An ePKPD model for sodium nitroprusside in adolescents

Background: Sodium nitroprusside (SNP) is commonly used to control blood pressure (BP) in children during surgery. Recently, the results from a clinical trial in children using SNP during surgery were used to develop a KPD model to help guide the dosing of SNP in this population. [1] The model revealed that there were two subpopulations of patients who have different sensitivities to SNP, but no correlations between the subpopulations and covariates were found. An extended PD (ePD) model of the haemodynamic system has been developed previously and contains appropriate control and feedback mechanisms. [2] This model was recreated and combined with a PK model for SNP and different pharmacological targets within the haemodynamic system were created and tested for effects comparable to the data.

Aim: The aim of this work is to develop an ePKPD (i.e. “semi-mechanistic”) model that can be used to describe the haemodynamic profile in adolescents upon infusion of SNP and test different suggested mechanisms for the effect of SNP.

Methods: The data arose from 88 adolescents and young adults, who were 13 years and over. The dosing of SNP was titrated in each subject to a predefined target mean arterial pressure ‘MAP’.  If the MAP was too low then the subject was discontinued from further dosing. A model describing the haemodynamic responses (MAP, systolic blood pressure ‘SBP’, diastolic blood pressure ‘DBP’, and heart rate ‘HR’) was recreated in MATLAB. Different targets of drug action were identified in the model that could be used to modulate the effect of SNP.  The two main targets were; (a) decreasing the peripheral resistance (PR), and (b) baroreceptor resetting. A simplified PK model for SNP was then added to the model. Three different scenarios for SNP mechanism of action were considered in the various model expressions; (a) affecting PR only, (b) affecting baroreceptors only and, (c) affecting both PR and baroreceptors (i.e. combined mechanism). Model PD parameters and baseline values for the haemodynamic variables were calibrated to the same initial values as the observed data. The model predictions (at a single SNP dose) were then overlaid on the data. No dose titration was performed based on model predictions.

Results: The overlaid profiles for model predictions and observed data show that the model can describe the central tendency of the data and that it can be used to evaluate the different mechanisms that were hypothesised about SNP mechanism of effect.  For the PR only mechanism, the model underpredicts MAP, SBP, & DBP at the early phase before reaching the steady state (SS) and overpredicts the SS level of HR. The baroreceptor resetting mechanism underpredicts the whole profile of HR. The combined mechanism was most flexible and aligned with the observed haemodynamic variables.

Conclusion: The results suggest the mechanistic plausibility of the proposed model, as reflected by the consistency of the model predictions with the observed data. Since the current data contains multiple dose titration steps in each individual including discontinuation of treatment, the next step requires the data to be modelled.


  1. Barrett, J.S., et al., A hemodynamic model to guide blood pressure control during deliberate hypotension with sodium nitroprusside in children. Frontiers in Pharmacology, 2015. 6.
  2. Chae, D., et al., Mechanistic Model for Blood Pressure and Heart Rate Changes Produced by Telmisartan in Human Beings. Basic & Clinical Pharmacology & Toxicology, 2018. 122(1): p. 139-148.