ALBERT

Description: Introduction: Polyneuropathy, organomegaly, endocrinopathy, monoclonal gammopathy, and skin changes (POEMS) syndrome is a rare paraneoplastic disease due to an underlying monoclonal plasma cell (PC) dyscrasia. Despite of dynamic symptoms associated with highly elevated VEGF, monoclonal PCs are thought to be quite small, and pathogenic significance of these PCs remains undetermined. In this study, we performed whole exome sequencing (WES), target sequencing, and RNA sequencing of PCs in patients with POEMS syndrome in order to define its genetic profiles. Methods: Patients diagnosed with POEMS syndrome at Chiba University Hospital from July 2014 to June 2016 were enrolled. DNA was extracted from either PCs which were isolated from patients' bone marrow (BM) using CD138 MACS (Miltenyi) or buccal cells as controls. WES and target sequencing were performed using HiSeq2500 (Illumina) and MiSeq (Illumina), respectively. The data of WES and target sequencing were analyzed by Empirical Bayesian mutation Calling (EBCall). Copy number was analyzed using the data of WES. RNA sequencing of PCs isolated by MACS and FACS sorting was conducted using HiSeq 1500 (Illumina). PCs from some patients diagnosed with multiple myeloma (MM) and monoclonal gammopathy of undetermined significance (MGUS) were also collected as controls for RNA sequencing. Results: Twenty POEMS patients (M:F 12:8, mean age 42.6, range 16-78; 15 newly diagnosed, 5 refractory or relapsed cases) were included in this study. Regarding the types of M protein, 55% (11/20) were IgA-λ, 25% (5/20) were IgG-λ, and each individual case of the following; IgA-λ+IgG-λ, BJP-λ, IgG-κ, and Castleman's variant with no M protein. The mean serum VEGF was 6,471 pg/ml (range 1,190-13,800), and the mean PCs percentage in the BM was 4.4% (range 0.8-10.5). WES was performed in 15 cases; a total of 359 somatic mutations in 334 genes were revealed in 93.3% of cases (14/15) with a mean number of 23.9 (range 0-119) in each. All these mutated genes were significantly enriched in several pathways related to cell adhesion. Importantly, frequently mutated genes in MM such as NRAS, KRAS, and TP53 were not identified. Among all mutations, 1.7% were frameshift insertions, 2.0% were frameshift deletions, 4.2% were stop gains, 0.8% were non-frameshift deletions, 60.2% were other non-synonymous single nucleotide variants (SNVs), 29.5% were synonymous SNVs, and 1.7% were splicing mutations which were within 2-bp of a splicing junction. Copy-number variations were detected in 33.3% of cases (5/10) including -13 (2 cases), +1q (2 cases), and hyperdiploidy (2 cases). To carry out target sequencing in all 20 cases, we defined 51 target genes which included recurrently mutated genes from our WES data, frequently mutated genes in hematopoietic and lymphoid tissues according to the database (COSMIC), and 15 frequently mutated genes in MM (NRAS, KRAS, TP53, BRAF, CDKN2C, FGFR3, BIRC3, DIS3, CYLD, KDM6A, LRP1B, FAM46C, COL6A3, DNAH5, and KRT6A). A total of 60 somatic mutations were revealed in 65% of cases (13/20), and 9 new somatic mutations were found in the cases in which WES was also performed. Ten recurrently mutated genes were identified; KLHL6 in 20% of cases (4/20), each of LTB, RYR1 in 15% of cases (3/20), and each of EHD1, EML4, HEPHL1, HIPK1, PCDH10, USH2A, and ZNF645 in 10% of cases (2/20). Among frequently mutated genes in MM, only 3 genes (FAM46C, LRP1B, and DNAH5) were mutated, each in a single case. We finally conducted RNA sequencing of the FACS-sorted PCs in 5 POEMS patients compared to 5 MGUS and 4 MM patients. Upregulated genes were significantly enriched in some gene sets, gene ontology terms, and pathways related to immune response and cell adhesion, whereas downregulated genes were related to tumorigenesis. Of note, VEGF was not significantly upregulated in POEMS patients. Principal component analysis distinguished the 3 disease groups of patients with marginal overlaps between POEMS and MGUS, and also MGUS and MM. Conclusions: Our data clearly demonstrate that the genetic profiles of PCs in POEMS syndrome are distinct from those in MM and MGUS. Notably, PCs may not be the main source of extremely elevated VEGF in POMES syndrome. On-going further investigation will help clarify the molecular pathogenesis of POEMS syndrome. Disclosures Ogawa: Takeda Pharmaceuticals: Consultancy, Research Funding; Sumitomo Dainippon Pharma: Research Funding; Kan research institute: Consultancy, Research Funding.

Description: [Backgrounds] POEMS syndrome is a rare plasma cell dyscrasia and its pathogenesis including the significance of monoclonal plasma cells in disease progression is poorly understood. Monoclonal plasma cells only produce λ immunoglobulins, and genes encoding immunoglobulin λ light chain (IGL) V regions are derived from IGLV1-44 or IGLV1-40 germline sequences. Here we analyzed the clonality in IGLV gene rearrangements using next generation sequencing (NGS) to evaluate the significance of monoclonal plasma cell clone size and follow its changes in clinical course to understand the pathogenesis of POEMS syndrome. [Methods] Patients who were diagnosed with POEMS syndrome between November 2006 and October 2015 at Chiba University Hospital were included in the study. As positive controls, 3 multiple myeloma (MM) patients with λ-type monoclonal light chain, and 9 negative control patients were also analyzed. NGS libraries were constructed from genomic DNA samples, extracted from bone marrow mononuclear cells. The IGLV1 and IGLV2 genes were amplified by polymerase chain reaction (PCR) using a 5' primer for the IGLV1/2 framework 3 (FR3) region and 3ʹ consensus primers for the IGLJ1/2/3 joining regions. Multiple samples were pooled, and paired-end 2 × 250 base pair sequencing reactions were carried out using an Illumina MiSeq sequencer. The closest matched germline sequences were determined using the ImMunoGeneTics database. Subsequently, frequencies of each clonotype that were characterized by a unique V-J rearrangement, conserved complementarity determining region 3 (CDR3) anchors and a unique CDR3 amino acid sequence, were calculated. [Results] Twenty-eight patients with POEMS syndrome were enrolled. All the patients had λ-type M protein. The median follow up time of the patients was 24.4 months (range, 3.7 - 113.9). Firstly we analyzed the usage of IGLV germline genes in each case. In 8 cases, the POEMS syndrome-specific germline sequences, IGLV1-40 or IGLV1-44, were dominant; accounting for more than 40% of all germline sequence usage. However, other samples showed minimal or no differences from controls, indicating that the clonal expansion of monoclonal plasma cells is generally low in patients with POEMS syndrome. Analyzing frequencies of the most dominant rearrangement in each germline, the clonal IGLV gene rearrangements of POEMS syndrome-specific germline sequences were significantly increased in 10 POEMS patients (35.7%; IGLV1-44: n = 8, IGLV1-40: n = 2). Significant increase of clone sizes were not directly linked to the initial disease status (vascular endothelial growth factor [VEGF] level and percentage of plasma cells in the bone marrow), overall survival and progression-free survival of POEMS patients. In 12 patients that we were able to follow their clinical courses, the clone size of IGLVgene rearrangements correlated with disease course assessed with serum VEGF level in most cases (n = 8), as they decreased with serum VEGF levels in disease remission and increased with re-elevation of serum VEGF in relapse cases. Clone sizes without significant increase at diagnosis were constantly flat even after achieving disease remission (n = 3). In one case, clone size with significant increase at diagnosis was unchanged even after serum VEGF level decreased (n = 1). Further observation is needed in this case. [Discussion] Considering the cases with significant increase of IGLV rearrangement clones, it was confirmed that clonal light chain gene expression is restricted to the IGLV1-44 and IGLV1-40 germline sequences, as previously reported. IGLV gene rearrangement clone was not detected as significant increase in some cases. Therefore, we speculated that there are certain numbers of patients with POEMS syndrome with extremely low frequency of clone cells. On the other hand, significant increase of clone size did not reflect disease status, suggesting that disease status is not regulated by tumor burden alone. By contrast, in cases with significantly increased clones, clone sizes changed dependiing on the disease status. These data do demonstrate that monoclonal plasma cells are related to the pathogenesis of POEMS syndrome. [Conclusions] Our analysis of IGLV gene rearrangements has demonstrated the association between the size of IGLV gene rearrangement clones and the clinical courses in POEMS syndrome. Disclosures No relevant conflicts of interest to declare.

Description: The expression of NR4A3, which is a member of the gene encoding NR4A orphan nuclear receptor subfamily, has been reported to be commonly silenced in blasts of patients with acute myeloid leukemia (AML), irrespective of karyotype. In line with this finding, Nr4a1-/-/Nr4a3-/- mice rapidly develop AML within one month following birth (Mullican et al., 2007). In addition, Nr4a1+/-/Nr4a3-/- and Nr4a1-/-/Nr4a3+/- mice show myelodysplastic/myeloproliferative neoplasms (Ramirez-Herrick et al., 2011), suggesting that NR4A3 functions as a tumor suppressor gene in myeloid malignancies. The extremely short latency of AML development in Nr4a1-/-/Nr4a3+/- mice indicates that silencing these tumor suppressors is sufficient to induce AML and that NR4A3 has a crucial role in the pathogenesis of AML. Thus, unveiling the molecular mechanism that regulates NR4A3 expression in AML would facilitate the development of novel therapies, including transcriptional reactivation of the gene. However, the therapeutic modalities targeting NR4A3 have been hindered by our minimal understanding of the mechanism underlying reduced NR4A3 expression, particularly in human AML cells. Abnormal epigenetic regulation is a common mechanism in the pathogenesis of several types of cancers. For instance, the expression of several tumor suppressor genes, such as p16 and MLH1, is repressed due to DNA hypermethylation at their promoter regions. Given that loss-of-function mutations in NR4A3 have not been reported in AML to date, we hypothesized that DNA hypermethylation contributes to a reduction in NR4A3 expression in AML. To test our hypothesis, we analyzed DNA methylation status of NR4A3 in human AML cells. We first compared the level of NR4A3 expression in eight human AML cell lines and two human primary AML samples, with that in CD34+ mononuclear bone marrow (BM) cells from healthy human controls. As expected, the expression of NR4A3 was markedly reduced in all of the AML cell lines and primary AML cells compared with that in the cells of the healthy controls. To evaluate the function of NR4A3 in human AML cells, we ectopically overexpressed NR4A3 in a human AML cell line (NB4 cells). The growth of NR4A3 -overexpressing NB4 cells was remarkably compromised compared with that of the controls, suggesting a tumor suppressive function of NR4A3 in both human AML and murine cells. To investigate the DNA methylation status of NR4A3, we performed bisulfite sequencing assays using eight human AML cell lines (HL60, NB4, Kasumi, TF1, U937, K562, MOLM13, and THP1) as well as CD34+ BM cells from healthy controls. Unexpectedly, a hypermethylated CpG site in the promoter region was not detected in any of the cell lines. However, the drastically or mildly methylated region including twenty eight CpGs was identified approximately 3 kb downstream of the transcription start site in six AML cell lines (97.5%, 78.3%, 77.1%, 89.9%, 95.2%, and 86.9% in HL60, NB4, Kasumi, TF1, U937, and K562, respectively) and two mixed lineage leukemia-related cell lines (31.0% and 53.6% in MOLM13 and THP1, respectively), whereas this site was hypomethylated in the controls (n = 2; mean, 12.7%; range, 7.1%-18.2%). To evaluate the contribution of this hypermethylated region to reduced NR4A3 expression, the six AML cell lines with heavily hypermethylated CpGs at NR4A3 and two human primary AML cell samples were treated with a DNA methyltransferase inhibitor (decitabine; DAC) for three or five days. DAC exposure inhibited cell growth and restored the expression of NR4A3 in all AML cell lines and primary cells in a dose- and time-dependent manner. Next, we examined the status of DNA methylation at the CpG site following DAC treatment with bisulfite sequencing assays. The frequencies of methylated CpG in HL60, NB4, and K562 cells was reduced from 97.5% to 53.6%, 78.3% to 68.5%, and 86.9% to 67.5% after DAC treatment, respectively. In contrast, the methylation status in Kasumi, TF1, and U937 cells did not significantly changed after DAC treatment. Our findings in the present study suggest that DNA hypermethylation may partially account for the transcriptional inactivation of NR4A3 in AML. However, the mechanism of reduced NR4A3 expression is complex and variable depending on the genetic background. We are currently working on a more detailed analysis of DNA methylation using human primary cells, by extending the regions for investigation, such as enhancer regions. Disclosures Nakaseko: Novartis: Honoraria, Research Funding, Speakers Bureau; Otsuka: Honoraria, Research Funding; BMS: Honoraria, Research Funding, Speakers Bureau; Pfizer: Honoraria, Research Funding, Speakers Bureau.

Description: Background: As recurrence of acute myeloid leukemia (AML) is difficult to predict, it is important to detect it by measuring minimal residual disease (MRD). PML-RARA, RUNX-RUNX1T1, CBFB-MYH11 are regarded as the reliable MRD markers. However, in AML with normal karyotype and many other forms, no MRD markers have been established. NPM1 mutations, occurring in approximately 30% of adult AML cases, and 50-60% of AML cases with normal karyotype, represent one of the most frequent mutations in AML. Recently, NPM1 mutation is reported to be useful in assessing MRD. We undertook a retrospective and prospective investigation of the usefulness of NPM1 mutation as an MRD marker in Japanese patients with AML. Methods: The subjects were 38 NPM1-mutated AML patients with first hematological remission at several hospitals related to our institution between 2001 and 2018. This study was approved by the ethics committee of Nippon Medical School and the informed consents were obtained from all patients, according to the Declaration of Helsinki. We analyzed peripheral blood cells or bone marrow cells at diagnoses, and evaluated only bone marrow cells after diagnoses. Detection of NPM1 mutation was carried out using allele-specific real time PCR following creation of a complementary primer. After dilution of the samples, sensitivity to TCTG, CATG, and CCTG was found to be 0.001%. The NPM1 mutant copies were qualified only at successful amplification of internal control. Results: The median age of the patients was 58 years (18-79 years). There were 32 cases with intermediate cytogenetic prognosis and 6 cases with unclear chromosomal profile. Of the 38 cases, 14 cases (37%) were FLT3-ITD-positive and allogeneic hematopoietic stem cell transplantation was carried out in 14 cases (37%). The base sequence was TCTG in 36 cases and CCTG in 2 cases. Persistence of NPM1-mutatation was present in 25 patients with first hematological remission (66%). Compared with patients with MRD negative, patients with MRD positive were associated with DNMT3A mutation (MRD positive 12/25 vs MRD negative 0/13, p=0.003). The rate of relapse in patients with MRD positive was significantly higher than those of in patients with MRD negative (MRD positive 76% vs MRD negative 23%, p=0.004). The rates of relapse free survival (RFS) and overall survival (OS) in patients with MRD positive were significantly lower than those in patients with MRD negative (RFS at 2 years: MRD positive 14% vs MRD negative 86% p=0.003; Figure 1, OS at 2 years: MRD positive 25% vs MRD negative 93%, p