Combining Cystatin C and Creatinine Increases eGFR Slope Accuracy in ADPKD

In autosomal dominant polycystic kidney disease (ADPKD) trials, creatinine-based estimated glomerular filtration rate (GFR, eGFR_cr) is a key end point; however, variability in creatinine measurements may affect reliability. We compared routine eGFR_cr to eGFR derived from batch-measured creatinine, cystatin C (eGFR_cys), or both (eGFR_cr-cys) using biobanked samples, against iothalamate-measured GFR (mGFR).

In the 260 patients with ADPKD from the DIPAK observational cohort, we compared 2021 Chronic Kidney Disease (CKD)-Epidemiology Collaboration (CKD-EPI)-based eGFR based on the following: (i) routine clinical creatinine, (ii) batch-measured creatinine, and (iii) batch-measured cystatin C to mGFR. Creatinine and cystatin batch measurements were performed on the same day to minimize variability. Cross-sectional performance was evaluated by bias (eGFR - mGFR), precision, accuracy (P30/P10), and CKD stage agreement. Longitudinal agreement between mGFR and eGFR slopes (ml/min per 1.73 m²/yr) was assessed using concordance correlation coefficients (CCC). Sample size requirements for hypothetical trials were calculated based on slope variability.

Cross-sectional performance of batch-measured eGFR_cr was largely similar to routine eGFR_cr. eGFR_cys and eGFR_cr-cys showed greater bias than eGFR_cr; eGFR_cr-cys showed the highest precision. Longitudinally, batch-measured eGFR_cr-cys slopes correlated more strongly with mGFR decline than eGFR_cr (CCC: 0.68; 95% confidence interval [CI]: 0.59-0.76] vs. 0.58 [95% CI: 0.47-0.67]; P = 0.02). eGFR_cr-cys slopes exhibited lower SDs than eGFR_cr and eGFR_cys. This reduction in slope variability translated to sample size reductions of 37% to 38% for clinical trials.

The use of eGFR_cr-cys enhances precision and reduces eGFR slope variability. Using combined creatinine-cystatin C estimates may optimize trial efficiency and design in ADPKD.