Proceedings of the 9th International Symposium on MDS and SAA in Childhood

2. Do Gata2 dynamics and levels matter during hematopoietic (stem cell, progenitor or other) development? C.S. Vink, A. Popravko, C. Eich, A. Maglitto, F.J. Calero-Nieto, W. Jawaid, X. Wang, S.A. Mariani, B. Göttgens, E. Dzierzak Centre for Inflammation Research, Queen’s Medical Research Institute, The University of Edinburgh, UK; Department of Haematology, Cambridge Institute for Medical Research, Wellcome Trust-MRC Cambridge Stem Cell Institute, University of Cambridge, UK Molecular and Cellular Pediatrics 2021, 8(Suppl 1):2.

Results: Here we describe the genetic landscape and somatic changes of 13 Spanish GATA2-deficient patients using WES. Among the 13 GATA2 mutations funded half were novel (4/9). Median age at the diagnosis was 35 years (range 6-75 years) and clinical manifestation range from immunodeficiency, MDS and AML; one patient was asymptomatic at the time of genomic analysis, however in the last follow-up showed absolute monocytopenia and monosomy 7. WES analysis detected mutations in relevant genes involved in the NOTCH pathway and somatic mutations in RUNX1, EZH2, ETV6, IDH2, TET2, GATA1 and STAG2 as already reported in previous studies. Next, we studied the impact of two of the most recurrent germline GATA2 mutations associated with MDS (R396Q and R389W) using a human iPSC-based disease model. Applying a CRISPR/Cas9-mediated genome editing strategy we generated two hiPSC-GATA2 mutant lines and differentiated them toward blood progenitors. Data collected from FACS analyses showed that heterozygous GATA2 mutations (R396Q and R398W) lead to a block of early hematopoietic progenitor maturation (CD34+CD43+CD45+) and to an augment of myeloid progenitors' compartment (CD33+/CD14+), which is in line with a low-risk MDS stage in childhood. The increased hematopoietic output of hiPSC-GATA2Mut lines could be related to the higher proliferation/survival of the emerging HPCs or CD33+cells. However, cell cycle analysis of CD34+ and CD33+ cells revealed no significant differences among hiPSC-GATA2 Mut and control. These data suggest a specific effect of GATA2 mutation on blood differentiation rather than on proliferation. Finally, we introduced GATA2 (R398W) mutation and selected somatic mutations (SETBP1 GoF and ASXL1 LoF) in healthy cord blood CD34+ cells by CRISPR/Cas9. Results from in vivo serial transplants of these cells into NSG-S mice suggest that germline GATA2 mutation alone is not sufficient to promote leukemia, however in combination with SETBP1+ASXL mutations a higher engraftment potential was observed in secondary transplants. Conclusions: This study broadens the genomic and clinical portrait of a Spanish cohort of GATA2 carrier. Moreover, our results show that germline GATA2 R398W and R396Q induce an accumulation of early myeloid progenitors in early stage of hematopoietic development and reveal how hiPSC-based hematopoietic differentiation represents a promising disease model to study GATA2 deficiency. Collectively, our results help to increase our understanding of the molecular mechanism underlying GATA2 deficiency.

7.
Pediatric myelodysplastic syndrome with inflammatory manifestations: Diagnosis, genetics, treatment, and outcome S. Barzilai-birenboim, A. Yanir Introduction: Inflammatory manifestations (IM) are well described in adult patients with myelodysplastic syndrome (MDS), but the presentation is highly variable, and no standardized treatment exists. This phenomenon is rarely reported in children. As more pediatric patients are hematopoietic stem cell transplantation (HSCT) candidates, the role of anti-inflammatory treatment in relation to HSCT should be defined. Aim: We report a series of five children from a tertiary center, describe the clinical presentation, molecular findings, and treatment options. Methods/Results: All patients presented with advanced MDS with blast percentages ranging 10-30%, all had severe IM. One patient had MDS secondary to severe congenital neutropenia, the other four patients had presumably primary MDS. All four were found to harbor a PTPN11 gene driver mutation, which is found in 35% of cases of juvenile myelomonocytic leukemia. The mutation was present in the myeloid lineage but not in T lymphocytes. Three children had symptoms of Behcet's-like disease with trisomy 8 in their bone marrow. All patients were treated with anti-inflammatory medications (mainly systemic steroids) in an attempt to bring them to allogeneic HSCT in a better clinical condition. All demonstrated clinical improvement as well as regression in their MDS status post anti-inflammatory treatment. All children have recovered from both MDS and their inflammatory symptoms post HSCT.

9.
Germline SAMD9L Triple-Allelic Mosaicism with Independent Segregation into Biallelic Offsprings: Molecular Support for Non-Mendelian Inheritance S. Sahoo 1 , C.Goodings 1 , M. Angeles Lillo 1 , T.Lammens 2 , J. van der Werff ten Bosch 3 , B.Strahm 4 , B. De Moerloose 2 , M.Wlodarski 1 Introduction/Aims: Single nucleotide variants (SNV) in humans are mostly biallelic; however, multiallelic variation (multiple different sequence variants observed at the same site) are estimated to affect 8% of all SNV sites. Known examples of multiallelic variation are the ABO and HLA systems. Despite multiple alleles in population, only two alleles are present at the same time in one individual, unless a chromosomal triploidy occurred. Here, building on the discovery of a family with triple allelic mosaicism at a single SAMD9L nucleotide position we aimed to decipher the underlying genetics and model the mutations in vitro and in vivo. Methods: We applied genomic methods (Sanger, whole exome sequencing, error-corrected deep sequencing, single-cell DNA sequencing (scDNAseq), digital-droplet PCR and TA-cloning), developed mutant HEK293 cell lines (via stable integration of mutant cDNA using piggyBac transposon system) and mutant inducible pluripotent stem cells (iPSC) using CRISPR-Cas9 editing, and created mouse models with constitutive Samd9L mutations. Results: We present a family with 2 germline SAMD9L mutations, where the index female patient carried heterozygous c.4534G>A (p.V1512M) mutation and presented with severe pancytopenia, del7q and marrow findings consistent with aplastic anemia, while the brother carried heterozygous c.4534G>T (p.V1512L) mutation and had refractory cytopenia with a milder clinical course. The asymptomatic mother who had normal diploid karyotype was found to be a mosaic for all 3 alleles (wildtype and 2 mutations), with wildtype allele present at~50% frequency in all tissues, while both mutant alleles showing different distribution depending on tissue origin (hematopoiesis: p.V1512M at~9% and p.V1512L at 41%, hair follicles: both at~25%). To understand why the mother had no abnormal phenotype, we performed scDNAseq in peripheral blood and identified UPD7q as a rescue event present in 18% of cells. We next modeled the mutations in vitro. Both mutations caused enhanced growth arrest in HEK293T cells, and hematopoietic differentiation of mutant iPSC yielded decreased CD45+CD18+ myeloid cells, with V1512M showing statistically more severe phenotype compared to V1512L in both cellular systems. Mice with heterozygous Samd9L V1512M mutation showed embryonic lethality exceeding 50% occurring between E9.5-E11; most newborns died by the age of 2weeks. Histopathological characterization of V1512M mice revealed hematologic (B-cell lymphopenia, anemia) and constitutional abnormalities (runtedgrowth, hydronephrosis, degeneration of thymus and ovary). In contrast, V1512L mice had no baseline phenotype. We generated a compound heterozygous intercross (V1512M/V1512L). These mice were active for 2months but begun dying between age 19 and 25weeks. The phenotype was similar (albeit milder) to V1512M mice. Further analysis of hematopoiesis in V1512L mice after different conditions (aging, chemical stress), as well as detailed characterization of marrow findings in V1512M/V1512L intercross, and survival analyses are underway.
Mono-Allelic germline disruptions of the transcription factor GATA2 result in a propensity for developing myelodysplastic syndrome (MDS) and acute myeloid leukemia (AML). How a partial loss of GATA2 enables leukemic transformation occurring years later in life, is unclear. This question is unsolved mainly as Gata2 heterozygote mice do not develop hematologic malignancies. Here we show that two different germline Gata2 mutations accelerate AML in mice expressing the human hematopoietic stem cell regulator ERG. Analysis of ERG/Gata2 het fetal liver and bone marrow derived hematopoietic cells revealed a distinct pre-leukemic phenotype characterized by enhanced transition from stem to progenitor state, increased proliferation, and a striking mitochondrial phenotype, consisting of highly expressed Oxidative-Phosphorylation related gene-sets, elevated oxygen consumption rates, and markedly distorted mitochondrial morphology. Importantly, the same mitochondrial gene-expression signature was observed in human AMLs harboring GATA2 aberrations. Similar to the observations in mice, non-leukemic bone marrows from children with germline GATA2 mutation demonstrated marked mitochondrial abnormalities. Thus, we observed the tumor suppressive effects of GATA2 in two germline Gata2 genetic mouse models. As oncogenic mutations often accumulate with age, Gata2 deficiency mediated priming of hematopoietic cells for oncogenic transformation may explain the earlier occurrence of MDS/AML in patients with GATA2 germline mutation. The mitochondrial phenotype is a potential therapeutic opportunity for prevention of leukemic transformation in these patients.

12.
Germline GATA2 mutations in familial MDS/AML: a manually curated online registry from the ERAPERMED GATA2-HuMo international consortium L. Kotmayer 1 , E. Kozyra 2 , A. Bekő 1 , T. László 1  Introduction: Pathogenic variants of the GATA2 transcription factor are one of the most common predisposing mutations of familial myeloid malignancies, including acute myeloid leukaemia (AML) and myelodysplastic syndrome (MDS). GATA2-associated familial AML/ MDS is frequently characterized by recurrent or atypical bacterial infections and lymphedema, but despite of the distinctive clinical manifestations, AML/MDS with GATA2 predisposition may go unrecognized due to incomplete penetrance and missing family history. Timely diagnosis is often further complicated by the lack of clear phenotype-genotype correlations. Although during the past few years a large number of GATA2 mutations have been identified by genetic testing, the number of sizeable cohort studies are still limited, making the clinical interpretation of these variants challenging. Aim: Our aim was to develop an online, public database of germline GATA2 mutations and associated clinical phenotypes based on international publications to aid the clinical recognition of patients harboring predisposing variants. With this database we also aim to scrutinize phenotype-genotype correlations based on pre-selected histopathological and clinical features. Methods: Familial AML/MDS cases with germline variants from 87 case reports and cohort studies published between 2009-2021 were identified via PubMed search and literature review. PubMed search was performed using a combination of keywords 'GATA2', 'deficiency', 'mutation', 'familial', 'acute myeloid leukaemia', 'myelodysplastic syndrome', 'Emberger-syndrome' and 'MonoMac'. Individual cases were uploaded to a self-developed online database, where clinical and pathological findings were classified based on the involvement of different organ systems. Accurate data submission was ensured by manual data entry and curation. Data visualization was performed using the Microsoft Power BI software. Results: To date, 412 cases with germline GATA2 variants were collected and uploaded to the database. Familial AML and MDS were described in 13.1% (54/412) and 56.8% (234/412) of the patients, respectively. Predisposition syndromes, such as Emberger-syndrome or MonoMac were identified in 10.2% (42/412) of the cases. Nearly onequarter (24.3%, 100/412) of the cohort had no symptoms associated with hematologic malignancies highlighting the incomplete penetrance of GATA2 predisposition. Clinical and histopathologic findings of symptomatic carriers were classified into 15 main categories including hematologic and immunologic features, lymphatic disorders, respiratory tract disorders and other malignancies. Twenty-six further subcategories allow for a more accurate data filtering. In total, 1,484 records are now included in the hierarchically structured dataset with a median 3 (range: 0-16) findings per case. After excluding overt myeloid malignancies, bacterial infections were found to be the most common symptoms of GATA2 deficiency. Most recurrent variants of GATA2 were T354M and R396Q, but further analyses are required to assess phenotype-genotype correlations.
Conclusions: Although recent studies have described a large number of germline GATA2 variants, assessment of the mutations' clinical significance is still an unmet need in the management of patients with familial AML/MDS. Here we developed the first online interactive, public database for the research of phenotype-genotype correlations in this entity. Manual curation and statistical analyses are still ongoing, preliminary release of the GATA2-HuMo germline GATA2 registry is expected by the end of 2021.

13.
Leukemogenesis in GATA2 haploinsufficient mice is a secondary event after bone marrow failure J. Introduction: Hematopoiesis is a tightly regulated process where transcription factors and cofactors ensure the proper proliferation, differentiation and survival of immature hematopoietic progenitor cells. Mutations in these genes can affect not only their expression levels but also their activity, disturbing hematopoiesis and leading to a high risk of developing bone marrow failure and/or leukemia. One important transcription factor governing the hematopoietic system is GATA2, which plays an essential role in the hematopoietic stem and progenitor cell (HSPC) development and differentiation. GATA2 belongs to the GATA transcription factor family and controls gene transcription from multiple target genes by binding to a consensus DNA sequence (of many enhancers, suppressors and promotors. More than 400 mutations were already described in GATA2, leading to a broad phenotypic spectrum in patients. While the hematological and non-hematological phenotypes caused by GATA2 haploinsufficiency are very variable, the risk of developing either MDS or AML is exceedingly high. However, it is still unknown whether leukemia emerges as a primary event or however, as a consequence of BM failure. Despite this high risk, the mechanisms underlying malignant transformation are poorly described. So, in order to predict, treat and prevent leukemic transformation in affected individuals, it is necessary to further understand the mechanisms underlying GATA2 syndrome. We hypothesized that GATA2-associated leukemia does not develop as a primary event but rather occurs as a secondary event after BM failure. Methods/Results: Using mice with a GATA2 haploinsufficient hematopoietic system, we generated a mouse model which developed leukemias spontaneously. We could observe that leukemias developed spontaneously after stem and progenitor cell transplantation into lethally irradiated WT hosts. About 15% of recipient mice succumbed due to leukemia. However, leukemias emerged exclusively in mice that developed bone marrow failure first (40% of all recipients). In contrast, 60% of recipients remained healthy. While no somatic alterations were detected in mice suffering from bone marrow failure alone, leukemia development was always associated with somatic mutations and/or chromosomal aberrations, indicating clonal evolution. These somatic aberrations were overlapping with those occurring in humans with GATA2 syndrome. Analysis of stem and progenitor cells isolated from recipient mice at early time points (i.e. prior development of bone marrow failure) showed profound transcriptomic changes, indicating that beside genetic alterations, epigenetic alterations might contribute to the different outcomes in our mouse model. Conclusion: We could conclude that GATA2 haploinsufficiency predisposes to bone marrow failure in mice, paving the way for secondary leukemia.
Introduction/Aim: Myelodysplastic syndromes (MDS) in childhood consist in a heterogeneous group of clonal hematopoietic disorders often associated with dysmorphic characteristics, genetic and MDSpredisposition syndromes. We present the 6-year experience on MDS, of the Greek study group for MDS/JMML/SAA, as a member of the EWOG MDS/SAA European working group. Methods: Thirty-two (32) MDS patients diagnosed during the period 10/2015 -7/2021 are presented. Appropriate detailed clinical, laboratory, cytogenetic and pathology investigations were performed. Diagnostic samples and clinical data were reviewed in national reference centers. The patients were treated in the paediatric hematologyoncology departments and the pediatric BMT-unit of Greece, and were consequently registered to EWOG-MDS. Results: The 32 patients (average age: 9,7 years, 20/32 male) were classified according to EWOG-criteria as following: i) Refractory Cytopenia of Childhood (MDS-RCC) 27/32 (2 out of 27 were secondary post chemotherapy due to prior malignancy), ii) MDS with excess blasts (MDS-ΕΒ) 5/32. At diagnosis: median WBC 3435/μl, median Hb 8.8 gr/dl and median PLTs 38000/μL. The bone marrow smear morphology at diagnosis revealed dysplasia of one/two/three hemopoietic lineages in 2/7/23 patients respectively. Seven patients presented with dysmorphic features or/and genetic stigmata (mainly in face and extremities). Cytogenetics: monosomy 7 in 3/32 patients (all RCC), trisomy 8 in 2/32 and inv17 in 1/32. The three patients with monosomy 7 and the one with inv17 , all RCC, were positive (4/32) for SAMD9/SAMD9L mutations. Twelve (12/32) patients are under watchful waiting (one of twelve with monosomy 7 and SAMD9L mutation), 18/32 underwent stem cell transplantation (SCT) due to MDS-ΕΒ (5/18), monosomy 7 (2/18), inv17 (1/18), or increase in transfusion needs (10/18 with MDS-RCC), while in 2/32 SCT is pending. Mean time from diagnosis to SCT was 5,5 months. Thirty (30/32) patients are alive (OS 94%). Out of the 18 transplanted patients 14 are in remission without transfusion needs, 1 is in remission after 2nd SCT (due to relapse) and 1 in search for a new donor after graft rejection.
Two patients died, one due to refractory disease after second SCT and one due to severe infection and cGvHD. Conclusions: MDS in childhood pose difficulties for accurate diagnosis and appropriate treatment. Participation in national collaborative groups and scientifically established international working groups such as EWOG-MDS/SAA contributes: a) to understand in depth the pathophysiology of diseases, b) to clarify, on the basis of molecular, genetic and clinical criteria, which patients benefit from SCT and which do not, and c) to investigate the optimal SCT timing or other appropriate therapeutic approach, at a time.

15.
Favorable outcome for young patients receiving allogeneic stem cell transplantation for RUNX1 germline mutation related diseases B.Strahm 1 , G. Introduction: Autosomal dominant mutations in RUNX1 gene are rare and have been associated to a familiar platelet disorder with a 35% risk of progression to acute myeloid leukemia (AML) in a lifetime and median age at transformation of 33 years. Haematopoietic stem cell transplantation (HSCT) is the treatment of choice for these patients, and a higher risk of relapse post SCT has been described in adults with RUNX1 related AML. Families harbouring specific mutations seem to be more prone to develop AML and a degree of anticipation is observed in pedigrees. Children who harbour a germ line mutation represent a yet to be studied group of patients, for which risk of malignant evolution as well as outcomes and risks of haematopoietic stem cell transplantation (HSCT) are unknown. Aims: This multicentre retrospective study describes HSCT characteristics and clinical outcomes of patients diagnosed with RUNX1 germline mutations during childhood who developed haematological diseases (from isolated thrombocytopenia to AML) and underwent HSCT by the age of 20. Methods/Results: We identified 25 patients from 23 families and 9 European countries meeting inclusion criteria. Median age at diagnosis was 7.9 years. 15 patients had a known familiar history of thrombocytopenia/RUNX1 related disease. At diagnosis 11/25 patients presented with thrombocytopenia or refractory cytopenia of childhood (RCC). With time 2/11 progressed to myelodysplasia with excess of blasts (MDS-EB) and 2/11 transformed into AML. 10/25 patients had MDS-EB at diagnosis, 2/10 progressed to MDS EB in transformation (MDS-EBt) and 2/10 to AML. For 5/25 patients AML was the presenting diagnosis. 4/25 patients had a RUNX1 mutation detected retrospectively after HSCT. 16 patients had normal karyotype, 2 had monosomy 7, 1 each had trisomy 8, deletion 5q, t(8;12) and a hyperdiploid karyotype. 2 patients (1 AML, 1MDS EBt ) received azacytidine, 8 patients (5 AML and 2 MDS-EB) received a combination of chemotherapy regimen or chemotherapy and azacytidine before HSCT. At the time of HSCT 5/8 patients who had developed AML were not in remission. Patients received their HSCT from MSD (5), MUD (12), MMUD (7) and MMFD (1). Bone marrow was the favourite source of stem cells (15/25) , followed by peripheral blood (7/25) and cord blood (3/25). 22/25 patients received a myeloablative conditioning regimen, which was busulfan (14) treosulfan (6) or total body irradiation based (2). 16/25 patients received serotherapy as part of the conditioning regimen. There were no primary graft failures. One patient diagnosed with MDS-EB died from transplant related mortality on day +27 (combination of severe hyperacute graft versus host disease and veno occlusive disease), 22/25 pts are alive with a median follow up of 3.5 years. Of note 4/5 patients who presented AML not in remission prior to HSCT (median blasts prior to SCT 25%) achieved a complete continuous remission following HSCT. Three patients relapsed following HSCT and underwent a second HSCT. One is alive in CR, 1 died from further disease relapse and 1 from transplant related mortality following second SCT. Conclusion: Stem cell transplantation is well tolerated in children with RUNX 1 germline mutation and outcomes are favourable even in patients transformed to AML. Natural history of the disease with identification of patients who are more prone to transform during the paediatric age will be important to select candidates for preemptive approach.

16.
Clinical application of transactivation assays for precise classification of germline RUNX1 missense variants associated with predisposition to MDS and AML M. Introduction: Familial platelet disorder with associated myeloid malignancies (RUNX1-FPD, FPDMM, FPD/AML, MIM 601399) is typically characterized by thrombocytopenia, functional platelet defects, and ã 40% risk to develop hematological malignancies, mainly MDS and/ or AML. Leukemic transformation requiring secondary somatic events can occur from early childhood to late adulthood. RUNX1-FPD is caused by heterozygous pathogenic germline variants in RUNX family transcription factor 1 (RUNX1). Based on present gene variant classification guidelines, rare RUNX1 missense variants must frequently be classified as variants of uncertain significance (VUS). Aim: To overcome this challenge, we have previously reported the development and application of a set of assays to characterize the functional impact of RUNX1 variants addressing its heterodimerization with CBFB, its phosphorylation, and its ability to activate transcription. In the present study, we evaluated the applicability of our transactivation assays to investigate RUNX1 variants in different regions of the protein by studying a set of nonsense variants distributed from the N-to the C-terminus of the protein. In addition, the transactivation assays were applied to investigate 11 RUNX1 variants detected in patients with thrombocytopenia, MDS, AML, or as secondary finding to independently validate our assays. Methods: We analyzed the ability of RUNX1 variants to activate transcription of RUNX1 target genes (i.e., CSF1R, ETV1, and MYL9) by applying luciferase reporter assays in HEK293T and HEL cells. As controls, wild-type (WT) RUNX1b, the known pathogenic variant Arg139Gln, and the benign variant Leu29Ser were investigated in parallel. Functional data were integrated in ACMG/AMP variant classification following the current expert guidelines of the ClinGen Myeloid Malignancies variant curation expert panel (MM-VCEP). Results: We showed that the majority of RUNX1 missense variants detected in RUNX1-FPD can functionally be addressed by our transactivation assays. Following the MM-VCEP expert guidelines, two VUS were reclassified to likely pathogenic and analyses supported the (likely) pathogenic classification of two additional variants. We demonstrated functionality of four VUS, but reclassification to (likely) benign was not possible following current classification guidelines. Hence, RUNX1-FPD suspicion was confirmed in three families with RUNX1-FPD-specific medical history. For variants identified in non RUNX1-FPD-typical families, no functional defects were observed. Conclusions: In the present study investigating 11 variants, the clinical utility of our transactivation assays was independently validated and illustrated in the context of seven index patients. Even though most missense variants detected in the context of RUNX1-FPD can functionally be addressed by our transactivation assays, additional assays are required to investigate C-terminal RUNX1 missense variants. Careful reevaluation of current variant classification guidelines should be considered to allow future reclassification of VUS to (likely) benign, if sufficient evidence for its functionality was observed. Applying functional assays supports final RUNX1 variant classification and can be essential for precise genetic diagnosis. It facilitates translation of genetic data into personalized medical care for index patients and their relatives who are at risk. MonoMAC and Emberger syndromes are GATA2 haploinsufficiency syndromes. These syndromes are characterized by monocytopenia, neutropenia, B-, dendritic-and NK cell lymphopenia and up to 80% of patients with innate GATA2 mutations develop myelodysplastic syndrome (MDS)/ acute myeloid leukemia (AML). Gata2 level is a key factor in embryonic hematopoietic stem cell (HSC) generation and maintenance. To study the effect of Gata2 haploinsufficiency on the mechanism of malignant transformation, we use both mouse and zebrafish Gata2 haploinsufficiency models. Gata2 heterozygous mutant mice survive to adulthood without developing any visible bone marrow phenotype but their phenotypic HSCs are lower in frequency, more proliferative and transcriptionally more committed to differentiation compared to WT. Furthermore, they show increased DNA damage indicating these HSCs experience proliferative stress. Upon aging and transplantation these HSCs repopulate the bone marrow, but are impaired in their contribution to hematopoiesis resulting in B-cell lymphopenia. In zebrafish, two orthologues of Gata2 exist: gata2a and gata2b, of which lineage analysis revealed that all hematopoietic cells once expressed Gata2b. We found that complete loss of gata2b results in a severe reduction in myeloid differentiation. Furthermore, gata2b +/zebrafish develop dysplasia of both erythroid and myeloid lineages as is found in patients with GATA2 mutations. Using single-cell RNA sequencing we aim to identify early onset markers of dysplasia and unveil the mechanism of malignant transformation. Our mouse and zebrafish models recapitulate characteristics of the GATA2 deficiency syndromes and it can give new insight into the pathophysiology of this disease.

18.
T lymphocyte profile and dynamics in patients with hepatitisassociated bone marrow failure M. Novakova 1#  #These authors contributed equally. Introduction: Hepatitis-associated bone marrow failure (HABMF) is a rare variant of bone marrow failure, which is typically diagnosed 2-3 months before failure itself after attack of hepatitis in patients that are seronegative for known hepatitis viruses. Although very little is known about its etiology, an autoimmune mechanism is presumed based on clinical response to immunosuppressive therapy (IST) and reports of oligoclonal T-lymphocyte expansion and skewing of CDR3 region length. Aim: Our aim was to uncover the role of T cell lymphocytes in pathogenesis by analysis of their mature stage, activation and T cell receptor beta repertoire (TRBV). Methods: We included 21 pediatric patients diagnosed with HABMF in Czech Republic in 2004-2021 and 1 patients diagnosed in Slovakia. The prevalence of HABMF in pediatric patients with BM failure in Czech Republic in this period (n=121; excluding patients with known genetic cause) was 17%. Interestingly, 5 out of total 22 HABMF patients had histopathological findings consistent with refractory cytopenia of childhood (RCC). We performed flow cytometry immunophenotyping in bone marrow (BM) and peripheral blood (PB) samples at diagnosis and at day 120 after initiation of immunosuppressive therapy (IST). TRB sequencing was performed following the SOP developed by the EuroClonality-NGS working group with DNA input normalized to the equivalent of 20 000 CD3+ cells per sample based on flow cytometry data. Bioinformatic analysis was performed with ARResT/Interrogate. WES was performed on Illumina NextSeq 500 and libraries prepared using the Agilent SureSelectXT Human All Exon V6+UTRs kit. Results: We found significant activation of CD8pos T cells based on the expression of HLA-DR and a decrease of naive forms CD27posC-D45RApos compared to non-HAAA RCC and AA patients. TRB repertoire analysis revealed expanded clones (>5%) of T lymphocytes in 8 diagnostic samples. The CDR3 sequences of expanded clones differed among patients. In one HABMF patient the expanded clone comprised 39% of all TRBV sequences and immunophenotypically belonged to CD8pos cells with highly abundant senescent CD57pos T cells. WES analysis did not reveal any previously published variant in genes associated with immune dysregulation, immunodeficiency or bone marrow failure. Rare (0.01% in gnomAD) heterozygous variants of unknown significance were found in ACD (in 2 patients), ASXL1, CXCR4, IFIH1 and SOS1 genes. Sixteen patients were treated according to standard European Working Group on MDS/SAA protocols with anti-thymocyte globulin (horse n=7; rabbit n=9), corticosteroids and cyclosporin A. Eight patients reached complete remission on IST by day 120 and 8 patients with insufficient response underwent MUD-HSCT. We did not observe any correlation of T cell populations at diagnosis or day 120 of IST (activated and exhausted T cells) and clinical response. Similarly, the presence of an immunodominant clone or oligoclonal TRB repertoire with lower diversity at diagnosis had no significant impact on the outcome of these patients in our cohort. Conclusions: In conclusion, we present the largest cohort of pediatric patients with HABMF analyzed by next-generation sequencing and flow cytometry so far. Even though our results show significantly increased activation of CD8pos T cells and presence of expanded clones, we did not confirm that these factors would be useful and significant for the prediction of treatment outcome. Large number of our patients does not fully respond to IST and require HSCT. Supported by AZV NV18-07-00430, NU20J-07-00028 a GAUK 534120

19.
The role of the telomere length measurement in children and adolescents with aplastic anemia enrolled in the EWOG-SAA study A. Introduction/ Aims: Although telomere length (TL) analysis is useful to support the diagnosis of dyskeratosis congenita (DC), previous reports have shown that some patients with aplastic anemia (AA) and other bone marrow failure syndromes have short TL as well. The aims of this study were to evaluate the role of TL analysis in AA by comparing TL in AA patients with healthy controls and DC patients, evaluating the incidence of short TL in AA patients and analyzing the correlation between TL and outcomes after immunosuppressive therapy (IST). Methods: Among 293 AA patients registered in the EWOG-SAA-2010 study between 2011 and 2020, TL in DNA extracted from granulocytes in bone marrow (BM, n=153) or peripheral blood (PB, n=40) was measured in 193 subjects using quantitative real-time PCR (adapted from RM Cawthon 2009). The results were compared with healthy pediatric controls (n=148) and DC patients (n=47). During the study period, 150 patients (M/F=92/58, med. age 10, range 1-18 years) received IST with anti-thymocyte globulin (ATGAM; n=110, thymoglobulin; n=40) and cyclosporine. The disease severity was very severe AA (VSAA), SAA and non-SAA (NSAA) in 127 (85%), 19 (13%) and 4 patients (3%), respectively. Variables associated with IST response were evaluated. Results: The median TL in BM granulocytes was significantly shorter in AA patients than in healthy controls: 1.10 (range 0.21-5.22, n=153) vs 1.24 (range 0.78-2.75, n=76, p<0.01), but significantly longer than that of DC patients (0.55, range 0.30-0.81, n=26, p<0.01). There was no significant difference in TL in PB granulocytes between AA group and healthy control (median 1.09, range 0.23-1.97, n=40 vs 1.16, range 0.59-2.97, n=72, p=n.s.) but TL in PB granulocytes was significantly shorter in DC patients (0.41, range 0.26-1.08, n=21, p<0.01). Short TL (<1.P) was noted in 22/153 BM samples (15%) and 2/40 PB samples (5%) in AA patients, but none of them had DC phenotype. In 14 of these patients, mutations in most common genes (DKC1, TINF2, RTEL1, TERT and TERC) were excluded (no data in 10 patients). One-hundred-fifty AA patients were treated with IST; 55 patients (37%) responded to IST at day 180, 59 (n=39%) had no response, 33 patients (22%) received early HSCT, 2 patient received second IST and one patient died early. The 5-year overall and failure-free survivals after IST were 92% and 27%, respectively (failures defined by no response, relapse, clonal disease, secondary therapy and death). Significant short TL (<1.P) was observed in 15 of 110 evaluated patients (14%) and only one of them responded to IST. In univariate analysis, low absolute neutrophil count (ANC), short TL and thymoglobulin were associated with poor response. In multivariate analysis, low ANC was the most significant risk factor for poor response (p= 0.015), followed by long interval between diagnosis and IST (p= 0.037), while impact of short TL on response was statistically not significant (p=0.08) probably due to a small number of patients with short TL. Conclusions: Although TL in BM granulocytes of AA patients was significantly shorter compared to healthy controls, clinical and genetic data did not indicate a high suspicion for unrecognized DC. However, further comprehensive genetic analysis is necessary to rule out mutations in any of the known or novel genes associated with DC. Furthermore, patients with short TL had a poor response to IST possibly indicating that short TL reflects a reduced stem cell reserve. These results might be used for improved therapy stratification in pediatric AA Introduction: Age-related B cells (ABCs) represent a small subpopulation of B cells. These cells express high levels of CD11c and CD19, they are CD21 negative and express T-bet in the nucleus. These T-bet+ ABCs are found increased in patients with autoimmune diseases (i.e. systemic lupus erythematosus, rheumatoid arthritis, and multiple sclerosis). Stimulation of B cells with antigens, Toll-like receptors and IFN-g leads to the formation of ABCs, which in turn "talk" to the T cells and stimulate them. Stimulation of T cells leads to IFN-g production, and this IFN-g may lead to further induction of T-bet expressing B cells (ABCs). Aplastic anemia is a rare disease characterized by immune dysregulation. T cells in aplastic anemia are characterized by various intrinsic defects leading to increased IFN-g levels and Fas-mediated apoptosis of hematopoietic stem cells. We and others, have previously shown that the transcription factor. Aim: In this study we wanted to examine the expression of ABCs in patients with aplastic anemia. Severity of aplastic anemia was defined based on standard criteria. We isolated peripheral blood mononuclear cells (PBMCs) from patients with aplastic anemia (n=11, age 8-50 years, 5 patients were children/adolescent) and five healthy, age-matched controls. Written informed consent was obtained from all study subjects. Cells were stained with the surface markers CD11c, CD19 and CD21, and subsequently analyzed using flow cytometry. Methods/Results: Patients with aplastic anemia at presentation showed increased numbers of circulating ABCs compared to healthy controls (2.71 ± 0.39% vs 0.43 ±0,09% respectively, p=0.002). (Fig. 1). Although the sample of patients is small to reach definitive conclusions, the number of circulating ABCs was directly correlated with aplastic anemia severity. Conclusions: Our preliminary results suggest that aplastic anemia patients show expanded ABCs compared to age-matched control subjects, and these numbers are related to disease status. Further analysis of a larger pool of subjects is underway along with the examination of the specific transcription factor Bcl-6, that is implicated in T-bet expression in ABCs. These results will reveal the role of ABCs in the immune pathogenesis of aplastic anemia. Introduction: Inherited bone marrow failure syndromes (IBMFS) occur in approximately 5%-30% of patients with bone marrow failure (BMF). Several investigators have demonstrated that telomere length (TL) is excessively short in patients with idiopathic aplastic anemia (AA) and non-DC IBMFS, including FA, SDS, and DBA. To assess the diagnostic value of TL, we measured TL in 133 patients with BMF and compared it in patients with DC, non-DC IBMFS, and AA. Methods: We retrospectively studied 133 patients (68 male and 65 female) with BMF in Japan between 2013 and 2018. We collected peripheral blood samples at diagnosis from all patients, measured TL from peripheral blood lymphocytes, and performed targeted sequencing analysis covering 184 genes associated with IBMFS. We divided 133 patients into three groups: DC, non-DC IBMFS, and AA. Results: Using targeted sequencing, we detected from 24 patients (18%) 31 pathogenic variants (3 nonsense, 14 missense, 5 frameshift, 4 splice site, and 4 deletion) of known causative IBMFS genes, including TINF2 (n = 6), TERT (n = 3), FANCA (n = 6), FANCG (n = 3), RPL5 (n = 2), RPS19 (n = 1), RPS17 (n = 1), SBDS (n = 1), and BLM (n = 1). Homozygous mutations were found in 3 patients (2 in FANCC and 1 in FANCA), compound heterozygous in 4 patients (2 in FANCA and 1 each in FANCG and SBDS), hemizygous in 3 patients in FANCA, and heterozygous in 14 patients (6 in TINF2, 3 in TERT, 2 in RPL5, and 1 each in RPS17, RPS19, and BLM). Out of the 133 patients, 11 were diagnosed with DC (8%), 15 with non-DC IBMFS (11%), and 107 with AA (81%). Of the 11 patients with DC, 9 were genetically diagnosed (6 in TINF2 and 3 in TERT), and those without diagnostic genetic mutations were diagnosed following clinical diagnostic criteria. The 15 non-DC IBMFS cases consisted of 9 FA, 4 DBA, 1 SDS, and 1 Bloom syndrome. Physical anomalies were observed in 11 of 15 (73%) patients. We compared the clinical characteristics of patients with DC, non-DC IBMFS, and AA. Median TL in the patients with DC, non-DC IBMFS, and AA were −3.50 SD (range, −5.73 to +0.83 SD), −1.89 SD (range, −4.74 to +2.05 SD), and −0.84 SD (range, −4.27 to +4.00 SD), respectively. Patients with DC showed significantly shorter TL compared with those with non-DC IBMFS (P = 0.031) and AA (P < 0.001). Furthermore, patients with non-DC IBMFS tended to show shorter TL than those with AA (P = 0.096). To validate the efficacy of TL measurement in diagnosing DC and IBMFS, receiver operating characteristic curves identified two cut-off values with the optimum sensitivity and false positive rate (1-specificity) combination, <−2.19 SD and < −1.71 SD, defined as "very short TL" and "relatively short TL," respectively. For the diagnosis of patients with IBMFS, the TL cut-off value at −1.71 SD (relatively short TL) yielded relatively high negative (0.921; 95% confidence interval [CI], 0.873-0.958) and moderately positive predictive values (0.432 [95% CI; 0.333-0.505]). Of the total cohort, 44 patients (33%) were classified with "relatively short TL", which was significantly more frequent (P < 0.001) in DC (10/11, 91%) and non-DC IBMFS (9/15, 60%) than in AA (25/107, 23%). Conclusions: This study confirms that a relatively short TL was found in a significant proportion of patients with DC and non-DC IBMFS, indicating the clinical diagnostic value of TL measurement in identifying patients who need further testing, particularly comprehensive genetic analysis.  Introduction: Human telomere biology disorders (TBD) are hereditary multisystemic disorders that are caused by premature telomere shortening due to loss-of-function mutations in telomereassociated genes, however many patients remain without a genetic resolve. Here, we characterize germline heterozygous missense variants in RPA1 identified in four unrelated probands presenting with varying phenotypes of TBD. RPA1 is an essential single-strand DNA-binding protein involved in DNA replication and repair.

Increased Age-related B cells in patients with
Aim: The aims of our study were to delineate the clinical phenotype of individuals with RPA1 variants, characterize the genetic landscape and mechanisms of somatic genetic rescue, and asses if RPA1 germline mutations are deleterious in biochemical and cellular systems.
Methods: All patients underwent detailed evaluation and had negative workup for inherited bone marrow failure (BMF) syndromes and TBD associated genes. Genomic studies (whole exome sequencing, ultradeep sequencing, single cell multi-omics, RNAseq, ddPCR-based haplotype phasing), telomere length assessment, biochemical assays (DNA and telomere binding studies, telomerase assay), and in vitro modeling using induced pluripotent stem cells (iPSC) were performed. Results: We found four unrelated TBD patients to carry three distinct (two confirmed de novo) missense heterozygous mutations in RPA1. All three mutations cluster to the DNA binding domain A (DBD-A), which has previously been shown to be the central functional domain involved in binding to single stranded DNA (ssDNA). The patients had heterogenous clinical manifestations unified by short telomeres and TBD features, including BMF, myelodysplastic syndrome, lymphopenia, pulmonary fibrosis, and mucocutaneous triad. Using forester resonance energy transfer (FRET) DNA binding assay, we show increased binding of germline RPA1 mutants to ssDNA and telomeric DNA, consistent with a gain-of-function effect. To understand the effect of RPA1 variants on telomere length regulation and hematopoietic development, we used CRISPR/Cas9-mediated mutagenesis to introduce patient p.E240K mutation (denoted as RPA1 E240K ) in a healthy donor iPSC line (RPA1 WT ). RPA1 E240K iPSC model exhibited premature telomere shortening at iPSC and hematopoietic progenitor level, as well as reduced hematopoietic differentiation. Finally, we identified somatic rescue events in hematopoiesis of patient with p.E240K mutation, which correlated with long term stable clinical phenotype. These included a somatic RPA1 p.K579* mutation (in cis with p.E240K) causing RNA degradation, and an independent clone with uniparental isodisomy 17p resulting in duplication of wild type RPA1 allele. Conclusions: We identified RPA1 mutations to be associated with telomere shortening in humans, which calls for careful consideration of RPA1 missense variants in the workup of patients with TBD phenotypes. The mutations are likely gain-of-function with increased DNA binding capacity, and negative selective pressure resulting in somatic rescue in hematopoiesis. We speculate that germline RPA1 alterations may be more common in human disease, given that somatic RPA1 mutations occur in~1% of cancers. Additional efforts are needed to not only find further pathogenic RPA1 mutations but to also elucidate the role of RPA1 in human telomere biology. Introduction: Next-generation sequencing (NGS) technology has revolutionized the genetic diagnosis of rare diseases, and variants in an increasing number of genes are known to cause non-malignant hematologic disorders. Defining a workflow for reliable and sensitive genetic diagnoses in these patients is essential to optimize patient management and treatment. In many cases, genetic diagnoses can be made with predefined sequencing panels. Other patients require whole-exome (WES) or whole-genome sequencing to define the causative genetic variant. However, for a significant proportion of patients, a clear genetic etiology cannot be identified. This may be due to a non-monogenic etiology but is also accounted for by limitations in the identification and interpretation of candidate pathogenic variants, including incomplete knowledge of gene-disease associations or technical shortcomings. As both the knowledge pool of molecular disease causes and bioinformatic means are rapidly advancing, increasing numbers of previously unsolved cases can be solved. Thus, it is imperative to revise previously unsolved sequencing data rigorously and constantly. Within the European Ribosomopathy Consortium (RiboEurope), we have conceptualized the strategic reanalysis of NGS data to identify the genetic etiology in previously unsolved cases.
Aims:To define and optimize a genetic diagnostic strategy for rare non-malignant hematologic diseases. Methods: 189 patients were analyzed using comprehensive panel sequencing (variable panels of 300-600 genes; 119 patients) or a more recently implemented pipeline based on WES followed by the targeted analysis of a constantly updated virtual gene panel (currently including more than 600 genes; 70 patients). In addition to SNPs and small indels, the workflow included copy number variant calling, deep-learning-based splice variant prediction, HPO terms and integration of publicly available databases (gnomAD). Candidate variants meeting basic criteria compatible with pathogenicity, such as a plausible allele frequency and effect on the gene product, were individually evaluated using literature, mouse phenotype data, prediction tools, and expert opinion. If, in the WES-based approach, no candidates were identified, the analysis was extended to include exomewide data. Regular updates and additions to the individual tools, annotations, and the virtual panel as well as the subsequent reanalysis of unsolved cases were an integral part of the pipeline. Results: We have successfully implemented a comprehensive, everevolving workflow for the diagnosis of patients with non-malignant hematologic diseases, integrating state-of-the-art NGS technology, bioinformatics tools, constantly updated clinical and molecular data, and expert opinion. The total solve rate in the cohort amounted to 40%, and was increased among patients analyzed with the more recently implemented WES-based pipeline, which allows a higher flexibility in the data analysis and reanalysis (43% as compared to 39% in physical panels). Due to the ongoing constant revision of the cohort using updated tools and reference information, the solve rate is expected to further increase in the future.
Conclusions: With this study, we present a comprehensive pipeline that increases the diagnostic capacities of NGS data for monogenic non-malignant hematologic disorders.

26.
A human model of clonal evolution in Fanconi anemia W. . Additional somatic mutations result in further dysregulation of self-renewal resulting in frank acute myeloid leukemia (AML), a significant source of morbidity and mortality in FA. This process of clonal evolution is a difficult clinical challenge in FA given the sensitivity of FA patients to genotoxic chemotherapy agents required to treat MDS and AML. Therefore, improved understanding of the process of clonal evolution of failing HSCs toward pre-leukemic and leukemic stem cells would benefit patients with FA and BMF more broadly through the derivation of improved surveillance strategies to guide the use of curative allogeneic HSC transplantation as well as the delineation of novel molecular vulnerabilities in FAassociated MDS and AML. Aims: To understand clonal evolution in FA in a human disease model. Methods: In this study, we used gene editing of human induced pluripotent stem cells from FA patients to gain understanding of clonal evolution in FA. We performed directed differentiation of these iPSCs to hematopoietic progenitor cells for analysis of DNA repair and blood stem cell properties.
Results: We find that somatic mutations commonly occurring in MDS and AML confer quiescent FA hematopoietic progenitors with aberrant self-renewal and that the incorporation of an activating oncogene mutation can trigger leukemogenesis. Gene expression analysis showed that FA MDS and AML cells express ectopic stem cell transcriptional signatures and lose signatures of inflammation and terminal differentiation typical of FA hematopoietic progenitors. Comparative functional analysis of human FA progenitors, MDS-like cells, and leukemia cells revealed that cell cycle checkpoints hyperactivated in progenitors were blunted in MDS and AML cells as a result of somatic mutations, and that AML cells continue to show genomic instability and accumulation of DNA damage despite ongoing proliferation.
Conclusions: Overall, these findings provide new insight into clonal evolution of FA hematopoietic progenitors to MDS and AML in a renewable human system and provide a platform for therapeutic discovery in myeloid neoplasms associated with FA. Our work provides a human based model system that can be adapted to the study of clonal evolution and leukemogenesis in many forms of pediatric BMF.  Methods: All patient samples had been obtained with informed consent, and the study had been approved by the local ethics committee. Genetic testing was done by Sanger sequencing for single gene analysis or target next-generation sequencing (NGS) on the MiSeq/ NextSeq (Illumina, USA) using Bone Marrow Failure Syndrome custom gene panel.

A Novel Classification of Hematologic Conditions in Patients with
Results: All 3 pts had profound macrocytic hypoproliferative anemia diagnosed at birth (patient 1), at 2 months (patient 2) or at 7 months (patient 3) requiring transfusions of packed red blood cells (RBCs). Bone marrow (BM) investigation at the first evaluation revealed normocellular marrow with no increase in blasts and red cell hypoplasia. The primary investigation ruled out the most common inherited bone marrow failure syndromes. Considering the age of manifestation, a preliminary diagnosis of Diamond-Blackfan anemia (DBA) was suspected. With NGS, patient 1 was found to have DNMT3 germline mutation p.(Arg882His), which is a hotspot mutation in the catalytic domain and linked to Tatton-Brown-Rahman syndrome. Besides anemia, a mild thrombocytopenia (57-108 X 10 9 L) was revealed at the age of 9 months. At age 12 months, he was started on prednisolone and his hemoglobin (Hb) level normalized. Prednisolone therapy was discontinued at age 6.7 years, and at age 11 years, Hb level started to decline again (7.3 g/dl). The BM aspirate showed a hypocellular marrow without an increase in blasts, marked myeloid dysplasia and a decrease in erythroid precursors and megakaryocytes. Bone marrow karyotype was normal. At age 13 years, patient had a disease progression in the form of profound neutropenia, high transfusion dependence of RBCs and platelets, and a reduction in number of B-and T-cells. Hematopoietic stem cell transplantation is planned. The non-hematological phenotype of patient 1 included mild dysmorphic facial features, excessive weight gain and febrile seizures. In patients 2 and 3, activating TP53 mutations were found: p.Ser362Alafs*8 and c.1100+1G>C, which cause truncation of the protein resulting in the loss of 32 residues from the C-terminal domain. For the diagnosis DBA both patients were started on L-leucine at the age of 2 years. They became transfusion independent and at the time of writing continue to do so 2.5 and 7 years after starting L-Leucine therapy. The non-hematological phenotype of both cases included mild microcephaly and short stature; in patient 2, cardiac anomalies and hepatic cavernous hemangioma with portal hypertension and moderate splenomegaly were also observed. Conclusion: The continuing effort to analyze clinical and genetic data on patients with unresolved bone marrow failure disorder is of great especially importance. It is the prerequisite for personalized surveillance and counseling of patients and carriers Introduction/Aims: Allogeneic hematopoietic stem cell transplantation (HSCT) is the only curative therapy for most patients with juvenile myelomonocytic leukemia (JMML). Relapse remains the major cause of treatment failure after HSCT, and therapy options for relapse are limited. The aim of this study is to analyze the outcome of JMML patients who received a second HSCT for relapse after the first grafting procedure. Methods: Seventy-eight patients registered in the EWOG-MDS studies received a second HSCT for relapse following a first HSCT performed between 07/1988 and 01/2020. After excluding 10 patients with insufficient data, 68 patients were included in the analysis. The median age at diagnosis was 3.0 years (0.2 -15.1) and at second HSCT 4.6 (1.0 -15.8) years. The intervals between first HSCT and relapse and between first and second HSCTs were 247 (15 -788) days and 352 (35 -907) days, respectively. A PTPN11 driver mutation was known in 42 patients (61%). Most patents (n=49) did not receive any anti-leukemic therapy prior to the second allograft, while 10, 9, 13 and 2 patients were treated with low-dose chemotherapy, azacitidine, donor-leukocyte infusion (DLI) and splenectomy, respectively. Only one patient (treated with azacitidine) was in clinical remission before the second HSCT. The majority of patients (n=54) had received a busulfan/cyclophosphamide/melphalan (Mel) regimen for first HSCT (others n=14). For second HSCT a total body irradiation based (n=28), treosulfan/fludarabine (Flu)/thiotepa (TT) (n=14) and Mel/Flu/TT (n=9) regimens were applied (others n=9). For second HSCT, 16 patients were transplanted from an HLA-matched sibling donor (MSD), 9 patients from a haplo-identical family donor and 43 patients from an unrelated donor (UD). The same donor was used for both first and second HSCTs in 31 children. Stem cell source was bone marrow, peripheral blood and cord blood in 37, 26, and 6 patients, respectively. Results: Engraftment was achieved in 61 of 68 patients, 1 patient encountered a secondary graft failure. The incidence of acute II-IV/III-IV and chronic graft-versus-host disease (GVHD) were 52% / 22% and 37%, respectively. With a median follow-up of 7.7 years, the 5-yearoverall survival, disease free survival, cumulative incidences of relapse and non-relapse mortality after second HSCT were 40%, 36%, 41% and 23%, respectively. Patients with older age at second HSCT (> 3 years) and early relapse (< 180 days) after first HSCT tended to have a worse outcome, while mutational subtype, type of donor, change of donor for second HSCT, and conditioning regimen did not affect outcome. Relapse after the second HSCT was the main cause of death (n = 23), while 17 children succumbed due to transplantationrelated causes; graft failure (n = 4), acute and chronic GVHD (n = 3 and 4), early transplant toxicity with multi-organ failure (n = 1), infections (n= 4) and chronic lung disease (n = 1). Poor prognostic features of age > 4 yrs/raised HbF and low platelet count were present across all 4 groups. Azacitidine was recommended as the preferred treatment option for all newly diagnosed children with JMML, 6/17 (35%) children received AZA alone, 11/17 received AZA sequentially with other therapies. In 7/17 (41%) children their clinical progression dictated the need to escalate to more aggressive chemotherapy regimens like FLA (Fludarabine/Cytarabine). Only 9/17 (52%) received a minimum of 3 cycles of Azacitidine. The mean interval from diagnosis to transplant was 6 months. Based on the response criteria described by Niemeyer et al, we achieved complete clinical remission in 2 patients (12%), stable disease in 2 patients (12%), partial clinical remission in 7 patients (41%), progressive disease in 5 patients (29%) and 1 patient with overt relapse (6%). All the patients underwent SCT. Out of 17 patients, 15 are currently alive with an OS of 88% with a median follow up of 5 years. Four patients relapsed (2 cPR, 1 cSD and 1 cRel), the event free survival is 82% (death related-disease progression, relapse, sepsis in the context of severe GVHD).
Conclusions: Treatment options recommended prior to transplant in JMML are variable. All 17 patients inour cohort received Azacitidine. However only 52% of the cohort received a minimum of 3 cycles of therapy. Escalation to aggressive chemotherapy was required in 41% of the cohort. Irrespective of the treatment given, 65 % of the cohort achieved cCR, cSD and pCR. Relapses occurred across all mutational subgroups. Interestingly, the 4 patients who relapsed did not have progressive disease at the time of HSCT. We did not measure mutation specific VAF at the time of HSCT. Our data highlights the challenges to recommending a uniform treatment strategy prior to transplant in JMML in view of the variable clinical course in these patients. There is an urgent need to standardize biomarkers of disease response prior to HSCT, which predict for post-transplant relapse. In the future, novel therapies in JMML will need to address the variability of disease progression and assessed using validated prognostic biomarkers of disease relapse.

32.
Using BH3 Introduction: Juvenile myelomonocytic leukemia (JMML) is a highly aggressive and life-threatening myeloproliferative neoplasm of early childhood. It is driven by oncogenic RAS signaling in addition to epigenetic deregulation. The only curative treatment until now is allogeneic stem cell transplantation (HSCT), bearing risks like graft failure or relapse. The DNA methyl transferase inhibitor Azacitidine is effective against JMML but unfortunately not all patients respond to this treatment. Therefore, alternative therapeutic strategies are urgently required. BH3-mimetics directly inhibit anti-apoptotic BCL-2 proteins (i.e. BCL-2, BCL-XL, MCL-1 and/or BCL-W) thereby inducing apoptosis at the mitochondrial level. These compounds can either be used alone or in combination with other anticancer drugs to achieve synergistic effects. Venetoclax, a specific BCL-2 inhibitor, was FDAapproved in combination with hypomethylating agents for the treatment of acute myeloid leukemia (AML) in adults. Here, we addressed the question if different BH3-mimetics, alone or in combination with azacitidine, can be used to fight JMML. Methods/Results: Our research questions were addressed with experiments done in vivo and in vitro, in a patient-derived xenograft (PDX) mouse model and in naïve JMML cells, respectively. PDX mice were treated with ABT737, an analogue of navitoclax known to inhibit BCL-2, BCL-XL and BCL-W. Treatment was initiated 8 weeks after xenotransplantation of splenic JMML cells. ABT737 was given for 28 days (75 mg/kg/d), and on day 29, mice were sacrificed and analyzed. One group of mice was treated with azacitidine as described earlier (i.e. 2 cycles à 5 days followed by 9-day breaks, 3 mg/ kg/d). Both ABT737 and azacitidine led to efficient leukemia depletion but only azacitidine selectively depleted leukemia-initiating cells (LICs). Next, we analyzed whether both drugs act synergistically and could be used at lower concentration when given in combination, with the aim to minimize any potential side effect. Indeed, a 28 days long treatment with 0.75 mg/kg/d azacitidine (5 days on, 9 days off) and 50 mg/kg/d ABT737 (daily) did not only completely reduce leukemia burden but also depleted all LIC as confirmed by serial transplantation experiments. To dissect the roles of BCL-2 and BCL-XL, both inhibited by ABT737, and to additionally analyze the role of MCL-1 in the survival of JMML cells, we performed in vitro experiments using the specific inhibitors ABT199, A1155463 and S63845, respectively. When used alone, only ABT737 and the MCL-1 inhibitor S53845 killed JMML cells efficiently. However, synergistic effects were noted between azacitidine and the BCL-XL inhibitor A1155463. Azacitidine treatment resulted in the upregulation of pro-apoptotic BCL-2 proteins (i.e. BIK, PUMA, BAD and BMF) and a slight downregulation of MCL-1 indicating that azacitidine and the BCL-XL inhibitor synergized at the level of the BCL-2 protein family.
Conclusions: To conclude, our study indicates that combining azacitidine with BH3-mimetics could be a new and promising way to treat JMML patients, especially the ones that became resistant to azacitidine monotherapy. Granulocyte DNA from bone marrow or peripheral blood collected at time of diagnosis was used for next-generation sequencing of the entire NF1 coding sequence (custom Ampliseq enrichment and Miseq, Illumina). Affymetrix Cytoscan HD arrays were applied to detect segmental deletions or copy number-neutral loss of heterozygosity. Among 25 JMML/NF-1 cases, 10 exhibited an NF1 loss-of-function mutation at near-100% variant allelic frequency (VAF) in combination with uniparental disomy involving almost the entire 17q arm, suggesting single mitotic recombination as the leukemiacausing event. Eight cases carried two independent pathogenic NF1 mutations at near-50% VAF each. In one of these cases, the two mutations were close enough to enable demonstration of in-trans configuration. However, germline and somatic events could not be distinguished in any of the 8 cases due to unavailability of nonhematopoietic or parental DNA. One case harbored three NF1 mutations at 53%, 28%, and 13% VAF, suggesting two co-existing subclones. Four cases exhibited NF1 microdeletions (3 typical, 1 atypical) in combination with a pathogenic NF1 mutation at near-100% VAF. Non-hematopoietic tissue available in 3 of these 4 cases showed the microdeletion but failed to display the mutation. In two cases, monoallelic evidence of NF1 deficiency (1 typical NF1 microdeletion, 1 pathogenic mutation) but no second event was detected. In the JMML-QN group, 3 cases of compound-heterozygous pathogenic NF1 mutations and 3 cases of microdeletions combined with hemizygous pathogenic NF1 mutations were found. Non-hematopoetic tissue was available in one case and harbored the microdeletion but not the mutation. In the absence of clinical NF-1 features, the findings in these 6 children may be explained by 1) postzygotic NF1 mosaicism, 2) onset of JMML before clinical manifestation of NF-1, or 3) double somatic NF1 hits. Two JMML-QN cases carried single NF1 mutations at VAF~50% or less, providing inconclusive evidence of driver function. There was no genetic evidence of NF1 involvement in the remaining 8 JMML-QN cases.

NF1 Tumor Suppressor Gene Inactivation in
Conclusions: We conclude that the clinical diagnosis is reliable in children with JMML/NF-1, and propose that it can be made on the basis of CALS and JMML alone as most patients are too young to display the entire spectrum of NF-1 features. Genetic work-up of children without clinical features of NF-1 and no canonical driver mutation unmasks involvement of NF1 in a significant number of cases. In suspected JMML-QN without identifiable NF1 lesion, other myeloproliferative neoplasms should be considered. * Equally contributing first authors Introduction: Juvenile myelomonocytic leukemia (JMML) is a rare myelodysplastic/myeloproliferative neoplasm of early childhood and is characterized by clonal growth of RAS signaling addicted stem cells. JMML subtypes are defined by specific RAS pathway gene mutations and display distinct gene, microRNA (miRNA) and long noncoding RNA expression profiles. Recently, circular RNAs (circRNAs), single stranded RNA molecules with 3'-and 5'-ends backspliced in a non-collinear manner, were discovered, and shown to play key roles in many cancer types, including leukemia. Aim: Our aim was to study the circRNA transcriptome in JMML and identify specific circRNA profiles associated with mutational subgroups and other disease characteristics. Methods: Total RNA from diagnostic bone marrow (BM) or peripheral blood (PB) samples from 31 children with JMML and 8 healthy donors (HD) was obtained from different institutions throughout Europe. None of the patients received prior treatment. A discovery cohort of 19 JMML patients (8 PTPN11, 5 KRAS, 4 NRAS, and 2 NF1 samples) and 3 HD was used for total paired-end RNA-sequencing profiling. CirComPara software, a bioinformatics pipeline which simultaneously uses 9 backsplice detection methods, was used to annotate and quantify circRNAs. Differential expression in pairwise comparisons between conditions or sample groups was assessed by DESeq2. Expression of 27 circRNAs was analyzed by qRT-PCR in an independent validation cohort of 12 JMML patients and 6 HD. CircRNA-miRNA-gene networks were reconstructed using custom cir-cRNA function prediction and gene expression data. Results: Circular RNAs expression profiles separated JMML samples from HD according to unsupervised principal component analysis, pointing toward a general dysregulation of circRNA expression in patients. Considerable heterogeneity was observed among JMML patients, and both PCA and clustering based on circRNA expression showed that KRAS, NRAS, and PTPN11 mutated patients clustered together, while the two NF1 samples clearly separated from both HD and the other molecular groups. We identified 119 circRNAs dysregulated in JMML, of which 90 up-and 29 down-regulated. Of those, 37 circRNAs had different expression levels in the entire JMML cohort compared to HD, whereas the others were only dysregulated in one or more molecular subgroups. Our data indicated also circRNA expression differences among molecular subgroups of JMML. Analysis in the validation cohort of 27 circRNAs with dysregulated expression in JMML confirmed the downregulation of circOXNAD1 and circATM, and a marked up-regulation of circLYN, circAFF2, and circMCTP1. Interestingly, we identified upregulated circMCTP1 to be linked to expression level of known tumor suppressor miRNAs in JMML. Conclusions: In conclusion, this study provides insight into the cir-cRNA transcriptome of JMML. Dysregulation of known and many unknown circRNAs, such as circMCTP1, suggests a role for circRNAs in JMML pathogenesis, and results of functional predictions indicated the genes and biological pathways possibly involved. This study paves the way for functional research on the impact of circRNA dysregulation in JMML biology and their diagnostic and therapeutic applications.

35.
Immunophenotypic changes in JMML patients treated with hypomethylating agents. A correlation with clinical response assessment A. Frisanco Oliveira 1 , A. Tansini  Introduction: 5-azacitidine has been considered to be safe and efficient medication for use as a bridge to hemotopoietic stem cell transplantation in JMML patients. Patients under clinical treatment can be stratified as complete or partial response, stable disease or disease progression. Recently, the features found in JMML immunophenotyping (FCI) at diagnosis have been described and so, they could be useful for examining these characteristics for determining response of the patients to treatment. Aim: to follow the immunophenotypic pattern of JMML patients during treatment with 5-azacitidine, with an already validated IFC protocol and also look for cytometric features that correlate with the type of clinical response. Methods: An 8-color antibody panel described for pediatric population with JMML at diagnosis was used. Patients treated with cycles of 5-azacitidine 75mg/m 2 subcutaneous for consecutive 7 days, each 28 days were evaluated at diagnosis and after 3 and 6 cycles of medication. The findings were compared with the type of clinical response. Results: 32 patients entered the study. Among them, only 28 patients could be analyzed by FCI after 3 cycles and 25 patients after 6 cycles. After treatment with 5-azacitidine, patients showed a reduction in CD34 + cells, with median CD34/CD117 + cells going from 3.35% to 2.8% after 3 cycles of medication. Values dropped to <2% in 12 and 22 patients after 3 and 6 cycles respectively, and this was observed only in patients responding to treatment. B cell progenitors (H1), that were decreased at diagnosis, had a further reduction after treatment. The monocytic population also decreased, from 11.91% to 7.5% and 6.4% after 3 and 6 cycles, respectively. Those presenting with complete response had an apparent increase in classical monocytes and a decrease in intermediate monocytes. Finally, the population of T lymphocytes, that were reduced at diagnosis, showed an important increase, especially in patients with a clinical response (p < 0.001); an increase of NK lymphocytes was also observed mainly after 3 cycles from 1.01% to 1.85%. Immunophenotypic changes including aberrant expression of CD7 in myeloid progenitors remained after 3 and 6 cycles of 5-aza. Aberrant CD7 expression in myeloid progenitors was associated with a worse response to treatment, as well as NF1.
Conclusions: FCI was a feasible tool in JMML patients after treatment with 5-azacitidine. Clinical response was associated with decrease of CD34 + myeloid progenitor and total monocytes and a rise in T and NK lymphocytes. But abnormal co-expressions remained, even in patients with complete clinical response and this speaks in favor of the persistence of the leukemic clone during the whole clinical treatment. The largest effect was seen after 3 cycles. CD34/117 + cell and T lymphocytes are variables that could be incorporated into the assessment of response to clinical treatment of JMML patients. A more detailed study of the several subsets of T cells should be included. The same MFC protocol could be used pre and post transplantation identifying patients with beneficial use of 5-aza after transplant, as it has recently been shown according to methylation status. Prospective studies are necessary to confirm the data with a larger number of patients, and including assessment of the pre and post transplantation status.

P2.
Emberger syndrome and GATA2 deficiency in 2 adolescents with advanced myelodysplastic syndrome -case report R. Balceiro 1,2 , A. Frisanco Oliveira 1,2  Introduction: GATA2 gene encodes a key transcription factor for normal hematopoiesis. It is highly expressed in immature hematopoietic cells, being crucial for proliferation and maintenance of the stem-cell pool. Several manifestations make up the spectrum of GATA2 deficiency (immunodeficiency, MDS, lymphedema, mycobacterial infections). Methods/Case report: Patient 1, age 17, male, history of 6 months with recurrent cellulitis in legs and local edema associated with pancytopenia. Bilateral hearing loss since 9 years. On physical examination, lymphedema in left leg and warts on his hands and arms. Hemoglobin 9.7 g/dL, hematocrit 31.4%, leukocytes 9.3x10 9 /L (neutrophils 3.72 lymphocytes 4.371, monocytes 0.372), platelets 764 x10 9 /L. Myelogram: normocellular with severe dyspoiesis and 8% blasts. Immunophenotyping: 7.9% CD34/CD117 cells, with abnormal expression of CD7 and CD36. Absence of B cell precursors. 0.16% B lymphocytes, 1.67% NK and 7.06% T lymphocytes, 0.13% monocytes. Bone marrow biopsy: 90% cellularity, with severe megakaryocytic dysplasia. Karyotype: 45,XY,. Next-generation Sequencing (NGS): mutations in ASXL1, SETBP1, GATA2. Final diagnosis: myelodysplastic syndrome with excess blasts and GATA2 deficiency with Emberger syndrome. He underwent hematopoietic stem cell transplantation (HSCT) from a haploidentical donor (father), with a myeloablative conditioning regimen consisting of busulfan, fludarabine and melphalan. Neutrophilic engraftment on D+18. The patient developed severe chronic graft-versus-host disease with involvement of the skin, mouth, eyes and gastrointestinal tract, refractory to therapy. Death on D+389 due to infectious complications. Patient 2, age 15, male, with anemia since 10 years. In 2018 he developed pancytopenia. On physical examination, edema in legs and warts on his hands. Hemoglobin 9.3 g/dL, hematocrit 30.4%, leukocytes 0,8 x10 9 /L (neutrophils 0.352, lymphocytes 0.24, monocytes 0.104), platelets 4 x10 9 /L. Myelogram: hypocellular with trilineage dysplasia and 13.8% of blasts. Immunophenotyping: 14.4% CD34/ CD117+ cells without abnormal expressions. Absence of B precursors.  [3]. NGS: mutations in ASXL1, SETBP1, GATA2 and VUS (variant of unknown significance) in IDH1. Final diagnosis: myelodysplastic syndrome with excess blasts and GATA2 deficiency with Emberger syndrome. He started azacitidine while waiting for bone marrow transplantation, but died due to severe infection during. Discussion: Both patients had GATA2 deficiency with Emberger syndrome (MDS and lymphedema) and monosomy of 7, as well as mutations in ASXL1 and SETBP1. The immunophenotyping of both is characteristic, with a reduction in precursors and B lymphocytes, in addition to monocytopenia. Somatic mutations in ASXL1 and SETBP1 are recurrent in patients with MDS and GATA2 deficiency. It is believed that they can contribute to clonal evolution in the context of leukemic transformation. Both cases were diagnosed with advanced MDS. Monosomy 7 is the main cytogenetic alteration found in these patients, especially adolescents. Bone marrow transplantation is the only curative option for patients with GATA2 deficiency. Conclusions: GATA2 deficiency is a rare, heterogeneous clinical condition. Early detection is important due to the risk of progression to Introduction: Precise and standardized phenotypic descriptions are of utmost importance to establish genotype-phenotype relations in rare diseases. Further, detailed and accurate descriptions are necessary to help geneticists and clinicians interpret functional relevance of genetic variants in rare hematologic disorders. The Human Phenotype Ontology (HPO) offers a unique catalogue of standardized phenotypic features in a hierarchical structure that is gaining interest in rare disease research. Yet, until today the applications in pediatric hematology especially with the focus on inherited Bone Marrow Failure syndromes (iBMFs) are scarce. IBMFS comprise a heterogeneous group of genetic disorders that share a malfunctioning bone marrow with a large spectrum of cytopenias, physical malformations and an increased risk of myelodysplastic syndromes as well as other malignancies. Within this group of diseases multiple gene defects have been identified, but a significant proportion of cases still remains genetically unsolved. Further, even patients with similar genetic defects show different clinical courses as well as treatment responses. Essentially, a clear genotype-phenotype relation is missing, and the phenotypic complexity amplifies diagnostic challenges. Aims: This study aims at using HPO terminology, protein-protein interaction (PPI) networks and deep clinical phenotyping to investigate genotype-phenotype relations in patients with iBMFs. Further, this study aims to improve the diagnostic efforts and variant calling, through phenotypic comparison of genetically unsolved with solved cases. Methods: We use network analysis for PPIs to investigate the proximity of known iBMF genes. Additionally, we systematically reviewed HPO diseases annotation for iBMFS and compared them to other rare diseases. Further, we performed deep clinical phenotyping to computationally align and compare different patient subgroups. Results: By comparing proximity of proteins encoded by known causative iBMF genes, we have found that these genes cluster significantly closer than expected by chance within our PPI network. Further, we have shown that iBMFs are sufficiently annotated with HPO terms and thus are suitable for bioinformatic application. Additionally, we have shown that the annotation of phenotypic features significantly improves disease ranking in rare diseases. We thus have annotated phenotypic features of a preliminary set of local patients with HPO terms and performed unsupervised clustering of patient subgroups. By comparing clinical groups with PPI clustering and pure phenotyping, we observed partial overlap between subgroups, yet due to a limited dataset, we were not able to cluster the unsolved cases. Conclusions: Our computational approach of combined network analysis and deep clinical phenotyping presents a feasible, objective and structured method to investigate genotype phenotype relations in iBMFS. To evaluate the survival rates after HSCT of the total cohort and according to the different molecular mutations. Methods: All the 27 patients diagnosed with JMML who underwent their first HSCT at our center, between January 2014 and July 2019, were selected for this retrospective study. Along with descriptive analysis, a measurement on the global survival and event-free survival of the total cohort and of each molecular marker group was done. The survival probability was estimated by a Kaplan-Meier curve and the comparisons were made by the Logrank test. The chosen significance level was 0.05 and the data was analyzed by the SPSS software, version 2.1. Results: Nine patients (33%) had PTPN11 mutation, four (15%) had KRAS, four (15%) had NRAS, one (4%) had both PTPN11 and NRAS mutations, six (22%) had been clinically diagnosed with neurofibromatosis type 1 (NF1) and 3 (11%) had no molecular markers. No germline mutations have been documented. Considering all analyzed patients, the 3-year overall survival (OS) and event-free survival (EFS) was 65% and 59.3%, respectively. Relapse was the major cause of death. According to the genetic subgroup, the estimate 3-year OS was: 61% for PTPN11-mutated JMML, 100% for NRAS, 50% for KRAS, 50% for patients diagnosed with NF1 and 100% for patients with no molecular markers (P= 0,153). The patient with simultaneous mutations of PTPN11 and NRAS died from disease relapse. Conclusions: PTPN11 mutation was the most common in our series, similar to that reported in the literature, however the proportion of children with NF1 was relatively high among our patients. We found no difference in survival between the different genetic subgroups, most likely due to the small sample size. Introduction: It controls gene transcription from multiple target genes by binding to the consensus DNA sequence (T/A(GATA)A/G) of many enhancers, suppressors and promotors. More than 400 different germline mutations have been already identified in GATA2, including missense, frameshift and splice-site-mutations within the zinc finger 2 (ZF2) domain as well as non-coding substitutions in the EBOX-GATA-ETS. This variability leads to a very complex and heterogeneous syndrome, characterized by variable cytopenias as well as immunodeficiency, lymphedema or deafness among other clinical problems. Even more, the risk of affected individuals to develop either MDS or AML is exceedingly high, however, the mechanisms underlying malignant transformation are only insufficiently studied. Methods/Results: Here, we hypothesize that predisposition mutations can lead to cellular deficiencies that are compensated by somatic events. Such somatic events, however, might be associated with oncogenic transformation. For this purpose, it is important to identify the genes and pathways affected by GATA2 haploinsufficiency and to analyze how their activation or inhibition, respectively, affects HSPC function and leukemogenic potential. Using Gata2 fl/+ LSK cells (Lin -, Sca-1 + , c-Kit + ) treated with recombinant CRE (i.e. TAT-CRE) to delete one Gata2 allele and forced to proliferate in vitro, we performed RNA-seq to analyze the transcriptome and compare them to WT cells treated accordingly. Among other genes, many MYC targets were found to be downregulated in GATA2 +/cells compared to WT cells. Since MYC is a crucial transcription factor that plays a role in the generation and maintenance of the hematopoietic system, we are studying whether lentiviral MYC overexpression in vitro in Gata2 +/cells is sufficient to compensate the defective stress response in Gata2 +/-HSPCs. This will be analyzed in apoptosis, proliferation and colony forming assays. Conclusions: We hypothesize that MYC overexpression can positively affect survival, proliferation and differentiation of HSPCs.  Introduction: Myelodysplastic syndromes (MDS) are rare in children (<5% of hematological malignancies). Monosomy of chromosome 7 (─7) or 7q deletion (del(7q)) are found in >20% and associated with a very poor outcome. We present 3 children with MDS and chromosome 7 aberrations. Methods/Case series: Case 1: 9yo boy was referred due to mild thrombocytopenia. During follow-up thrombocytopenia worsened and he developed mild neutropenia. Bone marrow (BM) evaluation was performed: hypocellular BM (<10% cellularity) with no dysplastic features nor excess blasts; karyotype: 46,XY,inv(9)(p12q13)[20] described as a non-pathological variant; FISH analysis for del(5q), -7, del(7q), +8, del(20q), -Y was normal. Given worsening blood counts (Hb 7.9g/dL, retic 17x10 9 /L, neut 0.52x10 9 /L, plat 11x10 9 /L) and exclusion of other entities he was diagnosed with severe aplastic anemia. No related HLA matched donor was found so he was treated with horse ATG + cyclosporine achieving partial remission and transfusion independence at 6 months. BM assessment after 1 year of treatment revealed: heterogeneous cellularity with erythroid dysplasia, <10% myeloid dysplasia, and a new karyotype aberration involving cr 7: 46,XY,der(7;9)(q10;-p10),+9,inv9(p12q13)c[2]/46,XY,inv(9)(p12q13) [18]. He was diagnosed with refractory cytopenia of childhood (RCC) and referred for unrelated donor hematopoietic stem cell transplantation (HSTC) achieving complete remission (CR) which he maintains. Case 2: 5yo girl referred due to pancytopenia (Hb 10.3g/dL, MCV 99fL, retic 83x10 9 /L, neut 1.1x10 9 /L, plat 143x10 9 /L). She had mild psychomotor development delay, microcephaly and hirsutism, but known BM failure syndromes (BMFs) such as Fanconi Anemia, SBDS or DBA were excluded. BM evaluation: normocellular, trilineage dysplasia and no excess blasts. BM karyotype: 45,XX,XX[4]. She was diagnosed with MDS with ─7 and underwent unrelated donor HSTC, achieving CR. She had an unrelated death 3 years later while in CR. Case 3: 4yo boy presenting with fever and pancytopenia (Hb 8.7g/dL, MCV 92fL, retic 28x10 9 /L, neut 1.0x10 9 /L, plat 113x10 9 /L). After excluding other causes, BM evaluation was performed: normocellular BM with dysplasia and 7% blasts. BM karyotype showed 45,XY,, confirmed by FISH. A customized NGS panel for BM failure and somatic myeloid variants showed 3 low frequency (VAF <50%) somatic mutations in RUNX1, PTPN11 and NF1. No germline variants or known predisposing conditions were found. He was diagnosed with MDS with excess blasts and ─7. Matched sibling donor HSCT was performed 3 months later, achieving CR. The patient had an early relapse (3 months) with progression to acute myeloid leukemia (AML) in spite of DLI, accompanied by refractory hepatic GVHD. He was scheduled for salvage therapy with azacytidine, but died of infectious complications before starting treatment. Conclusions: Pediatric MDS is associated with inherited BMFs or genetic predisposition syndromes in >30% of cases. We did not identify any known germline predisposing features in our small series. Given the very high risk of AML progression in monosomy 7 and 7q deletion HSCT remains the only curative therapy. Evolving high throughput technologies can help us better understand the nature of these complex and heterogeneous disorders and, together with novel therapeutic agents, allow us to improve therapeutic strategies. Consent to publish has been obtained. Introduction: The distinction between aplastic anemia (AA) and hypoplastic refractory cytopenia of childhood (h-RCC) in the absence of clonality markers is based on morphological criteria on bone marrow (BM) aspiration smears and biopsies. Aim: To assess hypocellularity and dysplasia in AA in distinction to h-RCC on BM smears and biopsies according to EWOG criteria. Methods: BM smears of children and young adults diagnosed in a single center from 2009 to 2019 with hypocellular BM failure syndrome were centrally reviewed by two examinators and compared with their concomitant BM biopsies. A score of hypocellularity was used on smears and biopsies, with values of 0, 1, 2, 3 and 4 indicating empty BM, severe, moderate, mild reduction of cellularity and normal cellularity, respectively. Definitive diagnosis of AA vs. h-RCC was based on all available data including morphology, cytogenetics, mutations, syndromic features, clinical presentation and course of the disease. Fanconi anemia was excluded in all cases. Results: Overall, we examined smears from 25 patients: 47 smears from 10 AA and 15 h-RCC patients at diagnosis, and 8 smears following immunosuppressive therapy (IST) from 5 AA patients. Median number of smears per patient was 2 (range 1-6). Male/female ratio was 21/4 (84%) and median age 6.8 y (range 1-21). AA vs. h-RCC patients had median values of neutrophils 0.44 vs. 1.2 G/L (P=0.04), Hb 8.2 vs. 10.7 g/dL (P=0.14), platelets 19 vs. 67 G/L (P=0.08), respectively. h-RCC was associated with genetic syndromes in Cornelia de Lange N=1, SAMD9/SAMD9L N=2, dyskeratosis congenita N=1.
Karyotype was -7 N=1, +21 N=1, normal N=23. Cellularity was 0/1/2/ 3/4 in 35%, 18%, 18%, 11% and 18% of BM smears, respectively. Overall, 80% of AA vs. 40% of h-RCC patients had ≥1 smear of null cellularity, respectively (P=0.11). Erythroid (E), granulocytic (G) and megakaryocytic (M) lineage was absent in 33%, 46% and 87% of AA patients, and in 9%, 9% and 16% of h-RCC patients, respectively. Increased mast cells were present in 47% of all patients, respectively, without difference between AA and h-RCC. Asynchronous nuclear and cytoplasmic maturation and double/triple nuclei were the more frequent signs of dyserythropoiesis. Dysgranulopoiesis with hypogranulation was observed only in h-RCC patients. Micromegakaryocytes and megakaryoblasts were rare abnormalities as compared with less specific changes of dysmegakaryopoiesis, like nuclear immaturity and abnormal maturation. Histology was assessed on 39 concomitant biopsies from 9 AA patients and 12 h-RCC patients. Severely reduced cellularity was more frequent in AA than in h-RCC (100% vs. 38% of patients, respectively, P=0.01). At least one occurrence of paratrabecular erythroblastic islands and micromegakaryocytes was observed in 33% vs. 100% (P=0.06) and 0 vs. 58% (P=0.04) of AA and h-RCC patients, respectively. Diagnostic discordance was noted between consecutive smears or between smear and concomitant biopsy in 2 and 1 patient, respectively, all with a final diagnosis of h-RCC based on histological evidence of erythroblastic islands. BM morphological and clinical/hematological response to IST, were concordant in AA patients with no response N=1, relapse N=1, and reponse N=3. Cellularity remained mildly reduced in responders beyond two years after IST.
Conclusions: h-RCC is characterised morphologically by less profound hypocellularity and more important dysplasia as compared to AA with, however, some important, in terms of diagnostic accuracy, overlap between these two distinctive BM failure syndromes. mutations in all pts were located in zink finger 2; there were 4 missense and 6 truncating mutations including 4 nonsense and 2 frameshift changes. Genetic analysis in non-hematological tissue demonstrated germline origin of the alteration in all six cases analyzed. All pts presented with a severely decreased absolute neutrophil count ranging between 0.32 -1.10 G/L. In addition, all had monocytopenia with an absolute monocyte count between 0.01 -0.18 G/L. None of the pts had a platelet count below 100 G/L (range 106 -469). Median hemoglobin level was 10.8 g/dl (range 7.2 -14.6), and one patient required red cell transfusions prior to diagnosis. In all pts, the blast percentage in the bone marrow was <5%, and in 8 out of the 10 cases the marrow was hypocellular at the first bone marrow evaluation. Five of the 10 pts had cytogenetic abnormalities (-7q/del, +8 trisomy) at the time of the first evaluation. Recurrent infections including one episode of a severe bacterial infection were observed in 8 pts. Eight out of the 10 pts had physical abnormalities like genitourinary tract abnormalities (n=3), lymphedema (n=2), cardiac malformation (n=2), biliary tract abnormality (n=1) and deafness (n=2). These data may indicate that genitourinary tract malformations are more common in the constitutional phenotype of GATA2 deficiency than previously appreciated. Four of the 10 pts received hematopoietic stem cell transplantation (HSCT). One patient was grafted from a matched unrelated donor (MUD), while the other 3 pts received a haploidentical graft since an MUD was not available. Two of the latter had primary graft failure requiring a second grafting procedure. At the time of writing, seven pts are alive, one patient died due to varicella zoster encephalitis, one patient died after the second HSCT due to infectious complications and another patient had disease progression and died due to infectious complications during intensive chemotherapy. Conclusion: In this retrospective study we expand the clinical and genetic phenotype of GATA deficiency by outlining previously undescribed constitutional abnormalities and novel mutations. Thus, our work underlines the broad heterogeneity of the predisposition syndrome. Introduction: Although Fanconi anemia (FA) is a rare mainly autosomal recessive genetic disease it is the common cause of inherited bone marrow failure revealing with pancytopenia and one of the most common genetic causes of hematologic malignancies such as myelodysplastic syndrome (MDS) and acute myelogenous leukemia (AML). Aim: We presents the patient with both Fanconi anemia and mosaic trisomy 8, that developed MDS, and progressed to chronic myelomonocytic leukemia (CMML). Methods/Results: 8 year old girl was admitted to Pediatric Oncology and Hematology Department for the diagnosis of pancytopenia lasting from about 4 months. One month earlier, the patient was hospitalized due to influenza B-related pneumonia and complications (respiratory failure and myocarditis). Patient's coexisting conditions included: chromosomal aberration of 8 chromosome, congenital inappropriate exit of the left coronary artery from the pulmonary artery (Bland -White -Garland syndrome), rotated ectopic left kidney in the pelvis, myopia and strabismus. Blood tests on admission revealed pancytopenia, bone marrow examination with low cell count, and dysplastic forms. Test with mitomycin confirmed mosaicism of 8 chromosome trisomy and spontaneous chromosome breakages. Karyotype of parents confirmed the carrier of monoallelic deletion in FANCA gene. Patient was qualified to allogeneic hematopoietic stem cell transplantation (allo -HSCT) from unrelated donor because of MDS and FA was diagnosed 2 months later, the girl developed a fever without symptoms of infection, in control morphology test: anemia, WBC 31 x 10 3 /μl, PLT normal In blood smearmonocytosis 1,0x10 6 /μl, in bone marrow smear: highcellular bone marrow with blast cell 20%, Progression to CMML was diagnosed. The patient was classified to azacytidine therapy. After first cycle the patient experienced complications -fatigue, loss of appetite, fever progressed to septic shock and developed severe acute respiratory distress syndrome (ARDS). Treatment with azacytidine has been stopped. After 3 months the girl developed pneumonia again with negative bacterial and fungal cultures. Despite multi-drug therapy, she died 8 months after diagnosis of MDS / FA . Conclusions: An earlier diagnosis of FA in a girl would be improve prognosis. Consent to publish has been obtained. 39500/μL and median HbF 13.7%. All patients presented with myeloid precursors in blood. Median percentage of blasts in blood was 4% (range 1-7%) and in bone marrow 9.3% (range 3-17%). Spontaneous growth or granulocyte-macrophage colony-stimulating factor (GM-CSF) hypersensitivity in colony assay was detected in all patients. All patients had splenomegaly, with respiratory symptoms 4/6 patients, rash 2/6 patients, dysmorphic features or/and genetic stigmata 4/6 (one with CBL syndrome's features, and one with Costellolike syndrome's features). Mutations in genes of GMCSF-receptor-RAS/ MAPK signaling pathway were detected in all patients. Cytogenetics and molecular findings and therapeutic approach is presented in the Table. The patient n.4 relapsed 12.3 months after SCT, received azaci-tidine+ventoclax followed by a second SCT. All patients are alive (OS 100%) in remission without transfusion needs under close follow up (median time from diagnosis 44months) Conclusions: JMML is characterized by special clinical, laboratory and molecular features. The pathogenesis lies in the deregulation of the RAS signaling pathway. The prognosis of JMML is strongly associated with genetic mutations. HSCT is the treatment of choice for most patients, but for some patients a watch-and-wait strategy is recommended. Current and further studies may identify the role of demethylatitng agents (eg azacitidin) and targeted therapy for JMML patients. In this direction and due to the rarity of the disease, international cooperation is fundamental.

P14.
Identification of novel therapeutic approaches in a xenograft model of juvenile myelomonocytic leukemia (JMML) J. Rajak 1 , N. Koleci 1 , Y. Wu 1 , T. Watrin 2 , S. Bhatia 2 , A. Borkhardt 2 , C. Introduction: Juvenile myelomonocytic leukemia (JMML) is a highly aggressive myeloproliferative disorder of early childhood and the only curative treatment is hematopoietic stem cell transplantation (HSCT). However, the main cause of treatment failure is relapse in up to 30% of patients urging the strong need for novel therapies. In JMML, malignant transformation is driven by constitutive activation of the RAS signaling pathway providing the rationale for its therapeutic inhibition. By using primary JMML cells and our patientderived xenograft (PDX) model that closely mimics human disease, we are testing different novel targeted therapies to identify novel treatment strategies able to cure JMML. Methods/Results: We performed drug screening assays on primary JMML samples with a library of anticancer compounds including both clinically approved and investigational drugs. Drug sensitivity results were heterogenous, but consistent sensitivity was observed to heat shock protein 90 inhibitors, histone deacetylase inhibitors, FLT3 tyrosine kinase inhibitors and polo-like kinase inhibitors. In addition, several other interesting compounds targeting (in)directly RAS signaling and tested in adult leukemias but not included in the drug screening assay, were tested in vitro for their cytotoxicity towards JMML cells. Our selection included Idelalisib (PI3K inhibitor), LY3009 (pan-RAF inhibitor), Pevonedistat (NEDD8 activation enzyme inhibitor) and Ruxolitinib (JAK1/2 inhibitor). JMML cells showed high sensitivity to Idelalisib and Pevonedistat. In contrast, LY3009 and Ruxolitinib did not influence or had little effects on cells, respectively. Our in vitro data on Pevonedistat as promising novel candidate will be further corroborated in engrafted mouse model of JMML. In vivo studies will be conducted in established JMML xenograft Rag2-/-γc-/mouse model that nicely imitates disease features. Newborn mice will be irradiated with 2.5 Gy and intrahepaticaly injected with 1x10 6 of JMML mononuclear cells. Eight weeks after transplantation mice will be subjected to Pevonedistat. Treated mice will be thoroughly analyzed by flow cytometry and histology. By serial transplantation assays, the effects of the drug specifically on leukemia-initiating cells will be tested. In a different line of investigation, the immunosuppressive effects of JMML cells will be analyzed, and checkpoint inhibitors (i.e. PD-L1 inhibitor, anti-CD47) will be used in vitro assays and, whenever possible, in our PDX mice. Conclusions: It is our primary goal to deplete JMML-initiating cells to prevent relapse after HSCT. Eventually, we plan to transfer our preclinical observations to phase I/II clinical trials and in that way improve care of JMML patients. In addition, our studies will contribute to a better understanding of pathogenetic mechanisms of JMML. Introduction: Bone marrow failure syndromes (BMFS) and myelodysplastic syndromes (MDS) are rare disorders that condition the normal differentiation of hematopoietic stem cells. The first group is characterized by inherited or acquired impaired production of blood cells; while the latter presents mainly with morphologic dysplasia of bone marrow cells. BMFS often present in childhood, whereas MDS are more common in adults; in children they are frequently associated with inherited forms of BMFS or predisposition syndromes. As some of these morphological changes overlap, definitive diagnosis may be hard to obtain and should be carried out by experienced teams in order to better predict risk of leukemic transformation and select the best treatment and follow up strategy. Aim: To review all BMFS and/or MDS Pediatric cases evaluated in a Pediatric Hematology (non-oncology) Unit of a level III Pediatric Hospital. Methods: Descriptive analytical retrospective study, covering a 10 year period (2010-2020). Inclusion criteria were BMFS and/or MDS diagnosis, with age of onset and follow-up period until 18 years old. No patient was excluded. The variables analyzed were demographic, clinical, laboratory and therapeutic. Statistical analysis was performed using SPSS 22®. Results:We reviewed a total of 38 cases, 55% male, with a median age of 6 years (IQR 11). 74% (28) were classified as BMFS and 26% (10) MDS. Regarding classification of BMFS (28): half of the patients (14) had congenital disease with predisposition to myelodysplasia; 57% (8) were male, median age of 3 years (IQR 4); 71% (10) had dysmorphisms. 36% had pancytopenia, 28% bicytopenia and in 36% only one affected lineage. We identified 5 patients with Fanconi