Interference of urinary albumin-to-creatinine ratio measurement by glycosuria: clinical implications when using SGLT-2 inhibitors

Abstract

AbstractAlbuminuria is used for chronic kidney disease (CKD) screening, diagnosis, staging, and monitoring. A change in albuminuria has been proposed as a surrogate outcome for CKD progression. High glucose concentration interferes with Jaffe serum creatinine assays but the extent to which glycosuria biases measurement of urinary albumin-to-creatinine ratio (uACR) is uncertain. Any interference would have implications as the use of sodium-glucose co-transporter-2 (SGLT-2) inhibitors increases. We performed laboratory-based interference studies on urine samples from 333 adults with CKD stages 3-4. Samples were separated into four aliquots: a reference aliquot and three aliquots spiked with increasing concentrations of glucose solution representing the range expected in patients taking SGLT-2 inhibitors (28, 111 and 333 mmol/L). uACR was assayed using Jaffe and enzymatic methods. Median (Q1-Q3) uACR in reference samples was 63 (17-150) mg/mmol. Glucose spiking did not interfere with uACR estimation using enzymatic creatinine assays. For the Jaffe assay, the presence of 28 mmol/L of glucose resulted in a -1.5% mean bias in uACR (95% confidence interval -1.9 to -1.1%) which increased to a -2.5% bias (−3.2 to -1.9%) at a concentration of 333 mmol/L. Overestimation of urinary creatinine concentration increased substantially with decreasing creatinine concentration (i.e. dilute urine). In this cohort, interference of the Jaffe assay by glucose spiking resulted in 2-5% of uACR samples having a ≥10% spurious reduction in uACR (on its original scale). Given the increasing use of SGLT-2 inhibitors, we suggest uACR measured using Jaffe creatinine assays should be avoided.