Splenic T 1ρ as a noninvasive biomarker for portal hypertension

There is a need for noninvasive methods for the diagnosis and monitoring of portal hypertension (PH). To 1) assess the correlation of liver and spleen T and T measurements with portal pressures in patients with chronic liver disease, and 2) to compare the diagnostic performance of the relaxation parameters with radiological assessment of PH. Prospective. Twenty-five patients (M/F 16/9, mean age 56 years, range 21-78 years) undergoing portal pressure (hepatic venous pressure gradient [HVPG]) measurements. 1.5T abdominal MRI scan, including T and T mapping. Liver and spleen T and T , radiological PH score, and (normalized) spleen length were evaluated. Spearman correlation of all MRI parameters with HVPG was assessed. The diagnostic performance of the assessed parameters for prediction of PH (HVPG ≥5 mmHg) and clinically significant PH (CSPH, HVPG ≥10 mmHg) was determined by receiver operating characteristic (ROC) analysis. The mean HVPG measurement was 7.8 ± 5.3 mmHg (PH, n = 18 [72%] including CSPH, n = 9 [36%]). PH score, (normalized) spleen length and spleen T significantly correlated with HVPG, with the strongest correlation found for spleen T (r = 0.613, P = 0.001). Spleen T was the only parameter that showed significant diagnostic performance for assessment of PH (area under the curve [AUC] 0.817, P = 0.015) and CSPH (AUC = 0.778, P = 0.024). Normalized spleen length also showed significant diagnostic performance for prediction of CSPH, with a slightly lower AUC (= 0.764, P = 0.031). The radiological PH score, T and T of the liver and T of the spleen, did not show significant diagnostic performance for assessment of CSPH (P > 0.075). Spleen T showed a significant correlation with portal pressure and showed improved diagnostic performance for prediction of CSPH compared to radiological assessment. These initial results need confirmation in a larger cohort. 1 Technical Efficacy Stage: 2 J. Magn. Reson. Imaging 2020;52:787-794.