MUSC Nephrology Proteomics Laboratory

Prognostic Markers in Postoperative Acute Kidney Injury

Acute kidney injury (AKI) after cardiac surgery occurs in about 7% of patients and is associated with a 20 fold increase in mortality. Among patients who require dialysis, the risk of death is increased 70-fold. The magnitude of renal injury is variable and difficult to predict. Risk factors for kidney injury have been described but they do not predict if a given patient will develop renal failure. No good biomarkers currently exist to predict the prognosis of patients with kidney injury. Effective biomarkers could suggest when early therapy may be beneficial. The absence of predictive biomarkers has also been an impediment to studies of new therapies for AKI. No therapies have been identified which are clearly beneficial in AKI. Prognostic markers could identify the subset of patients in whom testing of new therapies could be performed. A significant effort is being made to identify biomarkers that can predict the cause of kidney injury and biomarkers that can predict kidney injury in the early stages. In spite of their importance, both clinically and for the development and testing of new therapies, relatively little has been done to identify markers that can predict the magnitude and course of the disease in AKI.

It is becoming clear that biomarkers for complex diseases will consist of multiple analytes. Abundance of candidate markers change over time, differ with the magnitude of injury and are affected by concurrent disease processes. Therefore, measurements of single analytes are not interpretable. When the abundance is viewed in the context of other analytes that can be used to interpret the time or other variables, a better interpretation can be made. The complex relationship of the analytes as predictive markers makes the statistical and informatic analysis as important as the design of the study and the techniques of measurement and identification.

This study (CRISP Database Entry for R01DK080234) will identify prognostic markers in AKI after cardiac surgery. We will measure the abundance of candidate markers and use “unbiased” discovery techniques to discover novel markers. Urine will be collected from patients who show signs of AKI after cardiac surgery at five centers: The Medical University of South Carolina; The Ralph H Johnson VA Medical Center; George Washington University; Duke University Medical Center and a University of Tennessee-affiliated hospital in Chattanooga, Tennessee. The primary outcome variable is the requirement for renal replacement therapy. We will also predict secondary outcomes that are either clinically useful or would be useful research outcomes.

In the first aim we will measure candidate markers. The markers were chosen based on published literature and our own preliminary data. They include known candidate markers of tubular injury, inflammatory response, tubular function, recovery of function and progression to dialysis. In the second aim we will use two proteomic techniques, 2D electrophoresis with DIGE and MALDI polypeptide analysis to identify novel markers. In the third aim we will select the candidate markers to be combined in a final assay using a second set of patients that is independent of the set used in the first two aims. Finally, we will validate the markers and the algorithm used to identify them in a third set consisting of 590 new patients.


© 2012  Medical University of South Carolina | Disclaimer