(182q) A Modeling Framework to Characterize Kinetics, Efficacy and Toxicity of Hydroxyurea Based Treatment of Individual Sickle Cell Disease Patients

Hydroxyurea (HU) is the current FDA approved drug for sickle cell disease (SCD) treatment. Though several studies have indicated that the drug is effective in ameliorating disease pathophysiology, its use is limited due to myelosuppression and uncertainties associated with its long-term safety. Use of hydroxyurea presents major challenges to clinicians due to wide interpatient variability in pharmacokinetic-pharmacodynamic (PK-PD) profiles, cytotoxicity, lack of an effective biomarker to predict treatment efficacy and consequently lack of an optimal dosing regimen for individual patients.

A semi-mechanistic compartmental PK model using ODEs is developed that consists of three compartments: gastrointestinal tract, plasma and tissue where the drug gets metabolized to nitric oxide and its derivatives. Fetal hemoglobin (HbF) and mean cell volume (MCV) of red blood cells (RBCs) have been clinically used as biomarkers to indicate treatment efficacy. To capture dynamics of HbF, a signaling model using ODEs, based on the hypothesis that HU activates HbF through the NO-cGMP pathway, is formulated. To quantify toxicity and change in MCV of RBCs, erythropoiesis and leukopoiesis processes are modeled where stem cells undergo biochemical and morphological changes to produce fully functional erythrocytes and leukocytes in the blood circulation. The model is framed using a population balance modeling approach where all the cellular properties are lumped into two internal coordinates, cell maturation age and cell volume, then solved using the method of characteristics. The model predicts the volume distribution of RBCs and concentrations of leukocytes and erythrocytes of SCD patients with time under treatment with hydroxyurea. The PK model is integrated with the signaling model and the hematopoiesis model to characterize kinetics with both efficacy and toxicity. Furthermore, sensitive parameters are obtained using a global sensitivity analysis approach to obtain an individual patient's model. Model parameters are obtained after fitting the model predictions to an individual patient PK-PD variables i.e., drug concentration in the plasma, MCV, HbF, and blood cell count. Once the individual patient response is predicted, the optimal dose is obtained using model predictive control that keeps the model predictions within target values. Another major issue which plagues the entire treatment process is patients' non-compliance where they do not take the drug. It is difficult to differentiate non-compliance from treatment inefficacy because non-compliant patients might mislead clinicians into believing that treatment at the current dose level is insufficient. In this case, the dose might be increased but once the patient starts complying, the excess dose might prove to be harmful. Non-compliant patients are modeled and simulated to study the effect of missing the dose on patient PK-PD variables and based on that, an optimal dose for a non-compliant patient is calculated from the integrated model.

Hydroxyurea shows the potential for improvements in the health of patients with sickle cell disease, however, if it is not administered in the right amount, it leads to myelosuppression. Through our integrated PK-PD model, we seek to predict the individual patient trajectory and answer questions as to why some of the patients do not respond, why some of the patients respond well for the same amount of dose, and how to differentiate non-compliant patients from poor responders.