Questionable Validity of Creatinine-Based eGFR in Elderly Patients but Cystatin C Is Helpful in First-Line Diagnostics.

Abstract

The recommended chronic kidney disease (CKD) first-line diagnostic test is based on the creatinine-derived (estimated) glomerular filtration rate (eGFR). Cystatin C use may provide a better assessment. We compared creatinine- and cystatin C-derived eGFR determination as the first-line diagnostic test for 112 hospital patients aged > 60 years (median = 76 years). The patients were judged to not have CKD (no-CKD group) according to the first-line diagnostic recommendations (n = 61, eGFR (CKD Epidemiology Collaboration (CKD-EPI)) ≥ 60 mL/min/1.73 m2, total urine protein < 150 mg/g creatinine, urinary red/white blood cells not increased) or classified to be at risk for kidney insufficiency due to aortic valve dysfunction (at-risk group; n = 51). The accuracy of the eGFR values was evaluated retrospectively with the final case diagnoses. The eGFR (Caucasian, Asian, pediatric, and adult formula (CAPA)) was found to be linearly correlated to the eGFR (CKD-EPI) (R2 = 0.5, slope = 0.69, p < 0.0001). In 93/112 (>80%) cases, the eGFR (CAPA) yielded lower values (on average ≈-20%). In 55/112 (49%) cases, the cystatin C-derived CKD stage was lower. CKD reclassification from no-CKD to a kidney-insufficient state (i.e., CKD1/2 to CKD3a/b or 4) or reclassification to a more severe kidney insufficiency (i.e., CKD3a → 3b/4 or 3b → 4) was found in 41/112 (37%) cases. A worse CKD classification (no-CKD → kidney-insufficient) based on the eGFR (CAPA) was plausible in 30% of cases in light of the final case diagnoses. In elderly patients (>60 years), renal function appears to be systematically overestimated by the creatinine-based eGFR (CKD-EPI), indicating that, for this group, the cystatin C-based eGFR (CAPA) should be used as the first-line diagnostic test.