Article Abstract

Intravoxel incoherent motion magnetic resonance imaging in assessing histopathological features of gastric cancers: initial findings

Authors: Changfeng Ji, Ling Chen, Wenxian Guan, Tingting Guo, Qinglei Zhang, Song Liu, Jian He, Zhengyang Zhou

Abstract

Background: Intravoxel incoherent motion (IVIM) magnetic resonance (MR) imaging has been widely utilized in evaluating various tumors. The application of IVIM MR imaging in characterizing gastric cancer’s histopathological features has never been reported previously.
Methods: A total of 62 patients with gastric cancers were prospectively enrolled in this study. All patients underwent MR examination before surgery. All MR images were reviewed by 2 radiologists using software IDL 6.3 and an oval region of interest (ROI) was manually drawn on the specific slice showing the largest area of tumor. Four parameters were calculated automatically: apparent diffusion coefficient (ADC), pure diffusion coefficient (D), perfusion related fraction (f) and pseudo diffusion coefficient (D*). One-way analysis of variance (normality) or Kruskal-Wallis test (non-normality), independent samples t-test (normality) or Mann-Whitney U test (non-normality), Spearman correlation test and receiver operating characteristic (ROC) curve analysis were used for statistical analyses.
Results: Poorly/moderate-poorly differentiated gastric cancers showed significantly lower ADC and D values compared with moderately/well differentiated tumors (P=0.001, P<0.001, respectively). Significant differences were also observed for ADC and D values among different Lauren classifications (P=0.004, 0.048, respectively). No significant differences were observed for f and D* values among different differentiation degrees (P=0.774, 0.716, respectively) and Lauren classifications (P=0.633, 0.290, respectively). The ADC and D values correlated negatively with T stage (r=−0.363, −0.295, respectively), N stage (r=−0.265, −0.285, respectively) and overall stage (r=−0.485, −0.368, respectively) of gastric cancers, but no significant correlations were observed for f and D* values (all P>0.05). No significant differences were found for ADC, D, f or D* values among different status of vascular invasion and perineural invasion.
Conclusions: The IVIM-derived ADC and D values showed significant differences among different histological differentiation, Lauren classification and TNM stage of gastric cancers. They showed potential in assessing gastric cancer’s histopathological features.