Prediction of the clinical course of immune thrombocytopenia in children
by platelet kinetics
Abstract
Introduction: Childhood immune thrombocytopenia (ITP) is a rare
autoimmune disorder characterized by isolated thrombocytopenia.
Prolonged ITP (persistent and chronic) leads to a reduced quality of
life for children in many domains. To provide optimal support for
children, with ITP, it is important to be able to predict those who will
develop prolonged ITP. This study aimed to develop a mathematical model
based on platelet recovery that allows the early prediction of prolonged
ITP. Methods: In this retrospective study, we used platelet
counts from the six months following the diagnosis of ITP to model the
kinetics of platelet evolution using a pharmacokinetic-pharmacodynamic
model. Results: In a learning set (n=103), platelet counts were
satisfactorily described by our kinetic model. The K
heal parameter, which describes spontaneous platelet
recovery, allowed a distinction between acute and prolonged ITP with an
AUC of 0.74. In a validation set (n=58), spontaneous platelet recovery
was robustly predicted using platelet counts from 15 (AUC=0.76) or 30
(AUC=0.82) days after ITP diagnosis. Discussion: In our model,
platelet recovery quantified using the k heal parameter
allowed prediction of the clinical course of ITP. Future prospective
studies are needed to improve the predictivity of this model, in
particular, by combining it with the predictive scores previously
reported in the literature.