Detection of Acute Tubular Necrosis Using Blood Oxygenation Level-Dependent (BOLD) MRI

Background/aims: To date, there is no imaging technique to assess tubular function in vivo. Blood oxygen level-dependent magnetic resonance imaging (BOLD MRI) measures tissue oxygenation based on the transverse relaxation rate (R2*). The present study investigates whether BOLD MRI can assess tubular function using a tubule-specific pharmacological maneuver.

Methods: Cross sectional study with 28 participants including 9 subjects with ATN-induced acute kidney injury (AKI), 9 healthy controls, and 10 subjects with nephron sparing tumor resection (NSS) with clamping of the renal artery serving as a model of ischemia/reperfusion (I/R)-induced subclinical ATN (median clamping time 15 min, no significant decrease of eGFR, p=0.14). BOLD MRI was performed before and 5, 7, and 10 min after intravenous administration of 40 mg furosemide.

Results: Urinary neutrophil gelatinase-associated lipocalin was significantly higher in ATN-induced AKI and NSS subjects than in healthy controls (p=0.03 and p=0.01, respectively). Before administration of furosemide, absolute medullary R2*, cortical R2*, and medullary/cortical R2* ratio did not significantly differ between ATN-induced AKI vs. healthy controls and between NSS-I/R vs. contralateral healthy kidneys (p>0.05 each). Furosemide led to a significant decrease in the medullary and cortical R2* of healthy subjects and NSS contralateral kidneys (p<0.05 each), whereas there was no significant change of R2* in ATN-induced AKI and the NSS-I/R kidneys (p>0.05 each).

Conclusion: BOLD-MRI is able to detect even mild tubular injury but necessitates a tubule-specific pharmacological maneuver, e.g. blocking the Na+-K+-2Cl- transporter by furosemide.