Comparison of unenhanced and contrast-enhanced 3T magnetic resonance portovenography in children with extra hepatic portal venous obstruction.

Abstract

To evaluate role of 3-T magnetic resonance portovenography (MRPV) in children with extra hepatic portal venous obstruction (EHPVO) and compare unenhanced and contrast-enhanced sequences. 20 Children with EHPVO underwent MRPV using unenhanced [T2, T1 pre-contrast, Balanced turbo field echo (BTFE)] and contrast-enhanced (CE) modified DIXON (mDIXON) sequences. The images were evaluated for the patency of abdominal veins, hepatic and splenic parenchyma by two radiologists for interobserver agreement. BTFE and post-contrast mDIXON sequences performed best for evaluation of abdominal veins. Concordance between the BTFE and CE sequences was good to very good for both the radiologists for splenic vein (SV) evaluation. The concordance between the two sequences for evaluation of superior mesenteric vein and its confluence with the SV was less remarkable and varied from fair to good, while the interobserver agreement was very good to perfect. The interobserver agreement between the BTFE and CE sequence for the evaluation of left renal vein was very good, while that for IVC was perfect. The agreement between BTFE and CE sequence was perfect for the evaluation of right and middle hepatic veins, while the interobserver agreement was good to perfect. The interobserver agreement was poor for evaluation of hepatic parenchyma on BTFE sequence as compared to CE sequence, and moderate for splenic parenchyma. BTFE sequence is the single best unenhanced MR pulse sequence to detect all the vascular structures in children with EHPVO. CE-MRI is not superior to BTFE sequence and should be used at the discretion of the radiologist.