Spleen Stiffness on Magnetic Resonance Elastography for the Detection of Portal Hypertension: A Systematic Review and Meta-Analysis

Abstract

Background: Chronic liver disease, complicated by portal hypertension (PH), may alter the extra-hepatic hemodynamic condition and spleen stiffness (SS). We aimed to evaluate the diagnostic accuracy of MRI-based elastographic methods (MRE) for detecting PH.

Methods: Seven studies were included with reference to SS with regard to the prediction of PH patients. Major outcomes considered for data extraction were diagnostic parameters for MRI for concluding mild PH, clinically significant PH and severe PH. PubMed, Scopus, Google Scholar, Cochrane and Science Direct databases were used to extract the published literature through to May, 2021Using the Rayyan Zotero and R softwares.

Results: Out of 587 studies extracted, 7 were selected based on inclusion and exclusion criteria. A QUADAS-2 assessment showed that all studies were clear in terms of patient selection and reference standard. A funnel plot showed that all the selected studies were outliers, indicating a low level of accuracy for the studies included. Subgroup analysis, with reference to SS as a predictor of PH, revealed raw mean difference (RMD) of 7.78% (95% CI 5.23-10.34, P<0.01). The corresponding RMD observed for <60 years and >60 yr were 34.26% (95% CI 9.33- 59.20, I2=100%,  T2=646.7688, P=0), and 46.92% (95% CI 20-59.33, I2=97%, T2=1003.023, P=0) respectively. The specificity and sensitivity noted for MRI in determining SS were 0.721 and 0.747, respectively with an area-under the curve of 0.788. The estimated random effect models for specificity and sensitivity were 0.938 and 0.842, respectively.

Conclusion: The real-time MRE has acceptable specificity and sensitivity for diagnosing SS.