RESUMO
Objective:To investigate the feasibility of dynamic contrast-enhanced (DCE)-MRI in assessing the depth of invasion (DOI) of early tongue squamous cell carcinoma.Methods:From January 2016 to December 2020, a total of 100 patients with early tongue squamous cell carcinoma confirmed by postoperative pathology were retrospectively analyzed in the Ninth People′s Hospital, Shanghai Jiao Tong University School of Medicine. The study included 48 cases of T1 stage and 52 cases of T2 stage. All patients underwent routine MRI, DCE-MRI and contrast-enhanced T 1WI (CE-T 1WI) before surgery. The DOI was measured on images at different phases of axial DCE-MRI (30, 60 and 120 s after contrast injection) and CE-T 1WI (MRI-DOI) by 2 doctors independently. The intraclass correlation coefficient (ICC) was applied to evaluate the consistency of the measurements. MRI-DOI measured on DCE-MRI 30, 60, 120 s, CE-T 1WI and pathological DOI measured on biopsies were statistically analyzed with analysis of variance for repeated measurement and least significant difference t test. Pearson correlation coefficient was used to compare the correlation between MRI-DOI and pathological DOI. Receiver operation characteristic (ROC) curve analysis was used to explore the performance of MRI-DOI for clinical T1 and T2 staging. Results:There was a good consistency in MRI-DOI measured on DCE-MRI 30, 60, 120 s, and CE-T 1WI images with ICC of 0.752, 0.875, 0.883, and 0.841, respectively. The values of MRI-DOI were (8.35±3.52), (6.88±2.41), (7.52±2.65) and (8.60±3.39) mm, respectively, and pathological DOI was (5.75±2.01) mm. There was statistically significant difference in the overall comparison among different phases of MRI-DOI and pathological DOI ( F=69.25, P<0.001). MRI-DOI were significantly higher than pathological DOI ( P<0.05). All MRI-DOI measured on DCE-MRI 30, 60, 120 s and CE-T 1WI correlated positively with pathological DOI ( r=0.574, 0.851, 0.731, 0.663, all P<0.001). MRI-DOI derived from DCE-MRI 60 s showed the highest diagnostic efficiency for T1 and T2 staging (area under the ROC curve was 0.931, 95%CI 0.881-0.982). When the optimal cutoff value was 6.0 mm, the accuracy, sensitivity and specificity were 88.0%, 96.2% and 79.2%, respectively. Conclusions:DCE-MRI can be applied to assess DOI in early tongue squamous cell carcinoma. MRI-DOI based on DCE-MRI 60 s has the best correlation with pathological DOI and has a potential to predict clinical T staging in early tongue squamous cell carcinoma.