P0829THE ASSOCIATION BETWEEN URINARY NGAL AND EGFR SLOPE IN PATIENTS WITH CKD

Abstract

Abstract Background and Aims Urinary neutrophil gelatinase-associated lipocalin (uNGAL) is a new biomarker for renal tubular damage. The association between uNGAL and renal prognosis have been reported in many papers; however, it may often be affected by urinary tract infection (UTI), and its clinical value with consideration of UTI has not been well investigated in patients with chronic kidney disease (CKD). The aims of our study were to investigate the association between uNGAL and eGFR decline in CKD incorporating the effect of UTI. Method This was a retrospective observational cohort study at a single hospital in Japan. We included adult patients with the estimated glomerular filtration rate (eGFR) of 10 to 70 mL/min/1.73m2 from Jan 2017, who had at least one measurement of uNGAL. We used baseline uNGAL adjusted for urinary creatinine (Cr) as an exposure variable and divided the patients into quartiles. UTI was determined by a single evaluation of urinalysis at baseline. The repeated measured eGFRs were obtained from the electrical health record until dialysis initiation, loss to follow-up, or the end of the observation period whichever occurred first. We performed longitudinal analyses for eGFR decline using the mixed effects model and evaluated the longitudinal effect of uNGAL taking into account of UTI. All statistical analyses were done using STATA 13.1. Results In total, 281 patients with CKD were included. Mean age and eGFR at baseline were 70.2 years and 33.9 mL/min/1.73m2, respectively. The median [interquartile range (IQR)] of baseline uNGAL and urinary protein to Cr ratio (uPCR) were 49.6 [19.4 - 171.4] ug/gCr and 0.82 [0.18 - 3.26] g/gCr, respectively. Patients with UTI (15%) showed significantly higher uNGAL at baseline than those without (219 [89 - 425] vs. 39 [16 - 131]). During the mean follow up of 229 days, nine patients developed eGFR < 5 mL/min/1.73m2. Although a higher uNGAL level was significantly associated with lower eGFR levels at baseline (-3.3 [95% confidence interval (CI): -7.6, 0.93], -9.6 [-14.2, -4.9], and -16.3 [-21.4, -11.1] mL/min/1.73m2 in quartiles Q2-4 as compared to Q1); the eGFR trajectory did not differ with or without baseline UTI in the analytic model, while baseline proteinuria was significantly associated with a steeper slope of eGFR. Conclusion A higher uNGAL was significantly associated with lower eGFR levels at baseline; however, it was not associated with a steeper slope of eGFR in the longitudinal analysis, even with the consideration of the effect of UTI.