Magnetic resonance imaging in the diagnosis of acute renal allograft rejection and its differentiation from acute tubular necrosis: Experimental study in the dog

This study was designed to evaluate the potential utility of magnetic resonance imaging (MRI) for the diagnosis of acute renal allograft rejection and its differentiation from acute tubular necrosis (ATN). Eighteen canines were used. Five animals served as controls. ATN was induced in six animals by cross-clamping of the left renal artery for 90 minutes. In order to study acute renal allograft rejection, seven animals were subjected to exchange allograft transplantation of the left kidney. MRI was performed with a 0.35T superconductive magnet. A double spin-echo technique was used with varying TR and TE parameters. The spin echo images were analyzed for morphology, signal intensity, T1 and T2 relaxation times, and spin density. The most useful MRI criteria for the diagnosis of ATN and acute rejection were found to be the renal size, the intensity difference between cortex and medulla (corti- comedullary contrast), and the T1 relaxation time of the cortex. Normal kidneys showed maximal corticomedullary contrast (19% + /-2) on images obtained with TR = 0.5 sec and TE = 28 msec. Cortical T1 relaxation time was 551 msec + /-73. In the ATN group, the kidneys were slightly swollen (P = ns) and the corticomedullary contrast (11% + /-3) was reduced by 42% (P<.01). T1 of the cortex (689 + /-142) was increased by 25% (P<.10). In acute rejection, significant renal enlargement was noted (P<.05). The corticomedullary contrast (9% +/-3) was diminished by 53% (P <.01) while T1 of the cortex (802 msec + /-84) was increased by 45% (P <.01). Although the changes observed in acute rejection were more severe than in ATN, they were not specific since in one case of mild rejection, findings similar to those in ATN were observed.