White Matter Hyperintensities Quantification in Healthy Adults: A Systematic Review and Meta-Analysis

Luca Melazzini,Paolo Vitali,Emanuele Olivieri,Marco Bolchini,Moreno Zanardo,Filippo Savoldi,Giovanni Di Leo,Ludovica Griffanti,Giuseppe Baselli,Francesco Sardanelli,Marina Codari
DOI: https://doi.org/10.1002/jmri.27479
Abstract:Background: Although white matter hyperintensities (WMH) volumetric assessment is now customary in research studies, inconsistent WMH measures among homogenous populations may prevent the clinical usability of this biomarker. Purpose: To determine whether a point estimate and reference standard for WMH volume in the healthy aging population could be determined. Study type: Systematic review and meta-analysis. Population: In all, 9716 adult subjects from 38 studies reporting WMH volume were retrieved following a systematic search on EMBASE. Field strength/sequence: 1.0T, 1.5T, or 3.0T/fluid-attenuated inversion recovery (FLAIR) and/or proton density/T2 -weighted fast spin echo sequences or gradient echo T1 -weighted sequences. Assessment: After a literature search, sample size, demographics, magnetic field strength, MRI sequences, level of automation in WMH assessment, study population, and WMH volume were extracted. Statistical tests: The pooled WMH volume with 95% confidence interval (CI) was calculated using the random-effect model. The I2 statistic was calculated as a measure of heterogeneity across studies. Meta-regression analysis of WMH volume on age was performed. Results: Of the 38 studies analyzed, 17 reported WMH volume as the mean and standard deviation (SD) and were included in the meta-analysis. Mean and SD of age was 66.11 ± 10.92 years (percentage of men 50.45% ± 21.48%). Heterogeneity was very high (I2 = 99%). The pooled WMH volume was 4.70 cm3 (95% CI: 3.88-5.53 cm3 ). At meta-regression analysis, WMH volume was positively associated with subjects' age (β = 0.358 cm3 per year, P < 0.05, R2 = 0.27). Data conclusion: The lack of standardization in the definition of WMH together with the high technical variability in assessment may explain a large component of the observed heterogeneity. Currently, volumes of WMH in healthy subjects are not comparable between studies and an estimate and reference interval could not be determined. Level of evidence: 1 TECHNICAL EFFICACY STAGE: 1.
What problem does this paper attempt to address?