the PRISCILA project

Gestational diabetes (GDM) is the most frequent medical complication in pregnancy, increasing the risk of pregnancy complications, and leading to increased cardio-metabolic risk in the long-term for mother and child. The current management strategy for GDM is one-size-fits-all, despite the heterogeneity of GDM etiology. Moreover, current risk stratification strategies have limited value to inform clinical decisions. Detailed characterization of etiological subtypes of GDM are of paramount importance (1) to understand the underlying pathophysiology for differential risk for complications and (2) to inform precision approaches for risk stratification and management of GDM. I hypothesize that by careful characterization of different GDM subtypes through integration by artificial intelligence (AI) computation of a comprehensive dataset of clinical and biochemical characteristics, continuous glucose monitoring data, novel biomarkers, metabolomics and polygenic risk scores, a novel risk stratification can be discovered. In the PRISCILA project, I will therefore enrol a large cohort of GDM patients’ (including a large group with GDM diagnosed early in pregnancy), whereby prospective collection of consolidated and innovative risk predictors will take place in pregnancy and postpartum, in order to generate an unprecedented multidimensional dataset, and allowing for patients deep phenotyping. I will analyse data leveraging AI computation to find relevant associations with clinical outcomes, and compare a new AI integrated risk algorithm with current stratification strategy. This project will provide important novel knowledge on the characteristics, pathophysiology and risk for complications across different GDM subtypes. This will also be very insightful for other research fields to develop novel risk prediction models integrated with AI and precision markers. If successful, it will allow for a paradigm shift to better tailor the management strategy for GDM.

A large prospective cohort study will be performed to include 1000 women with GDM, of which at least 40% with early GDM. This will be a multicentric study with the participation with at least 14 Belgian hospitals. Women will be followed up from early pregnancy till one year postpartum, allowing for detailed prospective collection of clinical and biochemical data, including fetal ultrasound scan data, a 75g oral glucose tolerance test (performed in early and late pregnancy, and at 3 and 12 months postpartum), several validated indices of insulin sensitivity and β-cell function, HbA1c, lipid profile, continuous glucosa monitoring data, metabolomics, NIPT (Non-Invasive Prenatal Testing to calculate polygenic risk scores ) and Insulin-like growth factor binding protein 1 (IGFBP1).  GDM will be diagnosed in early pregnancy with higher cut-offs compared to later in pregnancy in line with the Flemish guidelines for screening for GDM. For uniformity and to reduce costs, laboratory analyses will be done centrally at the laboratory of UZ Leuven and blood samples will be stored temporarily in our established biobank.

For the data analyses, you will mainly focus on the novel biomarkers, the polygenic risk scores and metabolomics. As part of your PhD, you will assist with the study visits, by helping with the clinical examination, the administration of questionnaires, and the collection of blood samples from participants at UZ Leuven. Other duties include managing the database and coordinating the other Belgian centers participating in the study.

In addition, as part of your PhD, you will also be able to perform secondary analyses on several completed studies of the research group of K Benhalima (such as on the follow-up-study of the BEDIP cohort evaluating the long-term metabolic risk in mothers and offspring across different degrees of hyperglycaemia in pregnancy and secondary analyses on the CORDELIA trial, a RCT evaluating CGM compared to capillary glucose monitoring in pregnant women with GDM).