Our purpose was to investigate three-dimensional (3D) gadolinium-enhan
ced magnetic resonance angiography (Gd-MRA) in the evaluation of renal
transplant arteries. Eleven MR angiography examinations were performe
d in nine renal transplant patients. Gd-MRA, three-dimensional phase c
ontrast (3D-PC) post-gadolinium, and two-dimensional time-of-flight (2
D-TOF) MR angiography were performed and independently reviewed by thr
ee vascular radiologists who, for each MR angiography sequence, separa
tely graded occlusive disease in the ipsilateral iliac artery, the tra
nsplant artery anastomosis, and the transplant artery itself. The Gd-M
RA and SD-PC data were reviewed as maximum intensity projections (MIP)
reconstructed in standard planes, and the 2D-TOF data were interprete
d from source images. In addition, a single vascular radiologist prosp
ectively interpreted the Gd-MRA and 3D-PC data together, hereinafter G
d/PC, from MIP reconstructions for each case. In all of these patients
either surgical (n = 3) or angiographic studies (n = 8) were performe
d within 21 days following the MR examination, which served as a refer
ence standard to determine sensitivity and specificity. The sensitivit
y/specificity for the detection of significant stenosis were as follow
s: Gd-MRA, 67/88; 3D-PC, 60.3/76.6; 2D-TOF, 47/81; and Gd/PC, 100/100.
The kappa statistic (kappa) for interobserver agreement for the gradi
ng of stenoses by 2D-TOF, Gd-PC, and Gd-MRA was 0.48, 0.60, and 0.74,
respectively. The percentage of all vascular segments seen well enough
to grade (cumulative for all three observers) was 94%, 85%, and 79% f
or Gd-MRA, SD-PC, and 2D-TOF, respectively. The combination of Gd-MRA
and 3D-PC is a promising approach to the evaluation of transplant rena
l arteries. (C) 1997 Elsevier Science Inc.