Abstract
BACKGROUND: The concept of residual kidney function (RKF) is exclusively based upon urine volume and small solute clearance, making RKF challenging to assess in clinical practice. The aim of this study was to test the technical feasibility of obtaining usable sodium magnetic resonance imaging (23Na-MRI) kidney images in hemodialysis (HD) participants. METHODS: We conducted an exploratory prospective study to quantify the cortico-medullary sodium gradient in 17 healthy volunteers and 21 HD participants. Participants fasted for 8 h prior to their study visit. Urine samples were collected to measure urinary osmolarity, before MRI. Proton and sodium pictures were merged; regions of interest were delineated for the medulla and cortex when feasible. In cases where cortex could not be identified, we considered the corticomedullary gradient (CMG) to be no longer present, resulting in a medulla-to-cortex ratio of 1. RESULTS: Median (interquartile range) fasting medulla-to-cortex ratio was significantly higher 1.56 (1.5-1.61) in healthy volunteers compared with HD patients 1.22 (1.13-1.3), P < .0001. Medulla to cortex ratio and median urinary osmolarity were correlated (r = 0.87, P < .0001) in the whole population. We found a significant association between HD vintage and medulla-to-cortex ratio, whereas we did not find any association with urine volume. Sodium signal intensity distribution within healthy kidney describes two different peaks relating to well defined cortex and medulla, whereas HD participants displays only a single peak indicative of the markedly lower sodium concentration. LIMITATIONS: This study is only exploratory, with a modest number of patients. CONCLUSIONS: The application of kidney 23Na-MRI to the study of RKF in patients receiving maintenance HD is practical and provides a previously unavailable ability to interrogate the function of remnant tubular function.Clinical Trial Registration: NCT05014178.