Preliminary quantitative analysis of renal sinus fat dysfunction in T2DM patients using MRI fat fraction and R2* mapping

Yifan Dong ¹ Qinhe Zhang ¹ Xun Wang ¹ Yuhui Liu ² Qi An ¹ Ziting Zhang ¹ Lifang Hu ¹ Liangjie Lin ³ Ai Lian Liu ^1*

• ¹ Department of Radiology, the First Affiliated Hospital of Dalian Medical University, Dalian, China
• ² Department of Medical Imaging, Dalian Medical University, Dalian, Liaoning Province, China
• ³ Philips Healthcare, Beijing, China

To quantitatively analyze renal sinus fat (RSF) dysfunction in type 2 diabetes mellitus (T2DM) patients using magnetic resonance imaging (MRI) fat fraction (FF) and R2* mapping.The inpatients who underwent 1.5 T MRI examination (including MRI FF and R2* mapping) of the abdomen from January 2017 to December 2023 were enrolled. The RSF volume, FF and R2* of the right and left kidneys and the mean values were measured.Associations between mean FF and R2* value of RSF and T2DM were assessed with logistic regression. Receiver operating characteristic (ROC) curve was applied to calculate area under the curve (AUC) for the parameters to identify T2DM patients. Partial correlation coefficients after controlling for age, sex, and BMI were computed to analyze the correlations among the mean RSF volume, FF and R2*.Results: A total of 186 participants were finally enrolled in this study including 38 patients in T2DM group and 148 patients in non-T2DM group. Univariate logistic regression analyses showed the significant correlations of mean RSF FF (OR: 1.111, 95%CI: 1.054 -1.171), P < 0.001) and R2* (OR: 1.120, 95%CI: 1.013 -1.237), P = 0.027) with T2DM. Multivariate analysis showed that mean RSF FF (OR: 1.231, 95% CI: 1.098 -1.380) is independently associated with T2DM after adjusting for age, sex and BMI. The AUC of mean RSF FF was 0.701 (0.630 -0.766) with the sensitivity and specificity of 57.89% and 75.68%, respectively, when using 34.40% as the cut-off value. The AUC of mean RSF R2* was 0.616 (0.542 -0.686) with the sensitivity and specificity of 68.42% and 58.11%, respectively, when using 21.97 Hz as the cut-off value.Furthermore, mean RSF FF presents significantly higher diagnostic efficacy for T2DM than R2* (P < 0.05). And combining mean RSF FF and R2* improved the diagnostic performance (AUC = 0.729).Mean RSF FF and R2* were significantly associated with T2DM, and mean RSF FF was the independent risk factor of T2DM. This finding indicates the hypertrophy of adipocytes and excessive iron deposition and hypoxia in RSF, which may represent dysfunction of RSF for T2DM.