Hepatocellular carcinoma grading and recurrence prediction using T 1 mapping on gadolinium-ethoxybenzyl diethylenetriamine pentaacetic acid-enhanced magnetic resonance imaging.

The aim of the present study was to explore the value of T-1 mapping on gadolinium-ethoxybenzyl diethylenetriamine pentaacetic (Gd-EOB-DTPA)-enhanced magnetic resonance imaging (MRI) for grading hepatocellular carcinoma (HCC) and predicting its recurrence rate. A retrospective study was performed that included 75 patients (66 men and 9 women; mean age, 52.89 years; age range, 23-79 years) with HCC who had undergone Gd-EOB-DTPA-enhanced MRI with T-1 mapping before surgery. The T-1 relaxation time of the 81 lesions and non-tumorous liver parenchyma in 75 patients with HCC were measured before Gd-EOB-DTPA was injected and then at 5, 10 and 20 min after administration, respectively. T1[lesion (L)-hepatic parenchyma (H)]/H (%) was calculated as the increment rate of the T-1 value in the lesions relative to the non-tumorous liver parenchyma. One-way analysis of variance and Spearman's correlation analysis was used to compare the differences and relationship of T-1 mapping values among the three grades of HCC. A total of 81 lesions were divided into well-differentiated HCC (grades I; n=21), moderately differentiated HCC (grades II; n=40) and poorly differentiated HCC (grades III; n=20) according to the histopathology. The T1(L-H)/H (%) value among grades I, II and III HCC on pre-contrast results and on post-contrast results at the 5-, 10- and 20-min hepatobiliary phase (HBP) were significantly different (P<0.05), and T1(L-H)/H (%) was correlated with the histological grade of HCC at each time point (r=0.637, r=0.554, r=0.499 and r=0.560, respectively, P<0.001). A total of 41 recurrence cases [grade I (n=5), grade II (n=23) and grade III (n=13)] were verified by imaging (CT, MRI or ultrasound) or reoperation. Patients with grade III and grade II HCC had higher recurrence rates compared with that in patients with grade I HCC (P<0.05; median recurrence times were 258 days, 605 days and undefined, respectively). According to the optimal cut-off point for the T1(L-H)/H (%) of the three grades of HCC, patients with HCC in the low T1(L-H)/H (%) value group (<= 155.15%) had lower cumulative recurrence rates compared with that in the medium (T1(L-H)/H (%) >155.15% and T1((L-H)/H) (%) <= 241.20%) and high (T1(L-H)/H (%) >241.20%) value groups at the 20-min HBP (P<0.05; median recurrence times were undefined, 530 days and 447 days, respectively). These results indicate that the parameters of T-1 mapping would be beneficial for predicting the grading and recurrence of HCC.