Creatinine- and cystatin C-based estimated glomerular filtration rate slopes for the prediction of kidney outcome: a comparative retrospective study

Abstract

Background

Many studies have evaluated the usefulness of creatinine- (eGFRcr) and cystatin C-based estimated glomerular filtration rate (eGFRcys) at specific time points in predicting renal outcome. This study compared the performance of both eGFR changing slopes in identifying patients at high risk of end-stage renal disease (ESRD).

Methods

From 2012 to 2017, patients with more than three simultaneous measurements of serum creatinine and cystatin C for 1 year were identified. Rapid progression was defined as eGFR slope < − 5 mL/min/1.73 m²/year. The primary outcome was progression to ESRD.

Results

Overall, 1323 patients were included. The baseline eGFRcr and eGFRcys were 39 (27–48) and 38 (27–50) mL/min/1.73 m², respectively. Over 2.9 years (range, 2.0–3.8 years) of follow-up, 134 subjects (10%) progressed to ESRD. Both the eGFRcr and eGFRcys slopes were associated with a higher risk of ESRD, independently of baseline eGFR (hazard ratio [HR] = 0.986 [0.982–0.991] and HR = 0.988 [0.983–0.993], respectively; all p <  0.001). The creatinine- and cystatin C-based rapid progressions were associated with increased risk of ESRD (HR = 2.22 [1.57–3.13], HR = 2.03 [1.44–2.86], respectively; all p <  0.001). In the subgroup analyses, the rapid progression group, defined on the basis of creatinine levels (n = 503), showed no association between the eGFRcys slope and ESRD risk (p = 0.31), whereas the eGFRcr slope contributed to further discriminating higher ESRD risk in the subjects with rapid progression based on eGFRcys slopes (n = 463; p = 0.003).

Conclusions

Both eGFR slopes were associated with future ESRD risk. The eGFRcr slope was comparable with the eGFRcys slope in predicting kidney outcome.