Machine Learning Improves Risk Stratification in Myelofibrosis: An Analysis of the Spanish Registry of Myelofibrosis

SOHO State of the Art Updates and Next Questions | Diagnosis, Outcomes, and Management of Prefibrotic Myelofibrosis

Prefibrotic primary myelofibrosis (prefibrotic PMF) is a myeloproliferative neoplasm with distinct characteristics comprising histopathological and clinico-biological parameters. It is classified as a subtype of primary myelofibrosis. In clinical practice, it is essential to correctly distinguish prefibrotic PMF from essential thrombocythemia especially but also overt PMF besides other myeloid neoplasms. Risk stratification and survival outcomes for prefibrotic PMF are worse than that of ET but better than that of overt PMF. Rates of progression to overt PMF and blast phase disease are also higher for prefibrotic PMF than ET. In this review we first discuss the historical context to the evolution of prefibrotic PMF as an entity, its presenting features and diagnostic criteria. We emphasize the differences between prefibrotic PMF, ET, and overt PMF with regards to presenting features and disease outcomes including thrombohemorrhagic events and progression to fibrotic and blast phase disease. Next, we discuss the risk stratification models and contextualize these in the setting of clinical management. We share our view of personalizing treatment to address unique patient needs in the context of currently available management options. Lastly, we discuss areas of critical need in clinical research and speculate on the possibility of future disease course modifying therapies in prefibrotic PMF.

Refining risk prediction in pediatric acute lymphoblastic leukemia through DNA methylation profiling

Acute lymphoblastic leukemia (ALL) is the most prevalent cancer in children, and despite considerable progress in treatment outcomes, relapses still pose significant risks of mortality and long-term complications. To address this challenge, we employed a supervised machine learning technique, specifically random survival forests, to predict the risk of relapse and mortality using array-based DNA methylation data from a cohort of 763 pediatric ALL patients treated in Nordic countries. The relapse risk predictor (RRP) was constructed based on 16 CpG sites, demonstrating c-indexes of 0.667 and 0.677 in the training and test sets, respectively. The mortality risk predictor (MRP), comprising 53 CpG sites, exhibited c-indexes of 0.751 and 0.754 in the training and test sets, respectively. To validate the prognostic value of the predictors, we further analyzed two independent cohorts of Canadian (n = 42) and Nordic (n = 384) ALL patients. The external validation confirmed our findings, with the RRP achieving a c-index of 0.667 in the Canadian cohort, and the RRP and MRP achieving c-indexes of 0.529 and 0.621, respectively, in an independent Nordic cohort. The precision of the RRP and MRP models improved when incorporating traditional risk group data, underscoring the potential for synergistic integration of clinical prognostic factors. The MRP model also enabled the definition of a risk group with high rates of relapse and mortality. Our results demonstrate the potential of DNA methylation as a prognostic factor and a tool to refine risk stratification in pediatric ALL. This may lead to personalized treatment strategies based on epigenetic profiling.

Allogeneic hematopoietic cell transplantation in patients with CALR-mutated myelofibrosis: a study of the Chronic Malignancies Working Party of EBMT

Allogeneic hematopoietic cell transplantation (allo-HCT) is curative for myelofibrosis (MF) but assessing risk-benefit in individual patients is challenging. This complexity is amplified in CALR-mutated MF patients, as they live longer with conventional treatments compared to other molecular subtypes. We analyzed outcomes of 346 CALR-mutated MF patients who underwent allo-HCT in 123 EBMT centers between 2005 and 2019. After a median follow-up of 40 months, the estimated overall survival (OS) rates at 1, 3, and 5 years were 81%, 71%, and 63%, respectively. Patients receiving busulfan-containing regimens achieved a 5-year OS rate of 71%. Non-relapse mortality (NRM) at 1, 3, and 5 years was 16%, 22%, and 26%, respectively, while the incidence of relapse/progression was 11%, 15%, and 17%, respectively. Multivariate analysis showed that older age correlated with worse OS, while primary MF and HLA mismatched transplants had a near-to-significant trend to decreased OS. Comparative analysis between CALR- and JAK2-mutated MF patients adjusting for confounding factors revealed better OS, lower NRM, lower relapse, and improved graft-versus-host disease-free and relapse-free survival (GRFS) in CALR-mutated patients. These findings confirm the improved prognosis associated with CALR mutation in allo-HCT and support molecular profiling in prognostic scoring systems to predict OS after transplantation in MF.

AIPSS-MF machine learning prognostic score validation in a cohort of myelofibrosis patients treated with ruxolitinib

BACKGROUND

In myelofibrosis (MF), new model scores are continuously proposed to improve the ability to better identify patients with the worst outcomes. In this context, the Artificial Intelligence Prognostic Scoring System for Myelofibrosis (AIPSS-MF), and the Response to Ruxolitinib after 6 months (RR6) during the ruxolitinib (RUX) treatment, could play a pivotal role in stratifying these patients.

AIMS

We aimed to validate AIPSS-MF in patients with MF who started RUX treatment, compared to the standard prognostic scores at the diagnosis and the RR6 scores after 6 months of treatment.

METHODS AND RESULTS

At diagnosis, the AIPSS-MF performs better than the widely used IPSS for primary myelofibrosis (C-index 0.636 vs. 0.596) and MYSEC-PM for secondary (C-index 0.616 vs. 0.593). During RUX treatment, we confirmed the leading role of RR6 in predicting an inadequate response by these patients to JAKi therapy compared to AIPSS-MF (0.682 vs. 0.571).

CONCLUSION

The new AIPSS-MF prognostic score confirms that it can adequately stratify this subgroup of patients already at diagnosis better than standard models, laying the foundations for new prognostic models developed tailored to the patient based on artificial intelligence.

Artificial Intelligence-Assisted Diagnostic Cytology and Genomic Testing for Hematologic Disorders

Artificial intelligence (AI) is a rapidly evolving field of computer science that involves the development of computational programs that can mimic human intelligence. In particular, machine learning and deep learning models have enabled the identification and grouping of patterns within data, leading to the development of AI systems that have been applied in various areas of hematology, including digital pathology, alpha thalassemia patient screening, cytogenetics, immunophenotyping, and sequencing. These AI-assisted methods have shown promise in improving diagnostic accuracy and efficiency, identifying novel biomarkers, and predicting treatment outcomes. However, limitations such as limited databases, lack of validation and standardization, systematic errors, and bias prevent AI from completely replacing manual diagnosis in hematology. In addition, the processing of large amounts of patient data and personal information by AI poses potential data privacy issues, necessitating the development of regulations to evaluate AI systems and address ethical concerns in clinical AI systems. Nonetheless, with continued research and development, AI has the potential to revolutionize the field of hematology and improve patient outcomes. To fully realize this potential, however, the challenges facing AI in hematology must be addressed and overcome.

Association of Myelofibrosis Phenotypes with Clinical Manifestations, Molecular Profiles, and Treatments

Myelofibrosis (MF) presents an array of clinical manifestations and molecular profiles. The two distinct phenotypes- myeloproliferative and myelodepletive or cytopenic- are situated at the two poles of the disease spectrum and are largely defined by different degrees of cytopenias, splenomegaly, and distinct molecular profiles. The myeloproliferative phenotype is characterized by normal/higher peripheral blood counts or mildly decreased hemoglobin, progressive splenomegaly, and constitutional symptoms. The myeloproliferative phenotype is typically associated with secondary MF, higher JAK2 V617F burden, fewer mutations, and superior overall survival (OS). The myelodepletive phenotype is usually associated with primary MF, ≥2 cytopenias, modest splenomegaly, lower JAK2 V617F burden, higher fibrosis, greater genomic complexity, and inferior OS. Cytopenias are associated with mutations in epigenetic regulators/splicing factors, clonal evolution, disease progression, and shorter OS. Clinical variables, in conjunction with the molecular profiles, inform integrated prognostication and disease management. Ruxolitinib/fedratinib and pacritinib/momelotinib may be more suitable to treat patients with the myeloproliferative and myelodepletive phenotypes, respectively. Appreciation of MF heterogeneity and two distinct phenotypes, the different clinical manifestations and molecular profiles associated with each phenotype alongside the growing treatment expertise, the development of non-myelosuppressive JAK inhibitors, and integrated prognostication are leading to a new era in patient management. Physicians can increasingly tailor personalized treatments that will address the unique unmet needs of MF patients, including those presenting with the myelodepletive phenotype, to elicit optimal outcomes and extended OS across the disease spectrum.

Myelofibrosis and Survival Prognostic Models: A Journey between Past and Future

Among the myeloproliferative diseases, myelofibrosis is a widely heterogeneous entity characterized by a highly variable prognosis. In this context, several prognostic models have been proposed to categorize these patients appropriately. Identifying who deserves more invasive treatments, such as bone marrow transplantation, is a critical clinical need. Age, complete blood count (above all, hemoglobin value), constitutional symptoms, driver mutations, and blast cells have always represented the milestones of the leading models still used worldwide (IPSS, DIPSS, MYSEC-PM). Recently, the advent of new diagnostic techniques (among all, next-generation sequencing) and the extensive use of JAK inhibitor drugs have allowed the development and validation of new models (MIPSS-70 and version 2.0, GIPSS, RR6), which are continuously updated. Finally, the new frontier of artificial intelligence promises to build models capable of drawing an overall survival perspective for each patient. This review aims to collect and summarize the existing standard prognostic models in myelofibrosis and examine the setting where each of these finds its best application.