MRI-based kidney volume measurement in autosomal dominant polycystic kidney disease (ADPKD): Comparison of T2-weighted, ideal, and post-contrast T1-weighted sequences

Abstract

© JOURNAL OF THE MEDICAL ASSOCIATION OF THAILAND | 2020 Objective: To compare kidney volume (KV) measurements between, single-shot T2-weighted (sshT2W), IDEAL, and post contrast T1-weighted (T1W) images of autosomal dominant polycystic kidney disease (ADPKD) patients. Materials and Methods: KVs of ADPKD patients presenting at the nephrologic clinic between January 2014 and April 2015 were segmented from coronal sshT2W images, opposed phase IDEAL images, and post-gadolinium T1W images. Measurements were performed independently by three readers using a method involving manual segmentation with image thresholding. The mean KV differences between the evaluated magnetic resonance imaging (MRI) pulse sequences were compared using paired t-tests. Associations between KV and estimated glomerular filtration rate (eGFR) were determined by Pearson correlation coefficient (r). Inter-reader agreement in kidney volumetry was evaluated by intraclass correlation coefficient (ICC) for each comparison. Results: For the cohort of 28 ADPKD patients, the total kidney volume (TKV) and total cyst volume (TCV) measured on sshT2W images were significantly higher than those on post-contrast T1W and opposed phase IDEAL images. The mean differences were the smallest between sshT2W and post-contrast T1W images. The processing times for the sshT2W sequence were significantly less than for the other sequences. KVs showed significant negative correlation with renal function on all evaluated MR sequences. High levels of agreement between all pairings of the three readers were found for all measurements. Conclusion: Non-gadolinium MRI sequences, especially sshT2W sequences, are of sufficient quality for KV measurement in ADPKD, revealing the same negative correlation with eGFR as the other sequences, the smallest mean difference with post-contrast T1W images, and requiring less processing time.