SCD is a chronic life-threatening multisystem disorder, autosomal recessively inherited,
caused by the presence of abnormal hemoglobin S (HbS) resulting from the sickle mutation in
the HBB gene. In spite of being a single gene mutation disorder, SCD presents extreme
phenotypic variability that is incompletely understood. Several genetic and environmental
factors are supposed to have an impact on disease phenotype, clinical manifestations,
progression of organ damage and quality of life throughout the lifespan.
Although significant progress has been made over the past few decades in the highly complex
pathophysiology of SCD, the possibility of personalised medicine is still in its infancy.
There is a lack of markers of disease severity, prognosis, and response to treatment. In
particular, the heterogeneity of clinical expression of the disease along with long-term
chronic complications due to the increased lifespan of patients should be addressed by
innovative and personalised treatments. Furthermore, assessing the role of the novel
treatments both in regards of long-term efficacy and safety but also of cost/efficacy ratio
are required. The scarcity and fragmentation of SCD data prevent researchers from reaching
the critical numbers needed for basic and clinical research. Research and data-driven
solutions are therefore essential to improve the care of SCD patients and their quality of
life.
The availability of numerous treatment options as well as the high disease heterogeneity
highlight the need to address patients' severity profiles and offer the best care for each
affected individual. Developing the GENOMED4ALL AI algorithms for SCD will be of great
importance for the in-depth characterization and prediction of the diverse complications of
SCD. The primary endpoints of interest include:
Improving SCD classification
Develop a probability score to predict various patterns recognized by Artificial
Intelligence (AI) based analyzing brain magnetic resonance imaging (Radiomics)
To develop predictive risk scores for the occurrence of most prevalent and severe
clinical outcomes
To develop predictive risk scores over time for the appearance of most prevalent and
severe clinical outcomes.
RADeep will be used for standardization of existing clinical and laboratory data. A CRF was
developed, including just over 250 data elements. The GenoMed4All CRF builds on previous work
performed by RADeep and includes the "set of common data elements for rare disease
registration", which was released in December 2017 as result of a dedicated working group
facilitated by the Joint Research Centre (JRC). This approach will ensure interoperability
with other similar initiatives in Europe and will also enable the collected data to be reused
for future research studies.
Genome-wide Association Studies (GWAS) extends the concept of association studies to assay
hundreds of thousands of single-nucleotide polymorphisms (SNPs) simultaneously and provide a
cost-effective way to explore genetic variants across the whole genome. But despite
considerable interest in identifying genetic modifiers in SCD, the majority of previous GWAS
searched for genetic linkage and association with HbF levels, an established ameliorating
factor of disease severity. Addintionally, the utilization of data science and artificial
intelligence (AI) has been limited in SCD research. Therefore, the generation of GWAS data
combined with the use of the most recent imputation panel for imputation offers an
opportunity for the development of novel AI techniques and for novel discoveries in SCD.
Silent Cerebral Infarcts (SCIs) are a significant cause of morbidity in SCD: they affect 25%
of children by the age of 6 and 40% by the age of 18 with consequences on cognition,
schooling, working capacity and quality of life. Hence, one of the aims of the SCD clinical
case in GENOMED4ALL is the use of radiomics - quantitative method for the evaluation and
interpretation of medical images- and AI firstly to develop an automatic and uniform
identification and characterization of SCI on MRIs, secondly, to correlate imaging data with
other types of OMICS data in order to predict risk of occurrence and recurrence.
The deformability of red blood cells (RBC) from individuals with SCD is markedly abnormal,
regardless of genotype. Several studies reported some associations between the degree of
impairment of RBC deformability measured at steady state in SCD patients and the presence of
chronic complications, such as priapism, leg ulcers, glomerulopathy, etc. The recently
developed technique of oxygen gradient ektacytometry allows for a more comprehensive
functional characterization and rheological behavior of SCD RBCs over a range of oxygen
tensions to test whether the rheological changes could reflect clinical
severity/complications.Data on rheological characteristics of RBC on all patients in steady
state is going to be obtained through Laser Optical Rotational Red Cell Analyzer (Lorrca)
ektacytometer (RR Mechatronics).
By combining a large amount standardized multimodal (clinical, multi-omics, and imaging)
datasets, the investigators hypothesize that AI will allow to understand better SCD biology
and classification, enhance prognostic/predictive capacity of currently available tools and
apply treatments in a more targeted way, thus facilitating the implementation of personalized
medicine program across EU.