Detection of relationships between SUDOSCAN with estimated glomerular filtration rate (eGFR) in Chinese patients with type 2 diabetes

Abstract

ObjectiveSUDOSCAN (Impeto Medical, Paris, France) has been proved to be a new and non-invasive method in detecting renal dysfunction in type 2 diabetes mellitus (T2DM) patients. In this study, we sought to compare the result of diabetic kidney dysfunction score (DKD-score) of SUDOSCAN with estimated glomerular filtration rate (eGFR) by using quantile regression analysis, which was completely different from previous studies. MethodsA total number of 223 Chinese T2DM patients were enrolled in the study. SUDOSCAN, renal function test (including blood urea nitrogen, creatinine and uric acid) and 99mTc-diethylenetriamine pentaacetic acid (99mTc-DTPA) renal dynamic imaging were performed in all T2DM patients. DKD-score of SUDOSCAN was compared with eGFR detected by 99mTc-DTPA renal dynamic imaging through quantile regression analysis. Its validation and utility was further determined through bias and precision test. ResultsThe quantile regression analysis demonstrated the relationship with eGFR was inverse and significant for almost all percentiles of DKD-score. The coefficients decreased as the percentile of DKD-score increased. And in validation data set, both the bias and precision were increased with the eGFR (median difference, −21.2 ml/min/1.73 m2 for all individuals vs. −4.6 ml/min/1.73 m2 for eGFR between 0 and 59 ml/min/1.73 m2; interquartile range [IQR] for the difference, −25.4 ml/min/1.73 m2 vs. −14.7 ml/min/1.73 m2). The eGFR category misclassification rate were 10% in eGFR 0–59 ml/min/1.73 m2 group, 57.3% in 60–90 group, and 87.2% in eGFR > 90 ml/min/1.73 m2 group. ConclusionDKD-score of SUDOSCAN could be used to detect renal dysfunction in T2DM patients. A higher prognostic value of DKD-score was detected when eGFR level was lower.