Synthetic Data Generation By Artificial Intelligence to Accelerate Translational Research and Precision Medicine in Hematological Malignancies

Synthetic Histopathological Images Generation with Artificial Intelligence to Accelerate Research and Improve Clinical Outcomes in Hematology

Frequent Mutation of the Polycomb-Associated Gene ASXL1 in the Myelodysplastic Syndromes and in Acute Myeloid Leukaemia.

Association Between Gene Expression Profiles and Commonly Mutated Genes In The Hematopoietic Stem Cells Of Patients With Myelodysplastic Syndromes

Synthetic data are artificial data generated without including any real patient information by an algorithm trained to learn the characteristics of a real source data set and became widely used to accelerate research in life sciences. We aimed to (1) apply generative artificial intelligence to build synthetic data in different hematologic neoplasms; (2) develop a synthetic validation framework to assess data fidelity and privacy preservability; and (3) test the capability of synthetic data to accelerate clinical/translational research in hematology. A conditional generative adversarial network architecture was implemented to generate synthetic data. Use cases were myelodysplastic syndromes (MDS) and AML: 7,133 patients were included. A fully explainable validation framework was created to assess fidelity and privacy preservability of synthetic data. We generated MDS/AML synthetic cohorts (including information on clinical features, genomics, treatment, and outcomes) with high fidelity and privacy performances. This technology allowed resolution of lack/incomplete information and data augmentation. We then assessed the potential value of synthetic data on accelerating research in hematology. Starting from 944 patients with MDS available since 2014, we generated a 300% augmented synthetic cohort and anticipated the development of molecular classification and molecular scoring system obtained many years later from 2,043 to 2,957 real patients, respectively. Moreover, starting from 187 MDS treated with luspatercept into a clinical trial, we generated a synthetic cohort that recapitulated all the clinical end points of the study. Finally, we developed a website to enable clinicians generating high-quality synthetic data from an existing biobank of real patients. Synthetic data mimic real clinical-genomic features and outcomes, and anonymize patient information. The implementation of this technology allows to increase the scientific use and value of real data, thus accelerating precision medicine in hematology and the conduction of clinical trials.

Somatic mutations in genes that are involved in pre-mRNA splicing (i.e., spliceosome genes) occur in ~50% of patients with myelodysplastic syndromes (MDS). Commonly mutated spliceosome genes (e.g., SF3B1, SRSF2, and U2AF1) have variable prognostic significance in patients with myeloid malignancies. For example, SRSF2 and U2AF1 mutations are associated with worse and SF3B1 mutations with better clinical outcomes compared to patients without a spliceosome gene mutation. While multiple groups have shown that RNA splicing is altered in patients with spliceosome gene mutations, the contribution of these alterations to disease pathogenesis is largely unknown. Several groups have reported that RNA splicing and hematopoiesis are altered in mutant-expressing mouse models, providing direct evidence that spliceosome gene mutations contribute to some MDS phenotypes. Spliceosome gene mutations have also been detected in the blood cells of healthy elderly individuals with clonal hematopoiesis who are at an increased risk of developing hematopoietic cancer, further implicating these mutations as early genetic drivers in MDS. Consistent with this possibility, spliceosome gene mutations tend to occur in the founding clone (i.e., present in every tumor cell in a patient) of myeloid malignancies. Collectively, these observations indicate that targeting mutant cells could have a large clinical impact. Several observations indicate that spliceosome gene mutations may create a genetic vulnerability in cells that could be exploited therapeutically. Spliceosome gene mutations are typically heterozygous and mutually exclusive of each other in patients, implying either a redundancy in pathogenic function of mutant genes or that a cell cannot tolerate two spliceosome perturbations at once. Accumulating evidence now suggests that heterozygous mutant cells require appropriate expression levels of the nonmutant residual allele or other splicing factors to maintain cell viability. Therefore, we and others hypothesize that cells harboring spliceosome mutations will have increased sensitivity to pharmacologic perturbation of the spliceosome, providing a new treatment approach for patients with these mutations. To examine this, we used sudemycin compounds that bind the SF3B1 protein to modulate pre-mRNA splicing. We show that treatment of U2AF1 mutant mice with sudemycin results in attenuation of mutant-associated hematopoietic progenitor cell expansion compared to control treated mice. Collectively, data from our group and others suggest that cells expressing mutant spliceosome genes have an increased sensitivity to pharmacologic perturbation of splicing. Clinical trials are being designed and conducted to treat spliceosome mutant hematologic cancers with splicing modulator compounds. Citation Format: Matthew J. Walter. Spliceosome gene mutations in MDS: Biology and potential therapeutic strategies [abstract]. In: Proceedings of the Second AACR Conference on Hematologic Malignancies: Translating Discoveries to Novel Therapies; May 6-9, 2017; Boston, MA. Philadelphia (PA): AACR; Clin Cancer Res 2017;23(24_Suppl):Abstract nr IA11.

Somatic Mutations in Myelodysplastic Syndrome

ASXL1 Mutations Predict a Poor Response to Darbepoetin Alfa in Anemic Patients with Low-Risk MDS: A Multicenter, Phase II Study

Generation of Multimodal Longitudinal Synthetic Data By Artificial Intelligence to Improve Personalized Medicine in Hematology

Hereditary Predisposition to Acute Myeloid Leukemia in Older Adults.

Clonal Suppression of TP53 Mutant MDS and Oligoblastic AML with Hypomethylating Agent Therapy Improves Overall Survival

Detection of TP53 Mutations in Myelodysplastic Syndromes (MDS) and Acute Myeloid Leukemia (AML). a Comparison Between a Functional Method (FASAY) and Next Generation Sequencing (NGS)

6514 Background: Mutations in the tumor suppressor TP53 gene are common in elderly patients with myelodysplastic syndrome (MDS) and acute myeloid leukemia (AML) and confer resistance to conventional chemotherapeutic DNA damaging agents. Venetoclax (Ven) added to the hypomethylating agents (HMA) of Decitabine or Azacitidine is the current standard of care for elderly patients with AML and is frequently used in high-risk MDS (HR-MDS). Currently approved dosing schedules of HMA/Ven rely on cytotoxicity and have not improved outcomes in the TP53 mutant population. The efficacy and tolerability of metronomic weekly dosing of Decitabine and Ven in HR-MDS and AML were previously described (Goldfinger et al, Blood 2024). Mechanistically, metronomic dosing relies on terminal differentiation, rather than cytotoxicity making it an attractive regimen for TP53 mutant MDS/AML. Methods: Patients with histologically confirmed AML or MDS and a TP53 mutation received a once-weekly dose of decitabine 0.2 mg/kg subcutaneously and one dose of Ven 400 mg on days 1, 8, 15 and 22 of a 28-day cycle. Results: Between April 2020 and January 2025, 40 patients with TP53 mutated myeloid malignancies were treated with metronomic weekly low-dose Decitabine/Ven (14 AML, 26 MDS). Twenty-four patients were followed prospectively as part of a clinical trial (NCT05184842), and 16 were treated off-trial and had data collected retrospectively. Median age at diagnosis was 76.5 years, 13 (32%) were from minority backgrounds, 28 (70%) had complex cytogenetics and 31 (82%) had biallelic TP53 mutations (median VAF 36%). All AML patients were ELN-poor risk, 21 MDS patients (82%) were R-IPSS high or very high risk. The median time on therapy was 5.8 months, with 10 (25%) patients still on therapy at time of data cut-off. Four patients in the AML and five in the MDS cohorts were not evaluable (2 withdrew consent, 1 lost to follow-up and 6 did not have a BM biopsy for evaluation). Of the evaluable AML patients, 7 (70%) achieved a complete remission (CR), 3 (30%) did not respond. In the evaluable MDS patients, 9 (43%) achieved a CR and 3 (15%) a marrow CR, 4 (19%) with stable disease, 5 (24%) with no response. Of the 26 patients who were transfusion-dependent at the start of therapy, 15 (58%) became transfusion-independent. For the entire cohort (n=40), the median overall survival (OS) was 11.3 months. For the AML and MDS cohorts, the OS was 11.6 and 9.9 months, respectively. In patients who underwent allogeneic stem cell transplant (n=6), OS was 16 months. Non-heme therapy-related adverse events of ≥ grade 3 was seen in 13 (54%) of patients. Conclusions: In this cohort, of elderly patients with poor risk TP53 mutated MDS and AML the use of a non-cytotoxic dosing schedule of Decitabine and Ven resulted in over half the patients achieving a CR and transfusion independence. The median OS of 11.3 months compares favorably to currently approved cytotoxic dosing of HMA/Ven. Clinical trial information: NCT05184842 .

Development of a High-Resolution Melting Curve Analysis Screening Test for SRSF2 Splicing Factor Gene Mutations in Myelodysplastic Syndromes

Enasidenib: a magic bullet for myelodysplastic syndromes?

The Impact of PHF6 Mutations in Myelodysplastic Syndromes, Chronic Myelomonocytic Leukemia, and Acute Myeloid Leukemia