Acute kidney injury (AKI) is common after cardiac surgery, and associated with adverse patient outcomes. Urinary cystatin C (CysC) is a biomarker of proximal tubule function and may increase earlier in AKI than serum creatinine. In a recent study published by the American Journal of Kidney Diseases, Koyner et al discuss the role of urinary CysC as a marker of AKI. This trial was part of the larger TRIBE AKI Consortium in search for a better biomarker for AKI. Dr. Chirag Parikh (CP), corresponding author from Yale University Medical Center, discusses this very important trial with Dr. Tejas Desai (eAJKD), eAJKD Advisory Board member.
eAJKD: Did you have a hypothesis entering this study, or were you undecided about how urine CysC would perform as a predictor of AKI?
CP: Predictors of AKI are important to identify, as serum creatinine is not the optimal test for AKI. As a result, we started the Translational Research Investigating Biomarker Endpoints (TRIBE) Project, which is a multicenter study to identify biomarkers or proteins that could assist with clinical prediction of AKI in the perioperative period. We chose cardiac surgery for the setting because cardiac surgery has a high incidence of AKI and it is a common surgery with about 2 million procedures worldwide. Preliminary data about urine CysC showed it could be an exciting biomarker along with serum creatinine in terms of predicting kidney injury. There have been about 6 studies before TRIBE that looked at urine CysC, with variable results in settings like sepsis. In the perioperative setting, the results were slightly controversial: some studies have shown some benefit, whereas others have not. When we decided to look at urine CysC, we wanted to conduct the study in a multicenter format so that we could validate if CysC can be used for clinical prediction of AKI after cardiac surgery.
eAJKD: Why did you chose urinary CysC as the biomarker, and not urinary NGAL or IL-18?
CP: Earlier TRIBE studies around 2009-2010 looked at urinary NGAL and IL-18, which are both structural biomarkers of kidney injury. We wanted to look at urinary CysC as an additional biomarker because it’s a marker of proximal tubular function and not a structural marker of injury. Serum CysC is filtered by the glomerulus, and then completely reabsorbed and degraded by proximal tubule; the levels in the urine are almost undetectable in healthy participants. When CysC does appear in the urine, it denotes problem with proximal tubular function.
eAJKD: Why not use a combination of these markers?
CP: When we designed the study, we knew that urine IL-18 and urinary NGAL performed well in relaying information on tubular necrosis. By combining urinary CysC, the information is much more robust in predicting AKI.
eAJKD: Now that a large, multicenter, international trial has questioned the utility of urinary CysC as a biomarker for AKI, where do you feel the TRIBE AKI Consortium goes from here?
CP: The first phase of the TRIBE study was from 2005-2010 and we evaluated AKI as the main outcome. The second phase, over the next 5 years, is to measure additional serum and urine biomarkers. More importantly, we are going to look at more distant outcomes: we will be looking at mortality as a key outcome in the second phase. The questions we are now asking is whether these biomarkers give additional information on hard outcomes like mortality and cardiovascular death beyond what we understand about AKI.
eAJKD: What changes would you make to any future prospective study that looks at urinary CysC as an early biomarker for AKI?
CP: The results of urinary CysC have been disappointing as a clinical marker for AKI prediction. In the perioperative setting, urinary CysC did not pan out as an independent marker after we adjusted for clinical variables like age, cardio-pulmonary bypass time, or baseline co-morbidities, which are described in the article. At this point, urinary CysC should not be used to make decisions on dialysis or the progression of AKI. However, mechanistically it identifies patients with proximal tubular dysfunction after the surgery. Perhaps CysC should be combined with other markers of proximal tubular function to assess proximal tubular function.
Our article is an example of how large, definitive studies are necessary before we embrace new findings because small-center studies sometimes do not tell the whole story. There is also publication bias with some biomarkers. Evaluation of new molecules in a large, multicenter setting is essential before we advance them for clinical application. At the same time, if a marker does not work for one outcome, understanding what it means mechanistically should be carefully evaluated. These are the key lessons we have learned in our cohort.