussy, Villejuif, France
2Département De Genetique, Hôpital Pitié-Salpétrière, AP-HP, Paris, France
3Laboratoire d’Hématologie, Hôpital Robert Debré, APHP, Paris, France
4CNRS UMR7590, Sorbonne Université, Université Pierre et Marie Curie-Paris6-MNHN-IRD-IUC, Paris, France
5Département de Génétique, Hôpital Pitié-Salpétrière, Paris, France
6Service d’Hématologie, CHU Caen, Caen, France
7Service d’Hématologie et Oncologie pédiatrique, Hôpital Robert Debré, AP-HP, Paris, France
8CHU, Angers, France
9Service de Pédiatrie générale, CHU La Réunion, Site du Groupe Hospitalier Sud Réunion, Saint-Pierre, France
10CNRS UMR-9196, Gustave Roussy, Villejuif, France
11Inserm UMR1170, Gustave Roussy, Villejuif, France
12Registre des Neutropénies, Hopital Trousseau, AP-HP, Paris, France
1312. Unité d’Hématologie Pédiatrique, CIC 1401 INSERM CICP, CHU Bordeaux, Bordeaux, France
14Service d’Hématologie pédiatrique, CHU de Marseille Hopital La timone, Marseille, France
15Centre Hospitalier Côte Basque, Bayonne, France
16Service de Pédiatrie, Centre Hospitalier d’Argenteuil, Argenteuil, France
17Service de Neuropédiatrie et des Maladies Métaboliques, Hôpital Robert Debré, AP-HP, Paris, France
18Service d’hémato-oncologie pédiatrique, CHU Saint-Etienne, Saint-Etienne, France
19IHOPe, Lyon, France
20Service d’Hémato-oncologie pédiatrique, CHU Brest, Brest, France
21Service d’Immunologie clinique, Hôpital Saint-Louis, AP-HP, Paris, France
22Service d’Hématologie pédiatrique, CHU La Timone, Marseille, France
23Service d’Hématologie, CHU Angers, Angers, France
24Département de Génétique, Hôpital Bretagne-Atlantique, Vannes, France
25Service d’Hémato-oncologie pédiatrique, Hopital Jeanne de Flandre Chru de Lille, Lille, France
26Service d’Hémato-oncologie pédiatrique, CHU Hôpital Jeanne De Flandre, Lille, France
27Department of Pediatrics, National University hospital, Singapore, Singapore
28Service d’Immunologie Clinique, Hopital Saint-Louis, AP-HP, Paris, France
29IHOPE, Institut d’hématologie oncologie Pédiatrique, Lyon, France
30Service d’Hémato-oncologie pédiatrique, CHU de Toulouse, Toulouse, France
31Service d’Hématologie adultes, Hôpital Universitaire Necker-Enfants Malades, AP-HP, INSERM UMR 1163 et CNRS ERL 8254, Institut Imagine, Sorbonne Paris Cité, Université Paris Descartes, Paris, France
32Gustave Roussy, CNRS UMR-9196, VILLEJUIF, France
33Service d’Hémato-Oncologie Pédiatrique, Hopital Trousseau, AP-HP, Paris, France


Context: Congenital neutropenia (CN) is a heterogeneous group of diseases characterized by low neutrophil count, severe bacterial infections, increased risk of leukemic transformation and various extra-hematopoietic organ dysfunctions. Even if 24 different genes have been linked to the etiology of CN, in many patients the genetic causes of CN remain unknown.

Methods: Whole-exome sequencing (WES) was performed on a trio-based approach in 8 sporadic cases and 6 multiplex families. Sanger sequencing was performed in a second phase on 66 additional patients from the French CN registry. Structural and functional studies were performed using primary cells from patients including fibroblasts and hematopoietic cells. In vitro granulocytic differentiation was conducted by culturing CD34+ in serum-free medium with SCF, IL-3 and G-CSF for 21 days.

Results: WES analysis identified a heterozygous mutation in the SRP54 gene, encoding the signal recognition particle (SRP) 54 GTPase protein, in 3 sporadic cases and an autosomal dominant family. Considering these results we directly sequenced SRP54 in the French CN cohort and identified 13 additional sporadic cases and 2 multiplex families, thus reaching a total of 23 cases carrying a SRP54 mutation (19 families, Figure 1A). The Thr117del in-frame deletion was found in 12 probands. In all patients, neutropenia was profound (mean neutrophil absolute count 0.23x109/L), diagnosed in the neonate period or during childhood (mean age 4.2 months), and required long-term G-CSF therapy (mean dose 9 µg/kg/day). Bone marrow examination showed a maturation arrest at the promyelocytic stage. In contrast to CN associated with ELANE mutations, SRP54-mutated patients presented an important degree of dysgranulopoiesis (abnormal localization and number of mature granules) and enlarged endoplasmic reticulum (ER) in promyelocytes as well as dystrophic neutrophils (Figure 1B-1C). No evolution into acute myeloid leukemia was observed in this cohort after a median follow-up period of 14.8 years despite high doses of G-CSF. Six out of the 23 patients had extra-hematopoietic manifestations comprising severe neurodevelopmental delay (n=5) and/or exocrine pancreatic insufficiency (n=3), and/or bone abnormalities (n=2) (Figure 1A).

The SRP54 protein is a key component of the ribonucleoprotein complex SRP that mediates the co-translational targeting and the insertion of secretory and membrane proteins to the ER. We identified 7 distinct mutations that affect highly conserved residues within or interacting with G motifs involved in the GTPase activity of SRP54. Of note, 17 out of 18 patients with mutations located within the G1 element and predicted to affect the structure and/or stability of the NG domain presented only a severe neutropenia. In contrast, the other 5 patients with mutations affecting either the magnesium-binding or the guanine-binding site and/or implied in the heterodimeric association with the receptor of SRP54, were associated with Shwachman-Diamond-like syndrome features. During the in vitro granulocytic differentiation of both normal and mutated hematopoietic progenitors, we found a strong increase in SRP54 mRNA expression levels associated with a slight decrease in the protein level in patients compared to controls. Moreover, SRP54 mutations induced a major deleterious effect on proliferation (Figure 1D) and a delay in the differentiation associated with increased apoptosis. Using both in vitro-derived granulocytic cells and primary fibroblasts, we also observed that SRP54 mutations led to ER stress (increased eIF2a phosphorylation, XBP1 splicing, ATF4 and CHOP expression) and enhanced autophagy (higher levels of LC3-II and ULK1 expression).

Conclusions: This study thus identifies a novel pathological pathway implicating the co-translational process of protein targeting. This new genetic subtype which represents the second cause of CN in the French registry (prevalence 6.9%), is characterized by a promyelocytic maturation arrest with dysgranulopoiesis leading to a profound neutropenia, with a poor response to G-CSF, and in few patients, is associated with severe neurodevelopmental delay and exocrine pancreatic insufficiency.

High Throughput Sequencing in 3449 Patients with Bleeding and Platelet Disorders: Novel Gene Discovery and Robust Diagnosis

Number: 5

Program: General Sessions
Session: Plenary Scientific Session

Claire Lentaigne, BSc, MB, MRCP1*, Willem Ouwehand, MD, PhD2 and Thrombogenomics Consortium3*

1NIHR BRC Centre for Haematology, Imperial College London, London, United Kingdom
2NIHR, Cambridge, United Kingdom
3Thrombogenomics consortium, Cambridge, United Kingdom


Background: Bleeding symptoms are common and reported at some point by 25% of the population. Laboratory investigations frequently fail to provide an explanation for the symptoms or to identify those at significant risk of future bleeding. A similar problem is presented by patients with congenital thrombocytopenia which often cannot be distinguished from acquired disorders and making investigation of platelet function impractical. The broad heterogeneity of phenotypes associated with established disorders further complicates diagnosis. For these reasons, we have developed High Throughput Sequencing (HTS) approaches to investigate patients with undiagnosed bleeding and platelet disorders (BPDs) through the ThromboGenomics 79 gene panel HTS (TG) test and by whole genome sequencing (WGS).

Methods: BPD cases were recruited in 3 broad categories: Group 1: 1321 cases suspected to have a defect in one of the 79 known BPD genes (ie genes known to harbour variants responsible for a BPD); Group 2: 212 cases being prepared for elective surgery with self-reported bleeding symptoms; and Group 3: 1916 cases with a suspected inherited BPD, not thought to be caused by a variant in one of the 79 known BPD genes. DNA samples from the first two groups were sequenced by the TG test and from the third group by WGS as part of the first 35,000 samples for the 100,000 Genomes Project.

In total, 3449 DNA samples from BPD patients have been analysed by HTS. Laboratory and clinical phenotypes were recorded using Human Phenotype Ontology (HPO) terms. Variants were called and prioritised based on minor allele frequency, predicted impact and presence in the Human Gene Mutation Database and those variants present in the 79 known BPD genes were reviewed by a multi-disciplinary team (MDT) where pathogenicity was assigned to variants using ACGS criteria (Clearly pathogenic [CPV], Likely pathogenic [LPV] and Variant of unknown significance [VUS]).

Results: A mean of 6.6 variants were assessed at MDT for each case. In Group 1 pathogenic variants were observed in 713 of 1321 (54%) of cases and >40% were novel and labelled as LPV; ~20% of variants discussed were designated VUS. In sharp contrast, pathogenic variants were only identified in 6 of 212 (2.8%) Group 2 cases. In Group 3, pathogenic variants were identified in 106 of 1916 (12%) cases. In >300 cases with non-syndromic thrombocytopenia, CPV or LPV were identified in ~30%. In contrast, in 131 cases with a significant bleeding disorder but no identified laboratory platelet or coagulation abnormality causal variants were observed in only 5 cases.

Conclusion: These results demonstrate the success of HTS approaches in providing a genetic diagnosis for patients with well-defined inherited platelet or coagulation defects. In addition, 23 novel BPD genes were identified by analysis of the WGS data from Group 3 cases, including NBEAL2, RBM8A, SRC, DIAPH1, TPM4, ABCC4 and KDSR. Furthermore extensive replication was demonstrated in nearly 100 Group 3 cases for recent BPD gene discoveries by other groups (C6ORF25, CYCS, RNU4ATAC, STIM1, RASGRP2, ETV6, ACTN1). All MDT-reviewed variants and appended HPO terms are shared in ClinVar, supporting the international effort to improve reference catalogues, aligned to ASH’s Precision Medicine initiative.

Discussion: HTS has been particularly successful for thrombocytopenia, but the yield for bleeding disorders without numerical platelet or coagulation defects is low. Our results, and the observations by others are compatible with the notion that the genetic architecture of unresolved BPD cases is extremely diverse. Continuing efforts such as the 100,000 Genomes Project with sharing and transparency of data across projects will enable better interpretation of variants and the discovery of novel genes. Finally, we are annotating the non-coding space with data from recently identified GWAS variants for blood cell traits, BLUEPRINT and ROADMAP epigenome marks, and eQTLs for blood cell traits to interpret the regulatory regions of the genome and identify non-coding variants causing BPD and other unresolved rare diseases of the blood.

A Simulation Analysis to Evaluate the Effect of Prospective Biomarker Testing on Progression-Free Survival (PFS) in DLBCL

Result Type: Paper
Number: 419
Presenter: Edith Szafer-Glusman
Program: Oral and Poster Abstracts
Session: 627. Aggressive Lymphoma (Diffuse Large B-Cell and Other Aggressive B-Cell Non-Hodgkin Lymphomas)—Results from Retrospective/Observational Studies: Molecular Characterization of Diffuse Large B Cell Lymphoma

Edith Szafer-Glusman1*, Juan Liu2*, Franklin V. Peale Jr.1*, Thalia A. Farazi, MD, PhD1, Jill Ray, PhD1*, Carsten Horn3*, Mikkel Z. Oestergaard3*, Martin Kornacker3, Laurie H. Sehn, MD, MPH4, Günter Fingerle-Rowson3, Jeffrey M. Venstrom1*, Michelle Byrtek1* and Elizabeth Punnoose, PhD1*

1Genentech, Inc., South San Francisco, CA
2Roche (China) Holding, Ltd., Shanghai, China
3F. Hoffmann-La Roche Ltd., Basel, Switzerland
4Centre for Lymphoid Cancer, British Columbia Cancer Agency, Vancouver, BC, Canada


 Introduction: Novel treatment regimens that combine chemotherapy with targeted agents are being developed for DLBCL. Evaluation of these targeted agents in clinical trials may require prospective biomarker testing, such as immunohistochemistry (IHC), to select patients based on relevant molecular features. However, prospective biomarker testing takes days and there is concern that the delay in treatment may prevent enrollment of patients with the most aggressive disease, biasing the trial population. Here, using data from the Phase 3 GOYA trial (NCT01287741) in first-line DLBCL, we evaluate the impact on PFS of a treatment delay that simulates the additional screening time required for prospective IHC testing.

Methods: In GOYA, previously untreated DLBCL patients were randomized 1:1 to obinutuzumab or rituximab plus 6 or 8 cycles of CHOP. Median times from clinical diagnosis to initiation of screening (“DTIS interval”), and from initiation of screening to randomization (“screening time”) were determined in patients with IPI 2–5 (n=1124 patients with timeline information available, no prospective testing). A two-step analysis of PFS was conducted by dividing the evaluable patient population into three groups of varying DTIS intervals (Figure 1, Step 1), and then into additional groups of varying screening times (Figure 1, Step 2). Groups for screening times were 1) <6 days (based on median screening time of 6 days in GOYA; 2) between 6–9 days (based on median screening time of 9 days in Roche clinical trial with prospective IHC testing - NCT02366143); and 3) above 9 days. We asked whether the clinical outcome (PFS), assessed by Kaplan-Meier survival analysis, correlated with either the DTIS interval (Step1) or the screening times (Step 2).

Results: Median time from diagnosis to randomization was 24 days (range 2 to 1106 days, 95th percentile of 66 days). Patients with times from diagnosis to randomization below 15 days (n=250) had significantly worse outcome than patients that were randomized in 15 or more days (n=413 for 15–28 days and n=466 for >28 days; p<0.0001; Table 1; Figure 2 A). 3-year PFS was 56% for patients with <15 days, 70% for 15–28 and 73% for >28 days from diagnosis to randomization. In Step 1 of the two-step analysis, DTIS times of <8 days, 8–14 days, and >14 days were observed for 190, 228 and 706 patients, respectively. Patients with a DTIS interval of <8 days were enriched with BCL2-positive and Double Hit status (Table 2). Patients with a DTIS interval of <8 days had shorter PFS than patients with longer DTIS intervals (p=0.0011 Table 1; Figure 2 B); 3-year PFS was 55% for patients with <8 days, 66% for patients with 8–14 days and 72% for patients with >14 days from diagnosis to screening; no additional difference in 3-year PFS was observed in subgroups beyond 14 days. In Step 2, we asked if a longer screening interval (screening to randomization in Figure 1), simulating additional time for prospective testing, affected patient outcome. Of the 190 patients with the shortest DTIS interval (<8 days from diagnosis to screening) and the shortest PFS from Step 1, 70 patients had screening times of <6 days, 75 between 6–9 days and 45 above 9 days (max 26 days). All cohorts showed indistinguishable PFS curves (p=0.7652; Table 1; Figure 2C). Similarly, no differences in PFS were observed for groups of increasing screening times in the DTIS cohorts of 8–14 days and >14 days (not shown).

Conclusion: In GOYA, short PFS was associated with <15 days from diagnosis to randomization and <8 days from diagnosis to screening, possibly attributable to high-risk biology, such as high expression of BCL2 and DH. Despite seemingly expedited work-up, these high-risk patients do poorly, highlighting the need for targeted therapies and innovative trial designs. In this example, additional screening time similar to the time required for prospective testing did not adverselyaffect PFS. Our results may have implications for designing precision medicine trials in DLBCL.

The Prognostic Significance of the Complete IgH Rearrangement Pattern Using the BIOMED-2 Protocol in Diffuse Large B-Cell Lymphoma

Result Type: Paper
Number: 420
Presenter: Tomohiro Yabushita
Program: Oral and Poster Abstracts
Session: 627. Aggressive Lymphoma (Diffuse Large B-Cell and Other Aggressive B-Cell Non-Hodgkin Lymphomas)—Results from Retrospective/Observational Studies: Molecular Characterization of Diffuse Large B Cell Lymphoma

Tomohiro Yabushita, MD1*, Hayato Maruoka, PhD2*, Yoshimitsu Shimomura, MD1*, Nobuhiro Hiramoto, MD3*, Satoshi Yoshioka, MD, PhD1, Hisako Hashimoto, MD, PhD3* and Takayuki Ishikawa, MD, PhD1

1Department of Hematology, Kobe City Medical Center General Hospital, Kobe, Japan
2Department of Clinical Laboratory, Kobe City Medical Center General Hospital, Kobe, Japan
3Institute of Biomedical Research and Innovation, Kobe, Japan

 Introduction: In several B-cell malignancies, such as chronic lymphocytic leukemia and mantle cell lymphoma, the somatic hypermutation (SHM) status of immunoglobulin heavy chain (IgH) variable region (VH) significantly correlates with clinical outcome. Diffuse large B-cell lymphoma (DLBCL) usually contains a number of VH SHMs, with considerably variable frequency. However, there have been few reports about the association between VHSHM frequency and clinical outcome in DLBCL.

The BIOMED-2 multiplex PCR protocol is one of the standardized methods for detecting IgH gene rearrangements, and has been used to prove clonality in B-cell lymphomas. VH SHMs in the germinal center prevent proper annealing of the applied PCR primers, rendering clonal VH rearrangements undetectable by this protocol. Thus, the complete IgH rearrangement pattern produced by the BIOMED-2 protocol is expected to reflect the frequency of VH SHMs. In this study, we investigated the prognostic impact of complete IgH rearrangement pattern on patients with DLBCL, and analyzed the correlation between complete IgH rearrangement pattern and VH SHM frequency by Sanger sequencing.

Materials and Methods: We retrospectively identified 386 patients diagnosed with de novo DLBCL that received anthracycline-based immunochemotherapy at Kobe City Medical Center General Hospital from January 2005 to June 2015. We excluded 83 patients because fresh samples were not available for PCR analysis, and a total of 303 patients were enrolled. Based on the BIOMED-2 protocol, complete IgH rearrangement was detected by three PCR reactions using three VH forward primer sets: framework region (FR) 1, FR2, and FR3. The primary endpoint was 5-year overall survival (5-yr OS). The secondary endpoint was 5-year progression-free survival (5-yr PFS). OS and PFS were estimated using the log-rank test. Multivariate analysis using the Cox proportional hazard ratio was performed for 5-yr OS. For VH sequencing analysis, the cryopreserved PCR products of 42 available samples were sequenced directly using both forward and reverse primers, and for each sequenced sample, the percentage of VHidentity to the closest germline gene was calculated, using the immunoglobulin gene database (IMGT/V-QUEST).

Results: Of the 303 patients enrolled (median age: 69 years old), 183 patients (60.3%) had advanced stage disease and 148 (48.8%) had a high-intermediate or high International Prognostic Index (IPI) score. The median follow-up was 60.6 months (2.0–133.8 months). Complete IgH rearrangement was detected in 176 (58.1%), 172 (56.8%), and 113 (37.2%) of patients using the FR1, FR2, and FR3 primer sets, respectively. In total, 80 patients (26.4%) were positive for all three PCR reactions, 82 (27.1%) for two, 57 (18.8%) for one, and 84 (27.7%) had completely negative results. We compared the 5-yr OS and PFS among these subgroups (Figure 1), and observed that the probability of survival decreased with accumulation of positive PCR results. As shown in Figure 2, both the 5-yr OS and PFS were significantly worse for patients with all three reactions positive for complete IgH rearrangement (OS: 54.2% vs 73.2% p=0.002; PFS: 34.4% vs. 59.3%, p<0.001). In multivariate analysis adjusted by IPI score, complete IgH rearrangement detected by all three PCR reactions was significantly correlated with poor prognosis (hazard ratio: 2.13; 95% CI: 1.38–3.26; p<0.001). This classification also allowed subgroups with unfavorable prognosis to be identified, even among patients with a high-intermediate or high IPI score (5-yr OS: 28.0% vs. 62.9%, p<0.001). Moreover, VH sequencing analysis of 42 available samples confirmed that patients with complete IgH rearrangement in all three PCR reactions had significantly lower VH gene SHMs than other groups (median value of VH identity; 91.9% vs 84.0%, p<0.001).

Conclusion: Complete IgH rearrangement detected by all three primer sets in the BIOMED-2 PCR protocol was associated with poor prognosis in patients with DLBCL. Combined with the additional sequencing analysis, this suggests that the lower levels of VH SHMs may be associated with unfavorable prognosis in DLBCL.

Circulating Tumor DNA Is a Reliable Measure of Tumor Burden at Diagnosis of Diffuse Large B Cell Lymphoma: An International Reproducibility Study

Result Type: Paper
Number: 310
Presenter: David Kurtz
Program: Oral and Poster Abstracts
Session: 621. Lymphoma—Genetic/Epigenetic Biology: Molecular approaches to describe, monitor, and target lymphoma

David M. Kurtz, MD1, Michael Jin1*, Joanne Soo, BS1, Florian Scherer, MD1*, Alexander Craig, MPhil1*, Jacob J Chabon1*, Joseph G Schroers-Martin, MD1*, Chih Long Liu, PhD1*, Henning Stehr, PhD1*, Christine Schmitz, MD2*, Ulrich Duehrsen, MD3, Andreas Hüttmann, MD2, Anne Ségolène Cottereau4*, Michel Meignan4*, Olivier Casasnovas, MD5*, Jason R. Westin, MD6, Gianluca Gaidano, MD, PhD7, Davide Rossi, MD, PhD8, Mark Roschewski, MD9, Wyndham H. Wilson, MD, PhD10, Maximilian Diehn, MD, PhD11* and Ash A. Alizadeh, MD, PhD12

1Department of Medicine, Divisions of Hematology & Oncology, Stanford University Medical Center, Stanford, CA
2Department of Hematology, University Hospital Essen, Essen, Germany
3Hematology, University Hospital Essen, Essen, Germany
4Hôpital Henri Mondor, Créteil, FRA
5Department of Hematology, University Hospital of Djion, Dijon, France
6Lymphoma/Myeloma, MD Anderson, Houston, TX
7University of Eastern Piedmont, Novara, Italy
8Oncology Institute of Southern Switzerland, Bellinzona, Switzerland
9Lymphoid Malignancies Branch, Center for Cancer Research, National Cancer Institute, Bethesda, MD
10National Cancer Institute, National Institutes of Health, Bethesda, MD
11Department of Radiation Oncology, Stanford University Medical Center, Stanford, CA
12Department of Medicine, Divisions of Hematology & Oncology, Stanford University, Stanford, CA

 Background: Pretreatment tumor burden is a known prognostic factor in diffuse large B cell lymphoma (DLBCL). Components of the International Prognostic Index (IPI), including lactate dehydrogenase (LDH) and clinical stage, attempt to capture tumor burden but lack specificity. Tumor volume is an alternative metric for disease burden, but requires specialized radiographic interpretation. We recently described circulating tumor DNA (ctDNA) levels as a novel quantitative biomarker for disease burden in DLBCL (Scherer F and Kurtz DM, Sci Transl Med 2016). To validate the performance of ctDNA across a diverse cohort of patients and medical centers, we assessed pretreatment ctDNA levels in patients with DLBCL from six centers around the world.

Methods: We performed Cancer Personalized Profiling by Deep Sequencing (CAPP-Seq) on pretreatment cell-free and germline DNA from patients undergoing curative-intent anthracycline-based therapy. Somatic mutations were identified and variant allele fractions were determined. Levels of ctDNA were determined using the mean tumor allele fraction and the total concentration of cell-free DNA; absolute concentration of ctDNA was expressed as haploid genome equivalents per mL (hGE/mL). Finally, we correlated levels of ctDNA with clinical factors including stage, IPI, pretreatment metabolic tumor volume (MTV), and patient outcomes.

Results: A total of 183 patients from six centers were enrolled (Stanford University, n=52; University of Eastern Piedmont, n=36; MD Anderson Cancer Center, n=21; Hospital Le Bocage, Dijon, n=25; the National Cancer Institute, n=33; and University Hospital Essen, n=16). Pretreatment cell-free DNA was sequenced deeply; ctDNA was detectable prior to therapy in 97% of patients. The performance of ctDNA detection was consistent across cohorts, with >90% of patients from each center having detectable ctDNA. Concentration of ctDNA was not significantly different across cohorts, with a median of 295 hGE of tumor DNA / mL (one-way ANOVA, P=0.62; Fig 1A). Using pretreatment plasma, all but 5 patients had mutations identifiable via non-invasive genotyping without knowledge of tumor genotype. In 88 patients with tumor samples available, a median of 81% (interquartile range: 34-92%) of variants identified via non-invasive genotyping were confirmed by tumor sequencing on a per-patient basis.

We next sought to correlate levels of ctDNA with independent measures of tumor burden. We found a significant difference in ctDNA levels between patients in different IPI risk groups (one-way ANOVA, P<0.0001; Fig 1B). We also found that patients with higher-stage disease had higher ctDNA levels (P<0.0001); ctDNA levels also correlated with LDH (P<0.0001). We demonstrated a robust correlation between MTV and ctDNA level (P<0.0001; Fig 1C). This correlation was consistent across all three cohorts with MTV available (Essen: P=0.0005, r=0.80; Dijon: P<0.0001, r=0.73; Stanford: P=0.006, r=0.50). Finally, pretreatment ctDNA concentration was continuously associated with EFS and OS (P<0.0001 and P=0.0014). Using the median to divide patients into groups with high or low ctDNA, we found patients with high ctDNA levels had significantly inferior EFS compared to those with low ctDNA (P=0.005, HR 2.7, 95% CI 1.4-4.6; Fig 1D).

Conclusions: Circulating tumor DNA is a robust biomarker of disease burden in DLBCL in patients from multiple medical centers across North America and Europe. Patients from different centers have similar levels of ctDNA despite pre-analytical variation. Concentrations of ctDNA correlate with stage, IPI, and tumor volume; furthermore, ctDNA level prior to therapy is prognostic of clinical outcome. This study validates previous findings and provides evidence that ctDNA is a reliable biomarker in aggressive lymphomas that could be used to refine clinical estimates of disease burden including stage.

Genotyping of Classical Hodgkin Lymphoma on the Liquid Biopsy

Result Type: Paper
Number: 307
Presenter: Valeria Spina
Program: Oral and Poster Abstracts
Session: 621. Lymphoma—Genetic/Epigenetic Biology: Molecular approaches to describe, monitor, and target lymphoma

Valeria Spina, PhD1*, Alessio Bruscaggin, PhD1*, Annarosa Cuccaro2*, Maurizio Martini, MD, PhD3*, Martina Di Trani4*, Gabriela Forestieri1*, Martina Manzoni5,6*, Luca Nassi7*, Adalgisa Condoluci1,8*, Alberto Arribas, PhD1*, Silvia Locatelli, PhD4*, Elisa Cupelli2*, Luca Ceriani8*, Alden Moccia8*, Anastasios Stathis, MD9*, Clara Deambrogi, PhD7*, Fary Diop7*, Francesca Guidetti1*, Antonino Neri, MD5,6*, Bernhard Gerber, MD8*, Francesco Bertoni, MD1,8, Michele Ghielmini, MD8, Georg Stuessi, MD8*, Armando Santoro, MD4*, Franco Cavalli, MD8, Emanuele Zucca, MD8, Luigi Maria Larocca, MD3*, Gianluca Gaidano, MD, PhD10, Stefan Hohaus2*, Carmelo Carlo-Stella, MD4 and Davide Rossi, MD, PhD1,8

1Institute of Oncology Research, Bellinzona, Switzerland
2Institute of Hematology, Catholic University of the Sacred Heart, Rome, Italy
3Division of Pathology and Histology, Catholic University of the Sacred Heart, Rome, Italy
4Department of Oncology and Hematology, Humanitas Cancer Center, Humanitas University, Milan, Italy
5Department of Oncology and Hemato-Oncology, University of Milan, Milan, Italy
6Hematology Unit, Foundation Ca’ Granda IRCCS, Ospedale Maggiore Policlinico, Milan, Italy
7Division of Hematology, Department of Translational Medicine, University of Eastern Piedmont, Novara, Italy
8Oncology Institute of Southern Switzerland, Bellinzona, Switzerland
9Oncology Institute of Southern Switzerland, Bellinzona, Switzerland, Switzerland
10University of Eastern Piedmont, Novara, Italy


Introduction: In classical Hodgkin lymphoma (cHL), the low representation (~5%) of Hodgkin-Reed-Sternberg cells (HRS) challenged tumor genotyping on the tissue biopsy. Cell free DNA (cfDNA) is shed into the blood by tumor cells and can be used as source of tumor DNA for the identification of somatic mutations, track clonal evolution of tumors and detect minimal residual disease during therapy. AIMS. The study aimed at: i) showing that cfDNA mirrors the genetics of HRS cells in cHL patients; ii) characterizing the mutational profile of a large cohort of newly diagnosed and chemorefractory cHL; iii) identifying molecular prognostic subtypes; iv) early detecting residual disease during therapy; and v) longitudinally tracking tumor clonal evolution under different treatment modalities.

 Methods: The study included 80 newly diagnosed cHL and 32 chemorefractory cHL. The following biological material was analyzed: i) cfDNA from plasma collected at diagnosis, during ABVD courses, at refractory progression, before and during therapy with brentuximab or nivolumab; and ii) normal germline genomic DNA (gDNA) from granulocytes. For comparative purposes, paired tumor gDNA from microdissected HRS cells of 13 cases was also analyzed. A targeted resequencing panel optimized to include the coding exons and splice sites of 77 genes recurrently mutated in B-cell lymphomas was used for genotyping. Ultra-deep next-generation sequencing (NGS) of the gene panel was performed on NexSeq 500 (Illumina) using the CAPP-seq library preparation strategy (NimbleGen).

Results:. In cHL patients, cfDNA surrogated gDNA from HRS cells, since it harbored 87.5% of the tumor confirmed mutations. Genes recurrently affected by non-synonymous somatic mutations in >20% of cHL included STAT6 (37.5%), TNFAIP3 (35%), and ITPKB (27.5%) (Fig. 1A). Mutations clustered in major pathways, including NF-κB, PI3K-AKT, cytokine and NOTCH signaling, and immune evasion. ITPKB mutations: i) were quite specific for cHL, being rare or absent in other lymphomas; ii) caused the subcellular delocalization of the protein in primary HRS cells of mutated patients; iii) correlated with clues of PI3K-AKT signaling activation both at gene expression and protein levels; and iv) consistent with the positioning of ITPKB downstream PI3K in the pathway, associate with resistance to PI3K inhibitors. Mutations of CD58, encoding a co-stimulatory molecule for T-cells, associated with short PFS independent of interim PET/CT results, pointing to immune escape genetic lesions as biomarkers of aggressive disease (Fig. 1B). Newly diagnosed and chemorefractory cHL shared a largely overlapping mutational landscape. TP53 mutations were not enriched in refractory cHL as instead commonly found in other types of refractory B-cell tumors. Conversely, more TET2 mutations were documented in refractory cHL, including newly acquired mutation, thus signaling towards aberrant DNA methylation programming as a mechanism of resistance in cHL with potential therapeutic implications. By longitudinal analysis, in patients relapsing under/after chemotherapy or brentuximab vedotin, pre-treatment/relapse tumor pairs branched through the acquisition of phase specific mutations from an ancestral clone, that always persisted (Fig. 1C). Conversely, in patients maintained under nivolumab, clones were cyclically suppressed and replaced by completely novel clones. We utilized the change in circulating tumor cfDNA load from baseline to interim timepoint to predict the best response to ABVD and to complement interim PET/CT in anticipating cure. A drop of 100-fold or 2-log drop in tumor cfDNA after 2 ABVD courses associated with an eventual complete response and cure. All cured patients that were inconsistently judged as interim PET/CT positive turned out to have a >2 log drop in tumor cfDNA. A drop of less than 2-log in tumor cfDNA after 2 ABVD courses associated with an eventual progression. All relapsed patients that were inconsistently judged as interim PET/CT negative turned out to have a <2 log drop in tumor cfDNA (Fig. 1D).

Conclusions: Circulating tumor cfDNA allows to noninvasively detect tumor-specific mutations: i) identify prognostic subtypes; ii) early detect residual disease during therapy; and iii) longitudinally track tumor clonal evolution under different treatment modalities.

Development of a Dynamic Model for Personalized Risk Assessment in Large B-Cell Lymphoma

Result Type: Paper
Number: 826
Presenter: David Kurtz
Program: Oral and Poster Abstracts
Session: 627. Aggressive Lymphoma (Diffuse Large B-Cell and Other Aggressive B-Cell Non-Hodgkin Lymphomas)—Results from Retrospective/Observational Studies: PET-Guided Therapy and Prognostic Models in Diffuse Large B-Cell Lymphoma

David M. Kurtz, MD1, Florian Scherer, MD1*, Michael Jin2*, Joanne Soo, BS1, Alexander Craig, MPhil1*, Mohammad S Esfahani, PhD1*, Jacob J Chabon1*, Henning Stehr, PhD1*, Chih Long Liu, PhD1*, Robert Tibshirani3*, Lauren S. Maeda1*, Neel K. Gupta, MD1*, Michael Khodadoust, MD, PhD1, Ranjana H. Advani, MD4, Ronald Levy, MD4, Aaron Newman1*, Jason R. Westin, MD5, Gianluca Gaidano, MD, PhD6, Davide Rossi, MD, PhD7, Maximilian Diehn, MD, PhD8* and Ash A. Alizadeh, MD, PhD9

1Institute of Oncology Research, Bellinzona, Switzerland
2Institute of Hematology, Catholic University of the Sacred Heart, Rome, Italy
3Division of Pathology and Histology, Catholic University of the Sacred Heart, Rome, Italy
4Department of Oncology and Hematology, Humanitas Cancer Center, Humanitas University, Milan, Italy
5Department of Oncology and Hemato-Oncology, University of Milan, Milan, Italy
6Hematology Unit, Foundation Ca’ Granda IRCCS, Ospedale Maggiore Policlinico, Milan, Italy
7Division of Hematology, Department of Translational Medicine, University of Eastern Piedmont, Novara, Italy
8Oncology Institute of Southern Switzerland, Bellinzona, Switzerland
9Oncology Institute of Southern Switzerland, Bellinzona, Switzerland, Switzerland
10University of Eastern Piedmont, Novara, Italy


Background: Outcomes for patients with diffuse large B-cell lymphoma (DLBCL) remain heterogeneous. Several clinical and molecular risk factors have been described, both at time of diagnosis (i.e., International Prognostic Index and cell of origin) and during therapy (i.e., interim PET scans). However, these risk factors fail to consistently identify individuals destined for treatment failure. We recently described circulating tumor DNA (ctDNA) as a novel biomarker in DLBCL, demonstrating pretreatment ctDNA levels are prognostic of outcomes (Scherer F, Sci Transl Med 2016). A major advantage of ctDNA assessment is the ability to collect serial assessments over time. However, the best approach for integrating multiple ctDNA measurements with established prognostic factors remains unknown. Using the dynamics of ctDNA, we developed a method to quantify personalized disease risk as it changes during therapy.

Methods: We profiled 468 samples from 125 subjects collected during their first three cycles of immunochemotherapy using cancer personalized profiling by deep sequencing (CAPP-Seq). Early ctDNA dynamics were correlated with outcomes in a training cohort to define a response threshold that was tested in a validation cohort. Finally, we integrated ctDNA and clinical factors to dynamically assess disease risk over time in a continuous model.

Results: Prior to therapy, ctDNA was detectable in 98% of subjects. Pretreatment levels of ctDNA were prognostic of event-free survival in patients receiving either frontline anthracycline-based (n=92) or salvage regimens (n=33) (EFS: HR 2.7, 95% CI 1.5–4.7, P=0.0005), confirming prior results. In the training cohort, ctDNA levels changed rapidly, with a 2-log or 100-fold decrease after one cycle (early molecular response, EMR) stratifying outcomes (Fig 1A). In the validation cohort, patients achieving EMR had superior outcomes at 24-months to those who did not (Fig 1B) (EFS: HR 24, 95% CI 6.6–89, P<0.0001). These results remained significant in subgroups of patients receiving either frontline or salvage therapy. Circulating tumor DNA levels continued to decline during cycle 2, such that a distinct threshold after two cycles could also be trained. After two cycles, a 2.5-log decrease (major molecular response, MMR) also stratified patients for EFS (HR: 8.6, 95% CI 2.2-33, P=0.002).

Next, we integrated serial ctDNA measurements with established risk-factors to develop a model to predict an individual’s disease risk. This model – the Continuous Individualized Risk Index (CIRI) – provides a personalized estimate of disease risk over time. As more information becomes available during a patient’s course of disease, CIRI updates the disease risk, integrating the new information (Fig 1C). In patients receiving frontline therapy, CIRI outperformed the IPI for identification of 24-month EFS and OS (Fig 1D-E) (EFS24: AUC 0.64 vs 0.79; net reclassification improvement 0.47, P=0.02; OS: AUC 0.56 vs 0.84; net reclassification improvement 0.74, P=0.004).

Conclusions: Baseline and interim ctDNA measurements have prognostic significance in aggressive lymphomas. Integration of serial ctDNA measurements through a continuous, dynamic risk model can identify personalized outcome probabilities, yielding superior risk estimates. Dynamic risk assessment is potentially widely applicable and could guide future personalized therapeutic approaches.

Figure 1: A) A spider-plot depicts the dynamics of ctDNA during the first two cycles of therapy in 14 subjects. B) A Kaplan-Meier estimate depicts EFS for patients achieving or not achieving EMR. C) A schema for CIRI is shown. When a patient is diagnosed with DLBCL, the IPI is calculated, giving an initial estimate of risk. Additional pretreatment risk factors (i.e., cell of origin and pretreatment ctDNA) can be added, thereby updating the estimate of risk. As a patient undergoes therapy, further predictors of risk are obtained, such as ctDNA at cycle 2 and cycle 3, and interim imaging. These predictors can be used to update the patient’s estimated risk. Below is a CIRI risk model for two exemplar patients with the same initial pretreatment risk factors. Patient P214 (red) died of disease at day 210, while patient P224 (blue) is in a continued complete response. D) AUC for prediction of EFS and OS by IPI, EMR, interim PET scans, and CIRI. E) A Kaplan-Meier estimate demonstrates risk stratification of EFS by CIRI.

Genetically Defined Diffuse Large B-Cell Lymphoma Subsets Arise By Distinct Pathogenic Mechanisms and Predict

Number:  38
Program: Oral and Poster Abstracts
Type: Oral
Session: 621. Lymphoma—Genetic/Epigenetic Biology: Genomic characterization of lymphoma and beyond

Bjoern Chapuy, MD, PhD1, Chip Stewart, PhD2*, Andrew Dunford, MS2*, Jaegil Kim, PhD2*, Atanas Kamburow, PhD2*, Robert A. Redd, MS1*, Michael Lawrence, PhD2*, Margaretha G.M. Roemer, MSc1*, Amy Li3,4*, Marita Ziepert, PhD5*, Annette M. Staiger, PhD6,7*, Jeremiah Wala2*, Matthew D. Ducar, MS8*, Ignaty Leshchiner, PhD2*, Esther Rheinbay, PhD2*, Amaro Taylor-Weiner, BS2*, Caroline Coughlin, BS1*, Julian Hess2*, Chandra Pedamallu2*, Dimitri Livitz2*, Daniel Rosebrock2*, Mara Rosenberg2*, Adam Tracy2*, Heike Horn7*, Paul van Hummelen8*, Andrew L Feldman, MD9, Brian K Link, MD10, Anne J Novak, PhD11, James R. Cerhan, MD, PhD12, Thomas M Habermann, MD11, Reiner Siebert, PhD13*, Andreas Rosenwald14*, Aaron R. Thorner, PhD8*, Matthew Meyerson, MD PhD15*, Todd R. Golub, MD2*, Rameen Beroukhim2*, Gerald Wulf, MD16*, German Ott, MD6*, Scott J. Rodig, MD PhD17, Stefano Monti, PhD3,4*, Donna S. Neuberg, ScD18, Markus Loeffler, MD19*, Michael Pfreundschuh, MD 20, Lorenz Trümper, MD21*, Gad Getz, PhD22* and Margaret A. Shipp, MD1


1Dana-Farber Cancer Institute, Boston, MA
2Broad Institute, Cambridge, MA
3Section of Computational Biomedicine, Boston University School of Medicine, Boston, MA
4Bioinformatics Program, Boston University, Boston, MA
5Institute for Medical Informatics, Statistics and Epidemiology, University of Leipzig, Leipzig, Germany
6Department of Clinical Pathology, Robert-Bosch-Krankenhaus, Stuttgart, Germany
7Dr Margarete Fischer-Bosch Institute of Clinical Pharmacology Stuttgart, and University of Tuebingen, Stuttgart, Germany
8Center for Cancer Genome Discovery, Dana-Farber Cancer Institute, Boston, MA
9Department of Laboratory Medicine and Pathology, Mayo Clinic, Rochester, MN
10University of Iowa Hospitals and Clinics, Iowa City, IA
11Division of Hematology, Mayo Clinic, Rochester, MN
12Department of Health Sciences Research, Mayo Clinic, Rochester, MN
13Department of Human Genetics University of Ulm, Ulm, Germany
14Institute of Pathology, University of Wuerzburg and Comprehensive Cancer Center Mainfranken, Wuerzburg, Germany
15Department of Medical Oncology, Dana Farber Cancer Institute, Boston, MA
16Department of Hematology and Oncology, University Hospital of Göttingen, Göttingen, Germany
17Brigham and Women’s Hospital, Boston, MA
18Department of Biostatistics and Computational Biology, Dana-Farber Cancer Institute, Boston, MA
19Leipzig University, Faculty of Medicine, IMISE, Leipzig, DEU
20Department for hematology and oncology, Saarland University Medical School, Homburg, Germany
21Hematology and Medical Oncology, University Hospital Göttingen, Göttingen, Germany
22Cancer Program, Broad Institute of MIT and Harvard, Cambridge, MA


Diffuse large B-cell lymphoma (DLBCL) is a heterogeneous disease that largely arises from antigen-exposed B-cells that transit through the germinal center (GC). DLBCL is further classified into transcriptional subtypes – activated B-cell (ABC) and GC B-cell (GCB). ABC DLBCLs exhibit genetic alterations in NF-kB modifiers and proximal components of the B-cell receptor (BCR) pathway and perturbed terminal B-cell differentiation. GCB DLBCLs have reported alterations in chromatin-modifying enzymes, PI3K signaling and Gα-migration pathway components and frequent translocations of BCL2.

Despite the recognized molecular heterogeneity in DLBCL, previously published genomic analyses have largely focused on single types of alterations – mutations, copy number alterations (CNAs) or structural variants (SVs) – in smaller data sets with more limited clinical annotation. An unbiased comprehensive analysis of all three alteration types is needed to define discrete, clinically annotated subtypes of DLBCL.

We performed whole exome sequencing (WES) of 304 newly diagnosed DLBCLs, using an expanded bait set that captures recurrent SVs; 85% of study patients were uniformly treated with R-CHOP and had long-term follow-up. Somatic alterations (mutations, CNAs and SVs) and their clonality were determined with analytical pipelines developed at the Broad Institute. Notably, half of our DLBCL cohort lacked patient-matched normal samples, prompting the successful development of new methods to analyze tumor-only WES data. Significantly mutated genes (SMGs) were identified with MutSig2CV and recurrent CNAs were defined with GISTIC2.0.

With increased sample size and improved methodology, we identified 158 SMGs, CNAs and SVs; the 98 SMGs included ones previously unreported in DLBCL but described in other lymphoid malignancies or cancer. Additional insights into the putative biological function of newly identified alterations were obtained by overlaying the predicted protein changes onto their 3-dimensional protein structures.

We analyzed the mutational signatures in these DLBCLs and found that the majority of exome mutations were caused by spontaneous deamination at CpGs, a clock-like genetic signature associated with aging; only a minority of mutations were attributed to activation-induced cytidine deaminase (AID), an enzyme required for physiologic immunoglobulin receptor editing and aberrant somatic hypermutation.

IGHBCL2BCL6 and MYC were the most frequently rearranged genes (40%, 21%, 19% and 8%, respectively). There were 18 arm-level and 18 focal regions of copy gain and 2 arm-level and 32 focal regions of copy loss with frequencies ranging between 5-32%.

SMGs were significantly more likely to reside within focal CNAs (p=1x10-44), suggesting that these driver genes were perturbed by multiple mechanisms. Individual DLBCLs had a median of 17 genetic drivers, highlighting the need for a more comprehensive analysis. Therefore, we applied an integrated clustering approach to the recurrent mutations, CNAs and SVs and delineated 5 genetically distinct DLBCL subtypes and determined the likely temporal order of alterations within each cluster.

This unbiased approach identified a previously unappreciated favorable risk ABC subset with genetic features of an extrafollicular, possibly marginal zone origin; a more tightly defined poor risk group of ABC DLBCLs with frequent BCL2 gain and concordant MYD88L265P/CD79B mutations; a distinct poor risk subset of GCB tumors with BCL2 SVs and alterations of PTEN and epigenetic enzymes and a discrete group of good risk GCB DLBCLs with specific alterations in BCR/PI3K, JAK/STAT and BRAF pathway components and multiple histones; and a Cell-of-Origin-independent subset with biallelic inactivation of TP53, 9p21.3/CDKN2A and associated genomic instability. These findings likely explain the variability in outcome predictions with the binary ABC- vs. GCB-DLBCL transcriptional classification and the challenges of therapeutically targeting less well defined DLBCLs.

The genetically distinct DLBCL subsets provide a framework for assessing previously unrecognized heterogeneity in this disease, characterizing combinations of genetic alterations that drive DLBCL biology and guiding the development of rational single-agent and combination therapies in patients with the greatest need.

The T-Cell Receptor Repertoire Predicts Interim-PET in Patients with DLBCL Treated with R-CHOP: An Observational Study from a Prospective Clinical Trial

Result Type: Paper
Number: 825
Presenter: Mohamed Shanavas
Program: Oral and Poster Abstracts
Session: 627. Aggressive Lymphoma (Diffuse Large B-Cell and Other Aggressive B-Cell Non-Hodgkin Lymphomas)—Results from Retrospective/Observational Studies: PET-Guided Therapy and Prognostic Models in Diffuse Large B-Cell Lymphoma

Mohamed Shanavas, MD1,2, Mark Hertzberg, MBBS, PhD3*, Rodney J Hicks, MD4*, John F Seymour, MBBS, PhD5, Joshua W.D. Tobin, MD1*, Marina Mathews, PhD1*, Santiyagu Francis, PhD1*, Frank Vari, PhD1*, Maher K Gandhi, PhD6,7 and Colm Keane, MB, PhD1,7*

1University of Queensland Diamantina Institute, Brisbane, Australia
2Rockhampton Base Hospital, Rockhampton, Australia
3Prince of Wales Hospital, Sydney, Australia
4Peter MacCallum Cancer Centre, Melbourne, Australia
5Peter MacCallum Centre & Royal Melbourne Hospital, Melbourne, Australia
6University of Queensland Diamantina Institute, Brisbane, QLD, Australia
7Princess Alexandra Hospital, Brisbane, Australia


T-cell infiltration of the tumor microenvironment (TME) in DLBCL is a key determinant of response to chemo-immunotherapy (Keane, Lancet Haem 2015). We have previously shown that greater diversity of the T-cell receptor (TCR) repertoire within the TME is correlated with improved survival following R-CHOP in DLBCL (Keane, CCR 2017). There are limited data on the impact of the intratumoral TCR repertoire on interim-PET (iPET), the relationship between intratumoral and circulating TCRs, and on dynamic changes of the TCR during therapy. In this study, we interrogated the TCR repertoire in a subset of DLBCL patients treated on the prospective Australasian Leukaemia Lymphoma Group NHL21 study (Hertzberg, Haematologica 2017), in which all patients had 4x RCHOP prior to iPET risk stratification.

The CDR3 region of TCRβ chain underwent high-throughput unbiased TCRβ sequencing (Adaptive Biotechnologies). Metrics included: productive templates (total functional T-cells), productive rearrangements (functional T-cells with distinct specificity), productive clonality (repertoire unevenness due to clonal expansions), and maximal frequency clones (% most dominant single clone). Matched intratumoral diagnostic samples, blood at pre-therapy and post-cycle 4 (at the time of iPET) were tested. 42 patients (enriched for iPET+ cases) had sufficient material for testing.

Median age was 55 (range 22-69) years and 72% were males. IPI was low/intermediate/high in 13/63/25% respectively. Cell of origin (COO) by Lymph 2CX method (nanoString) was ABC in 30%, and GCB in 44%. 40% were iPET+. In tissue, there was a median of 4652 productive templates, translating into 2998 productive rearrangements identified.

Notably, the clonal repertoire of intratumoral TCRs in iPET+ patients was larger than iPET-ve patients (productive clonality 8.1 vs 5.1 x10-2p=0.04), whereas the numbers of functional T-cells did not vary between groups. Comparing the tumor with the blood samples showed a high, but variable, degree of overlap between peripheral blood and the TME – TCR repertoire. Median number of top 100 tumor tissue clones shared in peripheral blood was 53.5 (range, 1-97) in pre-therapy and 39.5 (range, 0-93) in post-therapy blood, indicating that the both the circulation and the tumor likely contribute to immune-surveillance.

In pre-therapy blood, the median productive templates and productive rearrangements were 44,950 (range, 6,003-273,765) and 29,090 (range, 5,190-152,706), and the median clonality was 8.5 (1.46-45.3) x 10-2. There were no differences between iPET+ and iPET-ve patients in these parameters. However, there was a marked change in T-cell composition between time points. Interestingly, in iPET-ve patients clonality measures were increased, with productive clonality 9.4 vs 14.4 x10-2p=0.03; and % maximum productive frequency 3.39 vs 5.89, p=0.04.

These findings demonstrate that the intratumoral TCR repertoire, and sequential blood sampling provide important information on outcome in DLBCL treated with RCHOP. A highly clonal T-cell repertoire in the TME was associated with iPET positivity after 4 cycles of R-CHOP. In line with findings in solid cancers treated with checkpoint blockade, development of clonal responses in peripheral blood was associated with iPET negativity. These findings indicate that clones expanded during therapy may be important in tumor clearance but that highly clonal T-cell responses in the tumor at diagnosis may hinder expansion of other T-cell responses to neoantigens. The circulating TCR composition is representative of the TME. These findings will assist the rationale design and therapeutic monitoring of novel immuno-therapeutic strategies.

The Tumor Microenvironment Is Independently Prognostic of Conventional and Clinicogenetic Risk Models in Follicular Lymphoma

Result Type: Paper
Number: 728
Presenter: Joshua Tobin
Program: Oral and Poster Abstracts
Session: 622. Lymphoma Biology—Non-Genetic Studies: Prognostic Biomarkers and Immune Mechanisms in Lymphoma

Joshua W.D. Tobin, MD1,2*, Colm Keane, MB, PhD1,2*, Peter Mollee, FRACP, FRCPA, MBBS, MMSc3,4, Simone Birch, MBBS1*, Clare Gould, MBBS2*, Jay Gunawardana, PhD5, Thanh Hoang2*, Ti Ma, BSci5*, Emad Uddin Abro, BSc, FRACP, FRCPA, MBBS5,6*, Mohamed Shanavas, MD5,7, Hyung Yoo8*, Li Li8*, Paul Scuffham9*, Valentine Murigneux5*, Lynn Fink, PhD5*, Nicholas Matigian5*, Frank Vari, PhD2*, Santiyagu Francis, PhD5* and Maher K Gandhi, PhD1,10

1Princess Alexandra Hospital, Brisbane, Australia
2University of Queensland Diamantina Institute, Brisbane, Australia
3Division of Cancer Services, Princess Alexandra Hospital, Brisbane, QLD, Australia
4School of Medicine, University of Queensland, Brisbane, Australia
5University of Queensland, Brisbane, Australia
6Mater Hospital Brisbane, Brisbane, Australia
7Rockhampton Base Hospital, Rockhampton, Australia
8Ochsner Health System, New Orleans, LA
9Griffith Univeristy, Brisbane, Australia
10University of Queensland Diamantina Institute, Brisbane, QLD, Australia


Follicular Lymphoma (FL) is the most common indolent Non-Hodgkin Lymphoma. Despite generally favorable survival outcomes, 20% of FL patients experience ‘Progression of Disease within 24 months’ (POD24) and subsequently have poor long-term overall survival (OS) (Casulo, JCO 2015). Unfortunately, POD24 has limited clinical value because it cannot guide up-front clinical decisions. Accurate pre-therapyprognosticators are vital for clinical trial design and are also increasingly being mandated by funding agencies for stratification of patients to emerging front-line treatments. The new ‘state-of-the-art’ prognosticators ‘m7-FLIPI’ and POD24-PI’ (Pastore, Lancet Oncol 2015; Jurinovic, Blood 2016) supplement clinical parameters with genetic mutational status. However, their applicability to population based cohorts including early-stage and asymptomatic patients remains unknown. Furthermore, there is significant heterogeneity of outcome within these prognostic groupings. The established biological and prognostic importance of the tumor microenvironment (TME) in FL suggests that prognosis would be enhanced by incorporating information on host immunity (Scott, Nat Rev Can 2014).

Forty-five pre-treatment FL biopsies were categorized into ‘hot’ or ‘cold’ immune nodes by multiplex immunofluorescent imaging and respectively characterized by concordant high or low expression of multiple immune effector and checkpoint-associated proteins. (Fig 1A). Consistent with these findings, gene expression using the Nanostring platform showed that immune effectors (CD4/CD8/TNFa/CD137/CD56) positively correlated with immune checkpoints (PD-1/PD-L1/PD-L2/TIM3/LAG3/CD163/CD68) indicative of an adaptive immune response. Additionally, high-throughput unbiased TCRb sequencing showed the intratumoral TCR repertoire was more clonal in ‘hot’ compared to ‘cold’ FL samples (p=0.024), indicative of a skewed T-cell immune response (Fig 1B).

We then applied these findings to an independent population-based cohort of 175 cases of FL from the rituximab era with long-term follow-up (median ~7 years), including advanced (n=137) and localized cases (n=38). The aims were to: a) identify new targetable immune parameters of prognostic importance in the rituximab-era; and b) compare and contrast these with published prognostic tools: FLIPI, FLIPI-2, m7-FLIPI, POD24-PI and ‘immune survival score’ (‘ISS’, Dave, NEJM 2004). OS was not only inferior in those experiencing POD24 (HR 4.88, p<0.0001, Fig 1C) but these patients had a >2-fold increase in 5-year patient health costs. Hence, POD24, as well as FFS and TT2T were therefore chosen as the primary outcome measures.

M7 mutation frequencies were similar to those previously published (Pastore, Lancet Oncol 2015). However, the prognostic utility of the m7-FLIPI could not be demonstrated, whereas the FLIPI, FLIPI-2, and POD24-PI retained their prognostic value. The POD24-PI was most predictive of FFS (p<0.0001, HR=3.54) and was most specific in identifying cases that experience POD24 (Sp=68%). The prognostic utility of the TME was then tested. Notably both the ISS (p=0.024, HR=1.74) and multiple immune genes not represented in the ISS including PD-L2, TIM3, LAG3, CD137, TNF and CD4 predicted FFS. PD-L2 demonstrated the strongest association with FFS (p<0.0001, HR=3.74, Fig 1D). It not only out-performed the ISS but was independent to the FLIPI and POD-24-PI. The prognostic significance of PD-L2 was validated in an independent population based cohort of uniformly R-CVP treated patients from an in-silico dataset with gene expression quantified using the Illumina DASL platform (Pastore, Lancet Oncol 2015).

We have validated the TME in predicting outcome in a population based cohort of FL patients with long-term follow-up treated in the rituximab era. Furthermore, we describe the role of PD-L2 as well as several additional pertinent, clinically-actionable markers of the TME which predict survival to conventional therapies in FL. Low expression of PD-L2 appears to be a surrogate of a broadly co-ordinated downregulation of the intratumoural response. These immune scores are independent of and additive to additive to the FLIPI and POD24-PI. Development of new prognostic models require the incorporation of host immunity along with clinico-genetic features to further improve the specificity, and to accurately risk stratify FL patients.