Molecular-Based Ecosystem to Improve Personalized Medicine in Chronic Myelomonocytic Leukemia - PubMed
5 hours ago
- #Molecular Prognostication
- #CMML
- #Personalized Medicine
- CMML is a rare myeloid neoplasm with poor outcomes, and limited molecular info has been used in its classification.
- Study analyzed 3013 patients retroactively and 516 prospectively, using multimodal data to develop molecular-based taxonomy and prognostication.
- Unsupervised clustering identified nine entities with distinct genomic features, including splicing, transcription, and signaling pathway aberrations.
- 15% of patients had molecular/clinical overlap with other myeloid neoplasms.
- The iCPSS integrates mutations in nine genes with hematologic and cytogenetic data, identifying five risk groups with distinct survival outcomes.
- iCPSS reassigned 55% of patients to different risk groups and outperformed existing prognostic models.
- iCPSS refined timing for allogeneic transplantation, changing strategy in 31% of cases and improving life expectancy.
- A federated learning platform enables continuous, privacy-preserving model updates across multiple centers.
- Molecular data enhances CMML classification, prognosis, clinical decisions, and trial design.