Abstract
Acute Kidney injury (AKI) is common and frequently fatal. Delay in plasma creatinine based diagnosis of AKI has compromised clinical trials of experimentally promising therapies of kidney injury. Therefore, a number of potential early urinary and plasma biomarkers of renal cell injury have been evaluated in several clinical and experimental studies. However, many uncertainties remain in the diagnostic and predictive capability of these biomarkers in various forms of AKI, especially in heterogeneous population.
The overall aim in this project was to analyse the diagnostic and predictive performance of some novel AKI biomarkers in a heterogeneous high-risk population. This project focused on the performance of plasma and urinary cystatin C (CysC) as markers of kidney function and injury respectively and both as predictors of mortality. This involved separate analysis of plasma and urinary CysC clinical data, and an experimental study investigating the performance of urinary CysC in the presence of albuminuria. Finally, the ability of fractional urinary biomarkers (fractional excretion of sodium (FENa) and urea (FEurea) to distinguish between so-called pre-renal AKI and established AKI was assessed.
Clinical data came from the EARLYARF study, which was initiated by Professor Zoltan Endre prior to commencement of my PhD project. I participated in data collection and data entry for the large part of this study and I was provided access to the database to address the objectives of my thesis. The hypotheses were explored by analysing data arising from this study. The experimental study of the effect of proteinuria and albuminuria on urinary CysC arose from the observation of cystatinuria in children with proteinuria. This was explored in an animal model of transient albuminuria, which I developed.
Plasma CysC has been proposed as an alternative to plasma creatinine as a measure of renal function. In Chapter 3, relative changes of plasma CysC and plasma creatinine were compared in critically ill patients. I was able to demonstrate that plasma CysC generally increased prior to plasma creatinine. Plasma CysC and creatinine were similarly moderately predictive of death or the need for dialysis. Plasma CysC was a more effective and earlier surrogate marker of decreased renal function than plasma creatinine in a general intensive care unit population.
In Chapter 4, the utility of urinary CysC as a diagnostic marker of AKI, and predictor of mortality in critically ill patients was evaluated. Urinary CysC was diagnostic of and predictive of death. Unexpectedly, it was also found to be diagnostic of sepsis. Concentrations of urinary CysC were significantly higher in the presence of sepsis (p<0.0001) or AKI (p<0.0001). There was no interaction between sepsis and AKI on the urinary CysC concentrations (p=0.53). Urinary CysC was independently associated with AKI, sepsis, and death within 30 days.
Low molecular weight (LMW) proteins, including albumin and novel urinary biomarkers of AKI such as CysC and neutrophil gelatinase-associated lipocalin (NGAL) are normally absorbed from the glomerular filtrate by receptor-mediated transport. The effect of albuminuria on urinary excretion of novel biomarkers was evaluated in Chapter 5. Bovine serum albumin (BSA) induced transient increases in albuminuria, proteinuria and cystatinuria. Clinical proteinuria in the EARLYARF patient cohort was associated with increasing CysC and NGAL concentrations. Proteinuria may increase the threshold of novel biomarkers for detection of AKI by increasing the excretion of LMW protein biomarkers.
The structural correlates of prerenal AKI are unknown. FENa and FEurea are often used to differentiate reversible prerenal from established AKI. However, FENa is misleadingly low in patients with AKI due to sepsis. In chapter 6, it was hypothesised that pre-renal AKI is reversible only because injury is less severe rather than because it is a unique physiological response to renal underperfusion. Patients were stratified according to duration of AKI: (i) recovery within 24 hours (AKI 24), (ii) recovery between 24 and 48 hours (AKI 24-48), (iii) no recovery within 48 hours or renal replacement therapy (AKI 48). Patients with no AKI within 48 hours served as controls. Pre-renal AKI-FENa and pre-renal AKI-FEurea were defined as recovery within 48 hours with on admission value of FENa <1 and FEurea <35 respectively. All urinary biomarkers increased with increasing duration and severity of AKI. CysC, NGAL and interleukin -18 (IL-18) concentrations were greater in both pre-renal AKI-FENa and pre-renal AKI-FEurea compared with controls. The result suggests that prerenal AKI represents the milder end of a continuum of renal cellular injury.