ABSTRACT:Children with ETV6::RUNX1 or high-hyperdiploid B-cell acute lymphoblastic leukemia (B-ALL) have favorable outcomes. The St. Jude (SJ) classification considers these patients low risk, regardless of their National Cancer Institute (NCI) risk classification, except when there is slow minimal residual disease (MRD) response or central nervous system/testicular involvement. We analyzed outcomes in children (aged 1-18.99 years) with these genotypes in the SJ Total XV/XVI studies (2000-2017). Patients with ETV6::RUNX1 (n = 222) or high-hyperdiploid (n = 296) B-ALL had 5-year event-free survival (EFS) of 97.7% ± 1.1% and 94.7% ± 1.4%, respectively. For ETV6::RUNX1, EFS was comparable between NCI standard-risk and high-risk patients and between SJ low-risk and standard-risk patients. Of the 40 NCI high-risk patients, 37 who received SJ low-risk therapy had excellent EFS (97.3% ± 2.8%). For high-hyperdiploid B-ALL, NCI high-risk patients had worse EFS than standard-risk patients (87.6% ± 4.5% vs 96.4% ± 1.3%; P = .016). EFS was similar for NCI standard-risk and high-risk patients classified as SJ low risk (96.0% ± 1.5% and 96.9% ± 3.2%; P = .719). However, EFS was worse for NCI high-risk patients than for NCI standard-risk patients receiving SJ standard/high-risk therapy (77.4% ± 8.2% vs 98.0% ± 2.2%; P = .004). NCI high-risk patients with ETV6::RUNX1 or high-hyperdiploid B-ALL who received SJ low-risk therapy had lower incidences of thrombosis (P = .013) and pancreatitis (P = .011) than those who received SJ standard/high-risk therapy. MRD-directed therapy yielded excellent outcomes, except for NCI high-risk high-hyperdiploid B-ALL patients with slow MRD response, who require new treatment approaches. Among NCI high-risk patients, 93% with ETV6::RUNX1 and 54% with high-hyperdiploid B-ALL experienced excellent outcomes with a low-intensity regimen. These trials were registered at www.clinicaltrials.gov as #NCT00137111 and #NCT00549848.

Relapse remains the major cause of mortality in pediatric acute myeloid leukemia (AML). Although initiating driver alterations are major determinants of clinical outcomes, the extent to which relapse emerges through clonal selection of pre-existing subpopulations or by the acquisition of de novo mutations remains unclear. We established a cohort of 39 diagnosis-relapse (D-R) and 2 relapse-relapse (R-R) pairs of pediatric AML. We comprehensively profiled the cohort using whole-genome sequencing (WGS) and target-capture sequencing. WGS (median coverage: 49x) and capture sequencing (450x) were performed to detect single-nucleotide variants (SNVs), insertions and deletions (indels), structural variants (SVs), and copy number variants (CNVs). Somatic alterations with variant allele frequencies (VAFs) exceeding background levels were regarded as real (P<0.05 by binomial tests), and coding genes were classified according to pathogenicity to assign pathogenic or likely pathogenic (P/LP) labels. Enrichment was assessed by Fisher's exact test. The study cohort represented major high-risk subtypes of pediatric AML, including KMT2A-rearranged (n=13, 32%) and NUP98-rearranged (n=6, 15%) AMLs. A total of 22,088 somatic alterations were identified across SNVs, indels, SVs, and CNVs. Among 142 P/LP mutations, WT1 (n=19) and FLT3 (ITD: n=9, TKD: n=7, other: n=3) were most commonly mutated, followed by NRAS (n=17) and SETD2 (n=9) mutations. The mutational burden was significantly higher at relapse (median 216 vs 376 mutations per sample at D-R, P=9.4×10⁻⁶), whereas the total number of P/LP mutations per sample did not significantly differ (median 4 at both, P=0.34). Among P/LP mutations, 39 had significantly higher VAF at relapse (P<0.05), with frequent involvement of FLT3 (n=7), WT1 (n=5), and SETD2, TP53, CREBBP, and NRAS (n=2 for each), indicating a growth advantage of clones with these alterations. Notably, 107 P/LP mutations were found at diagnosis, and only 35 P/LP mutations were specifically detected at relapse while undetectable at diagnosis, even by targeted sequencing, suggesting acquisition during or after therapy, or selection of rare pre-existing clones below the detection limit estimated to be 1%. Interestingly, clones with these mutations did not always expand at relapse, with different WT1 or FLT3 mutations depleted or enriched within an individual patient. In contrast, NRAS mutations were frequently depleted at relapse, suggesting that clones with NRAS mutations are sensitive to chemotherapy. We further investigated changes in clonal structures of the disease during progression. VAF at diagnosis and relapse showed clustering of mutations with specific patterns in each patient, and all relapses were clonally related to the paired AML at diagnosis. For example, initiating and category-defining alterations were identified with high VAF at both time points. In most cases (n=35, 85.3%), we observed that non-initiating mutations with high VAF (>20%) at diagnosis were frequently depleted to undetectable levels in the later time points. Contrarily, subclones marked by somatic mutations that were undetectable or with low VAF (<5%) at diagnosis became dominant in 38 cases (92.7%), indicating that leukemic cells were depleted down to only a few persistent genetic clones during remission. One exception is a case with FUS::ERG AML, which showed concordant VAF of somatic mutations at both diagnosis and relapse, including mutations with low VAF, suggesting that the entire population was resistant to chemotherapy with no therapy-induced bottleneck. Notably, mutations that mark subclones fated for relapse were mostly non-pathogenic or non-coding mutations, and 16 cases did not have any P/LP coding cooperating mutations that mark the relapsed clone. These data indicate that relapse of pediatric AML is driven by two mechanisms: one is a selection or expansion of clones with somatic alterations that confer chemo-resistance or growth advantage over the rest of the clones, and the other is a stochastic, non-genetic mechanism that likely confers a competitive advantage to certain clones, such as transcriptional heterogeneity. Bulk and single cell transcriptional analyses and clonal tracking using ultra-deep sequencing (~5000x) may reveal additional insights into the mechanisms of pediatric AML relapse.

ABSTRACT:Leukemias with NUP98 rearrangements exhibit heterogeneous phenotypes such as acute myeloid leukemia, T-cell acute lymphoblastic leukemia (T-ALL), or myelodysplastic syndrome/neoplasms associated with fusion partners, whereas the mechanism responsible for this heterogeneity is poorly understood. Through genome-wide mutational and transcriptional analyses of 177 NUP98-rearranged leukemias, we show that cooperating alterations are associated with differentiation status even among leukemias sharing the same NUP98 fusions, such as NUP98::KDM5A acute megakaryocytic leukemia with RB1 loss or T-ALL with NOTCH1 mutations. CUT&RUN profiling of in vitro cord blood CD34+ cell (cbCD34) models of major NUP98 fusions revealed that NUP98-fusion oncoproteins (FOs) directly regulate differentiation-related genes contributing to the disease phenotypes, represented by NUP98::KDM5A binding to MEIS2 or GFI1B for megakaryocyte (MK) differentiation. In patient samples, NUP98-FO binding patterns are heterogeneous, potentially shaped by somatic mutations and differentiation status. Using cbCD34 models and CRISPR/Cas9 gene editing, we show that RB1 loss cooperates with NUP98::KDM5A by blocking terminal differentiation toward platelets and expanding MK-like cells, whereas WT1 frameshift mutations skew differentiation toward dormant lymphoid-myeloid primed progenitor cells and cycling granulocyte-monocyte progenitor cells, providing evidence for NUP98-rearranged leukemia phenotypes affected by cooperating alterations. NUP98::KDM5A cbCD34 models with RB1 or WT1 alterations have different sensitivities to menin inhibition, suggesting that cellular differentiation provides stage-specific menin dependencies and resistance mechanisms that can be leveraged for future treatment strategies for NUP98-rearranged leukemia.

Acute Myeloid Leukemia (AML) is an aggressive hematologic cancer with poor treatment outcomes in pediatric patients, particularly due to high relapse rates and drug resistance. Standard treatment involves cytarabine (Ara-C), daunorubicin, and etoposide (ADE) but resistance to these drugs remains a significant challenge. To better understand the genetic factors influencing drug resistance and treatment outcomes, we conducted a pharmacogenomic analysis of genetic variants in AML-related and drug-metabolizing genes within the context of (ADE) chemotherapy regimen. We extracted SNPs in genes involved in pharmacological pathways of ADE as well as genes representative of myeloid panel from an existing set of genome-wide Illumina 2.5 Omni microarray data obtained from genomic DNA of 400 patients treated on multi-site AML02 [NCT00136084, n=167] and AML08 [NCT00703820, n=233] clinical trials. SNPs within +/-10 kb region of the gene locus were evaluated for associations with multiple clinical endpoints such as minimal residual disease after induction 1 (MRD1), event-free survival (EFS), and overall survival (OS), adjusting for genetic ancestry. Our analysis results showed that at a p value of <0.01, 215 SNPs in 52 genes were predictive of at least one of the outcomes. SNPs in solute carrier (SLC) transporters, including SLC22A1, SLC28A1, and SLC28A3, were linked to increased MRD1 positivity, indicating poor prognosis. SLC28A1 and A3 have been implicated in uptake of nucleoside analogs such as Ara-C. In contrast, SNPs in SLCO1B1 were predictive of better EFS and OS, suggesting that variations in drug uptake influences treatment outcomes. Five SNPs within the cytidine deaminase (CDA), involved in inactivation of Ara-C were associated with poorer EFS and OS. These SNPs were linked to higher CDA mRNA expression in blood, as per the GTEx database, and showed regulatory potential in chromatin states. SNP rs1048977 has been associated with reduced CDA enzyme activity, further corroborating its role in poor treatment outcomes. Among the myeloid panel genes SNPs in the AML-related gene RUNX1, CSF3R and CDKN2A were associated with poor clinical outcomes. SNPs in CSF3R and CDKN2A genes were predictive of poor EFS and OS. RUNX1 is a critical factor for myeloid differentiation, CSF3R is involved in neutrophil biology and activates the JAK/STAT pathway, while CDKN2A functions as a tumor suppressor and regulates the cell cycle. Dual mutations in CSF3R and CEBPA have been linked to high relapse rates in pediatric AML. In conclusion, our comprehensive pharmacogenomic analysis of genes involved in drug metabolism and AML pathogenesis has identified key genetic variants that influence treatment outcomes. These findings, once validated in larger cohorts, could offer new opportunities for personalizing chemotherapy and improving prognosis prediction in pediatric AML patients. Aneesha Nath, Richard Marrero, Xueyuan Cao, Jeffrey E. Rubnitz, Stanley Pounds, Jatinder K. Lamba. Single nucleotide polymorphisms located in PK/PD pathway genes of standard AML chemotherapy regimen predicts outcome in pediatric AML [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2025; Part 1 (Regular Abstracts); 2025 Apr 25-30; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2025;85(8_Suppl_1):Abstract nr 4358.

Abstract Cytarabine, daunorubicin, and etoposide (ADE) have been the standard backbone of induction chemotherapy regimen for patients with pediatric acute myeloid leukemia (pAML) for >5 decades. However, chemoresistance is still a major concern, and a significant proportion of pAML becomes resistant to ADE treatment and relapse, leading to poor survival. Therefore, there is a considerable need to identify mechanisms mediating drug resistance for overcoming chemoresistance. Herein, we performed synthetic lethal CRISPR/Cas9 screens using the ADE components to identify response markers. We further integrated significant markers in 3 independent pAML clinical cohorts treated with only an ADE regimen to identify drug response biomarkers with prognostic significance. We were able to identify several mediators that represent clinically and biologically significant marker genes for ADE treatment, such as BCL2, CLIP2, and VAV3, which are resistant markers to ADE, with high expression associated with poor outcomes in pAML treated with ADE, and GRPEL1, HCFC1, and TAF10, which are sensitive markers to ADE, with high expression showing beneficial outcomes. Notably, BCL2, CLIP2, and VAV3 knockdowns in their expression in AML cell lines sensitized the cells more to the ADE components, suggesting that these modulators should be further studied as potential therapeutic targets to overcome chemoresistance.

Introduction The 2022 WHO Classification of Pediatric Tumors marked a major step forward in molecular diagnostics. DNA-methylation profiling has emerged as a core diagnostic tool and shown to solve complex CNS and sarcoma cases. In AML, epigenomic patterns also reflect diagnostic genomic variants, but their clinical use is still emerging. The Acute Leukemia Methylome Atlas (ALMA) set a benchmark by correctly classifying >90% of 27 WHO 2022 AML subtypes using a decision-tree-based machine-learning model, though ~8% of cases remained mislabeled (PMID: 40730747). Here, we describe a transformer-based classifier that further improves AL subtype classification, aiming to enhance diagnostic precision in long-read WGS and methylation arrays. Methods DNA-methylation was profiled with Illumina 450K or EPIC arrays, covering 331556 CpG sites per sample. Processed data were obtained from the published ALMA dataset. For unsupervised learning, the cohort comprised of 3314 patient samples including AML, ALL, MDS, APL, MPAL and controls. For supervised learning, 2471 samples bearing WHO 2022 subtype labels entered stratified 5-fold cross-validation. Each methylome was compressed into a 64-dimensional latent vector by an autoencoder and then classified by a supervised transformer neural network; hyper-parameters were tuned within the training folds and locked before evaluation. Model performance was quantified with overall accuracy, sensitivity, specificity, weighted recall, macro and weighted F1-score. Generalizability was assessed on independent multi-center cohorts AML02 and AML08 trials (n=104), carrying confirmed WHO 2022 labels spanning six AML subtypes. Additionally, testing was performed in a long-read WGS cohort of 32 patient samples from UF Health Shands Hospital. Clinical truth for the UF cohort derived from cytogenetics, flow cytometry, and targeted NGS. Results In 5-fold cross-validation, the classifier correctly predicted 2456 of 2471 methylomes, yielding 99.4% overall accuracy. Leukemia-versus-control discrimination was near perfect, with 99.9% sensitivity (2218/2220 malignant) and 98.8% specificity (248/251 controls). Subtype-specific sensitivity (weighted recall) reached 99.3%, macro F1=0.994, and the median per-subtype AUPRC was 0.995. Twenty subtypes had complete concordance. Accuracy stayed high (0.95–0.99) for nearly all others, except for AML t(8;16) KAT6A::CREBBP with 11/12 correct (0.92) and MPAL t(v;11q23.3)/KMT2A-r with 5/6 correct (0.83). Altogether, 15 errors (0.61%) occurred: ten intra-leukemia subtype swaps and five control-vs-leukemia miscalls. This represents a >10-fold reduction versus the original ALMA model, which recorded 10.4% errors (accuracy=0.896, weighted F1=0.927) under the same 5-fold CV. External testing on the independent AML02/08 cohort (n=104) delivered 96.2% accuracy (100/104 correct), weighted F1=0.974, and precision=0.990, despite class imbalance in the test set. In real-world clinical testing using long-read WGS (n=32), all translocations detected by conventional cytogenetics were captured (t(8;21) n=1, t(6;9) n=1, t(16;21) n=1, KMT2A-r n=7) despite variable sequencing coverage (range 1.1x-23.8x). Eight post-therapy marrow samples were labelled “Otherwise-Normal Control”, coherent with morphological remission. Four discordant cases comprised of two Trisomy 21 AML patients and two B-ALL cases predicted to be AML. Five matched BM/PB pairs were correct and concordant. Conclusions By combining latent-space learning with a tabular transformer, the resulting classifier achieved 99% WHO-2022 concordance on >2400 leukemias and preserved ≥95 % accuracy in two multi-center external datasets, including low-coverage long-read WGS. The publicly available, open-source package (github.com/f-marchi/ALMA-classifier, v0.2.0) accepts both array and long-read WGS inputs, enabling rapid molecular classification in resource-limited settings or when time-critical diagnosis is needed. Ongoing work aims to expand the classifier capabilities to include lymphomas, Langerhans cell histiocytosis, trisomy-21 AML/ALL, chronic leukemias, multiple myeloma, and juvenile myelomonocytic leukemia.

Tandem duplications (TDs) in exons of upstream binding transcription factor (UBTF-TD) are a rare recurrent alteration in pediatric and adult acute myeloid leukemia (AML) and myelodysplastic syndrome (MDS)/neoplasm. Although recently identified, AML with UBTF-TD is now considered a distinct subtype of AML. To further our understanding of myeloid neoplasms with UBTF-TD, we analyzed clinical, morphologic, and immunophenotypic characteristics of 27 pediatric patients with UBTF-TD-positive myeloid neoplasm, including 21 diagnosed as AML and 6 as MDS. Our data demonstrated that UBTF-TD is frequently associated with cytopenia, hypercellular marrow with erythroid hyperplasia, and trilineage dysplasia. Blasts and maturing myeloid cells show a characteristic dysplastic feature with condensed eosinophilic cytoplasm. Blasts have a myeloid or myelomonocytic immunophenotype with a variably dim expression of CD34 and/or CD117, and except for CD7 expression lack a consistent pattern of aberrant lineage-specific antigen expression. Patients with MDS had a lower blast count in the peripheral blood (P = 0.03) and bone marrow (P <0.001) but otherwise had no significant differences in other hematological parameters. Three patients with MDS rapidly progressed to AML in 33, 39, and 210 days from the initial diagnosis and there was no difference in overall survival between patients with MDS and AML (P = 0.18). Our data suggest that MDS with UBTF-TD is prognostically equivalent to AML with UBTF-TD and thus should be considered as a continuum of the same molecularly defined myeloid neoplasm. These collective data also provide morphologic and immunophenotypic clues that can prompt screening for UBTF-TD in patients with MDS or AML.

Background: Toxicity due to antileukemic agents is common in acute myeloid leukemia (AML) patients and significantly impacts morbidity and mortality. Intensive chemotherapy regimens, such as the standard “7+3” induction protocol with cytarabine and daunorubicin, frequently result in severe neutropenia, markedly increasing susceptibility to infections (PMID: 38961525). Here we evaluated the impact of genetic variants on risk of severe febrile neutropenia post-induction 1 therapy. Methods: Febrile Neutropenia (FN) severity was graded based on the Common Terminology Criteria for Adverse Events (CTCAE), version 3.0. Grade 3 FN was defined as an absolute neutrophil count <1.0 × 10⁹/L with fever ≥38.5°C, without clinically or microbiologically documented infection. Grade 4 FN represented life-threatening complications. Patients enrolled in multisite trials AML02 and AML08 (N=400) with toxicity data were included in the study. Patients were stratified into two groups: those with severe FN (Grade ≥3) and those with low-grade (Grade 1–2) or no FN. Genome wide association study used the Illumina 2.5 Omni array, established quality control steps were used for data preparation. Logistic regression model, adjusted for ancestry, was used to test SNP associations with severe FN. Genome-wide significance was defined as P < 5 × 10⁻⁸, with a suggestive threshold at P < 1 × 10⁻⁴. Transcriptomic data from diagnostic specimen were available in 137 patients from AML02 trial and was used for evaluation with toxicity. Results: One of the top SNPs, rs8176592, (A>G) significantly associated with severe FN mapped to chromosome 2 in the regions with genes TFPI (tissue factor pathway inhibitor), CALCRL (calcitonin receptor-like receptor) and CALCRL-AS. Presence of G allele for the tag SNP rs8176592 associated with a significantly reduced risk of developing grade ≥3 FN (odds ratio [OR] = 0.46; 95% CI: 0.30–0.70; p = 9.79×10⁻⁵). rs8176592 is in intronic region of TFPI, CALCRL-AS1 and in promoter of CALCRL, it occurs in LD with several SNPs (~49 SNPs in LD at r2>0.95) spanning across the 3 genes including a mis-sense variant rs7586970 (T>C) in TFPI resulting in Asparagine to Serine change. The protective effect was dose-dependent: 83% of patients with homozygous genotype for the reference allele developed grade 3 or more FN, compared to 73% of heterozygous and 54% of patients with homozygous genotype for the variant allele (p <0.001). Further evaluation of using the GTEx database showed that for rs8176592, G allele was associated with significantly low levels of CALCRL and TFPI in several tissues and as a splice QTL for TFPI. Further evaluation of these genes showed CALCRL expression in AML02 cohort (N=137 with available transcriptomics data) to be associated with greater risk of grade 3 or more FN (p = 0.0004), though such an association with TFPI expression was not observed. These findings suggest that lower CALCRL and TFPI levels may protect against FN. TFPI is a known regulator of coagulation and has been implicated in inflammation through signaling mediated via activation of protease activated receptors, due to its dual anticoagulant and anti-inflammatory role it has been closely linked to the pathophysiology of sepsis; CALCRL which encodes a G-protein-coupled receptor has emerged as biologically relevant gene due to its role in vascular stability and inflammation, it is involved in adrenomedullin signaling which maintains vascular integrity and regulates immune responses during infection and sepsis. The presence of associated variants in both CALCRL and TFPI suggests that this genomic region may influence FN risk through coordinated modulation of vascular, inflammatory, and coagulative pathways. Conclusion: Our study reveals a common genetic variant in CALCRL and TFPI to be associated with substantially reduced risk of severe FN during induction therapy in pediatric AML. Future directions include independent validation in larger and more diverse AML cohorts, and functional validation studies to uncover the mechanistic role of CALCRL and TFPI in observed toxicity. Ultimately, these insights could support the development of personalized supportive care strategies in pediatric AML.

Pediatric acute myeloid leukemia (AML) exhibits distinct genetic characteristics, including unique driver alterations and mutations with prognostic and therapeutic significance. Emerging rare, recurrent genetic abnormalities and their associations with outcomes emphasize the need for high-throughput molecular diagnostic tools. Whole genome sequencing (WGS) reliably detects key AML biomarkers such as structural variants, mutations, and copy number alterations. Whole transcriptome sequencing (WTS) complements WGS by uncovering oncogene expression patterns, allele-specific expression, and gene expression signatures. In this study, we describe an integrated WGS and WTS clinical workflow for routine pediatric AML diagnosis and present a systematic evaluation of its application compared to conventional cytogenetics and standard molecular diagnostic methods. Our findings demonstrate that the integrated WGS and WTS (iWGS-WTS) approach improves the identification of clinically relevant genetic alterations, enhancing precise disease classification and risk assessment. Moreover, with advancements in workflow and bioinformatics pipelines, the testing turnaround time can be optimized to meet the demands of clinical decision-making, positioning iWGS-WTS as a practical and superior alternative to traditional diagnostic methods in pediatric AML management.

Despite the critical role of DNA methylation, clinical implementations harnessing its promise have not been described in acute myeloid leukemia. Utilizing DNA methylation from 3314 leukemia patient samples across 11 harmonized cohorts, we describe the Acute Leukemia Methylome Atlas, which includes robust models capable of accurately predicting AML subtypes. A genome-wide prognostic model as well as a targeted panel of 38 CpGs significantly predict five-year survival in our pediatric and adult test cohorts. To accelerate rapid clinical utility, we develop a specimen-to-result protocol that uses long-read nanopore sequencing and machine learning to characterize patients' whole genomes and epigenomes. Clinical validation on patient samples confirms high concordance between epigenomic signatures and genomic lesions, though uniquely rare karyotypes remained challenging due to limited available training data. These results unveil the potential for increased affordability, speed, and accuracy for patients in need of complex molecular diagnosis and prognosis.

Pediatric acute myeloid leukemia (AML) exhibits distinct genetic characteristics, including unique driver alterations and mutations with prognostic and therapeutic implications. Cytogenetics study, along with Next Generation Sequencing (NGS) panel testing, have long been the standard for molecular diagnosis of AML. While these approaches enable diagnosis and prognosis determination in most cases, they have limitations-particularly in detecting emerging rare, recurrent genetic abnormalities. In this study, we systematically reviewed our real-time clinical experience with the diagnostic workup of pediatric AML using an integrated whole genome and whole transcriptome sequencing (iWGS-WTS) approach and compared the test results obtained from various methodologies, including whole genome sequencing (WGS), whole exome sequencing (WES), whole transcriptome sequencing (WTS), iWGS-WTS, cytogenetics, and targeted panel NGS. Our findings demonstrate that the iWGS-WTS approach improves the identification of clinically relevant genetic alterations, enhancing precise disease classification and risk assessment. Additionally, the iWGS-WTS approach streamlines sample acquisition and reduces testing redundancy, positioning it as a practical and superior alternative to traditional diagnostic methods in pediatric AML management.

In recent years, extracellular vesicles (EVs) have gained significant interest in AML biology and prognostication. EVs are small circulating nano-sized particles implicated in cell-to-cell communication and transfer of miRNAs, proteins, metabolites/ligands across cells. In context of miRNAs, they have been shown to play a significant role in RNA silencing and gene regulation. Additionally, EV-encapsulated miRNAs profile is dependent on the cellular status. Furthermore, given the stability of miRNAs, the EV miRNA profiling is a highly attractive as biomarkers making them a promising area in oncology. With respect to AML, the circulating EVs have been implicated in bone-marrow microenvironment, immune system and contribute to drug resistance and hematopoiesis suppression. However, of the handful of studies published on exosomes/EVs in AML, the majority utilized AML cell lines and of the few on specimens from AML patients and none included pediatric AML patients. Thus, this is one of the first studies focused on investigating diagnostic and prognostic relevance of EV-derived miRNAs in pediatric AML. Serum samples collected at diagnosis from 34 pediatric patients enrolled in AML02 clinical trial were included in this study. EVs were isolated using Nano pom-poms, which employs antibody-functionalized nanographene magnetic particles for specific capture of EVs. Total RNA, including small RNAs, was extracted and miRNA libraries were prepared using the NEBNext® kit. After PAGE gel-based size selection and Bioanalyzer QC the libraries were sequenced using Illumina-based next-generation sequencing (NGS), and FASTQ files were processed using the nf-core/smrnaseq pipeline. This included adapter trimming, alignment, and annotation of mature miRNAs using miRBase. Raw miRNA counts were normalized to counts per million (CPM), using edgeR package. After excluding miRNA with missing values in 2 samples, Cox proportional hazard model was used for association of miRNAs with event free survival (EFS), overall survival (OS) and measurable residual disease 1 (MRD1). Of the 77 miRNAs meeting the minimum expression threshold, 11 were significantly associated with EFS, 3 with OS, and 2 with MRD1. High levels of EV-miR-222-3p were predictive of better EFS (HR= 0.81, 95% CI 0.67-0.97; p= 0.03 and OS (HR= 0.63, 95% CI 0.47-0.84; p = 0.002). Several members of let-7 family, including let-7d-5p, let-7f-5p, let-7g-5p were associated with OS and EFS. EV-let-7s have previously been shown to have a tumor suppressor role and in non-small-cell lung cancer have been shown to inhibit metastatic activity by inhibiting SUV39H2/LSD1/CDH1 axis. Additionally, miR-9-5p was associated with detrimental EFS (HR-1.25, p<0.02), and a previous report has implicated it in leukemogenesis. miR-191, miR-30d, miR-24, miR27b, miR-451a and miR-486 were predictive of better EFS (at p<0.05). Notably, miR-451a levels was also predictive of MRD1 negativity, while miR-10a-5p was associated with positive MRD1 status. This proof-of-concept pilot study provides the first evidence of EV-derived miRNAs from diagnostic specimens having the potential to serve as predictive biomarkers for clinical outcomes in pediatric AML. Several EV-miRs, including miR-9-5p, miR-486-5p, miR-222-3p and the members of let-7 family, were significantly associated with EFS, OS, or MRD1, highlighting their prognostic value. These findings support the hypothesis that EV-miRNAs contribute to leukemic progression and treatment resistance. Our ongoing studies in larger pediatric AML cohorts aim to further validate the EV-miR profiles and elucidate their mechanistic roles in disease progression and drug resistance.

Introduction: The dysregulation of DNA methylation in acute myeloid leukemia (AML) significantly changes the regulation and expression levels of genes involved in AML chemotherapy pharmacokinetics and pharmacodynamics and a panel of myeloid genes known to be associated with AML progression. The primary objective of this study was to evaluate the impact of the methylation levels of specific genes and their association with treatment outcomes among pediatric AML patients. Methods: The discovery cohort included 924 patients from TARGET database with methylation data obtained from the gene expression omnibus and expression data from genomic data commons. Cox proportional hazard analysis was performed to determine the association of DNA methylation of CpGs in the key myeloid panel genes (n= 107) and genes of relevance to cytarabine, daunorubicin and etoposide pharmacology genes (N= 65) with event-free survival (EFS) and overall survival (OS). Additionally, a Chi-square test was applied to assess the association between CpG methylation M value and status of Minimal Residual Disease after induction I (MRD1). A Bonferroni corrected p value of < 2.17×10-5was used as a significant association threshold. Significant results were validated in an independent patient cohort treated on AML02 trial (n=159). Results: Of the CpGs tested in the pertinent genes, 23 CpGs in drug pathway genes were associated with at least one endpoint at p <2x10−5. It included multiple CpGs (N=7) in ABCA3, were associated with clinical outcome endpoints. Of these for the most significant CpG (cg01278797) in ABCA3 greater methylation showed a consistent association with poor outcome in discovery (OS, HR= 1.43, p<0.0001, EFS HR= 1.35, p <0.0001, and a greater MRD1 positivity, p <0.001) and validation cohort (OS, HR= 1.81, p<0.0001, EFS HR= 1.61, p <0.0001, and a greater MRD1 positivity, p <0.01). For ABCA3, gene expression levels, higher expression was associated with worse OS (HR= 1.62, p <0.0001) and EFS (HR= 1.42, p=0.0002). Further, the methylation and expression of ABCA3 showed a positive correlation (p= 0.046), indicating possible epigenetic regulation. An etoposide metabolic pathway is mediated by myeloperoxidase (MPO), where hypermethylation of eight CpGs was associated with adverse clinical outcome. cg09421562 was the top significant CpG with higher methylation associated with poor outcome in discovery OS, HR= 1.22, 95% CI 1.15-1.29, p <10−11; EFS, HR= 1.15, 95% CI 1.10-1.20, p <10−9; MRD, p<10−7and validation cohort (OS, HR= 1.22, 95% CI 1.15-1.29, p <10−11; EFS, HR= 1.15, 95% CI 1.10-1.20, p <10−9; MRD, p<10−7. Additionally, increased MPO expression were related to better clinical outcomes in discovery (OS, HR=0.87, p <0.0001; EFS, HR= 0.89, p <0.0001) and validation cohorts (OS, HR=0.84, p =0.014; EFS, HR= 0.846, p =0.007). These CpGs and expression stayed significantly associate without come in multivariable analysis after adjusting for age, race, risk group ad WBC count. Other drug pathway related genes with methylation predictive of outcome included ABCC1 and NOS3. Among AML-relevant genes, 43 CpGs were associated with at least one outcome endpoint at Bonferroni corrected p<2.17×10-5. This included KIT, a receptor kinase and MPL, a myeloproliferative leukemia virus oncogene where hypermethylation of CpGs in these genes and the lower gene expression were associated with worse outcomes in both discovery and validation cohorts (OS, and EFS, p<0.001 for both). Methylation and gene expression for KIT and MPL showed significant inverse correlation as expected. Other genes with significant association of methylation and expression with outcome includedDNMT3B, CALR, DNMT3A, NOTCH1, PRPF8, RUNX1, GATA2, VWF. Conclusion: The study demonstrated that interpatient differences in DNA methylation levels of key genes related to AML drugs or disease pathogenesis affect clinical outcomes. This creates opportunities for outcome prediction and the strategic use of hypomethylating agents to enhance patient clinical outcomes.

BACKGROUND:Venetoclax is a potent, oral BCL-2 inhibitor approved as combination therapy for the treatment of adults with newly diagnosed acute myeloid leukaemia (AML) who are ineligible for intensive chemotherapy. This study evaluated the safety and preliminary efficacy of venetoclax alone or combined with chemotherapy in paediatric and adolescent/young adult patients with relapsed/refractory AML. PROCEDURE:In this phase 1, open-label, two-part, multicentre study, paediatric and adolescent/young adult patients (<25 years of age) with relapsed/refractory AML were treated with venetoclax alone or in combination with hypomethylating agents or cytarabine. The study is registered with ClinicalTrials.gov, NCT03236857. RESULTS:A total of 37 patients received treatment with either venetoclax as a monotherapy (n = 3) or in combination with decitabine (n = 5), azacitidine (n = 19), low-dose cytarabine (n = 1) or high-dose cytarabine (HDAC; n = 9). Febrile neutropenia (57%), hypokalaemia (38%), and thrombocytopenia (35% [thrombocytopenia, 19%; platelet count decreased, 16%]) were the most common grade 3/4 treatment-emergent adverse events. Across all venetoclax combinations, the overall response rate (ORR) was 24% (9/37), and the median duration of response was 2.6 months (95% CI, 0.5-7.9). Among the combinations, ORR was 44% with venetoclax plus HDAC and 21% with venetoclax plus azacitidine. In biomarker-evaluable patients, responses to venetoclax plus chemotherapy were observed in patients harbouring mutations across a range of functional classifications and heterogeneous BH3 family member dependencies. CONCLUSIONS:Venetoclax alone or combined with chemotherapy was well tolerated in paediatric and adolescent/young adult patients with relapsed/refractory AML, with promising, although transient, responses with venetoclax plus HDAC or azacitidine.

Tandem duplications (TDs) in exons of upstream binding transcription factor (UBTF-TD) are a rare recurrent alteration in pediatric and adult acute myeloid leukemia (AML) and myelodysplastic syndrome (MDS)/neoplasm. Although recently identified, AML with UBTF-TD is now considered a distinct subtype of AML. To further our understanding of myeloid neoplasms with UBTF-TD, we analyzed clinical, morphologic, and immunophenotypic characteristics of 27 pediatric patients with UBTF-TD-positive myeloid neoplasm, including 21 diagnosed as AML and 6 as MDS. Our data demonstrated that UBTF-TD is frequently associated with cytopenia, hypercellular marrow with erythroid hyperplasia, and trilineage dysplasia. Blasts and maturing myeloid cells show a characteristic dysplastic feature with condensed eosinophilic cytoplasm. Blasts have a myeloid or myelomonocytic immunophenotype with a variably dim expression of CD34 and/or CD117, and except for CD7 expression lack a consistent pattern of aberrant lineage-specific antigen expression. Patients with MDS had a lower blast count in the peripheral blood (P = 0.03) and bone marrow (P <0.001) but otherwise had no significant differences in other hematological parameters. Three patients with MDS rapidly progressed to AML in 33, 39, and 210 days from the initial diagnosis and there was no difference in overall survival between patients with MDS and AML (P = 0.18). Our data suggest that MDS with UBTF-TD is prognostically equivalent to AML with UBTF-TD and thus should be considered as a continuum of the same molecularly defined myeloid neoplasm. These collective data also provide morphologic and immunophenotypic clues that can prompt screening for UBTF-TD in patients with MDS or AML.

Introduction: Acute myeloid leukemia (AML) is a complex and highly heterogeneous hematologic malignancy with poor clinical outcomes. Despite improvements in supportive care, the standard treatment for AML has remained largely unchanged since the 1970s through use of cytarabine in combination with anthracyclines. The approval of only a limited number of new antileukemic agents since 2017 highlights the urgent need for innovative approaches to identify factors contributing to relapse, drug resistance, and therapeutic failure. This study aims to elucidate the prognostic and therapeutic relevance of proteomics in AML. Methods: Global proteomic analysis of leukemic cells obtained at diagnosis from 97 pediatric patients treated in the St. Jude AML02 clinical trial was performed using Tandem Mass Tag (TMT) labeling followed by two-dimensional Liquid Chromatography (LC/LC) and Tandem Mass Spectroscopy (MS/MS). Data were preprocessed with median normalization and log transformation and batch correction. Proteomic signatures were compared across patient subgroups, clinical characteristics, and AML-linked genetic alterations and with survival outcomes as event free survival (EFS), overall survival (OS) and measurable residual disease after induction 1 (MRD1). Results: Among the most significantly expressed proteins (p < 0.001), 25 proteins were associated with poor clinical outcomes and 16 proteins were associated with superior outcomes. Among the proteins predictive of poor outcome significant enrichment was observed in metabolic pathways. COX7B (cytochrome C oxidase subunit 7B), a component of mitochondrial complex IV involved in the oxidative phosphorylation involved in oxidative phosphorylation (OXPHOS) pathway was strongly associated with adverse prognosis (EFS, HR=2.6; OS, HR=3.8, both p<0.001). Given that leukemic stem cells (LSCs) depend on OXPHOS for survival making it a potential therapeutic target. These findings are consistent with previous reports showing that elevated mRNA levels of mitochondrial/OXPHOS genes to be associated with poor outcome in AML. SLC25A15, a mitochondrial amino acid and proton transport, was also associated with poor prognosis (EFS, HR=3.5, OS, HR=5.3 both p<0.001 and MRD1, OR=4.2, p=0.03). SCL25A15 is frequently upregulated in cancers and contributes to cancer progression and poor prognosis. Among other metabolic pathway enzymes associated with poor prognosis included fumarate hydratase with role in TCA cycle, lysophospholipase (shown to modulate LSCs response), 3-ketodihydrosphingosine reductase, dehydrogenase/reductase 13, mitochondrialaspartyl-tRNA synthetase 2, protein O-glucosyltransferase 1, protein O-fucosyltransferase 1 and glucosidase. Other proteins associated with poor prognosis included ODR4, belonging to G protein coupled receptor signaling (p < 0.001, EFS and OS) and NLE, involved in Notch signaling pathway with a role in hematopoietic stem cell homoeostasis. Among proteins predictive of better clinical outcome included GRB10, a growth factor receptor bound protein 10, ALOX5AP, Creatinine kinase B, microtubule associated protein, deoxyribonucleae etc. Conclusions: This study represents one of the first and most comprehensive global proteome profiling of diagnostic bone marrow samples from pediatric AML patients. We report several metabolic enzymes that are predictive of poor outcome and specifically COX7B and its role in OXPHOS pathway is highly relevant as LSC survival is dependent on OXPHOS highlighting its therapeutic potential. Understanding the molecular mechanisms behind these distinct protein patterns may guide the development of precision therapies as well as identify novel drug targets in AML. Our ongoing study is in the process of expanding this analysis to a full cohort of 448 patient samples across multiple clinical trials.

Genome-wide loss of DNA methylation accompanied by hypermethylation of promoter-associated CpG-islands is a hallmark of cancer. DNA methylation signatures have been used to identify biologically distinct subtypes of acute myeloid leukemia (AML), including several genomically defined subsets, suggesting that aberrant DNA methylation patterns may partially reflect underlying pathogenic mechanisms. The dependency of AML cells on methylation signatures for viability and/or differentiation status, however, remains largely unknown. Our prior retrospective analysis of diagnostic pediatric AML specimens identified an association between lower total genome-wide methylation (TGWM) burdens and improved outcomes. Further, increased levels of 5-methylcytosine may contribute to leukemic resistance to the cytosine-analog cytarabine through epigenetic repression of nucleoside transporters. Thus, we hypothesized that five days of epigenetic priming with azacitidine (AZA) or decitabine (DAC) before each course of chemotherapy would improve outcomes by lowering TGWM. The AML16 clinical trial tested this hypothesis by evaluating the tolerability and efficacy of randomly assigned AZA or DAC priming in 200 pediatric AML patients enrolled at 10 U.S. centers (NCT03164057). To establish tolerability, priming was limited to induction I and induction II during the initial phase of the study. Induction I included 5 days of priming followed by cytarabine, daunorubicin, and etoposide, and induction II included 5 days of priming before idarubicin, cytarabine, and fludarabine. Risk assignment and subsequent therapy were adapted based on molecular features and minimal residual disease (MRD) levels. Patients with high-risk (HR) disease and suitable donors underwent hematopoietic stem cell transplantation (HSCT) in first remission. Once priming prior to induction I and II were deemed tolerable, priming was expanded to all chemotherapy courses for patients not undergoing HSCT. A course was deemed tolerable if 15 or more of the first 25 evaluable patients completed it without experiencing an unacceptable adverse event (UAE), based on a multi-stage statistical design. The trial was approved by each participating IRB. Patients were assigned to AZA priming for two courses (AZA2, n=34), AZA priming for all courses (AZA4/5, n=68), DAC priming for two courses (DAC2, n=33), or DAC priming for all courses (DAC4/5, n=65). Eleven unique chemotherapy courses with priming were deemed tolerable; no course was deemed intolerable. Across all patients, epigenetic priming improved outcomes compared to historical controls, with 3-year event free survival (EFS) of 75.8%, 57.9%, and 61.2% for patients enrolled on AML16, AML08, and AML02, respectively (p=0.0002). Low-risk (LR) and intermediate-risk (IR) patients had excellent outcomes with 3-year EFS of 95.5% and 86.4%, and overall survival (OS) of 100% and 94.6%, respectively. HR patients had a 3-year EFS of 58.0% and OS of 63.3%. Among molecular subsets, patients with LR mutations, including RUNX1::RUNX1T1, CBFB::MYH11, CEBPA and NPM1c had 3-year EFS and OS uniformly ≥90%. KMT2A rearranged (KMT2Ar) patients had a 3-year EFS and OS of 64.1% and 66.7%, respectively; among them, those with KMT2A-MLLT3 had the best outcomes, with 3-year EFS and OS of 85.7%. The 3-year EFS was 69.6% and 82.1% (p=0.058) and 3-year OS was 74.4% and 88.5% (p=0.014) for AZA and DAC, respectively. LR and IR patients had excellent outcomes irrespective of the epigenetic priming agent, with 3-year EFS of 96.1%, 85.7%, 95%, 87.5% (p=0.6) and 3-year OS of 100%, 90.5%, 100%, and 100% (p=0.43) for LR AZA, IR AZA, LR DAC, and IR DAC, respectively. Among HR patients, DAC resulted in better outcomes, with a 3-year EFS of 67.3% vs. 50.5% with AZA (p=0.20) and OS of 70% vs. 53% (p=0.15). DAC was significantly better than AZA in a win-ratio (WR) analysis encompassing death, relapse, and response to induction I (WR = 1.77; 95% CI: 1.32, 2.39; p=0.0002). Five days of AZA and DAC priming reduced median TGWM by 0.39 and 0.50 M-value units, respectively (p=0.06). We conclude that epigenetic priming significantly improved outcomes in pediatric AML, with DAC demonstrating superiority over AZA. These results support the adoption of epigenetic priming as the standard of care for pediatric AML.

BACKGROUND:Acute lymphoblastic leukemia (ALL) is the most common childhood cancer, and although many patients respond to induction therapy, those who relapse or have refractory disease face a poor prognosis. Venetoclax has promising preclinical and clinical activity in ALL. Here, we report the safety and preliminary efficacy of venetoclax combined with chemotherapy in pediatric and adolescent/young adult patients with relapsed/refractory ALL. PROCEDURE:This phase 1, open-label, two-part, multicenter study evaluated venetoclax combined with chemotherapy in pediatric and adolescent/young adult patients (<25 years of age) with relapsed/refractory ALL. The study is registered with ClinicalTrials.gov, NCT03236857. RESULTS:Thirty-one patients were treated and received venetoclax monotherapy (n = 1), venetoclax plus dexamethasone and/or vincristine and/or pegasparaginase (VXL; n = 20) or venetoclax plus cytarabine and/or etoposide and/or pegasparaginase (n = 10). Patients were heavily pretreated, with a median of 3 prior lines of therapy. The most common grade 3/4 treatment-emergent adverse event was febrile neutropenia (55%). One fatal adverse event possibly related to venetoclax occurred. The overall response rate of treated patients was 42%, with all responding patients achieving complete remission/complete remission with incomplete marrow recovery. In biomarker-evaluable patients, responses to venetoclax plus VXL-based or cytarabine-based chemotherapy were observed in patients harboring a range of genetic alterations and heterogeneous BH3 family member dependencies. CONCLUSIONS:Venetoclax plus VXL-based or cytarabine-based chemotherapy was overall well tolerated, with promising preliminary efficacy.