Peritumoral Imaging Manifestations on Gd-EOB-DTPA-Enhanced MRI for Preoperative Prediction of Microvascular Invasion in Hepatocellular Carcinoma: A Systematic Review and Meta-Analysis

Abstract

PurposeThe aim was to investigate the association between microvascular invasion (MVI) and the peritumoral imaging features of gadolinium ethoxybenzyl DTPA-enhanced magnetic resonance imaging (Gd-EOB-DTPA-enhanced MRI) in hepatocellular carcinoma (HCC).MethodsUp until Feb 24, 2022, the PubMed, Embase, and Cochrane Library databases were carefully searched for relevant material. The software packages utilized for this meta-analysis were Review Manager 5.4.1, Meta-DiSc 1.4, and Stata16.0. Summary results are presented as sensitivity (SEN), specificity (SPE), diagnostic odds ratios (DORs), area under the receiver operating characteristic curve (AUC), and 95% confidence interval (CI). The sources of heterogeneity were investigated using subgroup analysis.ResultsAn aggregate of nineteen articles were remembered for this meta-analysis: peritumoral enhancement on the arterial phase (AP) was described in 13 of these studies and peritumoral hypointensity on the hepatobiliary phase (HBP) in all 19 studies. The SEN, SPE, DOR, and AUC of the 13 investigations on peritumoral enhancement on AP were 0.59 (95% CI, 0.41−0.58), 0.80 (95% CI, 0.75−0.85), 4 (95% CI, 3−6), and 0.73 (95% CI, 0.69−0.77), respectively. The SEN, SPE, DOR, and AUC of 19 studies on peritumoral hypointensity on HBP were 0.55 (95% CI, 0.45−0.64), 0.87 (95% CI, 0.81−0.91), 8 (95% CI, 5−12), and 0.80 (95% CI, 0.76−0.83), respectively. The subgroup analysis of two imaging features identified ten and seven potential factors for heterogeneity, respectively.ConclusionThe results of peritumoral enhancement on the AP and peritumoral hypointensity on HBP showed high SPE but low SEN. This indicates that the peritumoral imaging features on Gd-EOB-DTPA-enhanced MRI can be used as a noninvasive, excluded diagnosis for predicting hepatic MVI in HCC preoperatively. Moreover, the results of this analysis should be updated when additional data become available. Additionally, in the future, how to improve its SEN will be a new research direction.

Gadoxetic Acid–Enhanced Liver MRI

“…6). According to a meta-analysis, the pooled sensitivity and specificity of peritumoral arterial enhancement were 0.50 and 0.80, respectively, whereas these values for peritumoral hypointensity during the HBP were 0.55 and 0.87, respectively 19 . Thus, GA-MRI exhibits poor sensitivity, but good specificity for the identification of MVI.…”

Imaging prognostication and tumor biology in hepatocellular carcinoma

“… 54 , 55 Previous work has shown that HBP peritumoral hypointensity is highly correlated with MVI. 55 , 56 Kim et al 55 suggested that HBP peritumoral hypointensity without associated signal intensity changes on non-HBP sequences represents the change in function of hepatocytes surrounding the tumor. This explains why this imaging finding has a high specificity yet low sensitivity for MVI, as altered hepatocyte function is thought to occur later in the course of MVI development and expansion.…”

Hepatocellular Carcinoma: Optimal Radiological Evaluation before Liver Transplantation

“…In addition to better detection of nodules, in particular with hepatic arterial phase hypoenhancement and hypointensity on HBP acquisitions, HBP images have prognostic value in patients with HCC. Marked hepatobiliary phase hypointensity is correlated with HCC aggressiveness [19], and peritumoral hypointensity is associated with the presence of microvascular invasion (MVI) [20,21]. A recent meta-analysis has also reported an increased risk of tumor recurrence after resection or LT in the presence of hypointense HCC on HBP acquisitions [19].…”

“…The imaging features associated with MVI include nonsmooth tumor margin [78], incomplete tumor capsule [79], low ADC, large size, peritumoral APHE, and peritumoral HBP hypointensity [80]. Peritumoral hypointensity on HBP is the MRI feature that is most suggestive of MVI and can be useful to confirm MVI [21]. Numerous prognostic nomograms or feature clusters have been proposed to predict MVI [81][82][83].…”