ALBERT

Description: Immunotherapy has remarkably changed the treatment paradigm in hematologic malignancies and natural killer (NK) cell therapy represents an attractive option, as it has been feasible and safe in early clinical trials, without graft-versus-host effects. Nevertheless, the molecular markers determining cancer cell sensitivity or resistance to NK cells, especially in the context of tumor cell interaction with the bone marrow (BM) stromal microenvironment remain incompletely understood, but have major translational relevance since these tumor-stromal interactions have been known to attenuate the response of blood cancer cells to diverse classes of pharmacological agents. To address these questions, we performed NK cell treatment of a series of cell lines from hematologic malignancies by applying a pooled "DNA-barcoded" format of these cell lines (PRISM system). We specifically quantified the dose-dependent responses to primary NK cells for 70 molecularly-annotated blood cancer cell lines, including myeloid and lymphoblastic leukemia, diffuse large B cell lymphoma and 15 multiple myeloma (MM) lines, either in presence or in absence of BM stromal cells (BMSCs) and interferon gamma (IFNg), followed by integrated computational analyses to identify candidate molecular markers correlating with tumor cell sensitivity or resistance to NK cells. NK cell cytotoxicity, quantified by the relative abundance of barcodes in treated cells compared to controls, was correlated with the transcriptional, mutational and other molecular features of each of the 70 cell lines from publicly available databases. Furthermore, data from MM cell lines were compared to our genome-wide loss of function (LOF) and gain of function (GOF) CRISPR screen data in the MM cell line MM.1S. Two distinct clusters of cell lines, sensitive and resistant to NK cell treatment, were identified. Such clusters retained distinct pattern of in vitro resistance in the presence of stroma, while showing an overall markedly decreased NK cell responsiveness, which underscores the protective effect of stromal microenvironment in blood malignancies, regardless of the addition of IFNg. RNA-seq data showed no differences in dependencies between the two clusters and no distinct gene expression patterns at baseline that clearly allows to predict NK cell response, which underscores the heterogeneity of resistance patterns at single gene level, across different hematologic malignancies. However, when comparing baseline RNAseq data to data obtained from previous GOF and LOF CRISPR screens in MM.1S, surface antigens such as PVR, ULBP1, ULBP3 were more frequently downregulated, whereas MUC1 was upregulated in resistant cells clusters. An important observation is that gene lesions such as TP53, PTEN, MMSET, commonly associated with high-risk diseases, do not affect NK cell responses in the cell lines tested. Interestingly, a gene set enrichment analysis (GSEA) showed that the cluster of resistant cells displays upregulation of class I MHC complex, class II MHC complex binding, IL7 pathway and a downregulation of transmembrane receptor protein serine/threonine kinase signaling pathway. GSEA also showed that baseline state of IFN-JAK-STAT signaling correlates with BMSCs-induced NK cell resistance, a result further confirmed by addition of IFNgto tumor-NK cocultures in the absence of BMSCs. No significant differences in NK cell response were observed when comparing cell lines of different hematologic neoplasms, suggesting that candidate markers from these studies may be relevant across different hematologic malignancies. In conclusion, this is the first study of this size correlating the molecular annotation of different concurrently-treated hematologic cell lines with their response to a NK-based treatment in the context of BMSC interaction. This study of a large panel of pooled "DNA-barcoded" cell lines provided complementary and orthogonal information to our LOF and GOF screens, expanding our potential to identify and validate molecular markers for individualized use of NK cell-based therapies in hematologic malignancies. Disclosures Mustjoki: BMS: Honoraria, Research Funding; Novartis: Research Funding; Pfizer: Research Funding. Mitsiades:EMD Serono: Research Funding; Abbvie: Research Funding; Karyopharm: Research Funding; Sanofi: Research Funding; Arch Oncology: Research Funding; Fate Therapeutics: Honoraria; Ionis Pharmaceuticals: Honoraria; Takeda: Other: employment of a relative ; Janssen/Johnson & Johnson: Research Funding; TEVA: Research Funding.

Description: Immunotherapy is a promising approach to improve treatment responses in hematological malignancies. Accumulating evidence suggests a role for natural killer (NK) cells in controlling hematological malignancies. However, mechanisms regulating sensitivity or resistance of hematologic cancer cells to the effector function of NK cells are incompletely understood. Here, we performed genome-scale CRISPR-Cas9 loss-of-function screens to systematically map genes that regulate sensitivity of hematologic malignancies to NK cells. To screen for genes involved in the interaction between NK and cancer cells, we infected human cancer cells expressing Cas9 with a genome-scale lentiviral guide RNA library (Figure). The resulting pool of knockout cells was exposed to NK cells expanded from peripheral blood of healthy donors. Enriched and depleted knockouts were detected by next-generation sequencing of the integrated sgRNA cassettes, enabling identification of genes conferring resistance or susceptibility to NK cell-mediated lysis. The screens were performed in cell lines from diverse hematological malignancies, including chronic myeloid leukemia (CML), B cell acute lymphoblastic leukemia, diffuse large B cell lymphoma (DLBCL), and multiple myeloma. We recovered several known mechanisms of NK cell/cancer cell interactions, demonstrating feasibility of the screening approach. Loss of genes encoding components of the MHC class I complex (B2M, HLA-A, HLA-C, HLA-E) sensitized multiple cancer cell lines to NK cell lysis, consistent with missing-self recognition. Furthermore, knockout of IFN-JAK-STAT signaling mediators led to increased tumor cell lysis, suggesting that MHC class I induction in response to NK cell-derived IFN gamma enables NK cell evasion by tumor cells. We also identified genes essential for effective NK cell-mediated lysis. NCR3LG1, encoding the B7-H6 ligand for the NKp30 activating NK cell receptor, was essential for NK cell lysis of CML cells. In contrast, knockout of apoptotic mediators and TRAIL pathway components conferred resistance to NK cell cytotoxicity in DLBCL cells, indicating heterogeneity in NK cell/cancer cell interactions between cancer types. Our data support a view that distinct mechanisms regulate sensitivity to NK cell cytotoxicity in different hematologic cancers. Importantly, our results indicate that loss-of-function mutations in the antigen-presenting machinery and the IFN-JAK-STAT pathway sensitize tumors to NK cell effector function. As alterations in these genes are associated with resistance to T cell immunotherapies such as PD-1 blockade, NK cell-based therapies could be employed to overcome resistance in these patients. In summary, we suggest that systematic identification of mechanisms governing tumor immune susceptibility has the potential to uncover novel immunotherapy targets. Figure. Figure. Disclosures Kankainen: Medix Biochemica: Consultancy. Lee:Merck, Sharp, and Dohme: Consultancy; Courier Therapeutics: Equity Ownership, Membership on an entity's Board of Directors or advisory committees; CytoSen Therapeutics: Equity Ownership, Membership on an entity's Board of Directors or advisory committees, Research Funding. Mustjoki:Pfizer: Honoraria, Research Funding; Novartis: Honoraria, Research Funding; Celgene: Honoraria; Ariad: Research Funding; Bristol-Myers Squibb: Honoraria, Research Funding.

Description: Introduction: Genetic alterations of RUNX1 gene (mutations and fusion genes) are common in hematological malignancies including AML, ALL, MDS, and MPN. Mutations of RUNX1 occur in 10% of newly diagnosed AML patients and associate with inferior prognosis. Similarly, point mutations in RUNX1 are frequent in BP-CML, but little is known about their role in BP patients. Methods: Bone marrow samples from four BP patients with RUNX1 mutations were collected from Helsinki University Hospital, and samples from four other BP patients without RUNX1 mutations were used as controls. Samples were analyzed by exome and RNA sequencing. For drug sensitivity and resistance testing (DSRT), a high-throughput platform comprising 195 approved and investigational oncology drugs was used. To further study the role of RUNX1 mutations in BP-CML, Baf3/BCR-ABL1+ RUNX1-/- knock-out and RUNX1-/mut heterozygous deletion mutant models were created using CRISPR-CAS9 technology. The effects of the mutations in the cell line models were characterized by flow cytometry, RNA sequencing, and DSRT analyses. Results: The median age of RUNX1-mutated BP patients was 47 years (range 36-56 years) and included 3 patients with myeloid-BP and 1 with lymphoid-BP phenotype. RUNX1-mutated patients tended to have higher mutational load and co-occurrence with mutations to the BCOR, EZH2 and PHF6 genes in agreement with the mutational profile of RUNX1-mutated AML. Signature analysis of RUNX1-mutated BP patients showed dominance of signature 1 (age-related), signature 6, and 15 (DNA mismatch repair) correlating with the control group signature profiles. In addition, the RUNX1 mutated group showed specific enrichment of signature 9, related to somatic hypermutation and AID/RAG axis activity, that was not observed in the control group. RNA sequencing revealed that 388 genes were significantly differentially expressed between RUNX1-mutated and control BP patients (139 upregulated, 249 downregulated). Interestingly, the antigen processing and presentation pathway was among the top upregulated pathways in RUNX1 mutated patients with overexpression of CIITA and HLA-DR genes. Stem cell differentiation, complement, and hemostasis pathways were enriched in the downregulated gene set. Lymphoid-specific transcription factors (DNTT, PAX5 and VPERB1) in addition to CD19 and CD25 were significantly upregulated in the RUNX1-mutated group. DSRT data showed greater sensitivity to glucocorticoids, MEK-, mTOR- and VEGFR-inhibitors in RUNX1-mutated patients compared to the control group. Data from RUNX1-/- and RUNX1-/mut cell line models confirmed the induced expression of CD19 surface marker with RUNX1 mutation. DSRT results showed increased activity of MEK inhibitor and mTOR inhibitors in RUNX1-/- knock-out and RUNX1-/mut cell lines. Conclusions: RUNX1 mutations in BP-CML patients associate with distinct phenotypic and transcriptional features. DSRT results together with transcriptional data support the potential benefits of glucocorticoids and MEK and mTOR inhibitors as well as immunotherapy targeting CD19 and CD25 in this group of BP-CML patients. Disclosures Kankainen: Medix Biochemica: Consultancy. Heckman:Celgene: Research Funding; Orion Pharma: Research Funding; Novartis: Research Funding. Porkka:Novartis: Honoraria, Research Funding; Celgene: Honoraria, Research Funding. Mustjoki:Bristol-Myers Squibb: Honoraria, Research Funding; Pfizer: Honoraria, Research Funding; Novartis: Honoraria, Research Funding; Ariad: Research Funding; Celgene: Honoraria.

Description: Despite the advances in the treatment of acute lymphoblastic leukemia (ALL), a major fraction of adult patients still succumb to leukemia- or treatment-related events. In particular, the outcome of elderly ALL patients remains dismal. Our aim was to discover new or repurposed drugs for B-cell ALL in a clinically relevant ex vivo drug sensitivity testing platform. We analyzed 19 primary B-ALL samples using a well-established drug sensitivity and resistance testing platform and a drug panel including 65 drugs in five different concentrations. The main drug classes were glucocorticoids, MDM2 antagonists, and inhibitors of BCR-ABL1, VEGFR, BCL-2, BCL-XL, BET, MEK, JAK, Aurora kinase, PI3K, MTOR, IGF1R, ERK, STAT3, STAT5, HSP90 and NAMPT proteins. The samples were viably frozen bone marrow (BM) mononuclear cells collected at diagnosis. The cohort included both Philadelphia-positive (Ph+) (n=10) and Ph-negative (Ph-) (n=9) patients with a median age of 43 years (range 22-68). Cell viability (CellTiter-Glo) was measured after plating and after a three-day incubation with the drugs. A drug sensitivity score (DSS) was calculated from the viability readouts, which takes into account the area under the dose response curve, measuring both drug efficacy and potency. DSS values 〉10 are considered effective and 〉20 highly effective. As an overall view of drug sensitivity, a heatmap and dendrograms from DSS values are shown in Figure 1A. As expected, most patients were sensitive to glucocorticoids and tyrosine kinase inhibitors (TKIs) showed efficacy in Ph+ ALL. In addition, two Ph-negative patients were sensitive to TKIs, suggesting a Philadelphia-like disease. Drugs that showed pan-ALL efficacy included BCL-2 family inhibitors, idasanutlin (MDM2 inhibitor), luminespib (HSP90 inhibitor), daporinad (NMPRT inhibitor) and plicamycin (antineoplastic antibiotic). For the other drugs, only individual patients showed sensitivity, in line with the diverse molecular background of ALL. Strikingly, 17/19 (89%) of patients in our cohort were highly sensitive (DSS〉20) to navitoclax (a BCL-2, BCL-XL and BCL-W inhibitor), whereas the BCL-2-specific inhibitor venetoclax was effective only in a distinct subset of patients (Figure 1B). 6/19 (32%) of patients were highly sensitive (DSS〉20) to venetoclax and represented all risk classes based on age, white blood cell counts and karyotype, but interestingly, all were Ph-negative. Overall, response to venetoclax correlated with response to navitoclax (Spearman, r=0.85; P

Description: Chimeric antigen receptor (CAR) T-cell therapy has proven effective in relapsed and refractory B-cell malignancies, but resistance and relapses still occur. Better understanding of mechanisms influencing CAR T-cell cytotoxicity and the potential for modulation using small-molecule drugs could improve current immunotherapies. Here, we systematically investigated druggable mechanisms of CAR T-cell cytotoxicity using 〉500 small-molecule drugs and genome-scale CRISPR-Cas9 loss-of-function screens. We identified several tyrosine kinase inhibitors that inhibit CAR T-cell cytotoxicity by impairing T-cell signaling transcriptional activity. In contrast, the apoptotic modulator drugs SMAC mimetics sensitized B-cell acute lymphoblastic leukemia and diffuse large B-cell lymphoma cells to anti-CD19 CAR T cells. CRISPR screens identified death receptor signaling through FADD and TNFRSF10B (TRAIL-R2) as a key mediator of CAR T-cell cytotoxicity and elucidated the RIPK1-dependent mechanism of sensitization by SMAC mimetics. Death receptor expression varied across genetic subtypes of B-cell malignancies, suggesting a link between mechanisms of CAR T-cell cytotoxicity and cancer genetics. These results implicate death receptor signaling as an important mediator of cancer cell sensitivity to CAR T-cell cytotoxicity, with potential for pharmacological targeting to enhance cancer immunotherapy. The screening data provide a resource of immunomodulatory properties of cancer drugs and genetic mechanisms influencing CAR T-cell cytotoxicity.

Description: BACKGROUND A complex interaction between blasts and surrounding cells in the acute myeloid leukemia (AML) bone marrow (BM) microenvironment sustains blast proliferation and confers chemoresistance. T- and NK-cells have been shown to be dysfunctional in AML, which might be associated with immune evasion and poor prognosis. Here, we present a comprehensive analysis of the immune contexture of the AML BM at diagnosis and study its interaction with clinicopathological variables. METHODS Diagnostic BM biopsies (n=69) were collected from AML patients treated in the Helsinki University Hospital during 2005-2017 and age and gender-matched controls (n=12) to construct tissue microarrays (TMA). Using 8-plex immunohistochemistry (mIHC) and computerized image analysis, we determined cell abundance and immunophenotypic states of millions of immune cells. Immunoprofiles were integrated with a total of 120 clinicopathological variables including cytogenetics and molecular genetics, ELN (European Leukemia Net) risk classification, disease burden parameters, and patient demographics. RESULTS Unsupervised hierarchical clustering of the immunologic contexture defined by mIHC analysis grouped AML patients distinctly from control subjects (Fig 1a). By extracting significant differences (Mann-Whitney U test, q

Description: The second generation tyrosine kinase inhibitor dasatinib is a clinically approved drug for chronic myeloid leukemia (CML) as well as Ph+ acute lymphoblastic leukemia. Dasatinib is a dual-specific inhibitor of ABL and SRC family kinases but shows a broad inhibitory effect on various kinases, including Kit, EGFR, FAK and BTK. In addition to its antileukemic effects, dasatinib was shown to impact on the immune system, including alterations in B-cell development, T-cell differentiation, as well as a transient activation of quiescent hematopoietic stem cells (HSCs) in the bone marrow of wild-type mice. In CML patients treated with dasatinib, the development of pleural effusions and large granular lymphocyte (LGL) lymphocytosis during dasatinib treatment was associated with a favorable response. Non-hematologic adverse events included pulmonary artery hypertension and gastrointestinal symptoms. We investigated the influence of dasatinib (5 or 20 mg/kg p.o.) on the cellular composition of immune cells in our tetracycline-controlled transgenic BCR-ABL mice in comparison to control mice. We analyzed dasatinib-induced effects by flow cytometry in the bone marrow (BM) and spleen under steady-state conditions as well as after bone marrow transplantation. Furthermore, we conducted flow cytometry analysis of stem- and progenitor cells in the BM of BCR-ABL expressing mice treated with dasatinib or vehicle control and performed histophathology to evaluate effects on the BM, spleen, lung and intestine. Our results demonstrated that dasatinib dose-dependently increased Gr1/CD11b positive myeloid cells in the BM of normal mice, while B220 positive B cells and Ter119 positive erythrocytic cells were reduced. NK1.1 positive NK cells were decreased in BM but increased in the spleen. Interestingly, the lower dose (5 mg/kg) induced an increase of the percentage of CD41 positive megakaryocytic cells in BM and B cells in the spleen, while higher doses (20 mg/kg) decreased both percentages. The BCR-ABL-induced phenotype was strongly reversed by dasatinib in BCR-ABL expressing transgenic animals under steady-state conditions and in a transplantation setting: the oncogene-induced upregulation of granulocytes, megakaryocytes, and the reduction of erythroid cells as well as of B-lymphocytes was antagonized by dasatinib. Concurrently, a strong decrease of Linneg Sca1+ c-kit+ (LSK) HSCs and an increase of MEPs together with a reduction of GMPs were observed. Although the overall lymphocyte population was reduced in the BCR-ABL-expressing mice and replaced by granulocytes, we observed an increase of the proportion of CD3+, CD4+ and CD8+T cells among lymphocytes, and this increase was reverted by dasatinib treatment. Conversely, dasatinib did not significantly alter NK1.1 cells in these mice. Our observations in BCR-ABL-positive animals are in part contrary to the findings in wild-type animals, but this may be explained by the reversion of BCR-ABL effects in addition to the effects of dasatinib on normal hematopoietic cells. In addition to the above-mentioned changes, we observed a reduction of BCR-ABL mediated granulocyte infiltration of the small intestine indicating a beneficial effect in attenuating inflammation of the bowel. There was no significant alteration of the bronchial epithelium or bronchial arteries. Altogether, dasatinib robustly reduced the CML phenotype in vivo in our transgenic mouse model, and this effect included stem and progenitor populations and was stronger than the effects we have previously described for imatinib in this model. Thus, our transgenic CML mouse model is well suited to examine dasatinib effects on normal and malignant hematopoietic cells in vivo. Interestingly, there was a differential effect on myeloid cells in normal vs. CML mice, suggesting different targets in these cells. (CS and NC contributed equally to this study) Disclosures Braunschweig: Bristol-Myers-Squibb: Honoraria; MSD Sharp & Dome: Honoraria. Brümmendorf:Pfizer: Consultancy, Honoraria; Ariad: Consultancy, Honoraria; Bristol-Myers Squibb: Consultancy, Honoraria; Novartis: Consultancy, Honoraria, Research Funding; Patent on the use of imatinib and hypusination inhibitors: Patents & Royalties. Mustjoki:Bristol-Myers Squibb: Honoraria, Research Funding; Ariad: Research Funding; Pfizer: Honoraria, Research Funding; Novartis: Honoraria, Research Funding. Koschmieder:Bristol-Myers Squibb: Honoraria, Membership on an entity's Board of Directors or advisory committees, Other: Travel funding to conferences, Research Funding.

Description: Background Natural killer (NK) cell malignancies are rare lymphoid neoplasms characterized by aggressive clinical behavior and poor treatment outcomes. Clinically they are classified as extranodal NK/T-cell lymphoma, nasal type (NKTCL) and aggressive NK cell leukemia (ANKL). Both subtypes are almost invariably associated with Epstein-Barr virus (EBV). Recently, genomic studies in NKTCL have identified recurrent somatic mutations in JAK-STAT pathway molecules STAT3 and STAT5b as well as in the RNA helicase gene DDX3X in addition to previously detected chromosomal aberrations. Here, we identified somatic mutations in 4 cases of ANKL in order to understand whether these entities share common alterations at the molecular level. To further establish common patterns of deregulated oncogenic signaling pathways operating in malignant NK cells, we performed drug sensitivity profiling using NK cell lines representing ANKL, NKTCL and other malignant NK cell proliferations. We aimed to identify sensitivities to agents that selectively target components of pathways required for survival of malignant NK cells in an unbiased manner. Methods Exome sequencing was performed on peripheral blood or bone marrow of ANKL patients using the NK cell negative fraction or other healthy tissue as control. Profiling of drug responses was performed with a high-throughput drug sensitivity and resistance testing (DSRT) platform comprising 461 approved and investigational oncology drugs. The NK cell lines KAI3, KHYG-1, NKL, NK-YS, NK-92, SNK-6 and YT and IL-2-stimulated and resting NK cells from healthy donors were used as sample material. All drugs were tested on a 384-well format in 5 different concentrations over a 10,000-fold concentration range for 72 h and cell viability was measured. A Drug Sensitivity Score (DSS) was calculated for each drug using normalized dose response curve values. Results The ANKL patients displayed mutations in genes reported as recurrently mutated in NKTCL, such as FAS, TP53, NRAS, STAT3 and DDX3X. Additionally, novel alterations in genes previously implicated in the pathogenesis of NKTCL were detected. These included an inactivating mutation in INPP5D (SHIP), a negative regulator of the PI3K/mTOR pathway and a missense mutation in PTPRK, a negative regulator of STAT3 activation. Interestingly, the total number of nonsilent somatic mutations in 3 out of 4 ANKL patients (97, 82 and 45) was remarkably high compared to other hematological malignancies analyzed in our variant calling pipeline. Analysis of drug sensitivities in NK cell lines showed a close correlation between all cell lines and a markedly higher correlation with those of IL-2 stimulated than resting healthy NK cells, suggesting that malignant NK cells may share a common drug response pattern. Furthermore, in an unsupervised hierarchical clustering the NK cell lines formed a distinct group from other leukemia cell lines tested (Fig. A). Among pathway-selective compounds (namely, kinase inhibitors and rapalogs), the drugs most selective for malignant NK cells fell into two major categories: PI3K/mTOR inhibitors (e.g. temsirolimus, buparlisib) and inhibitors of aurora and polo-like kinases such as rigosertib and GSK-461364 (Fig. B). JAK inhibitors (e.g. ruxolitinib, gandotinib) and CDK inhibitors (e.g. dinaciclib) showed strong efficacy in both malignant NK cells and IL-2 activated healthy NK cells. Conclusions Our exome sequencing results suggest that candidate driver alterations affecting similar signaling pathways underlie the pathogenesis of ANKL as has been reported in NKTCL. Drug sensitivity profiling highlights the PI3K/mTOR pathway as a potential major driver of malignant NK cell proliferation, whereas JAK-STAT signaling appears to be essential in both healthy and malignant NK cells. Components of these pathways harbored mutations in our small cohort of ANKL patients and have been shown to be deregulated by mutations or other mechanisms in previous studies, underlining their importance as putative drivers. The systematic large-scale characterization of drug responses also identified these pathways as potential targets for novel therapy strategies in NK cell malignancies. Figure 1. (A) Unsupervised hierarchical clustering based on drug sensitivity scores (DSS) of NK, AML, CML and T-ALL cell lines. (B) Scatter plot comparing DSS of malignant NK cell lines (average) and healthy IL-2 stimulated NK cells. Figure 1. (A) Unsupervised hierarchical clustering based on drug sensitivity scores (DSS) of NK, AML, CML and T-ALL cell lines. (B) Scatter plot comparing DSS of malignant NK cell lines (average) and healthy IL-2 stimulated NK cells. Disclosures Mustjoki: Novartis: Honoraria, Research Funding; Bristol-Myers Squibb: Honoraria, Research Funding; Pfizer: Honoraria, Research Funding.

Description: INTRODUCTION The recent success of checkpoint blockade immunotherapies in diverse solid tumors has prompted the evaluation of these treatments in hematologic malignancies such as acute myeloid leukemia (AML). It is critical to identify the patient and disease subsets that could respond to such therapies. Infiltration of tumors by cytotoxic T lymphocytes (CTLs) has been associated with better prognosis and responses to checkpoint inhibition. We hypothesized that the presence of a substantial fraction of activated CTLs and natural killer (NK) cells in the blood and bone marrow samples of hematologic tumors could indicate a preexisting active immune response potentially targeting the tumor cells. Moreover, the density of the immune infiltrate could shape and be shaped by the expression of cancer-germline and leukemia-associated antigens (LAAs), antigen-presenting machinery (APM) and immunosuppressive genes by the tumor cells. Here, we examined these immunological properties of hematological tumors in large-scale gene expression datasets to identify immunologically active patient subsets. METHODS Curated set of 9,544 transcriptomes collected across 36 hematological malignancies (HEMAP), including 1,858 AML cases was utilized to identify subsets of patients with existing, potentially tumor-directed immune responses. Additional multi-omics datasets of 173 AML patients from The Cancer Genome Atlas (TCGA) were integrated to gain insight into the genetic landscape of immunologically active patients. Cytolytic activity (geometric mean of GZMA (granzyme A) and PRF1 (perforin) transcript levels, Rooney et al., Cell 2015) was used as a marker of immunologic activity. Cytolytic activity was correlated to the expression levels of all transcripts, gene sets from collections such as MSigDB and manually curated gene sets representing the APM (HLA-A, -B, -C, B2M), 145 known cancer-germline antigens as well as established LAAs such as WT1 and PRTN3. Furthermore, we used an in silico flow cytometry approach, CIBERSORT (Newman et al., Nat Methods 2015), to infer the relative fractions of 22 immune cell subpopulations from the gene expression data to dissect the immune cell composition of the samples. RESULTS Cytolytic activity showed high correlation with other transcripts expressed in activated CTLs and NK cells (e.g. GZMB, GNLY, KLRB1, CD8A, CD2; Spearman's R ≥ 0.7) as well as lymphocyte activation-related gene sets across both the HEMAP and the TCGA AML datasets, validating it as a robust and specific metric of active cellular immunity. When correlated to the CIBERSORT immune cell populations, cytolytic activity was positively associated with CD8 T cells and showed a negative correlation to the proportion of M2 macrophages. High levels of cancer-germline antigens were associated with decreased expression of components of the APM and low cytolytic activity, suggesting HLA downregulation as a mechanism of immune evasion by cancer-germline antigen-expressing tumor cells. We observed extensive heterogeneity in the cytolytic activity between different diseases and subtypes within the same disease, most prominently in AML. In AML patients, complex karyotype and unfavorable prognosis were correlated with high cytolytic activity, indicating biological similarity of the immune-infiltrated tumors. Furthermore, TP53 mutations, genome fragmentation and immune checkpoint transcripts such as CD274 (PD-L1), PDCD1LG2 (PD-L2), CTLA4 and LAG3 were enriched within the complex karyotype cluster in the TCGA AML dataset. In contrast, mutations in NPM1 and FLT3 showed a modest but significant negative correlation to cytolytic activity. CIBERSORT analysis revealed that AML cases with low cytolytic activity preferentially had enrichment of an eosinophilic phenotype in addition to increased M2 polarization of macrophages. CONCLUSIONS Using large-scale transcriptomics approaches, we were able to identify patient subsets with variable levels of immune cytolytic activity within hematologic malignancies. Furthermore, we identified connections between the cytotoxic immune response and genetic properties of AML tumors. These observations have potential clinical implications, as the choice of patients to clinical trials receiving immune checkpoint blockade immunotherapies would require careful consideration in light of the observed immunological heterogeneity. Disclosures Mustjoki: Bristol-Myers Squibb: Honoraria, Research Funding; Pfizer: Honoraria, Research Funding; Novartis: Honoraria, Research Funding; Ariad: Research Funding.

Description: INTRODUCTION Natural killer (NK) cell malignancies are rare aggressive neoplasms that are classified by the WHO as extranodal NK/T-cell lymphoma, nasal type (NKTCL) and aggressive NK-cell leukemia (ANKL). Recently, genome and exome level studies in NKTCL have shed light on the mutational spectrum of the disease. However, somatic mutations in ANKL have not been characterized. Here, we identified somatic mutations in 14 cases of ANKL to further clarify the genetic landscape underlying malignant NK cell proliferation. We compared the discovered variants to those detected in NKTCL to understand whether the two diseases harbor common molecular alterations. Moreover, we used high-throughput drug screening and RNA sequencing on NK cell lines derived from ANKL, NKTCL and other malignant NK cell proliferations to identify therapeutically actionable drivers of malignant NK cell growth. METHODS We performed whole-exome sequencing on genomic DNA extracted from peripheral blood or bone marrow samples of 14 ANKL patients. To compare the mutational profiles in ANKL and NTKCL, we re-analyzed the published whole-exome NKTCL data from Jiang et al. (Nat Genet 2015) using our somatic variant calling pipeline. For profiling of drug responses, we used a high-throughput drug sensitivity and resistance testing (DSRT) platform comprising 461 approved and investigational oncology drugs to screen the ANKL cell lines IMC-1, KHYG-1 and NK-92, NKTCL cell lines NK-YS and SNK-6 as well as DERL-7, KAI3, NKL, YT and IL-2-stimulated NK cells from healthy donors. All drugs were tested in 384-well format in 5 concentrations over a 10,000-fold concentration range for 72 h, cell viability was measured and normalized dose response curve values were used to calculate a drug sensitivity score (DSS) for each drug. Finally, we performed amplicon sequencing of known cancer driver genes and RNA sequencing on the cell lines and healthy NK cells to identify driver mutations and integrate gene expression profiles with drug sensitivity patterns. RESULTS We identified recurrent somatic activating mutations in STAT3 in 21% (3 of 14) of ANKL patients. Other mutated genes included RAS-MAPK pathway molecules (BRAF, NRAS, KRAS), protein phosphatases regulating JAK-STAT and PI3K-AKT-mTOR pathways (PTPRT, PTPRK, INPP5D) as well as several epigenetic modifiers (TET2, ARID2, KDM2B, SETD7, SETD2) and the tumor suppressor TP53. Interestingly, we detected mutations in genes recurrently mutated in NKTCL, such as the RNA helicase DDX3X and the cell surface receptor FAS. Re-analysis of the published NKTCL data revealed a high frequency of missense mutations in receptor type and non-receptor type protein phosphatases (e.g. PTPRC, PTPRR, PTPRT, PTPN1, PTPN2, PTPN3), many of which have established roles as negative regulators of JAK-STAT signaling. These findings potentially expand the subset of NK cell tumors where the JAK-STAT pathway is somatically activated and implicate deregulated JAK-STAT signaling as a major driver in these diseases. The malignant NK cell lines used in drug sensitivity profiling frequently harbored mutations in same genes and pathways, including STAT3 (N=3), STAT5B (N=1), DDX3X (N=2), KRAS (N=1), FAS (N=2) and several epigenetic modifiers, thus validating these cell lines as relevant disease models. The drug sensitivities in NK cell lines showed a high correlation and the cell lines formed a distinct group from other lymphoid and myeloid leukemia cell lines in unsupervised hierarchical clustering, suggesting an NK-cell specific drug response pattern. The most effective targeted drugs across all NK cell lines included HDAC inhibitors, inhibitors of Aurora and Polo-like kinases, JAK inhibitors, HSP inhibitors and CDK inhibitors as well as the Bcl-2 inhibitor navitoclax. Compared to other leukemia and lymphoma cell lines, JAK inhibitors, navitoclax and methotrexate emerged as the most NK-cell specific compounds. CONCLUSIONS Our genetic data demonstrate extensive heterogeneity in the mutational spectrum of ANKL and implicate JAK-STAT and RAS-MAPK signaling as well as disruption of epigenetic modifiers in the disease pathogenesis. Integrated drug sensitivity and gene expression profiling corroborates the JAK-STAT pathway as a major therapeutically actionable driver of malignant NK cell proliferation and identifies other potential novel targeted therapy options such as Bcl-2 inhibition in NK cell malignancies. Disclosures Suzumiya: Chugai: Honoraria, Research Funding; Toyama Chemical: Research Funding; Kyowa Hakko kirin: Research Funding; Astellas: Research Funding; Eisai: Honoraria, Research Funding; Takeda: Honoraria. Ohshima:Chugai: Research Funding, Speakers Bureau; Kyowa Kirin: Research Funding, Speakers Bureau. Mustjoki:Pfizer: Honoraria, Research Funding; Ariad: Research Funding; Novartis: Honoraria, Research Funding; Bristol-Myers Squibb: Honoraria, Research Funding.