Markus G. Seidel (Austria)
Medical University Graz
Pediatric Hematology-Oncology
Markus Seidel is a Professor of Translational Pediatric Hematology and Immunology at the Medical University of Graz, Austria. He studied medicine and specialized in pediatric hematology-oncology with a main interest in stem cell transplantation at the Medical University / St. Anna Children’s Hospital in Vienna, Austria. He worked as post-doc in Michael Freissmuth’s lab at the Institute of Pharmacology in Vienna in the field of MAPK signaling and in Tom Look’s lab in apoptosis signaling research at the Dana-Farber Cancer Institute (Peds Hem/Onc), Harvard Medical School, in Boston. His current clinical focus and research interests are inborn errors of immunity with immune dysregulation and cancer predisposition, and he is leading a multicenter, prospective registry and biomarker study for severe immune cytopenias and heading the outpatient clinic for pediatric hematology-oncology with stem cell transplantation in Graz.

Presenter of 1 Presentation

 Conference Calendar

UNSUPERVISED PHENOTYPE EXPRESSION PROFILING AND LONGITUDINAL MONITORING IN INBORN ERRORS OF IMMUNITY WITH IMMUNE DYSREGULATION BY MEANS OF THE IDDA2.1 ‘KALEIDOSCOPE’ SCORE

Session Type
Parallel Sessions
Session Name
0940 - Parallel Session 12: Artificial Intelligence in IEI (ID 75)
Date
Fri, 14.10.2022
Session Time
14:00 - 15:30
Room
Session Hall 02
Lecture Time
15:02 - 15:12
Presenter

Abstract

Background and Aims

Clinical scores (measures, indices, scales, or similar) may be used in inborn errors of immunity (IEI) to support making a diagnosis or to classify an IEI, to assess and monitor the disease severity over time, and to guide treatment decisions.

Methods

We developed a user-friendly 22-parameter score for evaluating the immune deficiency and dysregulation activity (IDDA) that includes graded organ involvement and disease burden, intended for prospective monitoring of all IEI with immune dysregulation.

Results

To extend the utility from LRBA deficiency (IDDA version 1), we included hemophagocytic lymphohistiocytosis into the parameter list; and we modified the calculation of the numerical score to correct for very low performance scales. A new accompanying feature, the kaleidoscope function, is enabled by plotting the IDDA2.1 parameters as radar chart or heatmap which illustrates phenotypical similarities and variances between different patients or conditions. The discriminative power of this method was confirmed by unsupervised hierarchical clustering of phenotype manifestation frequencies of 18 representative IEI (including predominantly antibody deficiencies and susceptibility to EBV and lymphoproliferation) in analogy to genotype expression arrays.

Conclusions

The IDDA2.1 kaleidoscope score may be used for prospective monitoring of patients with IEI with immune dysregulation, e.g., in patient registries or clinical trials. A recently launched ESID registry study will collect data and apply unsupervised machine learning algorithms to detect similarities of patterns in training cohorts consisting of patients with known monogenic IEI to assess potentially predictive values in diagnosis finding, complication monitoring, and to suggest phenotype-driven, “semi-targeted” therapy options for undiagnosed patients.

Hide