Uraemic Toxins in Chronic Kidney Disease Paediatric Patients: Kinetic Analysis

Overview

Children with chronic kidney disease (CKD) suffer from one of the most devastating diseases in childhood resulting in a lifelong need for health care, and a 3 times decreased life expectancy. In addition, they have important comorbidities that negatively impact on their quality of life and integration in society, jeopardizing their future even after a potential transplantation. Retention of uraemic toxins is accepted to play a major role in the pathogenesis of the comorbid conditions, but studies in children are lacking. Furthermore, there are currently no good tools to evaluate severity and monitor adequacy of treatment, resulting in suboptimal management. The overall scientific objective of this four years UToPaed IWT-TBM project is to provide the clinician with new diagnostic and therapeutic tools for the management of children with CKD, based on the improved understanding of uraemic toxicity. In the first part of UToPaed, the investigators will associate concentrations of a wide variety of uraemic toxins with different comorbidities in CKD children. In this second part, a kinetic analysis will be performed to unravel the distribution and transport of the different studied uraemic toxins in the body of the patient. The toxins of which concentrations are best correlated with comorbidities during the progress of CKD (UToPaed - part 1: observational study) and have representative kinetics will be selected as markers. These markers will be, together with the comorbidities, further tracked after interventions, i.e. starting on dialysis, transplantation, changes in dialysis strategy (UToPaed - part 3 - intervention study) in order to validate the different kinetic models. From the validated kinetic models (UToPaed - part 2 and 3), an open access user-friendly prediction simulator (PAEDSIM) based on patient characteristics and marker concentrations will be developed to optimise and individualise the dialysis therapy. By providing clinicians with more advanced and appropriate tools to improve management of all children with CKD, i.e. better assessment of the degree of renal dysfunction, better determination of the ideal time to start renal replacement therapy, and more accurate monitoring of dialysis adequacy, the investigators aim to improve neurocognitive and psychosocial functioning (short term), growth, maturation into puberty, and social integration (median term) and survival (long term).

This is an observational multicenter study in 20 children (≤ 18 years) with chronic kidney disease (CKD) stage 5D treated with haemodialysis. Distribution and transport of uraemic toxins inside the body is derived from a cross-sectional study. During a midweek haemodialysis session, blood is sampled from the dialyser inlet line at different time points (e.g. at 0, 15, 30, 60, 120, 240min) during the session to obtain the evolution of intradialytic concentrations as is needed for the kinetic analysis. To calculate dialyser clearance, blood is sampled simultaneously at the dialyser inlet and outlet in the first part of the session (e.g. at 30min). Total solute removal is measured by partial dialysate collection during the entire dialysis duration, using a validated sampling system. For each uraemic toxin under study, a kinetic model is calibrated simulating distribution and mass transport inside the patient's body. The body is characterised by a total distribution volume V (per toxin), consisting of one or more distinct compartments. In e.g. a 2-compartment model, one can distinguish a plasmatic or peripheral compartment, which is directly cleared by haemodialysis (i.e. dialyser clearance) or by renal or extrarenal clearances, and extraplasmatic compartment(s). Each compartment is assumed to be characterised by a homogeneous uraemic toxin concentration with variable inputs and outputs. The solute transport between two compartments is considered to be driven by concentration gradients (diffusion), and/or pressure gradients (convection) and is characterised by an intercompartment clearance. Presuming that removal and generation are in equilibrium in stable HD patients, solute generation rate in the interdialytic period is assumed equal to the total solute removal during the dialysis session. The patient-specific ultrafiltration rate is taken into account to change total distribution volume over time. The time variation of the compartment concentration is, for a particular toxin, determined by solving the mass balance equation for each compartment. The kinetic model iteratively solves these equations for the complete dialysis session time. Herewith, plasmatic volume, total distribution volume, as well as intercompartment clearance are calculated from fitting the solution to the measured patient's plasma concentrations. Such kinetic analyses result in the knowledge of all kinetic parameters for each studied toxin. These calibrated kinetic models for paediatrics are further used to simulate different dialysis strategies and, with it, look for the most optimal one, with the experience the investigators have from kinetic studies in adults. Herewith, per solute, inter- and intradialytic evolutions in concentration are calculated according to the mass balance equations per compartment using the derived kinetic parameters. In the interdialytic period, dialyser clearance is kept at zero, while solute generation is maintained constant, and the interdialytic volume gain is set equal to the intradialytic applied ultrafiltration rate. Starting from the intra- and interdialytic concentrations in steady state with a 3x4 hours dialysis schedule, intra- and interdialytic concentrations are calculated after mathematically altering several key characteristics of the dialysis regime. This is done for the individual paediatric patient data as well as for the average paediatric patient within its age category with data emanating out of our primary kinetic analysis. Possible strategies are: longer and/or more frequent dialysis with or without adapting blood and/or dialysate flow rates, increasing convection, haemodiafiltration (pre, post, or mixed dilution), or a combination accounting for different parameters. For each strategy, consecutive sessions are simulated until a new steady state of predialysis solute concentrations is reached (deviation between 2 consecutive sessions \<1%), paralleling the real in vivo effect of altering dialysis strategy. The different strategies are evaluated mathematically by comparing the calculated total solute removal during the first week with the new strategy, and the steady state time averaged as well as predialysis concentrations in the plasmatic volume. The calibrated kinetic modelling parameters and the results of the different dialysis strategies are compared for the different studied uraemic toxins. Accounting for the correlations with comorbidities (UToPaed-part 1), one (or more) uraemic toxin marker(s) are chosen. The kinetic models are further validated by quantifying uraemic toxin marker concentrations and comorbidities in individual patients after switching to different strategies as well as to the individualised optimal dialysis strategy based on the model (UToPaed- part 3).