# Prediction of the wall-invasion pattern of advanced gallbladder carcinoma using extracellular volume fraction

## TL;DR

This study shows that extracellular volume fraction can predict aggressive growth patterns in advanced gallbladder cancer, which may help guide treatment decisions.

## Contribution

The study introduces extracellular volume fraction as a novel imaging biomarker for predicting gallbladder cancer invasion patterns.

## Key findings

- ECV fraction was significantly higher in destructive growth type of gallbladder carcinoma.
- ECV values showed strong diagnostic ability with high ROC curve areas.
- Destructive growth type had significantly shorter recurrence-free survival.

## Abstract

This study aimed to evaluate the utility of extracellular volume (ECV) fraction for predicting wall-invasion patterns in advanced gallbladder carcinoma (GBCA).

Patients who had surgically resected GBCA at a single institution were retrospectively evaluated. All patients underwent computed tomography (CT) before the surgery. Based on pathological examinations, the wall-invasion pattern of GBCA was classified into two groups: infiltrative growth (IG, n = 19) and destructive growth (DG, n = 11).

ECV map was generated by inputting the patients’ hematocrit values and subtraction algorithms using pre-contrast and equilibrium phase images. CT parameters were evaluated by two radiologists (Rad1 and Rad2). The Mann–Whitney U test was performed to identify significant CT parameters for differentiating between the two groups. The diagnostic ability was measured using receiver operating characteristic (ROC) curve analysis. Recurrence-free survival (RFS) was estimated using the Kaplan–Meier method, and differences between the two groups were compared using the log-rank test.

Thirty patients (mean age, 75.5 years; 20 men) were evaluated. Mean ECV fraction of the DG-type (Rad1, 34.5%; Rad2, 34.1%) was significantly higher than that of the IG-type (Rad1, 28.5%; Rad2, 28.8%) (p < 0.05). The ECV values of the two radiologists indicated that the areas under the ROC curves for differentiation between the two groups were Rad1, 0.91 and Rad2, 0.84 (p < 0.05). Medium RFS of the DG-type (970 days) was significantly shorter than that of the IG-type (2200 days) (p < 0.05).

ECV fraction demonstrates potential as the most valuable predictor of the DG type of GBCA, which has a higher recurrence rate compared with the IG type. However, further large-scale multi-center studies are required to validate these findings.