Study of the natural history of diabetic kidney disease (DKD)

Introduction: Diabetes mellitus is a global public health problem affecting an increasing number of chronic kidney disease patients. The progression of diabetic kidney disease (DKD) is variable, and the natural course of the disease may be evolving. We aimed here to revisit the natural history of DKD and to evaluate the validity of a number of urinary biomarkers in predicting the progression of DKD by determination of their independent predictive values compared to more conventional risk factors. Methods: To address this, the effects of potential risk factors on the progression of DKD was studied retrospectively in a cohort of diabetic CKD patients followed at Sheffield Kidney Institute (n = 385). Another observational prospective study was performed on 102 T2DM with DKD, 21 diabetic patients without kidney disease and 21 healthy volunteers. Each provided fresh midstream urine sample with demographic and clinical parameters collected from the unit's electronic database. Commercial ELISA kits were optimised and used to measure different urinary proteins. After cytospin, urine cells were stained using anti- fibroblast specific protein-l (FSP-l) antibody and standard cytochemistry techniques to identify potential myofibroblasts. Finally, the urine matrix metalloproteinase activity was measured by using fluoresce in conjugate substrates. Results: In the retrospective DKD cohort study, baseline proteinuria, HbA 1 c and MAP as well as the presence of vascular co-morbidities were associated with a faster rate of annual eGFR decline. Whereas, in the prospective study, albuminuria was the weakest predictor of functional outcome (ROC = 64%) and transferrinuria was the strongest (ROC = 81.2%). Furthermore, the combined score of all measured proteins did not increase the predictive value considerably (86.3%) above that of urinary transferrin. Urine cytochemistry showed that FSP-I positive cells were detected in the urine of 61 % of DKD progressors group compared to 27% of non-progressors (P = 0.004). In contrast, all slides from DM patients and healthy volunteers were negatively stained for FSP-l. The predictive value of urinary FSP I positive cells (83.6%) was superior to albuminuria as well as podocyturia (66.2%). Finally. urine MMP activity was elevated in DKD. However, there was a significantly lower MMP activity in the progressors DKD patients. Conclusions: This thesis identified a number of key observations relating to the progression of DKD. Firstly, it showed that beside conventional clinical predictors of progression, vascular comorbidities were associated with a faster rate of progression of DKD in a largely older population suffering from T2DM. Urinary biomarkers analysis showed that peptiduria and in particular transferrinuria was highly predictive of DKD progression. For the first time, we show in this thesis the potential use of urine FSP-I positive cells (presumably kidney derived myofibroblasts) to predict DKD outcomes. Also, the lower urinary MMP activity predicted a faster rate of progression. In summary, this thesis successfully identifies a number of biomarkers that may prove highly valuable to predict DKD progression in T2DM affected by diffuse vascular pathology and co-morbidities.EThOS - Electronic Theses Online ServiceGBUnited Kingdo