Non-Obstructive Kidney Transplant Dysfunction: Magnetic Resonance Evaluation

Abstract

The value of magnetic resonance imaging in the differential diagnosis of non-obstructive dysfunction of renal allografts was studied in a series of 58 examinations at 0.5 T. Four parameters were evaluated: the corticomedullary differentiation; the relative thickness of the cortex; the evolution, with echo number, of the relative signal intensities of kidney parenchyma and adjacent fatty tissue on images generated by a long time to repeat multiecho sequence; and the proximal vascularization. The loss of corticomedullary differentiation is the major finding in acute rejection, but it is not specific as it is also observed in chronic rejection and in the much rarer acute glomerulonephritis. Thickening of the cortex is helpful for the detection of rejected transplants with visible corticomedullary delineation (26% of the cases). Uncomplicated acute tubular necrosis appears as a normal transplant. New immunosuppressive medications, especially cyclo- sporin, have improved the results of kidney transplanta- tion, but they have also added to the difficulty of monitoring the post-transplantation course: cyclosporin nephrotoxicity must now be added to acute tubular necrosis (ATN) and rejection as causes of graft dysfunction. Moreover, cyclosporin decreases the clinical and functional manifestations of acute rejection. Radionuclide scintigraphy and sonography are estab- lished non-invasive modalities in the evaluation of post- transplant renal failure. Sonography is clearly the ideal technique for recognizing urinary obstruction. How- ever, both sonography and scintigraphy lack sensitivity and specificity in the early detection of acute rejection. Experience with magnetic resonance imaging (MRI) is still limited, and the present study was undertaken to assess its usefulness in the evaluation of kidney transplant patients with non-obstructive renal dys- function. Materials and methods Technique Magnetic resonance examinations were performed on a 1.5 T superconductive system (Philips SI5 Gyroscan), operating at 0.5 T. A surface coil was used for signal detection and the regular body coil for spin excitation. Both coils were electronically decoupled (Boskamp, 1985). During initial studies, a rectangular surface coil was used; this was subsequently replaced by a sandwich coil that provided good magnetic coupling with both walls of the hemipelvis. Although the prone position generates less movement artefact, patients were positioned supine, as this position proved to be better