Abstract

Infliximab is an antibody that is approved for treatment of the inflammatory bowel diseases Crohn’s disease and ulcerative colitis. It is important to predict the course of the depletion of Infliximab in the body to time the regular infusions that patients get. The base model is a two-compartment-model and three parameter identification approaches are compared: identification of each infusion period separately, identification of each patient separately and identification of all meas-urement points at once. The best results provides the approach to identify the parameters for each infusion period separately. Wanting to improve the quality of the results, an extended model with continuous parameter antibodies to Infliximab (ATI) is considered, but the mean error is higher than for the base model. We conclude that the presence of antibodies to Infliximab in the body carries more weight than the actual number of antibodies. Finally, a model with an additional parameter, number of previous infusions with Infliximab (PRIORIFX), is created. This model could not improve the results of the base model. This indicates that the number of previous infusions with Infliximab is not significant for the model.