Minimal apparent diffusion coefficient in predicting the Ki-67 proliferation index of pancreatic neuroendocrine tumors

Purpose To investigate the diagnostic performance of the minimal apparent diffusion coefficient (ADC min ) to distinguish between pancreatic neuroendocrine tumors (Pan-NETs) with low and high Ki-67 proliferation index values and to evaluate the relationship between ADC min and the Ki-67 proliferation index. Methods Pre-operative magnetic resonance imaging data and postoperative Ki-67 proliferation index data of 42 patients with primary neuroendocrine tumor of the pancreas from November 2014 to March 2021 were included in this retrospective study. According to the Ki-67 proliferation index value, Pan-NETs were divided into a high-expression group (Ki-67 ≥ 10%, n = 17) and low-expression group (Ki-67 < 10%, n = 25), and mean ADC (ADC mean ) and ADC min values were compared between groups using receiver operating characteristic (ROC) curves to evaluate the performance of ADC mean and ADC min in judging the expression level of Ki-67 proliferation index. The relationship between ADC min and the Ki-67 proliferation index was also evaluated. Results The ADC min was significantly higher in the low-expression group ( Z = − 3.537, p < 0.01). The area under the ROC curve (AUC) for ADC min was 0.825, which was higher than that for ADC mean (0.781). Using 1.32 × 10 –3 mm 2 /s as the optimal discriminating threshold, the sensitivity, specificity, accuracy, and positive and negative predictive values of the two groups were 80%, 88.2%, 83.3%, 90%, and 75%, respectively. The ADC min of Pan-NETs showed a significant negative correlation with the Ki-67 proliferation index ( r s = − 0.634, p < 0.001). Conclusion The ADC min is a potential imaging biomarker, which may be helpful for non-invasive preoperative prediction of the Ki-67 proliferation index of Pan-NETs and the subsequent planning of appropriate treatment.