Abstract
Background/Aim: The neuroendocrine tumor (NET) proliferation-based grading system (ENETS/WHO) for gastroenteropancreatic (GEP) tumors has proved reliable for prognostic stratification. To date, concerns exist regarding Ki-67 heterogeneity within the tumor and little is known on whether grade varies between primary and secondary sites. As tumor heterogeneity may have a significant impact on clinical management, our aim was to retrospectively evaluate Ki-67 on a series of GEP NETs in order to establish whether there is variability in different samples of the same lesion or between primary and metastatic disease (local/distant, synchronous/metachronous). Methods: Sixty patients with multiple samples of tumor were accrued from a total of 338 GEP NETs; 44 of them also had tissue from local/distant metastases and a further 5 had multiple metastatic foci from unknown primary tumors. Immunohistochemistry for Ki-67 was performed on all paraffin blocks from both primary and metastatic tumors. Results: Intratumor Ki-67 heterogeneity sufficient to change grade at first diagnosis was seen in 3/60 cases (5%). Out of 49 patients with primary NETs and/or multiple metastases, discrepancy in grade between sites was identified in 19 (39%) cases and in particular in 11/47 (23%) and in 10/12 (83%) patients with synchronous and metachronous metastases, respectively (p = 0.0002). Change in grade was more frequent in distant compared to locoregional metastases (p = 0.024) and in particular in distant sites other than the liver (p = 0.006). Conclusions: NETs show frequent differences in grade between primary sites and their synchronous/metachronous metastases; assessment of Ki-67 at all sites may prove to be significant for patient management.