Validating AI in Haematological Diseases

The Disease

Myelodysplastic syndromes (MDS) are a group of bone marrow failure disorders that typically affect the elderly. Patients suffer from blood cytopenia (low blood cell counts), since their bone marrow is no longer able to produce enough healthy blood cells. The disease is also known as a form of blood cancer, and in some patients can evolve into acute myeloid leukemia (AML), which is usually fatal if not treated.

Validation

Prevention based on Genomic ScreeningInvestigate factors that influence the development of MDS, enabling early-stage identification of individuals at risk. Omics-based Classification and PrognosisPersonalized predictive models through integration of comprehensive genomic and clinical information. Omics-based Clinical Decision MakingAI-based algorithms to stratify the individual probability of response to specific treatments. Drug RepurposingBuild a rationale for drug repurposing in specific subsets of MDS.

Database

Targeted mutation screening (NGS on 50 genes related to clonal haematopoiesis) from 11,000 healthy elderly subjects (>65 yo) from population-based studies. Information on targeted mutation screening (NGS including 60 genes related to MDS) from 6700 MDS patients; data on RNA-sequencing of hematopoietic progenitors from 1200 MDS patients; whole exome/genome sequencing from 900 MDS patients. Information on targeted mutation screening (NGS including 60 genes related to MDS) from 1500 MDS patients who received transplantation. WES, Gene Expression, clinical data on haematologicalmalignancies downloaded from TCGA, GEO and MILE databases,including more than 2400 patients with HD.

Clinical Partners

The Disease

Multiple Myeloma (MM) is a type of bone marrow cancer originating in plasma cells, a type of white blood cell responsible for producing antibodies to fight off infections. In patients with MM, cancerous plasma cells accumulate in the bone marrow and produce abnormal proteins instead, which can lead to decreased blood cell numbers, bone and kidney damage.

Validation

Understand Disease ComplexityDescribe the different layers of MM heterogeneity integrating baseline genomic and imaging data. Identify Evolution DynamicsDefine the quantitative and qualitative dynamics of the disease in time. Study Risk ProgressionDevelop a prognostic risk score for the baseline and the disease remaining after therapy. Integrate Radiomics and RadiogenomicsDevelop and validate a model to predict treatment response and determine progression-free survival.

Database

Baseline clinical data and whole genome Copy Number Alterations (CNAs) landscape data (as detected by SNPs array) from 1,000 patients. Baseline clinical and genomic data (coming from both BM clone and liquid biopsy) and imaging data (PET-CT and WB-MRI) at baseline, as well as molecular (NGS) and imaging (PET-CT and WB-MRI) minimal residual disease (MRD) data from 200 patients. Clinical and genomic data from 750 patients available from public repositories (1A12 release CoMMpass database). Data from 300 newly diagnosed patients who will receive induction therapy followed by high dose melphalan and lenalidomide maintenance.

Clinical Partners

The Disease

Sickle Cell Disease (SCD) is a group of hereditary red blood cell disorders. It is a rare, chronic and life-threatening disease. In patients with SCD, red blood cells become C-shaped in resemblance to a sickle, the farming tool the disease is named after. Sickle cells die early and tend to clog the blood flow when going through small blood vessels, so patients usually suffer from low red blood cell counts, infections, acute chest syndrome and strokes.

Validation

Identify gene mutations associated to high levels ofinflammation markers (C-reactive protein - CRP)SCD patients will be examined for correlations between previously identified genetic inflammatory risk profiles CRP level, aiming to develop high inflammation prediction models.AI allocation of SCD patients to a sickling risk profile based on their genomic profile Understand which genetic loci (GWAS) are associated with SCD patient-specific blood rheology and the point of sickling (PoS).Develop a combined model from inflammation and sickling risks to predict clinical outcomeUsing the extent of renal damage (nephropathy) expressed as microalbuminuria as gold standard, together with other known genetic modifiers of SCD severity.AI-based RadiomicsBuild a probability score using AI-based brain MRI image analysis to predict incidents of silent infarction (MRI based outcome) in young SCD patients.

Database

MRI data file (DICOM, CTI, SIGNA etc.), Lorrca Oxygenscan raw data file (RTS,TXT), GWAS data file, laboratory parameters (RTS, XLS etc.), clinical parameters and baseline characteristics (RTS, XLS etc.)

Clinical Partners