Noninvasive Identification of Renal Hypoxia in Experimental Myocardial Infarctions of Different Sizes by Using BOLD MR Imaging in a Mouse Model

Purpose To test the feasibility of using blood oxygen level-dependent (BOLD) magnetic resonance (MR) imaging to measure alterations in renal oxygenation in a mouse model with experimental myocardial infarctions (MIs) of different sizes. Materials and Methods The study was approved by the local animal ethics committee. One hundred eighty-nine male C57BL/6 J mice were randomly subjected to MI surgery (with different locations of left anterior descending coronary artery occlusion) or sham surgery, defined as the exposure of the heart but no ligation. Mice with MI underwent late gadolinium enhancement imaging 1 day after occlusion to confirm infarct size. Mice were sorted into three groups: those with large MI (n = 48), those with small MI (n = 48), and those with sham operation (n = 36). Renal BOLD MR imaging was performed before and 1, 7, 14, 28, and 60 days after MI, and histologic analysis of renal hypoxia-inducible factor-1α (HIF-1α) and kidney injury molecule-1 (KIM-1) was performed to evaluate tissue hypoxia and kidney injury in subgroups imaged at each time point. The relationships between the BOLD R2* and HIF-1α expression and between HIF-1α and KIM-1 expression were assessed. Statistical analyses were performed with one-way analysis of variance or the Kruskal-Wallis test and Spearman correlation test. Results A significant elevation in R2* was detected in the MI groups compared with the sham group in the cortex (P < .001 for large MI vs sham group; P = .007 for small MI vs sham group) and medulla (P < .001 for large MI vs sham group; P = .003 for small MI vs sham group) on day 60, and R2* was higher in the large MI group than in the small MI group (P < .001). Renal HIF-1α expression was increased after MI and showed linear correlation with R2* in the cortex (R² = 0.56) and medulla (R² = 0.63). In addition, an increase in renal KIM-1 was observed in the MI groups compared with the sham group on day 60 (sham group, 53.9 × 10³ arbitrary units [au] ± 35.2; large MI group, 389.3 × 10³ au ± 99.8; and small MI group, 185.8 × 10³ au ± 91.9; P < .001 for large MI group vs sham group; P = .037 for small MI group vs sham group), and renal KIM-1 showed a positive correlation with HIF-1α (R² = 0.68). Conclusion The magnitude of renal hypoxia with MIs of different sizes can be noninvasively measured with BOLD MR imaging, and increased renal hypoxia is a potential risk factor for progressive tubulointerstitial injury in mouse kidneys. ^© RSNA, 2017 Online supplemental material is available for this article.