AIMS

Serum creatinine and albuminuria are the core of most CKD prediction and progression risk models. Several biomarkers have been introduced to improve these results such as beta-2-microglobulin (B2M) and cystatin C (CysC). Nevertheless, few clinical comparisons of these biomarkers are available. We have compared serum B2M levels with albuminuria, CysC levels, and the CKD-EPI GFR equations.

DESIGNS AND METHODS

A sample of 434 patients were studied: 234 males and 200 females, the mean age was 58.3 ± 15.0 years, and 33.4% have diabetes mellitus. In all patients, plasma B2M, CysC, creatinine, and urinary albumin excretion were analyzed. EGFR was calculated using CKD-EPI equations for creatinine, CysC, and creatinine-CysC. The risk of death and CKD progression was evaluated using ROC curves and Cox proportional hazards survivorship models.

RESULTS

For mortality, the highest area under the curve (AUC) was for CysC (0.775, 0.676-0.875). The lowest sensitivity was shown by eGFR (creatinine) (0.298, 0.195-0.401, p < 0.001), eGFR (CysC) (0.216, 0.118-0.314, p < 0.001), and eGFR (creatinine + CysS) (0.218, 0.124-0.312, p < 0.001). For progression to advanced CKD, the highest AUC was for CysC (0.908, 0.862-0.954). The lowest sensitivity was shown by eGFR (creatinine) (0.184, 0.106-0.261, p < 0.001), eGFR (CysC) (0.095, 0.048-0.14, p < 0.001), and eGFR (creatinine+ CysC) (0.087, 0.040-0.134, p < 0.001). CysC, after age, was the second-best marker of life risk. Contrariwise, for CKD progression, CysC, and albuminuria were the best markers.

CONCLUSIONS

The best biomarker of mortality and risk of progression to CKD was CysC. Albuminuria and B2M were the next best options to be used. The lowest sensitivity was shown by estimated eGFR.