Endoscopic Ultrasound-Guided Fine Needle Aspiration of Pancreatic Neuroendocrine Tumors: Is Accurate Grading Based on the 2010 ENTS/WHO Criteria Possible on Cytologic Specimens?
Jessica M Farrell, Grace E Kim, Laura Tabatabai. UCSF, San Francisco, CA
Background: The natural history of pancreatic neuroendocrine tumors (panNETs) is highly variable, and one of the most controversial aspects in the diagnosis of these tumors is prediction of their clinical behavior. Some panNETs that behave in a malignant fashion may have deceptively bland cytologic features and, conversely, some with previously described "malignant" criteria may not always prove lethal. Various classification schemes have been proposed for panNETs, the most recent of which includes assessment of mitotic count and/or Ki-67 labeling index (WHO 2010 and European Neuroendocrine Tumor Society proposals). We undertook this study to determine if FNA cytology samples can be accurately graded comparable to histologic material based on the 2010 ENTS/WHO criteria.
Design: The archival files were retrospectively searched for panNET cases diagnosed on endoscopic-ultrasound guided fine needle aspiration (EUS-FNA). Those cases with adequate cell block and correlating histological material were retrieved for inclusion in the study. Paraffin-embedded formalin-fixed cell blocks and representative tissue blocks from the surgical specimens were immunostained with Ki-67 using standard techniques. The cytological samples were assessed for positive staining in tumor cells, and the proliferative index was calculated by dividing the number of positive tumor cells by the total number of tumor cells present in the cell block. The correlating histological material was graded in a blinded fashion using the ENTS guidelines, and the results compared to those derived from the cytologic evaluation.
Results: Ten out of 11 cases (91%) had equivalent grading scores when comparing Ki-67 proliferation rates in cell block material to histological material. Of these, 5 of 11 cases were interpreted as grade 1 (Ki-67 index ≤2%), and another 5 of 11 cases were interpreted as grade 2 (Ki-67 index >2-20%). One case was interpreted as grade 2 on cell block material and grade 1 on histological evaluation.
Conclusions: Although larger studies with inclusion of high grade tumors are needed for more accurate validation, the results of this study indicate there is reliable correlation of tumor grade based on ENTS criteria when comparing EUS-FNA samples to histological material. Routine calculation of the proliferation index and subsequent grading of panNETs on cytologic material can be a useful tool in predicting the clinical behavior of these tumors and to guide further treatment and patient management.
Tuesday, March 20, 2012 8:30 AM
Platform Session: Section F, Tuesday Morning