HLA Loss Research Articles

Association of HLA loss of heterozygosity with allele-specific neoantigen expansion in response to immunotherapy.

e18030 Background: Human leukocyte antigen loss of heterozygosity (HLA LOH) restricts immune recognition of tumors by limiting the major histocompatibility complex (MHC) presentation of neoantigens to T cells and correlates with reduced response to immune checkpoint blockade therapy (ICB) in non-small cell lung cancer. To explore the mechanism behind the impairment of HLA LOH on ICB, we analyzed the relationship between the antigen presentation pathway, neoantigen presentation and response to ICB in a head and neck squamous cell carcinoma (HNSCC) cohort. Methods: Following baseline sample collection, a cohort of 14 HNSCC patients recieved a single dose of PD-1 inhibitor. The primary tumor mass was definitively resected approximately one month later. If resection was impractical, a second biopsy was taken. Response to therapy was evaluated using RECIST criteria. Each pre- and post-intervention tumor sample and normal PBMC sample were profiled using Personalis’ ImmunoID NeXT Platform, an HLA-enhanced exome/transcriptome platform. HLA LOH was detected using a digital PCR validated machine learning algorithm (DASH). Neoantigen presentation was computationally predicted using a machine learning algorithm (SHERPATM) trained on mono-allelic immunopeptidomics data. Results: We found that 50% of the HNSCC cohort had HLA LOH, a larger percentage than in a large pan-cancer cohort (23%, n=611) and a distinct HNSCC cohort (40%, n=20). Further, two patients had B2M LOH and one patient had a deleterious mutation in an HLA allele. Despite the high frequency of somatic alteration in the antigen presentation pathway, we did not find an association between HLA LOH and ICB response. However, if HLA LOH was still shaping tumor evolution in response to ICB, we would expect to see immune pressure against subclonal tumor populations with neoantigens presentable by the retained HLA alleles but not the deleted HLA alleles. Indeed, we found that significantly more novel post-treatment neoantigens were predicted to bind to deleted HLA alleles compared to their homologous alleles (p=0.045). Conclusions: Given the high prevalence of HLA LOH across tumor types, a greater understanding is needed regarding the impact of HLA LOH on tumor evolution during ICB treatment. Though HLA LOH does not correlate with response to ICB, the consistent shift in neoantigen composition suggests that it acts as an evolutionary force in resistance to response during immunotherapy.

Clinical Utility of the Detection of the Loss of the Mismatched HLA in Relapsed Hematological Patients After Haploidentical Stem Cell Transplantation With High-Dose Cyclophosphamide.

Haploidentical hematopoietic stem cell transplantation (Haplo-HSCT) with high-dose cyclophosphamide (PTCy) has resulted in a low incidence of graft-vs.-host disease (GVHD), graft failure, and non-relapse mortality. However, post-transplantation relapse remains a common cause of treatment failure in high-risk patients. Unraveling the mechanisms of relapse is therefore crucial for designing effective relapse treatment strategies. One of these mechanisms is the loss of the mismatched HLA on the recipient's leukemic cells. To study the incidence and clinical relevance of this phenomenon, we analyzed 181 patients treated with Haplo-HSCT with PTCy (2007–2019), of which 37 relapsed patients after transplantation. According to the kit employed for HLA-loss analysis, among 22 relapsed patients, we identified HLA loss at relapse in 6 of the 22 patients (27%) studied. Based on the results obtained, the genomic loss of HLA was more common in females than males (66 vs. 33%) and HLA-loss relapses occurred later than classical relapses (345 vs. 166 days). Moreover, the patients with HLA-loss had a greater presence of active disease at the time of transplantation and had undergone a larger number of treatment lines than the group with classical relapses (66 vs. 43% and 66 vs. 18%, respectively). Four of these relapses were studied retrospectively, while two were studied prospectively, the results of which could be considered for patient management. Additionally, two relapsed patients analyzed retrospectively had myeloid neoplasms. One patient had not undergone any treatment, and three had undergone donor lymphocyte infusions (DLIs) and chemotherapy. All presented severe GVHD and disease progression. In contrast, the two patients studied prospectively had a lymphoid neoplasm and were not treated with DLIs. One of them was treated with chemotherapy but died from disease progression, and the other patient underwent a second Haplo-HSCT from a different donor and is still alive. We can conclude that the detection of HLA-loss at the onset of relapse after Haplo-HSCT with PTCy could help in clinical practice to select appropriate rescue treatment, thereby avoiding the use of DLIs or a second transplantation from the same donor.

346 - Detection of HLA Loss of Heterozygosity by an MHC Targeted STR Assay

346 - Detection of HLA Loss of Heterozygosity by an MHC Targeted STR Assay

A loss-of-function mutation in exon1 of limited HLA class I alleles is common in patients with aplastic anemia

Leukocytes that lack HLA allelic expression (HLA-LLs) caused by a copy-number-neutral loss of heterozygosity in the short arm of chromosome 6 (6pLOH) and somatic mutations in HLA class I genes are commonly identified in patients with acquired aplastic anemia (AA), although the exact mechanisms underlying the HLA loss and HLA class I allele repertoire likely to acquire loss-of-function mutations remain unknown due to the limited number of AA patients that have been studied for loss-of-function mutations in HLA class I genes. We identified a common nonsense mutation at codon19 (c.19C>T, p.R7X) in exon1 (Exon1mut) of different HLA-A and HLA -B alleles in HLA-LLs from AA patients. Screening of 353 Japanese patients with AA using a novel droplet digital PCR assay revealed Exon1mut in 101 (29%) of the patients. Exon1mut occurred in only 12 different HLA-A (n=4) and HLA-B (n=8) alleles that corresponded to 4 HLA supertypes (A02, A03, B07, and B44), suggesting that limited autoantigens presented by these HLA class I alleles to T cells are involved in AA development. These findings provide insight into the immune pathophysiology of BM failure and contribute to identify candidate autoantigens in AA.

Leukemia Relapse after Allogeneic Hematopoietic Stem Cell Transplantation: From Recapitulation/Acquisition of Leukemogenic Hits to Immune Escape Due to Somatic Class I/ II HLA Mutations

Curative potential of allogeneic hematopoietic cell transplantation (aHCT) in myeloid malignancies is principally related to the graft versus leukemia (GvL) effect exerted by donor-derived immune effectors on leukemic cells. However, different pathways may drive post-transplant relapse, including perturbation/attenuation of T cell-mediated GvL responses. For example, decreased expression of HLA alleles has been described in post-aHCT relapse.1 This could be due to down-regulation of HLA or 6p HLA locus deletion. However, the occurrence of chromosome 6p uniparental disomy (UPD) along with del6p (first described by our group in the context of aplastic anemia)2 suggests that HLA loss of heterozygosity (LOH) and inherent loss of one allele (presenting immunodominant peptides) may be a significant contributor to leukemic relapse either directly, causing decreased HLA expression or as additive effect to HLA down-regulation. In relapses after haploidentical or mismatched aHCT, LOH of mismatched allele has been described.3 Here, we hypothesize that not only copy neutral LOH or haploinsufficient HLA expression but also defective function due to HLA mutations may lead to immune escape from GvL. Such a mechanism would involve the loss of an allele responsible for the presentation of an immunodominant antigenic peptide. Moreover, unlike relapse due to the acquisition of additional myeloid mutations, HLA mutant relapse would also acquire resistance to donor lymphocyte infusion (DLI) therapy. In order to dissect the immunogenomic mechanisms leading to leukemic immune-evasion, we performed a comprehensive genetic analysis of specimens sequentially collected from a cohort of patients with acute myeloid leukemia (AML) and myelodysplastic syndromes (MDS) relapsed after aHCT. Specifically, we applied a deep-targeted NGS panel to study HLA region along with 173 genes known to have a role in leukemogenesis and cancer ontogeny. So far, 57 paired/serial biospecimens from 25 transplanted patients have been analyzed (25 samples at AML/MDS diagnosis, 25 at the moment of relapse after aHCT and 7 samples at relapse after chemotherapy) Overall, we found the acquisition of 8 disruptive HLA somatic mutations in 6 patients at post-transplant relapse (24%), 4 in class I and 4 in class II loci. None of those events were found in samples at diagnosis or at post-chemotherapy relapse, suggesting the possibility of an "immune-escape relapse". Those somatic hits accounted for 4 intronic indels, 1 frameshift insertion, one splicing site and 2 point mutations in 3′ and 5′ untranslated regions (UTRs). Median variant allele frequency (VAF) was 17% (range 2-58%). Of note is that all HLA mutant patients received a matched aHCT (4 from related and 2 from a 10/10 matched unrelated donors) suggesting that this mechanism is independent from the deletion and the loss of an immune privileged mismatched allele. Interestingly, median time to relapse was 514 (range 119-935) days for HLA-mutated patients vs 126 (62-543) for HLA wild type cases (p=.00042), consistent with the hypothesis that the establishment of immune-tolerance, and the presence of a GvL effect (less likely in early relapses) are required for the selection of those mutations. When somatic genotype of these patients prior and after transplant was studied, we found that post-aHCT relapses were associated in most cases with a new genomic configuration with either loss of previous events or acquisition of new subclonal mutations in myeloid or cancer related genes (in 17/25 cases). However no difference was seen in terms of number and patterns of new events among the HLA mutated and HLA wild type patients. Results shown here represent an important proof-of-concept for the role played by somatic mutations in HLA genes in the setting of post-aHCT AML/MDS relapses. These events, in analogy to the deletion or copy neutral LOH, may promote immune escape relapse resistant to immunologic manipulations and may coexist or be an alternative pathway to the progression/relapse characterized by acquisition of myeloid subclonal driver mutations. In that context, HLA mutations would be considered facilitator lesions. Disclosures Hsi: CytomX: Consultancy, Honoraria; Eli Lilly: Research Funding; Abbvie: Research Funding; Miltenyi: Consultancy, Honoraria; Seattle Genetics: Consultancy, Honoraria. Hamilton:Syndax Pharmaceuticals: Consultancy, Honoraria. Carraway:Abbvie: Other: Independent Advisory Committe (IRC); BMS: Consultancy, Other: Research support, Speakers Bureau; Novartis: Consultancy, Speakers Bureau; Jazz: Consultancy, Speakers Bureau; Stemline: Consultancy, Speakers Bureau; Takeda: Other: Independent Advisory Committe (IRC); ASTEX: Other: Independent Advisory Committe (IRC). Majhail:Incyte: Honoraria; Mallinckrodt: Honoraria; Nkarta Therapeutics: Honoraria; Anthem, Inc.: Consultancy. Maciejewski:Alexion, BMS: Speakers Bureau; Novartis, Roche: Consultancy, Honoraria.

Clonal Hematopoiesis By HLA Class I Allele-Lacking Hematopoietic Stem Cells and Concomitant Aberrant Stem Cells Is Rarely Associated with Clonal Evolution to Secondary Myelodysplastic Syndrome and Acute Myeloid Leukemia in Patients with Acquired Aplastic Anemia

[Background] HLA-class I allele-lacking (HLA[-]) leukocytes are detected in approximately 30% of patients with acquired aplastic anemia (AA), and are thought to represent the involvement of cytotoxic T lymphocyte attack against hematopoietic stem cells (HSCs) in the development of AA, based on the high response rate to immunosuppressive therapy (IST) in patients with such aberrant leukocytes. Similar to glycosylphosphatidylinositol-anchored protein (GPI-AP)-deficient (GPI[-]) leukocytes in patients with paroxysmal nocturnal hemoglobinuria (PNH), HLA(-) leukocytes in AA patients are often clonal or oligoclonal and expand to account for more than 50% of the total leukocytes. Despite such overwhelming proliferation, somatic mutations in driver genes as well as telomere shortening that portend clonal evolution are rarely detected in HLA(-) granulocytes, suggesting the genetic stability of HLA(-) HSCs and the persistence of the immune pressure on HSCs that favors expansion of HLA(-) HSCs (Imi, et al. Blood Adv). However, recent studies from the United States have shown a higher incidence of clonal evolution to secondary myelodysplastic syndrome (MDS) and acute myeloid leukemia (AML) in AA patients with HLA(-) leukocytes than in those without such leukocytes, a finding inconsistent with the results of our previous study. Given the high prevalence of HLA(-) leukocytes in AA patients, it is critical to determine whether or not the presence of the aberrant leukocytes is associated with clonal evolution. We therefore addressed this issue by studying the prognosis of a large number of AA patients with or without HLA(-) leukocytes who had been followed for a long term period. We also studied the clonal composition of granulocytes in AA patients with HLA(-) cells, wherein aberrant clones other than HLA(-) cells might be responsible for clonal evolution to MDS/AML. [Methods] We retrospectively analyzed the clinical characteristics of 633 AA patients and peripheral blood samples were examined for the presence of HLA(-) leukocytes using a high-sensitivity flow cytometry (FCM) assay, droplet digital PCR, single-nucleotide polymorphism arrays, or next generation sequencing (NGS) between 2010 and 2020. GPI(-) cells were detected using a high-sensitivity FCM assay as previously described. [Results] HLA(-) granulocytes were detected in 127 (20.1%) of the 633 patients with a median clone size of 16.9% (range, 0.04%-100%); the aberrant granulocytes accounted for greater than 50% of the total granulocytes in 29 (22.8%) of 127 patients. Eighty-nine (70.0%) of the 127 patients possessed aberrant clones other than HLA(-) clones, which included 0.005% to 91.6% GPI(-) cells (n=86), del(13q) cells (n=3), t(1;10) cells (n=1), t(9;13) cells (n=1), inv12 cells (n=1), and trisomy 8 cells (n=1). The prevalence of GPI(-) cells was not significantly different between patients with and without HLA(-) cells (67.7% vs 65.4%). Eighty-five of 102 (83.3%) patients with HLA(-) cells responded to IST, whereas 231 of 318 (72.6%) without HLA(-) cells responded (p<0.05). In 13 patients who had been in hematological remission for more than 7 years, HLA(-) cells and other concomitant aberrant cells accounted for >90% of granulocytes, suggesting that these few escape clones were enough to sustain the hematopoietic function of the patients. The prognosis survey revealed no clonal evolution to MDS/AML in any of the 127 AA patients with HLA(-) leukocytes after a follow-up period of the median 5 years. In contrast, 15 of 234 (6.4%) patients without HLA(-) cells who were trackable evolved to MDS/AML during a median 5 year follow-up. [ Conclusions] The presence of HLA(-) leukocytes and concomitant aberrant clones was not associated with clonal evolution to MDS/AML in Japanese AA patients, even in those possessing a large (>50% of the total granulocyte) HLA(-) cell population. The discrepancy between our results and the data from the United States may be due to the difference in the race and mechanism underlying HLA loss. These data suggest that HSC clones that escape immune attack, such as HLA(-) and GPI(-) clones, are healthy enough to support hematopoiesis for a long term in AA patients. Disclosures Ishiyama: Novartis: Honoraria; Alexion: Research Funding. Yamazaki:Novartis: Honoraria; Kyowa Kirin: Honoraria, Research Funding. Ogawa:Eisai Co., Ltd.: Research Funding; Sumitomo Dainippon Pharma Co., Ltd.: Research Funding; Chordia Therapeutics, Inc.: Membership on an entity's Board of Directors or advisory committees, Research Funding; Asahi Genomics Co., Ltd.: Current equity holder in private company; Otsuka Pharmaceutical Co., Ltd.: Research Funding; KAN Research Institute, Inc.: Membership on an entity's Board of Directors or advisory committees, Research Funding. Nakao:Alexion: Research Funding; Kyowa Kirin: Honoraria; Novartis: Honoraria; Symbio: Consultancy.

Off-the-shelf EBV-specific T-cell Immunotherapy for EBV-associated PTLD.

Off-the-shelf EBV-specific T-cell Immunotherapy for EBV-associated PTLD.

Abstract 4537: SBT6050, a HER2-directed TLR8 ImmunoTAC™therapeutic, is a potent human myeloid cell agonist that provides opportunity for single agent clinical activity

Abstract SBT6050, a novel therapeutic comprised of a potent toll-like receptor (TLR) 8 agonist conjugated to a HER2-directed monoclonal antibody that binds an epitope distinct from trastuzumab, is designed for systemic delivery and tumor-localized activation of human myeloid cells in the presence of moderate and high HER2-expressing tumor cells. Many solid tumors, including those expressing HER2, are refractory to immunotherapy due to immune-suppressive mechanisms, loss of HLA, low neoantigen availability, and/or minimal T cell infiltrates. These tumors frequently contain abundant populations of tumor-associated myeloid cells. Activation of these cells through TLR agonism has emerged as a promising approach in overcoming resistance mechanisms to current cancer immunotherapies. Unlike other endosomal TLRs such as TLR7 and TLR9, TLR8 is highly expressed in human myeloid cells known to be prevalent in human tumors such as conventional DCs and macrophages. TLR8's restricted myeloid cell expression removes the risk of inducing T cell death or tumor cell proliferation as described for other innate immune activators such as STING. Agonism of TLR8 in human myeloid cells activates a broad spectrum of anti-tumor immune mechanisms, including proinflammatory cytokine production, repolarization of suppressive myeloid cells and the priming of CTL responses. These functions cannot be replicated by a potent TLR7 agonist or with clinical agents such as resiquimod that agonize TLR7 and only weakly engage TLR8. Here we present data demonstrating the superiority of SBT6050 at activating human myeloid cells compared to HER2 antibody conjugates that use either a selective TLR7 agonist or resiquimod. In vitro studies with human immune cells show that SBT6050 potently induces multiple anti-tumor immune activities, including proinflammatory cytokine and chemokine production, inflammasome activation, direct activation of DCs and indirect T and NK cell cytolytic activity. Our data indicate the favorable profile of SBT6050 is likely due to efficient engagement of TLR8 in conjugate form and TLR8's unique expression profile, and not due to differences in the potency of the small molecule payloads. Using an SBT6050 mouse surrogate we show in vivo evidence for intratumoral TLR agonist activation of several myeloid driven anti-tumor pathways leading to curative single agent efficacy in both a T cell-excluded syngeneic tumor model and a xenograft model lacking T, B and NK cells. Collectively, these data demonstrate that SBT6050 displays an attractive functional profile unachievable with agonist-based antibody conjugates directed against other innate immune receptors. These data also highlight the potential for SBT6050 to be clinically active as a monotherapy. SBT6050 is projected to enter the clinic in 2020 for patients with moderate or high HER2-expressing tumors. Citation Format: Michael R. Comeau, Ty Brender, Monica Childs, Jamie Brevik, Damion Winship, Heather Metz, Jenny Chang, Jeffrey Adamo, Ben Setter, Hengyu Xu, Li-Qun Fan, Brenda Stevens, Sean W. Smith, Phil Tan, Robert DuBose, Yvette Latchman, Peter Baum, Valerie Odegard. SBT6050, a HER2-directed TLR8 ImmunoTAC™therapeutic, is a potent human myeloid cell agonist that provides opportunity for single agent clinical activity [abstract]. In: Proceedings of the Annual Meeting of the American Association for Cancer Research 2020; 2020 Apr 27-28 and Jun 22-24. Philadelphia (PA): AACR; Cancer Res 2020;80(16 Suppl):Abstract nr 4537.

Abstract 3547: Frequent genomic alterations to evade the immune system in colorectal cancer with POLE gene mutation

Abstract Introduction POLE-mutated tumor is a rare subtype of colorectal cancer (POLE-CRC), whose pathogenesis has been poorly understood because of its rarity. In this study, we have investigated the clinical and genetic features of POLE-CRC in the largest cohort of this unique subtype of CRC ever analyzed. Methods We included a total of 3,265 patients who had been treated at the Cancer Institute Hospital of Japanese Foundation for Cancer Research between 2004 and 2017, which were screened for POLE mutations by targeted-panel sequencing. The POLE-mutated samples were further analyzed by whole exome sequencing. All samples were well-annotated for clinical information. Results In total, 44 samples showed prominent hypermutation with a median of 5,346 (range: 837-16,990) with the predominance of COSMIC signature 10 associated with defective POLE functions. By contrast, copy number variations (CNV) in POLE-CRC was much less common compared to non-hypermutated CRC. In accordance with previous reports, mutational hotspots in the POLE gene were found at codons 411, 286, 459, and 456. Patients with POLE-CRC were significantly younger (median of 50 years), compared with non-hypermutated CRC (median of 65 years, p<0.01) and MSI-CRC (median of 70 years, p<0.01). They also showed a significantly favorable prognosis than non-hypermutated CRC patients (p<0.01). Driver genes in 44 POLE-CRC cases were investigated using dNdSCV. A total of 9 genes showed significantly deviated dN/dS (q-value<0.01), of which APC (43/44; 97.7%) was most frequently mutated, followed by PIK3R1, PIK3CA, PTEN,TP53, ARID1A, B2M, TAP1 and CD58. Among these, two genes, B2M and TAP1 are components of the antigen presentation machinery (APM). We also found frequent LOH (8/44; 18%) affecting the 6p arm, containing HLAs, TAP1 and TAP2, also suggesting the role of compromised APM. Functional impact of core genetic events was investigated by immunohistochemistry of HLAs, B2M, TAP1 and TAP2. All samples with B2M biallelic mutations showed complete loss of B2M and HLAs. All samples with TAP1 biallelic mutations showed complete loss of TAP1 and TAP2 even with intact HLA expression, suggesting a defect in the loading of antigens onto HLAs. Also of note is frequent loss-of-function mutations in CD58, a molecule involved in NK-cell recognition. Loss of intact CD58 might play a role in the evasion from NK-cell surveillance for tumors with defective HLA presentation. Altogether, 75% (33/44) of POLE-CRC harbored alterations to evade the immune system though loss of relevant HLA or APM components as well as CD58. Conclusions POLE-CRC is a CRC subtype with a favorable prognosis compared with other CRC subtypes and is characterized by high frequency of genetic alterations affecting APM, which is implicated in immune evasion of this unique subtype of CRC with an extremely high mutational burden. Citation Format: Yoshikage Inoue, Nobuyuki Kakiuchi, Kenichi Yoshida, Yasuhide Takeuchi, Yusuke Shiozawa, Yuichi Shiraishi, Kenichi Chiba, Tetsuichi Yoshizato, Hiroko Tanaka, Ai Okada, Satoshi Nagayama, Satoru Miyano, Yoshiharu Sakai, Seishi Ogawa. Frequent genomic alterations to evade the immune system in colorectal cancer with POLE gene mutation [abstract]. In: Proceedings of the Annual Meeting of the American Association for Cancer Research 2020; 2020 Apr 27-28 and Jun 22-24. Philadelphia (PA): AACR; Cancer Res 2020;80(16 Suppl):Abstract nr 3547.

Abstract 6678: Sensitive HLA loss of heterozygosity detection reveals allele-specific neoantigen expansion as resistance mechanism to anti-PD-1 therapy

Abstract Loss of heterozygosity (LOH) in the HLA locus is increasingly being recognized as an important mechanism of immune escape and a proposed biomarker for immunotherapy response. Neoantigens that bind to a deleted HLA allele will no longer be presented to the immune system, potentially allowing subclones with these deletions to escape immune surveillance. Despite interest in the field, few methods exist to detect HLA LOH, and their sensitivity is not well understood. Moreover, the mechanistic impact of HLA LOH in response to immune checkpoint inhibitors (ICI) remains unexplored. Here, using a novel tool to detect HLA LOH, DASH (Deletion of Allele-Specific HLAs), we reveal allele-specific neoantigen expansion in response to ICIs in a head and neck carcinoma cohort and the widespread occurrence of HLA LOH across several tumor types. We performed exome sequencing with ImmunoID NeXT, which enhances coverage of the HLA locus, on tumor and normal samples from 260 patients to create a training dataset for our model. For each patient, we mapped reads to each of their allele-specific HLAs and manually annotated LOH. Then, using purity, ploidy and two novel features (normalized b-allele frequency and allele-specific coverage ratios), we trained an XGBoost model. To evaluate our tool, we compared DASH predictions on held out tumor samples to deletion calls from a standard copy number tool in the regions flanking each HLA gene (91% concordance, 0.73 F1-score), ascertained our sensitivity by diluting cell line data with known HLA LOH to imitate variable purity (100% accuracy in samples above 17% purity) and confirmed the functional impact of HLA LOH using immunopeptidomics data of tumor samples (average of 47% fewer unique peptides binding to lost alleles than kept). To explore the mechanistic impact of HLA LOH in response to ICIs, we studied a cohort of nine head and neck carcinoma patients who received a single dose of nivolumab, and sequenced pre- and posttreatment tumor biopsies for each patient. With DASH, we detected HLA LOH in four of the patients, pretreatment. For these patients, we found a significant posttreatment expansion of neoantigens predicted to bind to the deleted HLA alleles in comparison to the pretreatment biopsy (p=0.046, Wilcoxon signed-rank), revealing the evolutionary force of HLA LOH as a resistance mechanism during ICI therapy. To assess the pervasiveness of HLA LOH across tumor types, we applied DASH to over 500 pretreatment tumors across 13 tumor types and found highly variable frequencies of HLA LOH across the tumor types. In summary, we developed a sensitive method to detect HLA LOH and exposed neoantigen expansion to deleted HLA alleles in response to ICI therapy, emphasizing the limitations of deleted alleles to ignite an immune response. Moreover, we found widespread occurrences of HLA LOH across tumor types, highlighting the importance of accurate HLA LOH detection as a pan-cancer biomarker. Citation Format: Rachel Marty Pyke, Charles Abbott, Dattatreya Mellacheruvu, Simo V. Zhang, Nikita Bedi, A. Dimitrios Colevas, John Sunwoo, John West, Richard Chen, Sean Michael Boyle. Sensitive HLA loss of heterozygosity detection reveals allele-specific neoantigen expansion as resistance mechanism to anti-PD-1 therapy [abstract]. In: Proceedings of the Annual Meeting of the American Association for Cancer Research 2020; 2020 Apr 27-28 and Jun 22-24. Philadelphia (PA): AACR; Cancer Res 2020;80(16 Suppl):Abstract nr 6678.

Abstract 2085: Precision neoantigen discovery using a pan-allelic machine learning model for enabling the development of composite biomarkers and personalized immunotherapy

Abstract Background: Technologies for neoantigen discovery are critical for developing more advanced, composite biomarkers for immunotherapy as well as personalized cancer therapies. Precision neoantigen discovery entails comprehensive detection of tumor specific genomic variants and accurate prediction of MHC presentation of epitopes originating from such variants. Here we present our pan-allelic machine learning model for predicting MHC class I presentation and identifying potentially immunogenic patient-specific neoantigens. We then apply our predictions to develop composite biomarkers that can stratify patients by response to immunotherapy. Methods: Mono-allelic cell lines were generated by transfecting a single allele of interest into HLA-null K562 cell lines. Immunoprecipitation mass spectrometry (IPMS) was performed as follows: W6/32 antibody was used for immunoprecipitation of HLA complexes, followed by elution of bound peptides and identification using liquid chromatography-mass spectrometry. Machine learning models were implemented to predict neoantigen presentation. Our prediction model, integrated into the ImmunoID NeXT platform, was used with other genomic features to generate composite neoantigen-based biomarkers. Results: We generated a high quality and unambiguous immunopeptidomics training dataset by performing IPMS on ~60 mono-allelic cell lines, with ongoing efforts to expand it to ~100 alleles. These alleles, selected to optimize both allelic diversity and population coverage, enable accurate and comprehensive modeling of MHC ligand processing and presentation. Our advanced prediction model combines multiple modelling strategies, including deep neural networks, convolutional neural networks and gradient boosted decision trees. New features that model antigen-processing were implemented using large scale public and private datasets to improve presentation specificity (60% PPV in dataset containing 999-fold decoys). As a result, our pan-allelic model has significantly higher specificity across a range of sensitivity values in comparison to NetMHCPan 4.0 and generalizes to both trained and untrained alleles. Our comprehensive validation strategy includes: evaluation of overall performance of the model on an independent, multi-allelic immunopeptidomics dataset generated from tumor samples; validation of top ranking tumor specific neoantigens nominated using an integrated patient-centric model that incorporates HLA loss of heterozygosity using targeted proteomics (Parallel Reaction Monitoring); and evaluation of our model's utility to predict neoantigens that drive immunogenic responses to tumors. Finally, a composite neoantigen-based biomarker score calculated using our model stratifies patients by response to immunotherapy. Conclusions: In summary, we present here a pan-allelic MHC presentation prediction model trained on a large mono-allelic data set and evaluated using tumor samples and known immunogenic peptides. Through integration with ImmunoID NeXT, our platform enables precision neoantigen discovery by comprehensively surveying neoantigens and accurately predicting MHC presentation. These methods can significantly enhance the creation of composite biomarkers and applications in personalized immunotherapy. Citation Format: DATTATREYA MELLACHERUVU, Rachel Marty Pyke, Charles Abbott, Nick Phillips, Rena McClory, John West, Richard Chen, Sean Michael Boyle. Precision neoantigen discovery using a pan-allelic machine learning model for enabling the development of composite biomarkers and personalized immunotherapy [abstract]. In: Proceedings of the Annual Meeting of the American Association for Cancer Research 2020; 2020 Apr 27-28 and Jun 22-24. Philadelphia (PA): AACR; Cancer Res 2020;80(16 Suppl):Abstract nr 2085.

Abstract PR10: Analytical validation and preliminary clinical utility of PD-L1 and HLA I expression profiling of circulating tumor cells using automated exclusion-based sample preparation technology

Abstract Only a subset of patients with non-small cell lung cancer (NSCLC) respond to PD-L1 targeted therapy, even after solid tumor biopsy screening for PD-L1 expression. Downregulation of HLA I may contribute to this high prevalence of resistance, warranting its inclusion in attempts to increase response rates. While it is unclear if loss of HLA I occurs as a de novo or acquired resistance mechanism to immunotherapy, liquid biopsies could detect dynamic changes in expression levels, enable early detection of resistance, and accelerate the shift to alternate, more effective treatments. To quantify the expression of both PD-L1 and HLA I on circulating tumor cells (CTCs), we have developed automated exclusion-based sample preparation technology (ESP), which leverages the forces of surface tension and magnetism at the microscale for high-efficiency, high-throughput cell capture and manipulation. On-chip fluorescent antibody staining is performed for CTC identification, enumeration, and analysis of protein expression. A quantitative readout of protein expression is performed using a refined fluorescence microscopy approach wherein images of cells are acquired and analyzed with automated algorithms for uniformity and objectivity. To evaluate precision, accuracy, and analytical sensitivity of both ESP processing and quantitative microscopy, we used standardized cell lines, synthetic beads, and patient sample replicates. Calibration beads demonstrate high precision between multiple users for the quantitative microscopy (0.9% CV max of all wavelengths between 3 different analysts). ELISA beads with serial dilutions of protein standards demonstrate the accuracy of the quantitative microscopy (R2 values of 0.99 for both proteins). Cell lines demonstrate the precision of ESP for protein staining (3.4% CV max of two different cell lines tested on three different days). Patient sample replicates demonstrate precision for patients with as few as 3 CTCs per sample (6% CV max between triplicates). Early clinical utility of this assay was demonstrated on a preliminary patient cohort of 21 patient time points from 8 unique patients with NSCLC treated with immunotherapies. Phenotypic analysis for CTC expression of PD-L1 and HLA I with this assay demonstrated a sensitivity of 94% and specificity of 75% for detecting either resistant or stable disease during treatment with immunotherapies. Preliminary evaluation of patients with clear cell renal cell carcinoma demonstrates the feasibility of applying this approach to other cancer types. By confirming high performance, we ensure compatibility for clinical laboratory implementation and eventual application to better predict and detect resistance to PD-L1 targeted therapy in patients with NSCLC. The preliminary patient cohort suggests this diagnostic approach can detect fluctuations in expression of PD-L1 and HLA I in CTCs that also correspond to patient response to immunotherapy and warrants its pursuance in larger cohorts of patients with NSCLC and other tumor types. This abstract is also being presented as Poster B38. Citation Format: Jennifer L. Schehr, Zachery D. Schultz, Camila I. Hernandez, Matthew C. Mannino, Jay W. Warrick, Ticiana A. Leal, David J. Beebe, Joshua M. Lang. Analytical validation and preliminary clinical utility of PD-L1 and HLA I expression profiling of circulating tumor cells using automated exclusion-based sample preparation technology [abstract]. In: Proceedings of the AACR Special Conference on Advances in Liquid Biopsies; Jan 13-16, 2020; Miami, FL. Philadelphia (PA): AACR; Clin Cancer Res 2020;26(11_Suppl):Abstract nr PR10.

A frequent nonsense mutation in exon 1 across certain HLA-A and -B alleles in leukocytes of patients with acquired aplastic anemia.

Leukocytes that lack expression of HLA alleles are frequently detected in patients with acquired aplastic anemia (AA) who respond to immunosuppressive therapy, although the exact mechanisms underlying the HLA loss and HLA allele repertoire likely to acquire loss-of-function mutations are unknown. We identified a common nonsense mutation at codon 19 (c.19C>T, p.R7X) in exon 1 (Exon1mut) of different HLA-A and -B alleles in HLA-lacking granulocytes from AA patients. A droplet digital polymerase chain reaction assay capable of detecting as few as 0.07% Exon1mut HLA alleles in total DNA revealed that the mutation was present in 29% (101/353) of AA patients, with a median allele frequency of 0.42% (range, 0.071% to 21.3%). Exon1mut occurred in only 12 different HLA-A (n=4) and HLA-B (n=8) alleles, including B*40:02 (n=31) and A*02:06 (n=15), which correspond to four HLA class I supertypes (A02, A03, B07, and B44). The percentages of patients who possessed at least one of these 12 HLA alleles were significantly higher in the 353 AA patients (92%, P<0.001) and in 83 AA patients with copy number neutral loss of heterozygosity in chromosome 6p (100%, P<0.001) than the percentage (81%) in 18,604 Japanese healthy individuals. Eighty-two percent (37/45) of AA patients with Exon1mut responded to immunosuppressive therapy. Small populations of leukocytes that lack particular HLA-A or B alleles due to Exon1mut are common in AA patients. The detection of Exon1mut using a droplet digital polymerase chain reaction assay without the need for HLA typing may serve as a powerful tool for diagnosing the immune pathophysiology of patients with bone marrow failure.

From the Editor’s Desk…

From the Editor’s Desk…

Abstract PR12: Genomic landscape of immunotherapy resistance in lung cancer

Abstract Immune checkpoint inhibitors (ICI) have emerged as an important modality in lung cancer treatment. However, studies have shown that only a fraction of patients will experience any clinical benefit when treated with ICI, and a better understanding of the mechanisms of resistance is still needed. In this study, we sought to characterize the prevalence of known mechanisms of primary and adaptive resistance in the context of the greater immunogenomic landscape in a real-world cohort of 100 non-small cell lung cancer patients treated with ICI. Targeted panel DNA sequencing and whole-transcriptome RNA sequencing were performed on a primary tumor biopsy for each patient, the clinical notes were abstracted and structured, and the digital image of the H&amp;E was processed through a tumor-infiltrating lymphocyte (TIL) analysis pipeline. We evaluated each tumor for evidence of HLA loss of heterozygosity (LOH), mutations in the antigen presentation or interferon gamma signaling pathways, expression of alternative checkpoint molecules, and immune infiltrate exclusion. We found that over 25% of patients in the cohort showed signs of at least one of these mechanisms of ICI resistance. We evaluated the real-world progression-free survival (rwPFS) for these patients and found that the majority experienced a progression event within 6 months of treatment start. This study suggests that primary resistance to ICI affects a significant population of lung cancer patients. This abstract is also being presented as Poster B22. Citation Format: Denise Lau, Michelle Stein, Jason Perera, Madhavi Kannan, Aly Khan. Genomic landscape of immunotherapy resistance in lung cancer [abstract]. In: Proceedings of the AACR Special Conference on Tumor Immunology and Immunotherapy; 2019 Nov 17-20; Boston, MA. Philadelphia (PA): AACR; Cancer Immunol Res 2020;8(3 Suppl):Abstract nr PR12.

Mechanisms of Leukemia Immune Evasion and Their Role in Relapse After Haploidentical Hematopoietic Cell Transplantation.

Over the last decade, the development of multiple strategies to allow the safe transfer from the donor to the patient of high numbers of partially HLA-incompatible T cells has dramatically reduced the toxicities of haploidentical hematopoietic cell transplantation (haplo-HCT), but this was not accompanied by a similar positive impact on the incidence of post-transplantation relapse. In the present review, we will elaborate on how the unique interplay between HLA-mismatched immune system and malignancy that characterizes haplo-HCT may impact relapse biology, shaping the selection of disease variants that are resistant to the “graft-vs.-leukemia” effect. In particular, we will present current knowledge on genomic loss of HLA, a relapse modality first described in haplo-HCT and accounting for a significant proportion of relapses in this setting, and discuss other more recently identified mechanisms of post-transplantation immune evasion and relapse, including the transcriptional downregulation of HLA class II molecules and the enforcement of inhibitory checkpoints between T cells and leukemia. Ultimately, we will review the available treatment options for patients who relapse after haplo-HCT and discuss on how a deeper insight into relapse immunobiology might inform the rational and personalized selection of therapies to improve the largely unsatisfactory clinical outcome of relapsing patients.

Genomic and immune profiling of pre-invasive lung adenocarcinoma

Adenocarcinoma in situ and minimally invasive adenocarcinoma are the pre-invasive forms of lung adenocarcinoma. The genomic and immune profiles of these lesions are poorly understood. Here we report exome and transcriptome sequencing of 98 lung adenocarcinoma precursor lesions and 99 invasive adenocarcinomas. We have identified EGFR, RBM10, BRAF, ERBB2, TP53, KRAS, MAP2K1 and MET as significantly mutated genes in the pre/minimally invasive group. Classes of genome alterations that increase in frequency during the progression to malignancy are revealed. These include mutations in TP53, arm-level copy number alterations, and HLA loss of heterozygosity. Immune infiltration is correlated with copy number alterations of chromosome arm 6p, suggesting a link between arm-level events and the tumor immune environment.

Exploiting an Anti-CD3/CD33 Bispecific Antibody to Redirect Donor T Cells Against HLA Loss Leukemia Relapses

Background Genomic loss of mismatched HLAs ("HLA loss") represents a frequent modality by which acute myeloid leukemia (AML) evades immune recognition from donor T cells after partially HLA-incompatible allogeneic hematopoietic cell transplantation (allo-HCT). One important consequence of this post-transplantation relapse mechanism is that infusions of lymphocytes from the original donor become ineffectual, prompting the search for alternative therapeutic options. Here, to circumvent the loss of physiological T cell receptor-HLA interactions in these patients, we tested the ability of an anti-CD3/CD33 bispecific antibody (BsAb) to re-target donor T cells towards HLA loss relapses. Methods For short-term in vitro experiments, T cells were co-cultured with the MOLM-13 AML cell line or with primary patient blasts for 96 hours in presence or absence of an anti-CD3/CD33 BsAb. As readouts, we measured T cell activation (as surface expression of CD25 and CD69) and the absolute counts and relative proportion of effectors and targets. For long-term in vitro experiments, we established mixed lymphocyte cultures (MLCs) of T cells purified from two patients after haploidentical HCT and primary AML blasts obtained from the same patients at the time of diagnosis. After sequential stimulations, the co-cultures were tested against targets of interest, with or without addition of the BsAb. Functional readouts were T cell degranulation (measured as CD107a expression), antigen-specific activation (as CD137/41-BB expression) and target-specific cytotoxicity (measured by time-lapse live cell imaging over a 48 hour time span). For in vivo experiments, human leukemic cells were infused intravenously into non-irradiated NSG mice, followed by intraperitoneal infusion of T cells and daily administration of the BiTE compound. Results First, we retrospectively analyzed immunophenotypic data of 36 AML patients who experienced HLA loss relapses at our Institution, documenting robust expression of CD33 on the surface of the relapsed leukemia in 35 of them (97%; Figure 1A). By short-term co-culture experiments we titrated the BsAb concentration to be used for subsequent in vitro assays to 100 ng/ml, and the most informative effector:target ratio to 1:3. Then, we established MLCs by stimulating T cells collected from two patients after partially HLA-incompatible allo-HCT with AML blasts collected from the same patients at the time of diagnosis. In both cases, donor-derived T cells robustly responded against the patient blasts both in term of degranulation (Figure 1B) and of antigen-specific activation (Figure 1C). As expected, when we tested the same T cells against the patient leukemia at time of HLA loss relapse, we detected no T cell-mediated responses. Noticeably, when the BsAb was added, in both cases we detected a strong response not only against the diagnosis but also against the HLA loss variants, indicating that T cells were effectively re-targeted towards leukemic cells. Similar results were obtained also by live cell imaging, measuring target cell apoptosis over 48 hours of recording: also in this assay, in fact, donor T cells recognized and killed leukemia at diagnosis (45% of detection area positive for apoptosis dye) and failed to recognize its HLA loss relapse counterpart (32% of area positive for apoptosis dye). Addition of the BsAb to the co-cultures had a minor effect on recognition of the original disease (45% of area positive for apoptosis dye) but drove dramatic cell death of HLA loss blasts (80% of area positive for apoptosis dye), demonstrating that the BsAb induced not only T cell activation but also and most importantly target cell killing (Figure 1D). Finally, we modeled the BsAb activity in vivo, showing that, whereas the sole infusion of human T cells is not able to prevent the outgrowth of leukemia in the bone marrow of NSG mice, addition of the bispecific antibody leads to effective disease clearance (Figure 1E). Conclusions Our results demonstrate that anti-CD3/CD33 BsAbs can effectively redirect donor T cells against HLA loss leukemia variants, resulting in their rapid and effective killing. Taken together, these promising findings strongly support translation of this approach to ad hoc designed early-phase clinical trials, to provide a rational therapy for this increasingly recognized but still treatment-orphan modality of post-transplantation relapse. Figure 1 Disclosures Subklewe: Janssen: Consultancy; Miltenyi: Research Funding; Pfizer: Consultancy, Honoraria; Oxford Biotherapeutics: Research Funding; Gilead: Consultancy, Honoraria, Research Funding; Celgene: Consultancy, Honoraria; Morphosys: Research Funding; Roche: Consultancy, Research Funding; AMGEN: Consultancy, Honoraria, Research Funding. Vago:Moderna Therapeutics: Research Funding; GenDx: Research Funding.

Mapping the Evolution of the Mutational Landscape of Acute Myeloid Leukemia from Diagnosis to Post-Transplantation Relapse and Its Interplay with Immune Evasion Mechanisms

Background Over the last decade, the advent of next-generation sequencing (NGS) has dramatically improved our understanding of the genomic landscape that characterize acute myeloid leukemia at time of diagnosis, leading to the identification of new prognostic subsets and therapeutic targets. Still, few studies have addressed how this landscape changes upon treatment, and in particular after allogeneic hematopoietic cell transplantation (allo-HCT). Moreover, the links between the mutational profile of AML and mechanisms of post-transplantation immune evasion remain to date largely unknown. Methods Taking advantage of the HaloPlex High Sensitivity (HS) technology, which allows a more precise definition of mutations and clonal architecture through the use of unique molecular identifiers, we designed and tested a NGS panel covering genes and miRNAs known to be involved in the pathogenesis of myeloid malignancies (n=112), in the DNA damage response (n=50) or in immune-related processes (n=30). By using an Illumina HiSeq2500n instrument,we sequenced a total of 200 AML samples collected from 131 patients, including 84 diagnoses, 45 relapses after chemotherapy (CT) and 71 relapses after allo-HCT, 31 of which characterized by genomic loss of HLA (Vago et al, N Engl J Med, 2009). Variant calling was performed using a pipeline based on the FreeBayes algorithm. Results Sequencing coverage was uniform across all target amplicons, with a 855x average depth-of-sequencing, and a mean of 122 unique barcodes for each region. Among the 84 diagnosis samples we identified 253 oncogenic mutations, with a median of 3 mutations per patient (range 0-8). Both the frequency and the prognostic value of mutations detected at diagnosis were in line with the largest published dataset on AML genomics (Papaemmanuil et al, N Engl J Med, 2016; R2= 0.88, p&amp;lt;0.0001). The overall burden of oncogenic mutations did not change significantly neither at relapses after CT nor after allo-HCT. Analyzing the profile of known oncogenic drivers of diagnoses and of the two relapse cohorts we documented high similarity between the three time-points, with significant differences limited to FLT3-ITD (present in 29% patients at diagnosis vs 49% and 48% at relapses after CT and allo-HSCT, p=0.02 and p=0.01 respectively) and WT1 mutations (5% vs 20% and 28%, p=0.006 and p&amp;lt;0.0001 respectively; Figure 1A). Also when comparing post-transplantation HLA loss relapses with their "classical" counterparts we could find no significant difference in the frequency or co-occurrence pattern of known AML driver mutations. Still, in HLA loss relapses we documented a significantly higher frequency of variants of unknown oncogenic potential befalling in genes implicated in the DNA damage response, including those belonging to the Fanconi Anemia pathway (35% of mutations in HLA loss vs 10% in classical relapses, p&amp;lt;0.0001; Figure 1B). This intriguing observation suggests that, in these cases, double strand breaks might be more frequently repaired through homologous recombination, and that this in turn might result in a higher likelihood of copy-neutral loss of heterozygosity events, the molecular mechanism at the basis of HLA loss. Finally, thanks to our panel, we were able to track possible routes of leukemic clonal evolution between diagnosis and post-transplantation relapse: interestingly, in several cases emergence of an HLA loss subclone out-competed clones carrying known oncogenic drivers, suggesting that after allo-HCT the immune privilege might outperform the proliferative advantage (Figure 1C). Discussion Taken together, our data highlight a complex interplay between the genomic landscape of AML and the environmental selective pressure represented by the graft-versus-leukemia effect. Whereas in fact some genomic alterations that carried little consequence before transplant become able in this setting to promote specific patterns of immune escape, others, that had driven AML biology before, turn out to be dispensable after allo-HCT. Understanding and tracking these evolutionary dynamics may provide new tools and targets to treat, or even prevent, AML post-transplantation relapses. Figure 1 Disclosures Sala: Gilead Company: Honoraria. Gentner:Genenta Science: Consultancy, Equity Ownership, Research Funding. Vago:GenDx: Research Funding; Moderna Therapeutics: Research Funding.

Integrated Epigenetic Profiling Identifies EZH2 As a Therapeutic Target to Re-Establish Immune Recognition of Leukemia Relapses with Loss of HLA Class II Expression

Background It is becoming increasingly recognized that evasion from immune control represents one of the main drivers of acute myeloid leukemia (AML) relapse after allogeneic hematopoietic cell transplantation (allo-HCT). In particular, alterations in the antigen processing and presentation machinery represent one of the most effective strategies enacted by tumor cells to avoid recognition from T cells. Whereas it is now well recognized that genomic loss of HLA is frequently at the basis of post-transplantation relapse, it was only recently reported that up to 40% of AML relapses display transcriptional downregulation and complete loss of surface expression of HLA class II molecules without any genetic lesion explaining this phenotype (Christopher et al, N Engl J Med, 2018; Toffalori et al, Nat Med, 2019). This led us to investigate the links between epigenetic changes, immune evasion and post-transplantation relapse. Methods Starting from primary AML samples pairwise collected from patients at diagnosis and relapse with non-genomic loss of HLA class II expression, we generated Patients-Derived Xenografts (PDXs) into NOD-SCID γ-chain null mice. Leukemic cells expanded in the mice and their original human counterparts were analyzed for surface expression of selected immune-related markers (HLA class I and II, PD-L1, B7-H3), and characterized for changes in gene expression (by RNA-Seq), DNA methylation profile (by RRBS), histone modifications associated with active promoters (H3K4me3) or regulatory elements (H3K27ac) (by ChIP-seq) and chromatin accessibility (by ATAC-Seq). The results obtained by all these approaches were integrated by Multi-Omics Factor Analysis (MOFA), followed by Gene Set Enrichment Analysis (GSEA). Finally, the immunological effects of epigenetic drugs and recombinant immune-modulatory cytokines on primary and PDX-derived AML samples were tested in ex-vivo short-term cultures on a layer of mesenchymal stromal cells. Results We verified that PDXs faithfully recapitulate immune-related differences between diagnosis and post-transplantation relapse, including loss of expression of HLA class II molecules (Figure 1A). Integration of all the high-throughput technologies by MOFA evidenced that the differences between diagnosis and post-transplantation relapse samples were mostly explained by changes in chromatin accessibility and histone marks, and largely unrelated to the DNA methylation profile (Figure 1B). We documented that the gene sets that emerged upon integrating epigenetic analyses by MOFA matched our previously published immune-related relapse signature (Toffalori et al, Nat Med, 2019) but also and most intriguingly lists of genes known to be targeted by EZH2, the enzymatic subunit of the PRC2 chromatin repressor complex (Figure 1C). These computational analyses were supported by the evidence of a relapse-specific closed chromatin status of HLA class II genes and their regulators (Figure 1D). To revert these epigenetic changes, we inhibited EZH2 with tazemetostat (EPZ-6438), an epigenetic drug currently being tested in early-phase clinical trials for lymphomas, in two AML relapses with non-genomic loss of HLA class II expression. EZH2 inhibition reduced the levels of the repressor mark H3K27me3 (Figure 1E), increased the surface expression of HLA class II molecules on leukemia cells (Figure 1F) and ultimately improved leukemia recognition by CD4+ T cells (Figure 1G). Notably, these effects were even more pronounced when EZH2 inhibition was combined with IFN-g treatment (Figure 1F,G), suggesting synergism between this epigenetic compound and cytokines released by immune cells upon target recognition. Conclusions Our results provide mechanistic insights into epigenetic regulation of HLA class II downregulation in leukemia and a strong therapeutic rationale to test EZH2 inhibition as an innovative strategy for the treatment of AML post-transplantation relapses. Figure 1 Disclosures Vago: GenDx: Research Funding; Moderna Therapeutics: Research Funding.