Classification Of Acute Myeloid Leukemia Research Articles

What is new in acute myeloid leukemia classification?

Recently, the International Consensus Classification (ICC) and the 5th edition of the World Health Organization classification (WHO2022) introduced diagnostically similar yet distinct approaches, which has resulted in practical confusion. This review compares these classification systems for acute myeloid leukemia (AML), building up on the revised 4th edition of WHO (WHO2016). Both classifications retain recurrent genetic abnormalities as a primary consideration. However, they differ in terms of blast threshold. The ICC mandates a minimum of 10% blasts in the bone marrow or peripheral blood, whereas the WHO2022 does not specify a blast cut-off. AML with BCR::ABL1 requires > 20% blast count in both classifications. In WHO2022, AML with CEBPA mutation requires > 20% blasts. TP53 mutation, a new entity is exclusive to ICC, diagnosed with > 20% blasts and variant allele frequency > 10%. AML with myelodysplasia-related changes is defined by cytogenetic or gene mutation-based criteria, not morphological dysplasia. Eight genes were common to both groups: ASXL1, BCOR, EZH2, SF3B1, SRSF2, STAG2, U2AF1, and ZRSR2. An additional gene, RUNX1, was included in the ICC classification. AML cases defined by differentiation (WHO2022) and AML not otherwise specified (ICC) are categorized as lacking specific defining genetic abnormalities, WHO2022 labels this as a myeloid neoplasm post cytotoxic therapy (MN-pCT), described as an appendix after specific diagnosis. Similarly, in ICC, it can be described as "therapy-related", without a separate AML category.

Advances towards genome-based acute myeloid leukemia classification: A comparative analysis of WHO-HAEM4R, WHO-HAEM5, and International Consensus Classification.

Two recent guidelines, the 5th edition of the World Health Organization Classification of Haematolymphoid Tumours (WHO-HAEM5) and the International Consensus Classification (ICC), were published to refine the diagnostic criteria of acute myeloid leukemia (AML). They both consider genomic features more extensively and expand molecularly defined AML subtypes. In this study, we compared the classifications of 1135 AML cases under both criteria. According to WHO-HAEM5 and ICC, the integration of whole transcriptome sequencing, targeted gene mutation screening, and conventional cytogenetic analysis identified defining genetic abnormalities in 89% and 90% of AML patients, respectively. The classifications displayed discrepancies in 16% of AML cases after being classified using the two guidelines, respectively. Both new criteria significantly reduce the number of cases defined by morphology and differentiation. However, their clinical implementation heavily relies on comprehensive and sophisticated genomic analysis, including genome and transcriptome levels, alongside the assessment of pathogenetic somatic and germline variations. Discrepancies between WHO-HAEM5 and ICC, such as the assignment of RUNX1 mutations, the rationality of designating AML with mutated TP53 as a unique entity, and the scope of rare genetic fusions, along with the priority of concurrent AML-defining genetic abnormalities, are still pending questions requiring further research for more elucidated insights.

Epigenetic Classification of Acute Myeloid Leukemia Revealed By Genome-Wide Chromatin Profiling

Background Acute myeloid leukemia (AML) is a clinically and molecularly heterogeneous disease and is predominantly defined by genetic abnormalities and morphology. However, many lines of evidence suggest the role of epigenetic alterations in the pathogenesis and the classification of AML, whereas no large-scale comprehensive analyses of both genetic and epigenetic alterations in AML have been reported. Methods We enrolled a total of 1,564 primary AML cases (524 from Japanese and 1,040 from Swedish cohorts), whose diagnostic samples were subjected to ATAC sequencing together with targeted capture sequencing using an in-house gene panel including 331 known AML driver genes and an additional 1,158-1,317 probes for copy number detection. We also performed RNA (N=1,398) and H3K27ac-targeted ChIP sequencing (N=80). Normal bone marrow samples from 25 donors as well as 15 remission samples were analyzed by ATAC-seq as a control. Results ATAC-seq revealed approximately 300,000 reproducible peaks in primary AML. Most peaks were found in gene-distal regions, and explained a large variance across samples. Among all peaks, 72% were not detected in control samples and thus, specific to AML. Analysis of the correlations between ATAC-seq accessibility and gene expression revealed a subset of regulatory elements that regulate nearby gene expression; each peak was linked to a median of 1 gene (0-31), while each gene was associated with a median of 11 peaks (0-110). We next estimated the cellular composition of the bulk AML samples by deconvolution analysis using ATAC-seq data. The predominance of hematopoietic stem cell (HSC), monocyte, and erythroid signatures was associated with French-American-British (FAB) subtypes of minimal differentiation/without maturation subtypes (M0/M1), myelomonocytic/monoblastic/monocytic subtypes (M4/M5), and erythroid subtype (M6), respectively, suggesting bulk ATAC-seq of AML samples reflected the differentiation potential of leukemic cells. The ATAC-based clustering identified 11 distinct epigenetic subgroups, including three well-known genetic classes defined by gene fusions, such as PML::RARA, RUNX1::RUNX1T1, and CBFB::MYH11. By contrast, some genetic classes defined by a single mutation, such as NPM1 or CEBPA mutation, were further classified into multiple distinct epigenetic subgroups. For example, the NPM1-mutated subtype was separated into four subgroups based on the varying degrees of ATAC signatures associated with HSC and monocytes. Each subgroup of NPM1-mutated AML showed different clinical characteristics such as white blood cell counts, age, and prognosis, and co-mutation pattern in FLT3, DNMT3A, and IDH1/ IDH2 mutations. Gene ontology analysis revealed that the subgroup of NPM1-mutated AML with high HSC signature showed upregulated TGF-β signaling, while one with monocytic signature showed upregulated TNF-α signaling and interferon-γ response. Next, we evaluated transcription factor (TF) binding sites in a genome-wide manner, using TF motifs and foot printing in upregulated ATAC-peaks in each subgroup and revealed distinct profiles of activated TFs in each epigenetic subgroup of NPM1-mutated AML. In the subgroup characterized by high HSC signature, HSC-related TFs that belong to MECOM, HOX, and RUNX families, were more active. By contrast, other TFs, such as IRF, CEBP, and JUN/FOS families, were more enriched in the subgroup showing high monocyte signature. By developing a prediction model for these subgroups of NPM1-mutated AML based on the gene expression profile, we reproduced these subgroups and validated clinical and molecular features in an external Beat AML cohort. Utilizing ex vivo drug sensitivity dataset in the Beat AML, we identified candidate drugs showing subgroup-specific effectiveness such as panobinostat for NPM1-mutated AML with monocytic differentiation. Finally, ChIP-seq analysis identified a number of super-enhancers in the vicinity of leukemia-related TFs and oncogenes. Conclusion Through comprehensive profiling of chromatin accessibility and gene mutations/expression in a large cohort of AML patients, we demonstrate the epigenetic heterogeneity and identify novel AML subgroups that had unique epigenetic and clinical features. Our findings highlight the role of epigenetic alterations in AML pathogenesis and impact on molecular classification of AML.

Optical Genome Mapping Combined with High-Throughput Sequencing Is Effective for the Diagnostic and Prognostic Genomic Classification of Acute Myeloid Leukemia and Myelodysplastic Neoplasms

Cytogenetic and molecular analyses are crucial for the diagnosis, choice of treatment and prognostic evaluation of acute myeloblastic leukemia (AML) and myelodysplastic neoplasms (MDS). Optical genome mapping (OGM) is a recent technology enabling a genome-wide high-resolution detection of structural abnormalities (SVs) and copy number variations (CNV) developed by the Bionano® company. The clinical value of OGM in AML and MDS and its positioning in relation to conventional karyotype and FISH (K&F) and targeted high throughput panel sequencing (HTS) has been little evaluated. Here, the primary objective was to evaluate prospectively the ability of OGM to apply the diagnosis (WHO 2022) and prognostic cytogenetic classification of AML (ELN 2022) and MDS (IPSSm) in comparison with K&F and HTS. The secondary objectives were to evaluate the proportion of informative cases in intention-of-analysis and among them the level of concordance of OGM versus K&F, the amount of additional information (OGM vs KF), and the potential reclassification (OMS 2022, ELN 2022, IPSSm) in addition to K&F and a targeted HTS panel (i.e. including ASXL1, CEBPA, DNMT3A, IDH1, IDH2, FLT3, KRAS, NPM1, NRAS, RUNX1, SRSF2, TP53, WT1) used in real-life. A two-stage prospective selection of bone marrow or blood samples was carried out at Nantes University Hospital between May 2022 and June 2023. A 1st collection of samples from 67 patients (pts) with AML (n=58) or MDS (n=9) was selected and cryopreserved according to pre-analytical criteria (cell count, time after collection). A 2nd final selection of 39 pts including 34 AML and 5 MDS was carried out according to their cytogenetic and/or molecular data, in order to evaluate the performance of OGM in different profiles of clinical interest (Table 1). Of these 39 intention-to-analyze patients, 35 were informative for OGM comparison with K&F (3 OGM failures, 1 K not realized). Interpretation criteria with Bionano access software considered SVs and CNVs >200 and >500 kbp respectively, or smaller if a gene of interest was found. Further details will be presented at the conference. OGM showed concordance with K&F in 30/35 (86%) of informative pts, but 5 had subclonal (<20%) gains undetected i.e. +8, +21, +6p or poorly detected i.e. t(1;15) with centromeric break. None of these OGM false negatives modify the diagnosis or prognosis. Additional information was found with OGM in 19/35 pts, with a median of 1 cryptic anomaly/patient (range = 1-13). Cryptic abnormalities were identified in 9/18 pts with normal K (7 AML, 2 MDS), including 5 pts with pathogenic SV, notably KMT2A with partial tandem duplication (PTD) and TET2 deletion, and 4 pts with SV of unknown significance (-6p, +4q, +15q, +19p, +20q). Finally, the WHO diagnosis was confirmed with OGM in all informative pts but not modified in any of them. OGM confirmed ELN prognostic classification of all the AML cases without modification. Two AML cases had >3 anomalies with OGM (range = 5-6) of which >2 were cryptic without complex karyotype (CK) (size = 226-1221 kbp, VAF% = 3-100) including gains (3q, 11q, Xp) or losses (3p, 4p, 21q) of undetermined or pathogenic ( RUNX1 deletion) significance. In this context of multiple cryptic anomalies, there is no precise definition of abnormal or CK with OGM. One is a MDS-related AML cases that also had MDS-associated somatic mutations (i.e. ASXL1, SRSF2 ELNadverse) and the other was an AML with NUP98 rearrangement and FLT3-ITD mutation (i.e ELN intermediate). OGM confirmed the IPSSm prognostic cytogenetic category in 3 pts and modified it in 2/5 MDS pts (40%). Among the two latter, K was normal (e.g. favorable) but OGM identified cryptic KMT2A- PTD of poor prognosis. These two SMD-IB2 changes the IPSSm risk category from Moderate-High or High to Very-High risk group. The combination of OGM and HTS techniques provided a more accurate prognostic classification of MDS cases. Confirmations of cryptic pathogenic anomalies detected by OGM using a 2nd technique will be presented at the congress. OGM technology is powerful for identifying pathogenic anomalies, but the criteria for abnormal K and CK need to be clarified in order to apply international diagnosis and prognosis classifications. To this end, the OGM+HTS combination is highly effective. The positioning of OGM as a complement and/or replacement for conventional cytogenetic and molecular tools, and its impact on therapeutic choice, still needs to be evaluated in AML and MDS.

Advances in Acute Myeloid Leukemia Classification: A Comparative Analysis of Who-HAEM4R, Who-HAEM5, and International Consensus Classification

Introduction Two guidelines were recently published to refine the diagnostic criteria of acute myeloid leukemia (AML): the 5th edition of the World Health Organization Classification of Haematolymphoid Tumors (WHO-HAEM5, PMID: 35732831) and the International Consensus Classification (ICC; PMID: 35767897), and both expanded molecularly defined AML subtypes. In this study, we utilized a cohort of 1135 AML cases to examine their performance. Methods From Sep 1, 2018 to Dec 31, 2022, a consecutive cohort of 1135 cases diagnosed with de novo AML according to WHO-HAEM4R (PMID: 27069254) were enrolled, with a median follow-up of 20 months. All underwent high-throughput sequencing (PMID: 34135310, 29932212). Results When classified according to WHO-HAEM4R, 55% of the 1135 cases were diagnosed as AML with recurrent genetic abnormalities, 16% were AML with myelodysplasia-related changes (AML-MRC), 2% were therapy-related AML, 4% were AML with germline predisposition, and 24% fell into AML not otherwise specified (AML, NOS). Major changes in WHO-HAEM5 include the inclusion of AML subtypes of KMT2A rearrangement (KMT2A-r), MECOM-r, and NUP98-r; adding AML with other defined genetic alterations (AML-ODGA) as a basket entity; and annulling the entity of AML with mutated RUNX1 (AML-RUNX1). The entity of AML with CEBPA mutation (AML-CEBPA) has changed to include not only biallelic (biCEBPA) but also single bZIP (smbZIP-CEBPA) mutations. AML-MRC in WHO-HAEM4R is redefined as AML, myelodysplasia-related (AML-MR), with a mutation-based definition based on an 8-gene panel (without RUNX1). When comparing WHO-HAEM4R and WHO-HAEM5 (Fig. 1), AML with KMT2A-r increased from 2% to 10% because WHO-HAEM5 no longer restricts the partner genes of KMT2A. AML with NUP98-r and AML-ODGA in WHO-HAEM5 account for 6% and 4%, respectively. AML-MR increased from 12% AML-MRC to 21% benefited from the 8-gene panel-based definition. AML-RUNX1 has been annulled in WHO-HAEM5 and 63% cases of this subtype were reclassified as AML-MR, and 30 cases with RUNX1 mutations as the only basis for classification were reclassified as AML defined by differentiation. In total, AML without defining genetic abnormalities dropped from 24% to 11%. When comparing WHO-HAEM4R and ICC (Fig. 1), AML with mutated TP53 (AML-TP53) is a new category in ICC and accounted for 4% of AML, 79% of them were previously classified as AML-MRC in WHO-HAEM4R. AML-MR in WHO-HAEM5 is divided into two entities in ICC: AML with MR gene mutations (including RUNX1) and AML with MR cytogenetic abnormalities, comprising 13% and 5% of cases, respectively. ICC only includes in-frame bZIP CEBPA mutations. Thus 6 cases classified as AML-biCEBPA based on WHO-HAEM4R fell into AML, NOS under ICC since their CEBPA mutations were not in bZIP. Conversely, 21 cases with smbZIP-CEBPA and diagnosed as AML, NOS according to WHO-HAEM4R were classified as AML-CEBPA under ICC. Cases classified as AML-CEBPA based on WHO-HAEM5 tended to have a better overall survival (OS) than cases classified as AML-biCEBPA under WHO-HAEM4R because AML-smbZIP-CEBPA was added while cases with MR gene mutations which had a worse prognosis fell into AML-MR under WHO-HAEM5. Despite ICC only included in-frame bZIP-CEBPA mutations, the prognosis of AML-CEBPA diagnosed under ICC was not superior to that under WHO-HAEM5. This might be because cases with both MR cytogenetic abnormalities and CEBPA mutations were reclassified as AML-CEBPA under ICC's hierarchical principle. ICC distinguished AML with KMT2A::MLLT3 from AML with KMT2A-r, while WHO-HAEM5 only included AML with KMT2A-r entity. Survival analysis showed significantly better OS in AML with KMT2A::MLLT3 than AML with KMT2A-r. ICC separated AML-TP53 from AML-MR due to the notorious prognosis of TP53 defects, and cases classified as AML-TP53 showed the worst OS in this cohort. But prognostic analysis showed no difference between AML-TP53 and cases with concurrent TP53 mutation and other AML-defining genetic abnormalities. The rationality of considering AML-TP53 as a distinct entity remains debatable. Conclusions Both new classification systems consider genomic features more extensively and enhance the precise classification of AML. Discrepancies between WHO-HAEM5 and ICC, like classifying AML with RUNX1 mutations and the rationality of defining AML-TP53 as a unique entity, remain open questions that require further research for clearer answers.

Data-Driven Multiparameter Flow Cytometry (MFC) Classification of Blast Differentiation in Elderly Patients with Acute Myeloid Leukemia (AML)

Background: The identification of the nearest normal counterpart of blasts has been part of the patients' diagnostic workup and contributed to the sub-classification of AML. While there is a progressive displacement of morphologically defined AML subtypes by those with recurrent genetic abnormalities, recent studies in younger patients eligible for intensive chemotherapy suggested that a phenotypic sub-classification of AML might be prognostic. However, an antigen-based assessment of blast differentiation is in part complex and subjective, and to our knowledge has not been systematically investigated in elderly AML. Aim: Develop a data-driven approach to facilitate and systematize the phenotypic dissection of older AML patients ineligible to receive intensive chemotherapy. Methods: The study included 252 patients (>65yo) that were randomized to azacitidine vs low-dose Ara-C plus fludarabine in the PETHEMA-FLUGAZA trial. MFC was performed using the EuroFlow AML panel in bone marrow aspirates. The percentage of expression of each marker in hematopoietic progenitor cells (HPC), granulocytic, monocytic and erythroid precursors, as well as in maturing granulocytes, monocytes and nucleated red blood cells (NRBC) was determined in 15 healthy adults. The data were merged and a matrix was obtained for each subject. The percentage of expression of each marker was determined in patients' blasts. After projecting each case onto the matrix generated from healthy adults, the distance between blasts and normal populations was calculated, and the shortest distance identified the corresponding differentiation stage. Blasts were isolated according to patient-specific aberrant phenotypes for DNA and RNA co-isolation and subsequent next-generation sequencing (n=179/252) and RNA-seq (n=190/252). Results: Of the 252 patients, 64% were classified with an HPC-like AML phenotype, 27% a granulocytic-precursor, 2% monocytic-precursor, 2% erythroid-precursors, <1% maturing-granulocytic, 4% maturing-monocytic and <1% maturing-NRBC. Cell types from the same lineage were regrouped into granulocytic-, monocytic and erythroid-like AML. Overall, there was a 46% concordance between phenotypic and FAB classification. Interestingly, patients classified as HPC-like AML treated with azacitidine (and not those receiving semi-intensive chemotherapy) showed slightly longer overall survival when compared to those with maturing AML subtypes. We next investigated the relationship between phenotypically-defined stages of blast arrest and the presence of underlying mutations. Interestingly, there was scarce overlap in the top three recurrently mutated genes across the HPC- ( SRSF2, TET2, RUNX1), granulocytic- ( ASXL1, FLT3, TET2), monocytic ( FLT3, NF1, TET2), and erythroid-like ( TP53, DNMT3A, NF1) AML subtypes. Conversely, EPAS1, FLT3 and NF1 were more frequently mutated ( p<.05) in erythroid- and monocytic- vs HPC-like AML. Amongst the maturing subtypes, monocytic-like AML was characterized by frequent SH2B3 and absent TP53 mutations ( p<.05). The comparison between HPC-like vs all maturing subtypes identified several deregulated genes associated with different stages of hematopoiesis (i.e., HOXA3, ANKRD28, CLEC2B, HSD17B11, SYPL1, HBD, S100A8, ITGAM). Granulocytic-like AML patients showed deregulation of 200 genes; 44 of which were upregulated and enriched in granulocytic-related pathways. Only 23 genes were deregulated upon comparing the transcriptional profile of HPC- vs monocytic-like blasts. A total of 415 genes were deregulated between HPC- and erythroid-like blats; 269 of which being overexpressed and associated with pathways such as transferrin transport, erythrocyte development and differentiation. Accordingly, the transcriptional profile of blasts corroborated our data-driven approach. Conclusions: This study builds upon recent interest in non-genetic drivers of AML heterogeneity and classification systems based on the partial recapitulation of myeloid development, and shows that these concepts may be applied in elderly patients. Our proof-of-concept data-driven approach has the potential to be incorporated into flow cytometry analytical software towards a more objective dissection of AML subtypes based on blast differentiation, which could potentially yield prognostic information complementary to other routine diagnostic procedures.

Unraveling the Impact of 2022 Classifications on Secondary Acute Myeloid Leukemia: Assessing the True Qualification Power of Diagnostic Qualifiers

Background: In 2022, two new acute myeloid leukemia (AML) classifications (World Health Organization, WHO and International Consensus Classification, ICC) and a new edition of the European LeukemiaNet (ELN) prognostic system have been released. These advances have been driven by the need to incorporate molecular characterization into the diagnosis and management of AML. Although important similarities in the diagnostic classifications have been recently recognized by diverse validation studies, the different approach to define AML ontogenesis (primary vs secondary) remains a central issue of the classification discrepancy. While for the WHO 2022 a prior diagnosis of myelodysplastic syndrome (MDS) or MDS/myeloproliferative neoplasm (MPN) is sufficient to define AML myelodysplasia-related (MR), the ICC no longer recognizes the AML arising from MDS (sAML) as a disease entity, and instead introduces a list of “diagnostic qualifiers”, also including therapy-related AML (tAML). Nevertheless, whether these “diagnostic qualifiers” represent a mere “descriptive” information or impact on prognosis and outcome is not well defined. Methods: We studied a cohort of 936 AML patients collected through a multicenter international collaboration across 5 academic centers. An integration of clinical and cytogenetic/molecular data was used to apply the WHO 2016, WHO 2022 and ICC diagnostic classifications and the 2017 and 2022 edition of ELN prognostic risk stratification criteria. Results: Overall,220 cases were classified as secondary AML (155 sAML and 65 tAML). As per the2016 and 2022 WHO editions, the majority of sAML cases, accounting for 98.1% and 95.5% respectively, fell into the myelodysplasia-related changes (MRC) AML and AML-MR categories due to the “disease defining” presence of an antecedent MDS or MDS/MPN. According to the ICC, 81.9% of sAML cases were defined by the presence of a MDS-related genetic profile, belonging to AML with mutated TP53 (AML- TP53) and AML with MDS-related gene mutations (AML-MDSgene) and cytogenetic abnormalities (AML-MDSk) subcategories. The tAML group showed a greater diagnostic heterogeneity, with a higher frequency of KMT2A-related categories (10.8% and 7.7% according to WHO 2022 and ICC, respectively) and AML- TP53 for the ICC (21.5%, Fig. 1A). In terms of outcomes, according to the ELN 2022 risk classification the sAML and tAML groups were classified into favorable risk 5.8% and 13.9%, intermediate 12.3% and 36.9%, and adverse 81.9% and 49.2%, respectively. Notably, while the ELN 2022 edition was effective in stratifying the group of tAML patients receiving conventional treatment (p=0.002), no statistically significant differences were noticed when sAML patients were stratified based on the same ELN 2022 risk categories ( Fig. 1B). We then compared our sAML cohort to AML cases with ICC MDS-related profile not fulfilling the anamnestic criterion of a prior MDS diagnosis (233 cases, 32.54% - AML- TP53/MDSgene/MDSk subcategories), in terms of molecular characteristics and outcome. Remarkably, both groups showed a similar prevalence of MDS genes, except for RUNX1, which was more frequent in the AML cases with the ICC MDS-related profile (14%) than in sAML (6.3%, p =0.034). Finally, both groups showed similar outcomes. Conclusions: Here, we demonstrated the overarching importance of the biological ontogenetic profile in AML (>80% of sAML presented an MDS-related profile) and of the risk stratification according to the new classifications (>80% of sAML belonged to adverse ELN 2022 risk). As to outcomes, 2022 ELN edition confirmed its stratification capability for tAML, while sAML were not adequately stratified. This evidence supports the idea that therapy-related AML should be acknowledged and classified as “second” neoplasms, rather than being categorized as secondary AML. Considering the overlap in AML of MDS-related profile regardless of the presence of an antecedent diagnosis of MDS as “diagnostic qualifier”, these cases could potentially benefit from treatments specifically tailored for MDS-related conditions or those typically reserved for secondary AML, such as CPX-351.

Mutational Profile and Dynamics of PPM1D-Mutant Clones in the Spectrum of Myeloid Disorders

Introduction Mutations in PPM1D a regulator of DNA damage response, are enriched in patients with clonal hematopoiesis (CH) exposed to cytotoxic treatment or with therapy-related myeloid neoplasm. However, their role in leukemic transformation is unclear. Here, we describe a large cohort of patients with PPM1D mutations from CH to acute myeloid leukemia (AML) to understand the molecular landscape of mutant PPM1D related disease and the longitudinal dynamics of CH under treatment. Methods We included, in a non-sequential cohort, 96 PPM1D mutated patients treated at Gustave Roussy with CH, clonal cytopenia of unknown significance (CCUS), myelodysplastic syndrome (MDS), myeloproliferative neoplasm (MPN) or AML. A 77 gene panel NGS analysis was performed. An additional 16 PPM1D mutated AML patients from ALFA trials (1200, 0701,0702) were included. 10 patients with ovarian cancer from this cohort had sequential blood samples (OvBIOMark trial) available prior to hematologic evaluation, from the diagnosis or first relapse of their cancer through the course of therapy. Those samples were analyzed with a UMI based 18 gene panel (HaloplexHS, Agilent). Overall survival (OS) analyses were performed with the Kaplan-Meier method from the time of diagnosis until death from any cause. Results Among the 112 patients with PPM1D mutations, 23 (21%) had CH, 36 (32%) CCUS, 19 (17%) MDS, 25 (22%) AML (including 4 relapses), and 9 (8%) MPN. Median age was 65 [range, 21-88] years with 67% of females. Only 18% of the patients had no previous cancer history. Most frequent primary cancers were gynecological cancer (27%), lymphoid malignancies (22%), and breast cancer (17%). Median time between primary cancer diagnosis and hematologic assessment was 5.3 [range, 1.2-8.5] years. Eighty six percent of patients had one PPM1D mutation, 9% had 2, and 5% had 3 or more (restricted to CH and CCUS). Median variant allele frequency (VAF) was 3% [0.2-38], 3% [1-31], 2.4% [1-41], 23% [1-50], and 4% [1-42] in CH, CCUS, MDS, AML and MPN, respectively. Among CH/CCUS patients, PPM1D was the sole detected somatic mutation in 39% (23/59), compared to 9% (5/53) in MDS/AML/MPN patients (odds ratio=6, p=0.0004); 3/5 of the latter had complex karyotype. The most frequently co-mutated genes were DNMT3A (29%) and TP53 (25%), uniformly across all conditions ( Fig 1A). MDS-related gene mutations, RUNX1 (7%), ASXL1 (4%), SF3B1 (4%), SRSF2 (4%), U2AF1 (4%), were specific to AML/MDS/MPN. Among 28 patients with both PPM1D and TP53 mutations, PPM1D mutations were dominant or co-dominant in 64% (18/28), and secondary in 36% (10/28). IPSS-M risk of evaluable MDS was high/very high in 54% (7/13) of patients. ELN 2022 classification of AML was adverse in 68% (17/25) of patients. AML treatment options included best supportive care for 8% of patients (2), 5-Azacytidine for 32% (8), intensive chemotherapy for 60% (15). With a median follow up of 2.4 years, the median OS was 4.6 (CI95%; 4.1-NA), 1.31 (CI95%; 0.53-NA), 1.27 (0.43-NA), 0.66 (CI95%; 0.42-1.26) and 20.4 (CI95%; 20.4-NA) years for CH, CCUS, MDS, AML and MPN, respectively. TP53 mutation status did not stratify OS. Four ovarian cancer patients with CH/CCUS transformed to MDS/AML with a median of 5.3 [1-12] years. At transformation, 3/4 had a stable PPM1D mutation, 1/4 an increase in PPM1D VAF (2 to 28%), and 3/4 had TP53 mutations before and at transformation. We analyzed 67 timepoints from 10 ovarian cancer patients during therapy (median 7 per patient). 10/10 received alkylating agents and 5/10 PARPi. The median number of mutations per patient was 4 and 6 at baseline and last follow-up, respectively. PPM1D-mutated clone size increased from 0.4% [0.1-3] at baseline to 7% [2-30] at last follow-up. Beyond clonal expansion, PPM1D VAF dynamics showed non-linear changes related to alkylating agents exposure: 2/10 patients had a continuous expansion, 2/10 had expansion then contraction (GR-1/3, Fig 1B), and 6/10 had expansion then stabilization (GR-2, Fig 1B). Conclusion We described a large cohort of PPM1D mutated patients from CH to AML. Their prognosis was poor independently of TP53 mutation at AML/MDS stage. PPM1D mutations were frequently part of the dominant clone. To confirm the clonal architecture, single cell sequencing analysis in 8 AML/MDS patients are ongoing. Trajectory of PPM1D mutations in ovarian cancer patients revealed a non-linear alkylating agent dependency which warrants further investigation.

Next Generation Sequencing and Cytogenetics in Acute Myeloid Leukemia - Therapeutic and Prognostic Impact: A Retrospective Cohort from a Private Centre of Reference in Latin America

Introduction: Acute Myeloid Leukemia (AML) is a prevalent and severe hematological disorder. The emergence of Next-Generation Sequencing (NGS) and karyotiping, has demonstrated associations between chromosomal abnormalities and mutations with worse treatment response and mortality in AML. Current classification of AML is based on its cytogenetic and molecular profile, as defined by World Health Organization (WHO) 2022 and European Leukemia Net (ELN) 2022. Despite its evident importance in AML, application of these diagnostic methods poses new challenges. In Latin America, particularly, there is a clear limitation of data and studies regarding molecular and cytogenetic analysis in AML patients, due to the relatively short time these techniques have been employed and the few accredited centers able to perform them. Therefore, this study primarily aims to provide an in-depth evaluation of the application of NGS and cytogenetic analysis in AML, identifying the impact of somatic mutations on AML mortality and progression free-survival based on the experience of a reference center in Latin America. Secondarily, we aim to determine the molecular and cytogenetic profile of these population and analyze the impact of age, sex and comorbidities in AML prognosis. Methods: This is an unicentric, retrospective and non-interventional cohort performed in Brazilian Hospital Albert Einstein. Inclusion criteria were adult patients (aged ≥18 years) diagnosed with AML between May 2017 and August 2022, who underwent myeloid panel testing through NGS and cytogenetic evaluation at diagnosis. Independent variables analyzed were age, gender, prognostic risk according to ELN 2022, indication of bone marrow transplantation and molecular and cytogenetic risk at diagnosis. High cytogenetic risk was defined by the presence of at least one of the high-risk karyotype abnormalities described in ELN 2022 and high-risk NGS was defined by mutation in the genes ASXL1, BCOR, EZH2, RUNX1, SF3B1, SRSF2, U2AF1, ZRSR2, TP53 or FLT3. The primary outcomes evaluated were overall survival (OS) and progression-free-survival (PFS), estimated by the Kaplan-Meier method. Secondary outcomes were disease relapse and non-relapse-related mortality (NRM). Results: From a total of 168 selected patients, 64 filled out elegibility criteria and were included in the study. Mean age at diagnosis was 62,7 years old with a equal proportion of both sexes. 7 patients (10,9%) were classified as favourable risk according to ELN 2022, 13 (20,3%) as intermediate risk and 44 (68,8%) as high risk. 42 (65,6%) had a standard-risk karyotype, whereas high-risk karyotype was found in 22 (34,4%) patients. 47 (73,4%) individuals presented high-risk NGS mutations, with a higher prevalence of gene rearrangements in RUNX1 (20,1%), FLT3 (18,7%), TP53 (15,6%) and SRSF2 (15,6%). Only 17 (26,6%) patients reported exclusively standard-risk or no mutations. After a median follow-up was 26 months, estimated OS was 73% in 1 year and 54% in 2 years. 1 and 2 years PFS was 63% and 45%, respectively. 11% relapsed after 1 year and 27% after 2 years. The incidence of NRM was 25% and 28% after 1 and 2 years, respectively. High-risk karyotype was associated with a signficant reduction in overall survival (hazard ratio [HR] 2,27, 95% CI 1,09-4,70; p=0·028) and progression-free-survival (HR 2,15, 95% CI 1,23-4,11; p=0·02). On the other hand, there was no statistically significant association between high molecular risk by NGS and OS (HR 1.45, 95% CI 0.59-3.58; p=0.41) or PFS (HR 1.43, 95% CI 0.65-3.13; p=0.37). Progressive age has discreetly impacted in OS (HR 1,05, 95% CI 1,02-1,08; p=00005), PFS (HR 1,03, 95% CI 1-1,05; p=0,025) and NRM (HR 1,06, 95% CI 1,02-1,11; p=0,002). No differences in outcomes were found between sex, ELN 2022 risk groups, and treatment with bone marrow transplantation. Conclusions: Next-Generation Sequencing, combined with cytogenetic analysis, has improved precision in prognosis prediction of AML patients of each patient, allowing safe implementation of appropriate individualized therapy. This study showed and association between high-risk cytogenetics and worse prognosis in AML, but no association was observed between molecular risk and AML morbimortality. Further studies with a bigger population and longer follow-up should be conducted in Latin America in order to better clarify and delve deeper into the results reported in this cohort.

Pre-Transplant Somatic Co-Occurring Mutations (by next generation sequencing) in Acute Myeloid Leukemia: Frequency and Impact on Clinical Outcomes after Allogeneic Hematopoietic Cell Transplantation - a Large Study on Behalf of the EBMT Acute Leukemia Working Party

Background: Acute myeloid leukemia (AML) is a very heterogeneous hematological malignancy, which includes numerous genetically defined subsets. The genomic classification of AML with the identification of mutations in transcription factors, epigenetic modifiers, spliceosome, cohesin complex, and signaling pathways has led to a more accurate risk stratification model. The main genetic aberrations included in the European LeukemiaNet (ELN) 2022 classification are NPM1 (risk group according to karyotype and FLT3-ITD status), FLT3-ITD (intermediate risk), bZIP in-frame mutated CEBPA (favorable risk), and RUNX1, ASXL1, BCOR, EZH2, RUNX1, SF3B1, SRSF2, STAG2, U2AF1, ZRSR2 and TP53 mutations (all belonging to the adverse risk group). In the context of allogeneic hematopoietic stem cell transplantation (allo-HSCT), the frequency and prognostic value of different gene-gene interactions has not been studied and may differ from that of patients treated with chemotherapy alone. We aimed at evaluating, through the European Society for Blood and Marrow Transplantation (EBMT) database, the frequency and impact of different recurrent somatic mutations, either alone or in association, on prediction of relapse and survival in patients receiving allo-HSCT. Methods: This is a retrospective, registry-based, multicenter analysis from the EBMT with the approval of the EBMT Acute Leukemia Working Party. Adult patients aged more than 18 years with a diagnosis ofAML who received an allo-HSCT between 2013-2022, with an available genetic profile determined at diagnosis by next generation sequencing (NGS) were included. Results: We identified 836 allografted AML patients who had NGS performed at diagnosis. Most of these patients had de novo AML (88%), with a median age of 53 years (range: 18-78 years). Karyotype was favorable in 7%, intermediate in 69% and adverse in 24% of patients. At transplant, 74% of patients were in first remission (CR1) and 13% in CR2. The most frequent detectable mutations by frequency were DNMT3A (33%), FLT3 (29%), NPM1 (29%), TET2 (28%), NRAS (23%), RUNX1 (22%), WT1 (22%), BCOR (19%), ASXL1 (17%), IDH2 (17%), IDH1 (15%), SRSF2 (13%), STAG2 (12%), CEBPA (11%), TP53 (10%), KRAS (10%), and PTPN11 (10%) . By multiple correspondence analysis, two independent groups of co-occurring mutations were identified, the first group included DNMT3A, NPM1 and FLT3, the second group included ASXL1, SRSF2 and RUNX1. Outcome analysis was performed on the subset of 298 patients allografted in CR1 with available data for the aforementioned six genes ( DNMT3A, NPM1, FLT3, ASXL1, SRSF2 and RUNX1). Most of these patients had de novo AML (90%), with a median age of 53 years (range: 19-75 years). Patients received primarily reduced intensity conditioning (58%) and peripheral blood stem cells (93%) from matched sibling donors (35%), matched unrelated (28%), and haploidentical donors (21%). Seventy percent of these patients had intermediate-risk cytogenetics, while 27% were classified as adverse-risk. Median follow up calculated by the reverse Kaplan-Meier method was 2.5 years. Overall, the 2-year relapse incidence (RI), leukemia-free survival (LFS) and overall survival (OS) were 24%, 62% and 69%, respectively. When outcome analysis was performed according to the presence or absence of single mutations, none of the six mutations significantly affected RI or LFS. The 2-year OS was positively affected by the presence of NPM1 mutation (78% vs 66%; p=0.02) and FLT3 (79% vs 65%, p=0.01) but not significantly affected by the other 4 mutations. When mutations were investigated in groups, the 2-year RI, LFS and OS were 24%, 70% and 78%, respectively, for patients with NPM1 mutation regardless of other co-mutations, 35%, 56% and 69% for patients with FLT3-ITDand/or DNMT3A mutation, wild type NPM1, regardless of other co-mutations, 17%, 70% and 74% for patients with RUNX1 and/or ASXL1 and/or SRSF2 mutation without FLT3-ITD and with wild type NPM1 and wild type DNMT3A and 20%, 56% and 61% for patients with all six genes unmutated. Conclusion: NGS at diagnosis can be extremely useful in risk stratification of AML patients undergoing allo-HSCT, potentially allowing adequate post-transplant interventions. Notably, the 2-year LFS of 70% for patients harboring RUNX1 and/or ASXL1 and/or SRSF2 mutation indicates that allo-HSCT can overcome the adverse risk associated with these somatic mutations at diagnosis.

Analysis of Post-Transplant Clinical Outcomes and Prognostic Factors in FLT3-ITD Mutated Acute Myeloid Leukemia Patients Under ELN 2022 Risk Stratification

Background In 2022 European LeukemiaNet (ELN) recommendations for diagnosis and management of acute myeloid leukemia (AML), one of the most important changes made to the genetic risk classification of AML is that the FLT3-ITD allelic ratio (AR) is no longer considered in the risk classification; consequently, AML with FLT3-ITD (without adverse-risk genetic lesions) are now categorized in the intermediate-risk group, irrespective of the allelic ratio or concurrent presence of NPM1 mutation. Aims To investigate the survival outcomes of FLT3-ITD mutated AML patients underwent allogeneic hematopoietic stem cell transplantation (allo-HSCT), based on prognostic stratification defined by the 2017 and 2022 ELN guidelines, and to analyze the risk factors affecting transplant prognosis. Methods A total of 163 FLT3-ITD mutated AML patients underwent allo-HSCT were enrolled in this study. Bone marrow aspirates were obtained, cytogenetic and molecular characteristics were collected at the time of diagnosis. Genetic risk category was classified according to the 2017 and 2022 edition of the European LeukemiaNet (ELN) recommendations. HLA-haploidentical donor transplants were performed if no HLA matched sibling or unrelated donor was available. Patients received myeloablative conditioning with a modified Bu/Cy/Flu/Ara-c (busulfan/cyclophosphamide/fludarabine/cytarabine) or MCBC (melphalan/cladribine/busulfan/cyclophosphamide) preparative regimen. The therapeutic process and clinical outcomes were retrospectively analyzed. Results We assessed 163 adults with FLT3-mutated AML (median age 37 years) allografted between 2017 and 2022 in our clinical center. The median follow-up of alive patients was 24.9 months. The 2-year relapse-free survival (RFS) and overall survival (OS) were 67.2% and 71.9%, respectively. The 2-year incidences of relapse and non-relapse mortality were 27.9% and 12.5%, respectively. There was no significant difference in survival outcomes between the intermediate and high-risk groups after transplantation, both defined by the 2017 and 2022 ELN guidelines. High FLT3-ITD allelic ratio (≥0.5) at initial diagnosis was associated with elevated WBC counts ( P<0.001), but did not affect the response ratefollowing induction chemotherapy ( P=0.059) . Patients with high allelic ratio were more likely to benefit from maintenance therapy after transplantation ( P=0.048 for OS). In univariate analysis, presence of NPM1 mutation and the allelic ratio of FLT3-ITD did not affect any of the transplant outcomes. In multivariate analysis, MRD-negative at transplantation, occurrence of chronic GVHD and posttransplant maintenance therapy improved OS and RFS. Conclusion Allo-HSCT can overcome the poor prognosis in patients with FLT3-ITD mutated AML. Neither the presence of NPM1 mutation nor the FLT3-ITD allelic ratio affected survival. Prevention of disease relapse after transplantation remains a challenge. The main prognostic factors affecting transplant outcomes included pre-transplant MRD status and the application of maintenance therapy. Thus, introduction of FLT3 inhibitors as post-transplant maintenance therapy and the development of highly sensitive MRD assays will further reduce disease recurrence and improve survival.

Treatment of Adverse-Risk and Refractory/Relapsed Acute Myeloid Leukemia (AML) Patients with FLAG-IDA ± Venetoclax and CLAG-M: A Monocentric Experience

Background: ELN 2022 risk classification of AML includes three risk categories and guides treatment choice. Adverse-risk patients, encompassing secondary (sAML) and therapy-related AML (tAML), show lower response rates to upfront traditional intensive chemotherapy (IC) combining anthracycline and cytarabine (3+7 regimen). In refractory/relapsed (R/R) cases, no standard IC approach is established, in particular as a bridging therapy to allogeneic stem cell transplantation (ASCT) that, whenever feasible, remains the only curative strategy. Recently, alternative IC such as CLAG-M (cladribine, cytarabine, G-CSF and mitoxantrone) and FLAG-IDA (fludarabine, cytarabine, G-CSF and idarubicin) were considered for adverse-risk and R/R patients, and were associated with encouraging results including our recent experience with CLAG-M. Specifically, the addition of the bcl2 inhibitor venetoclax (VEN) to FLAG-IDA was associated with deep remissions and a high rate of transition to transplantation (56%). As publications on intensive approaches in these high-risk patients remain sparse in the literature, we report our single-center experience with FLAG-IDA ± VEN compared to our updated experience with CLAG-M in this setting. Methods: We retrospectively collected data on patients with adverse and R/R AML who were treated with either CLAG-M or FLAG-IDA ± VEN at our center from January 2015 to July 2023. We assessed data for baseline characteristics, response to administered treatment as well as univariate and multivariate analyses for specific hazard of survival. Results: Fifty-two patients were enrolled. Thirty patients were treated with CLAG-M (median age 64.0 y [33, 79] and 22 with FLAG-IDA ± VEN (median age 55.5 y [26, 79], p 0.008), 55% of whom received VEN. The distribution of AML types as de novo, sAML and tAML was as follows: 50%, 30% and 20% in CLAG-M; 81.8%, 13.6% and 4.5% in FLAG-IDA ± VEN (ANOVA, p 0.04). Overall, 15 (28.8%) and 37 (71.2%) were newly diagnosed (ND-AML) and R/R cases, respectively: 11 (36.7%) and 19 (63.3%) in CLAG-M; 4 (18.2%) and 18 (81.8%) in FLAG-IDA ± VEN (p 0.115). In R/R, median time to relapse (in months) after first remission was 22 [6, 97] in CLAG-M and 15.0 [4, 36] in FLAG-IDA ± VEN, and median number of prior lines was 1 [0, 4] and 1 [0, 2], respectively. As for ELN 2022 risk groups, 14 (46.7%) and 12 (54.4%) were classified as adverse in CLAG-M and FLAG-IDA ± VEN , respectively (p 0.811). Gene mutations comprised, respectively, the following high-risk defining alterations: 2 (6.7%) and 2 (9.1%) RUNX1; 3 (10%) and 6 (27.3%) ASXL1; 7 (23.3%) and 3 (13.6%) TP53. After induction course, 19 (63.3%) patients in CLAG-M and 19 (86.3%) in FLAG-IDA ± VEN obtained a complete remission (CR). Early deaths (0-30 days) were observed in 4 (13.3%) and 1 (4.5%) patients, respectively. Overall, observed treatment-related toxicities were mild: no unusual infections were seen; median duration of neutropenia (< 0.5 G/L) in days was 30.0 [12, 212] in CLAG-M and 25 [20, 85] in FLAG-IDA ± VEN (p 0.34); median duration of thrombocytopenia (< 100 G/L) was 33.0 [22.0, 212] and 27.5 [22.0, 187] (p 0.38), respectively. Thirteen patients in each group (CLAG-M 43.3%, FLAG-IDA/VEN 59.1%; p 0.086) proceeded to ASCT with a median time in days to transplantation of 100 [31, 212] and 103 [36, 124], respectively. After median follow-up of 365 days [11, 2640] in CLAG-M and 410 [26, 3100] in FLAG-IDA/VEN, 8 (26.7%) and 13 (59.1%) patients were alive (p 0.066). We then performed hazard-risk analyses. In the univariate analysis, these variables were associated with survival: female gender (HR 0.4 [0.17, 0.94], p 0.03); complex karyotype (HR 2.78 [1.07-7.17], p 0.03); ELN 2022 adverse/intermediate (HR 3.82 [0.90-16.21], p 0.05); TP53 mutation (HR 2.84 [1.20-6.71], p 0.013) and ASCT (HR 0.34 [0.16-0.71], p 0.003). In the multivariate analysis, the following variables remained independently associated with survival: ASCT (HR 0.27 [0.12-0.60], p 0.001) and TP53 mutation (HR 3.15 [1.24-8.0], p 0.01). Conclusion: In this cohort of high-risk patients, FLAG-IDA ± VEN and CLAG-M induced high remission rates than prevoiusly reported. These regimens were associated with limited toxicity and bridged to transplant 50% of patients. In addition, ASCT offered a survival benefit, specifically in the FLAG-IDA ± VEN group, although numbers were limited. The presence of a TP53 mutation was associated with inferior outcomes.

The Impact of DNMT3A Mutation on Survival of AML Patients Receiving Allogeneic Hematopoietic Cell Transplantation in First Remission Depends on the Karyotype and Co-Occurring Mutations: On Behalf of the EBMT Acute Leukemia Working Party

DNMT3A gene mutations are more frequent in cytogenetically normal acute myeloid leukemia (AML) and frequently associated with NPM1 mutation and internal tandem duplication (ITD) of the FLT3 gene. However, DNMT3A mutation is not included as a distinct prognostic group in the recent European Leukemia Net (ELN) 2022 genetic risk classification of AML. In the context of allogeneic hematopoietic cell transplantation (allo-HCT), the prognostic value of DNMT3A is still not well studied. We evaluated, through the European Society for Blood and Marrow Transplantation (EBMT) registry, the impact of DNMT3A mutation, according to the karyotype, and NPM1 and FLT3-ITD mutation status, on post-transplant outcomes in AML patients receiving allo-HCT in first complete remission (CR1). We identified a total of 1374 adult AML patients (51% female, median age 57 years; range 18-78 years) with ELN-2022 intermediate- or poor-risk cytogenetics and available DNMT3A, FLT3ITD and NPM1 mutation status, who underwent their first allo-HCT in CR1 between 2013 and 2021. Of these, 631 (46%) patients had DNMT3A mutation. Patients with mutant DNMT3A were more likely to be female (57% versus 46%, p<0.001), to have normal cytogenetics (75% versus 54%, p<0.001), FLT3-ITD co-mutation (50% vs 23%, p<0.001), NPM1 co-mutation (60% vs 21%, p<0.001), and to receive a myeloablative conditioning and in vivo T cell depletion. After a median follow-up of 2.3 years, the 2-year leukemia free survival (LFS) and overall survival (OS) were 62% and 69%, respectively. The subsequent analysis was split according to baseline karyotype (normal or abnormal). Among 870 patients with normal cytogenetics, 472 (54%) patients had DNMT3A mutation. The 2-year relapse incidence (RI), LFS and OS were 24%, 63% and 70% respectively, in patients with DNMT3A mutation compared to 18%, 69% and 74% in patients with wild type DNMT3A. Among these patients with normal karyotype, multivariable analyses (MVA) were done within the 4 groups of FLT3-ITD and NPM1 mutations. Within the group of 324 patients with both FLT3-ITD and NPM1 positives, 244 were triple-positive (75%), LFS and OS were not significantly different with and without DNMT3A mutation (2y LFS: 62% vs 63%, HR: 0.98, p=0.95; 2y OS: 70% vs 73%, HR: 1.17, p=0.58). Conversely, among 146 patients with NPM1 mutation without FLT3-ITD, LFS was significantly lower for 91 patients with DNMT3A mutation (2y LFS: 68% vs 87%, HR: 2.70, p=0.02) but OS was not significantly different (2y OS: 79% vs 87%, HR: 2, p=0.13). On the other hand, among 318 with wild type NPM1 and without FLT3-ITD, LFS and OS were not significantly different for 108 patients with DNMT3A mutation compared to 210 patients with wild type DNMT3A (2y LFS: 62% vs 67%, HR: 1, p=0.98; 2y OS: 65% vs 71%, HR: 1.02, p=0.94). Conversely, among 82 patients with FLT3-ITD and wild type NPM1, 2-year LFS and OS were 53% and 58% respectively, for 29 patients with DNMT3A mutation compared to 71% and 76% respectively, for 53 patients with wild type DNMT3A. The number of patients didn't allow to perform MVA in this group. Finally, among 504 patients with abnormal cytogenetics, 159 (31%) patients had DNMT3A mutation. On MVA, DNMT3A mutation did not significantly affect LFS or OS. Poor-risk cytogenetics was associated with worse LFS, OS and increased RI. Older age negatively affected LFS and OS. In conclusion, in AML patients allografted in CR1, the impact of DNMT3A mutation depends on the karyotype and co-occurring mutations. Our data suggest that DNMT3A mutation negatively affects post-transplant survival selectively in patients with normal karyotype and either NPM1 mutation without FLT3-ITD, or patients with normal karyotype, FLT3-ITD and wild type NPM1. DNMT3A mutation shows no impact on post-transplant outcomes in other settings.

Harnessing Artificial Intelligence for Risk Stratification in Acute Myeloid Leukemia (AML): Evaluating the Utility of Longitudinal Electronic Health Record (EHR) Data Via Graph Neural Networks

Background: AML is a life-threatening disease, and to determine which patients need allogeneic stem cell transplantation, hematologists risk-stratify each case. However, standard risk stratification using the European LeukemiaNet (ELN) criteria is focused on baseline mutations and chromosomal aberrations, and the risk estimate is not updated during a patient's course. In other blood cancers, recalculating the risk with treatment response data can help guide the need for more intensive therapy (Kurtz, et al, Cell, 2019). Furthermore, deep learning graph neural networks (GNN) applied to EHR data have strong predictive power in a hematology context (Fouladvand, et al, J Biomed Inform, 2023). Thus, we evaluated the power of a GNN to predict survival in AML using longitudinal EHR data, specifically with labs and histological features that are not included in the ELN but may capture the treatment response. Methods: Patients who were seen at the Stanford Cancer Institute, had EHR data available within six months of diagnosis, and were diagnosed with AML between June 1998 and January 2021 were included in this retrospective analysis. The GNN was trained to predict survival at two years from diagnosis using the first six months of clinical data. Patients were excluded if they were lost to follow-up before two years or died before six months. Data were collected from structured databases associated with Stanford's EHR, except that diagnosis dates were from Stanford's Cancer Registry, and survival data was supplemented with other databases including the Social Security Death Index. Dysplasia, bone marrow cellularity, and bone marrow blast percentages from pathology reports (“pathology report data”) were extracted using text processing algorithms and weakly supervised machine learning (Ratner, et al, ArXiv, 2017). To represent time series information, we framed each patient's timeline as a network (or “graph”) of events. The primary GNN model was a heterogenous graph transformer classifier with two node types: complete blood count (CBC) data and pathology report data (Hu, et al, ArXiv, 2020). Data from the same week were assumed to be from the same timeframe and connected with bidirectional edges. Data separated by longer time periods were connected with unidirectional edges of a separate edge type. The independent test dataset consisted of patients whose ELN 2022 classification was available, and to train the model, the remaining data were divided into train/validation splits of 0.9/0.1. Results: Of the 2,535 patients with survival data, 1,029 met inclusion criteria. Table 1 summarizes the data available in the EHR for each variable, and nearly all patients had CBC and pathology report data. The area under the receiver operating characteristic (AUROC) using the ELN 2022 criteria for predicting survival in the test dataset was 0.79. The AUROC curve for the GNN model was comparable at 0.76, despite not using any variables from the ELN criteria, and the model effectively stratified patients' disease into high- and low-risk in the independent test dataset (hazard ratio [HR] 3.0, log-rank p = 0.0009). Interestingly, despite not having access to mutation or cytogenetic data, the high-risk cases were enriched in known high-risk mutations, like TP53 and RUNX1, and in high-risk chromosomal aberrations, like 5q deletion (Table 1). Although the model predictions correlated with the ELN criteria in some ways, they also stratified the ELN intermediate-risk AML cases into high and low risk (HR 6.1 for model-predicted high risk among ELN intermediate cases, p = 0.07). Conclusions: Risk stratification using artificial intelligence and longitudinal data from the EHR performed comparably to the ELN 2022 criteria and has the potential to further stratify the ELN categories. The model performed well despite only using histological features and lab values, which are more readily available and more frequently updated than next-generation sequencing results. In the future, this approach may further improve with a larger sample size and additional variables, such as measurable residual disease and treatment information. Given the heterogeneity and increasing complexity of AML classification, leveraging artificial intelligence to assist with classification will be crucial, and these results are a step towards a future where data are automatically extracted from the EHR and used for continuously updated risk stratification.

The role of flow cytometry in the classification of myeloid disorders.

The World Health Organization classification (WHO-HAEM5) and the International Consensus Classification (ICC 2022) of myeloid neoplasms are based on the integration of clinical, morphologic, immunophenotypic, and genomic data. Flow cytometric immunophenotyping (FCIP) allows the identification, enumeration, and characterization of hematopoietic cells, and is therefore apowerful tool in the diagnosis, classification, and monitoring of hematological neoplasms. The vast majority of flow cytometry (FCM) studies in chronic myeloid neoplasms focus on its role in myelodysplastic neoplasms (MDS). FCM can also be helpful for the assessment of myeloproliferative neoplasms (MPN) and MDS/MPN, including the early detection of evolving myeloid or lymphoid blast crisis and the characterization of monocytic subsets. The classification of acute myeloid leukemia (AML) is primarily based on cytogenetic and molecular findings; however, FCIP is needed for subclassification of AML, not otherwise specified (NOS; ICC)/AML defined by differentiation (WHO-HAEM5). The main role of FCM in AML remains in making arapid diagnosis and as atool for measurable residual disease monitoring. Machine learning and artificial intelligence approaches can be used to analyze and classify FCM data. This article, based on an invited lecture at the 106th Annual Meeting of the German Society of Pathology in 2023, reviews the role of FCM in the classification of myeloid neoplasms, including recent publications on the application of artificial intelligence.

КОРОНАВИРУСНАЯ ИНФЕКЦИЯ COVID19 У ПАЦИЕНТОВ С ОСТРЫМ МИЕЛОИДНЫМ ЛЕЙКОЗОМ: ОБЗОР ЛИТЕРАТУРЫ И СОБСТВЕННЫЕ ОПЫТ

The presented literature review presents studies, including our own, on the peculiarities of the course of coronavirus infection in patients with hematological malignancies, in particular in patients with acute myeloid leukemia (AML). A modern classification of AML is presented. It has been shown that in patients with AML, the addition of coronavirus infection SARS-CoV-2 COVID-19 significantly worsens the prognosis of the patient’s life, which makes significant changes in overall survival and mortality. Factors worsening the prognosis of life in patients with AML are advanced age, comorbidities, AML out of remission, the critical course of SARS-CoV-2 COVID19, interruption of treatment without subsequent resumption. In patients with AML and SARS-CoV-2 COVID19, the most optimal treatment tactic is temporary discontinuation of the course of chemotherapy followed by resumption after the regression of the viral infection (both at the onset of the disease and when chemotherapy is already underway). If it is not possible to postpone the course of chemotherapy, treatment can and should be carried out with the provision of adequate accompanying therapy simultaneously with ethological and pathogenetic therapy for SARS-CoV-2 COVID19. Even in the presence of hematological contraindications in patients with AML, SARS-CoV-2 COVID19 treatment should be carried out in full according to clinical guidelines for the treatment of SARS-CoV-2 COVID19.

Assessing the expression of differentiation antagonizing non-protein coding RNA (DANCR) in newly diagnosed Egyptian acute myeloid leukemia patients

BackgroundRole of Long non-coding RNAs in cancer research in the recent years have been highlighted with evidence to their involvement in cancer disease pathogenesis and progression. One of these emerging long non-coding RNAs is differentiation antagonizing non-protein coding RNA (DANCR). DANCR distinct expression in different cancers and implication in tumor signaling pathways made it a promising therapeutic target for cancer.The purpose of this study was to evaluate DANCR expression in de novo acute myeloid leukemia (AML) patients and to assess DANCR expression in relation to cytogenetics and French American British (FAB) AML classification as well as correlate DANCR expression with patients’ response to treatment.The present study included 60 newly diagnosed AML patients and 30 healthy subjects as controls. Relative DANCR expression was done using real time qPCR method.ResultsDANCR was significantly downregulated in AML patients compared to controls (p = 0.038). In addition, DANCR showed significantly lower expression in M4 and M5 compared to M0, M1, and M2 groups (p < 0.001). Furthermore, DANCR expression was significantly downregulated in cytogenetically normal AML patients compared to the controls (p = 0.011).ConclusionSignificant downregulation of DANCR in AML suggests a potential tumor suppressor role and variable expression of DANCR among AML subtypes suggests that DANCR action may be different among AML subtypes. Also, M1 subtype patients with higher DANCR expression were less refractory to treatment and therefore less resistant to cytarabine.

Cytogenetic profile of 1791 adult acute myeloid leukemia in India

BackgroundCytogenetic analysis continues to have an important role in the management of acute myeloid leukemia (AML) because it is essential for prognostication. It is also necessary to diagnose specific categories of AML and to determine the most effective form of treatment. Reports from South Asia are few because the availability of cytogenetic services is relatively limited. MethodsWe performed a retrospective analysis of the cytogenetic findings in adults with AML seen consecutively in a single centre in India. The results were categorised according to the 2022 World Health Organisation (WHO), International Consensus Classification (ICC) and European LeukemiaNet (ELN) classifications.ResultsThere were 1791 patients aged 18–85 years (median age 42, 1086 males). Normal karyotypes were seen in 646 (36%) patients. The 1145 (64%) abnormal karyotypes comprised 585 (32.7%) with recurrent genetic abnormalities (RGA), 403 (22.5%) with myelodysplasia-related cytogenetic abnormalities (MRC), and 157 (8.8%) with other abnormalities. There were 567 (31.7%) patients with solitary abnormalities and 299 (16.7%) with two abnormalities. Among the 279 (15.6%) patients with ≥ 3 abnormalities, 200 (11.2%) had complex karyotypes (CK) as per the WHO/ICC and 184 (10.3%), as per the ELN definition. There were 158 (8.8%) monosomal karyotypes (MK). Patients with normal karyotypes had a higher median age (45 years) than those with abnormal karyotypes (40 years, p < 0.001), and those with ≥ 3 abnormalities (43 years), than those with fewer abnormalities (39 years, p = 0.005). Patients with CK (WHO/ICC) and monosomal karyotypes had a median age of 48 years. Those with RGA had a lower median age (35 years, p < 0.001) than MRC (46 years) or other abnormalities (44 years). The t(15;17) was the most common abnormality (16.7%),followed by trisomy 8 (11.6%), monosomy 7/del 7q (9.3%), t(8;21) (7.2%), monosomy 5/del 5q (6.7%) and monosomy 17/del 17p (5.2%). ConclusionOur findings confirm the lower age profile of AML in India and show similarities and differences with respect to the frequencies of individual abnormalities compared to the literature. The frequencies of the t(15;17), trisomy 8 and the high-risk abnormalities monosomy 7 and monosomy 5/del 5q were higher, and that of the inv(16), lower than in most reports.

Interaction between myelodysplasia-related gene mutations and ontogeny in acute myeloid leukemia.

Accurate classification and risk stratification are critical for clinical decision making in patients with acute myeloid leukemia (AML). In the newly proposed World Health Organization and International Consensus classifications of hematolymphoid neoplasms, the presence of myelodysplasia-related (MR) gene mutations is included as 1 of the diagnostic criteria for AML, AML-MR, based largely on the assumption that these mutations are specific for AML with an antecedent myelodysplastic syndrome. ICC also prioritizes MR gene mutations over ontogeny (as defined in the clinical history). Furthermore, European LeukemiaNet (ELN) 2022 stratifies these MR gene mutations into the adverse-risk group. By thoroughly annotating a cohort of 344 newly diagnosed patients with AML treated at the Memorial Sloan Kettering Cancer Center, we show that ontogeny assignments based on the database registry lack accuracy. MR gene mutations are frequently observed in de novo AML. Among the MR gene mutations, only EZH2 and SF3B1 were associated with an inferior outcome in the univariate analysis. In a multivariate analysis, AML ontogeny had independent prognostic values even after adjusting for age, treatment, allo-transplant and genomic classes or ELN risks. Ontogeny also helped stratify the outcome of AML with MR gene mutations. Finally, de novo AML with MR gene mutations did not show an adverse outcome. In summary, our study emphasizes the importance of accurate ontogeny designation in clinical studies, demonstrates the independent prognostic value of AML ontogeny, and questions the current classification and risk stratification of AML with MR gene mutations.

Applicability of 2022 classifications of acute myeloid leukemia in the real-world setting.

The increasing knowledge of molecular genetics of acute myeloid leukemia (AML) necessitated the update of previous diagnostic and prognostic schemes, which resulted in the development of the World Health Organization (WHO), the International Consensus Classification (ICC), and the new European LeukemiaNet (ELN) recommendations in 2022. We aimed to provide a real-world application of the new models, unravel differences and similarities, and test their implementation in clinical AML diagnosis. A total of 1001 patients diagnosed with AML were reclassified based on the new schemes. The overall diagnostic changes between the WHO 2016 and the WHO 2022 and ICC classifications were 22.8% and 23.7%, respectively, with a 13.1% difference in patients' distribution between ICC and WHO 2022. The 2022 ICC "not otherwise specified" and WHO "defined by differentiation" AML category sizes shrank when compared with that in WHO 2016 (24.1% and 26.8% respectively, vs 38.7%), particularly because of an expansion of the myelodysplasia (MDS)-related group. Of 397 patients with a MDS-related AML according to the ICC, 55.9% were defined by the presence of a MDS-related karyotype. The overall restratification between ELN 2017 and ELN 2022 was 12.9%. The 2022 AML classifications led to a significant improvement of diagnostic schemes. In the real-world setting, conventional cytogenetics, usually rapidly available and less expensive than molecular characterization, stratified 56% of secondary AML, still maintaining a powerful diagnostic role. Considering the similarities between WHO and ICC diagnostic schemes, a tentative scheme to generate a unified model is desirable.