No abstract

Purpose: To identify the MRI-hallmarks of liver metastatic neuroendocrine tumors (mNETs) with different localization of primary tumor.Material and methods: 75 liver mNET patients were enrolled in the study. The hepatic metastasis patients were divided into two groups: with pancreatic mNETs (n = 37) and with gastrointestinal tract (gut) mNETs (n = 38), including those of a stomach, small and large bowel, and appendicular primary. All patients underwent abdominal contrast-enhanced MRI with the measurement: the number and the maximum size of the lesions, the presence and size of avascular zones in the lesions, the presence of MRI signs of hemoglobin deg-radation products. In the region of interest, which corresponded to a rounded section in the solid portions of metastases, were measured quantitative indicators of signal intensity on T2-weighted images (WI), native and post-contrast T1-WI, the degree of accumulation of MR contrast agent (MRCA) and its washout, the value of apparent diffusion coefficient (ADC). A total of 171 lesions were assessed. The data were compared in the varying localization of the primary NET groups of patients.Results: The study demonstrated that the solid portion of the gut mNETs compared with that of the pancreatic mNETs are characterized by lower ADC-value (p = 0.0102, medians: pancreatic mNETs — 1036 × 10–3 mm2/s, gut mNETs — 846 × 10–3 mm2/s), less active accumulation of MRCA on the arterial (p = 0.0002, medians: pan-creatic mNETs — 1.48, gut mNETs — 1.24) and venous (p = 0.0026, median: pancreatic mNETs — 2.22, gut mNETs — 1.9) phases of contrast enhancement, longer washout of MRCA (p = 0.0057, median: pan-creas mNETs — 0.92, gut mNETs — 0.98). Based on regression-factor analysis, a model for determining the localization of primary tumors based on MRI signs of liver mNETs was created with an accuracy of 93.8 %.Conclusion: Gut mNETs compared with that of the pancreatic mNETs are characterized by lower ADC-value, less active accumulation and longer washout of MRCA. The data can be used to draw up a personalized examination plan of patient with liver mNETs from the unknown primary.