MRI-based kidney volume measurements in ADPKD: reliability and effect of gadolinium enhancement

To evaluate the inter- and intrareader reliability and the effect of gadolinium enhancement on kidney volume measurements obtained from pre- and postgadolinium T1 MR images in patients with autosomal dominant polycystic kidney disease (ADPKD).

Twenty subjects were randomly selected with approximately equal frequency from three kidney-size groups. Pre- and postgadolinium 3D T1 (pre-T1, post-T1) MR images were obtained. The stereology method was applied to segment and measure kidney volumes. The measurement process was repeated at two-wk intervals by two radiologists. Reliability was assessed with correlation coefficients. Intra- and inter-reader bias and measure differences were assessed with paired T-tests. The size effect on the pre- and post-T1 measurements was evaluated with one-way ANOVA.

The intra- and inter-reader reliability was extremely high in all measurements. No systematic intrareader bias but a small inter-reader bias for the post-T1 measurements was observed. All kidney volumes measured on the pre- and post-T1 images were highly correlated with each other for both readers. The post-T1 volumes were significantly higher than pre-T1 volumes. While the post-pre volume differences were relatively constant across the three kidney-size groups, the post-pre percent volume differences were significantly smaller as the size of the kidney increased.

Kidney volume measurements can be made with minimum intra- and inter-reader variability on both pre- and post-T1 MR images. Kidney volumes measured on the pre-T1 were smaller than those on post-T1, and percent differences between pre-T1 and post-T1 kidney volumes decreased with increasing kidney size.