Differentiating Esophageal Small Cell Carcinoma from Esophageal Squamous Cell Carcinoma based on DWI, DKI, and IVIM: A prospective study

Abstract Background Magnetic resonance imaging (MRI), including diffusion-weighted imaging (DWI), intravoxel incoherent motion imaging (IVIM), and diffusion kurtosis imaging (DKI), is a widely used approach for preoperative evaluation, pathological grading, and evaluation of the efficacy of chemoradiotherapy in patients with esophageal cancer. Yet, these methods have not been tested for differentiating esophageal small cell carcinoma (SCCE) from esophageal squamous cell carcinoma (ESCC). Purpose To evaluate the diagnostic value of IVIM, DKI, and DWI in distinguishing SCCE from ESCC. Study type: A prospective study. Populations: Fifty patients with esophageal carcinoma were selected (19 with SCCE and 31 with ESCC). Sequence: All subjects underwent sagittal IVIM, sagittal DWI, and sagittal DKI scans. Assessment: IVIM, DWI, and DKI parameters, including true diffusion coefficient (D), perfusion fraction (f), pseudo-diffusion coefficient (D * ), apparent diffusion coefficient (ADC), mean kurtosis (MK), and mean diffusivity (MD) were measured for each patient. Statistical tests: The bland-Altman method was used to analyze the repeatability and inducibility of imaging parameters; ANOVA test (normal distribution) or Kruskal-Wallis H test (non-normal distribution) was used to compare differences in parameters. ROC curve was used to analyze the diagnostic efficacy of each parameter in distinguishing esophageal SCCE from ESCC. Results Except for D*, ADC, D, and MD of ESCC were higher than those of esophageal SCCE, while f and MK were lower (all P < 0.05). The AUC of MK, MD, ADC, and D values was higher than that of the f value (P < 0.05), while the difference in the AUC values of MK, MD, ADC, and D was not statistically significant (P > 0.05). Conclusion IVIM and DKI can provide more information than DWI in differentiating SCCE from ESCC. MD, MK, D, and ADC feature a higher diagnostic efficacy.