As we all know, there are significant limitations to the use of creatinine as an estimate of GFR. Creatinine is freely filtered at the glomerulus, but in addition is also secreted to some degree, meaning that creatinine clearance can overestimate GFR. Furthermore, since creatinine is produced by muscle, individuals with high muscle mass may have a calculated GFR which is low despite actually preserved renal function. Another failing of creatinine-based GFR calculations is highlighted by the ability of creatinine to predict all-cause mortality: although high creatinines are associated with a higher risk of mortality, individuals with extremely low creatinines, perhaps reflective of a poor nutritional status or low muscle mass--also show poor mortality outcomes.
In a recent article in JASN by Astor et al, the investigators provide evidence that cystatin C does a better job of predicting mortality than creatinine. Cystatin C is a cysteine protease inhibitor which is freely filtered at the glomerulus, but not secreted like creatinine is. Furthermore, cystatin C production is independent of muscle mass, making it less susceptible to the limitations of creatinine. One of the main strengths of the study is the it uses data from NHANES, which is generated from a large and ethnically diverse U.S. population with a long follow-up time; it also contains a large number of individuals with only mildly reduced GFRs in the CKD3 range. The apparent superiority of cystatin C to predict mortality and provide more accurate assessments of GFR compared to creatinine suggest it could eventually play a major role in routine lab assessment in many of our patients.