Prediction of the Renal Elimination of Drugs With Cystatin C vs Creatinine: A Systematic Review

Abstract

Serum cystatin C has been proposed as a kidney biomarker to inform drug dosing. We conducted a systematic review to synthesize available data for the association between serum cystatin C and drug pharmacokinetics, dosing, and clinical outcomes in adults (!18 years). PubMed, Ovid MEDLINE, Ovid EMBASE, EBSCO CINAHL, and Scopus were systematically searched from 1946 to September 2017 to identify candidate studies. Studies of cystatin C as a predictor for acute kidney injury or for management of contrast-associated acute kidney injury were excluded. Also, studies were excluded if drug concentrations were unavailable and if a reference standard for drug dosing (eg, serum creatinine) was not concurrently reported. The outcomes of interest included drug clearance (L/h), concentrations (mg/L), target level achievement (%), therapeutic failure (%), and drug toxicity (%). We included 28 articles that evaluated 16 different medications in 3455 participants. Vancomycin was the most well-studied drug. Overall, cystatin Cebased estimated glomerular filtration rate (eGFR Cystatin C ) was more predictive of drug levels and drug clearance than eGFR Creatinine . In only one study were target attainment and outcomes compared between 2 drug-dosing regimens, one based on eGFR Creatinine-Cystatin C and one dosed with the Cockcroft-Gault creatinine clearance equation. Compared with eGFR Creatinine , use of eGFR Cystatin C to predict elimination of medications via the kidney was as accurate, if not superior, in most studies, but infrequently were data on target attainment or clinical outcomes reported. Drug-specific dosing protocols that use cystatin C to estimate kidney function should be tested for clinical application.

Cystatin C: A Primer for Pharmacists

“…Other drugs evaluated included digoxin, dabigatran, perindopril, carboplatin, and topotecan. Overall, eGFR cysC predicted observed drug clearance and serum concentrations of medications as well as, if not better than, SCr in the majority of the identified studies [25]. The literature was most robust for vancomycin, which was evaluated in 12 studies that included 1806 patients [26][27][28][29][30][31][32][33][34][35][36].…”

“…The association between cysC and the corresponding eGFR and drug pharmacokinetics has been described in a number of studies [25]. A systematic review that explored the predictive relationship between cysC and drug clearance identified 28 studies published between 2004 and 2017 that included 3455 patients treated with 16 different renally-eliminated medications [25]. Antimicrobials were the most commonly studied and included vancomycin, aminoglycosides, beta-lactams, tenofovir, and teicoplanin.…”

“…The literature was most robust for vancomycin, which was evaluated in 12 studies that included 1806 patients [26][27][28][29][30][31][32][33][34][35][36]. In six of these studies that reported vancomycin clearance (L/h), eGFR cysC better predicted drug clearance than eGFR creatinine (R 2 , eGFR cysC = 0.70-0.85; eGFR creatinine = 0.37) [25]. In the remaining six studies, trough levels were the outcome of interest and were better predicted by eGFR cysC than eGFR creatinine [pooled R 2 (N = 191): eGFR cysC = 0.54 (95% CI, 0.43-0.64) vs. eGFR creatinine = 0.26 (95% CI 0.12-0.39)] [25].…”

“…In six of these studies that reported vancomycin clearance (L/h), eGFR cysC better predicted drug clearance than eGFR creatinine (R 2 , eGFR cysC = 0.70-0.85; eGFR creatinine = 0.37) [25]. In the remaining six studies, trough levels were the outcome of interest and were better predicted by eGFR cysC than eGFR creatinine [pooled R 2 (N = 191): eGFR cysC = 0.54 (95% CI, 0.43-0.64) vs. eGFR creatinine = 0.26 (95% CI 0.12-0.39)] [25]. Across all 28 included studies, equations for SCr-and cysC-based estimates of kidney function varied [25].…”

“…In the remaining six studies, trough levels were the outcome of interest and were better predicted by eGFR cysC than eGFR creatinine [pooled R 2 (N = 191): eGFR cysC = 0.54 (95% CI, 0.43-0.64) vs. eGFR creatinine = 0.26 (95% CI 0.12-0.39)] [25]. Across all 28 included studies, equations for SCr-and cysC-based estimates of kidney function varied [25]. Since 2017, additional literature has been published that corroborated these findings [37][38][39].…”