Clinical biobanking and the distinguishing risk of progressive chronic kidney disease study

Abstract

Chronic kidney disease (CKD) is a major health and economic burden within Australia. Estimated glomerular filtration rate (eGFR) and albuminuria (Alb) are used to determine if a patient has CKD, but these routine clinical kidney function measurements are inadequate when predicting progressive CKD. This thesis addressed two themes: firstly, clinical CKD biobanking within Australia; and secondly, improving clinical and research outcomes for CKD patients by developing a prognostic clinical tool.Four aims were completed. i) To establish a clinical biobank, termed the CKD Biobank, within Queensland, Australia and develop this into a bioresource. 100 CKD patients, consisting of patients with diabetic nephropathy, vascular diseases, genetic kidney diseases, glomerulonephritis, and a range of other primary kidney diseases, were recruited. In approximately 80% of these patients, baseline bio-specimens were collected alongside clinical data.ii) To address the need for a better definition of progressive CKD and to identify a robust definition of progressive kidney function decline. A systematic review was conducted to identify definitions of progressive kidney function decline currently used in research. A range of definitions was identified and subsequently investigated within the CKD Queensland Registry. A ≥30% decline in eGFR from baseline was found to be robust in distinguishing between patients who experienced a progressive decline in kidney function and those who did not.iii) To use the CKD Queensland Registry and CKD Biobank to identify biomarkers of progressive CKD. Discovery-based and hypothesis-based approaches were used, within the CKD Queensland Registry and the CKD Biobank, respectively, to identify novel and emerging biomarkers of progressive CKD. Several biomarkers were identified that characterised the pathophysiological mechanisms of CKD, with varying capacity to distinguish between progressive and non-progressive CKD.iv) To develop a novel prognostic clinical tool, termed the “distinguishing risk of progressive CKD” (DROP CKD) tool, for differentiating between progressive and non-progressive CKD patients at baseline biomarker measurements. The DROP CKD tool was a predictive model calculated by linear discriminant analysis and constructed from biomarkers that differed between progressive and non-progressive CKD patients. A step-backwards approach was used for biomarker selection to maximise accuracy of the DROP CKD tool.Three hypotheses were investigated. i) A biobank, designed for a specific application that overcomes the ethics, governance, and quality control hurdles, will prove a valuable bioresource. The CKD Biobank has proved a useful bioresource, not only for this thesis, but for several other research projects. The CKD Biobank was utilised for investigating progressive CKD biomarkers and developing the DROP CKD tool for predicting progressive CKD. This Biobank was used, additionally, to supply bio-specimens from healthy controls for a discovery-based project to identify metabolites with prognostic capabilities in kidney cancer, and bio-specimens from diabetic nephropathy and glomerulonephritis CKD patients for research concerning presence of coagulation proteases in urine.ii) Progressive and non-progressive CKD will be characterised by biomarkers of kidney function, tissue injury, inflammation, oxidative stress, tissue repair, fibrosis, and comorbidities of CKD within the venous blood or urine of CKD patients. Compared to non-progressive patients, progressive patients of the CKD Queensland Registry and the CKD Biobank were characterised as having reduced kidney function, being anaemic, having an altered electrolyte-water balance, metabolic acidosis, dyslipidaemia, and as having mineral and bone disease. Progressive CKD patients of the CKD Biobank were further characterised by tissue injury, inflammation, and hypercoagulability while progressive CKD patients of the CKD Queensland Registry were only additionally characterised by eosinophilia.iii) A panel of the biomarkers, measured at baseline, will accurately predict progressive CKD. Biomarkers that were observed as differing between progressive and non-progressive CKD patients of the CKD Biobank were utilised to develop the DROP CKD tool. Through the step-backwards method, a biomarker panel consisting of eGFR, serum creatinine, cystatin-c, urea, tumour necrosis factor (TNF)-α, TNF Receptor-I, TNF Receptor-II, stem cell factor, tryptase, neutrophil gelatinase-associated lipocalin, tissue factor, bicarbonate, calculated osmolality, and haematocrit were observed as having an accuracy of 95.5% when predicting progressive CKD based on baseline biomarker expression. Moreover, the DROP CKD tool was more accurate than traditional kidney function measurements for determining progressive CKD.This thesis attempted to expand upon CKD biobanking capabilities and improve clinical and research outcomes for CKD patients by developing a novel prognostic clinical tool for predicting progression of CKD. Several avenues of research are still required to produce translatable results that will improve patient outcomes. Defining progressive CKD requires further investigation to identify a definition that robustly characterises a “progressive” decline in kidney function, accounting for the non-linear nature of eGFR decline and associating with meaningful clinical outcomes without underdiagnosing progression. Additional biomarker discovery is needed to identify novel biomarkers which can characterise progressive CKD. Finally, pre-existing and novel biomarkers need to be combined into a panel that can predict progressive CKD. This panel must be assayed in a minimally-invasive, efficient, manner, and communicate clinically-meaningful information to inform clinical management of CKD patients.