ProFiLE:Standardized multi-parameter flow cytometry for early diagnosis, risk assessment and optimized treatment in primary immunodeficiency (ProFiLE)

Primary Immune Deficiency diseases (PIDs) comprise a heterogeneous group of life-threatening genetic disorders of the innate and adaptive immune system. To date, more than 400 different PID subtypes with more than 300 PID causing genes have been unraveled. However, there is still a lot of underdiagnosis due to the clinical heterogeneity in disease presentation even among patients with the same genetic pathogenic variant. Further complicating the diagnosis is the overall low and variable prevalence rate of PIDs between 1 in 2 000 to 1 in 100 000 in Caucasians, depending of the PID subtype. The recent evolutions in genetic testing have led to an exponential increase in the identification of PID-causing gene defects. Nevertheless, an underlying disease-causing gene is detected in only 1020% of the PID patients. As a result, a large number of patients remain undiagnosed or remain undefined PID. Beside genetic testing, immunophenotyping using multi-parameter flow cytometry is an important tool in the diagnostic and prognostic work-up of PID. Despites its crucial role, flow cytometric analysis in PID diagnostics is still faced with several challenges (e.g. standardization and reference values are lacking, lack of clinical validation on large datasets, difficult data analysis). The aim of this project is to clinically validate and optimize the use of standardized multiparameter immunophenotyping in the context PID as a tool for early diagnosis, risk assessment and personalized treatment. To this end we will apply the standardized Euroflow strategy in a consecutive patient series with suspicion of PID/known PID diagnosis in two university hospitals in order to make it routinely applicable across different labs in Belgium, and subsequently across Europe.  In addition, we will evaluate the clinical added value of using new and automated software tools based on (un)supervised machine learning techniques.