Grading of pancreatic neuroendocrine neoplasms using pharmacokinetic parameters derived from dynamic...

The present study aimed to evaluate the diagnostic efficacy of pharmacokinetic parameters derived from dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) in prospective evaluation of pancreatic neuroendocrine neoplasms (pNENs) grading. A total of 25 histologically proven patients with pNENs (30 lesions in total) who underwent DCE-MRI were enrolled. Lesions were divided into G1, G2 neuroendocrine tumor (NET) and G3 NET/neuroendocrine carcinoma (NEC) groups based on their histological findings according to 2017 World Health Organization Neuroendocrine Tumor Classification Guideline. In addition, the same numbers of tumor-free regions were selected using as normal control group. For each group, pharmacokinetic DCE parameters: volume transfer constant (K^trans); contrast transfer rate constant (k_ep); extravascular extracellular space volume fraction (v_e); and plasma volume fraction (v_p) were calculated with Extended Tofts Linear model. Receiver operator characteristics analysis was conducted to assess the diagnostic efficacy of these parameters in pNENs grading. There were significant differences of K^trans, k_ep, v_e and v_p between tumor-free areas and G1, G2 NET (P<0.001). The K^trans and k_ep of G1 NET were significantly lower compared with those of G2 ones (P<0.005). The area under the curve of K^trans and k_ep in differentiating G2 from G1 NET were 0.767 and 0.846, respectively. When K^trans was >0.667 and k_ep >1.644, the sensitivity of diagnosing G2 NET was the lowest (53.85%), but the specificity was the highest (93.75%). When K^trans was >0.667 or k_ep >1.644, the sensitivity of diagnosing G2 NET was 92.31%, but the specificity was 75.00%. Pharmacokinetic parameters of DCE-MRI, particularly the quantitative values of K^trans and k_ep, are helpful for differentiating G2 NET from G1 ones.