Comparison of methods for proliferative index analysis for grading pancreatic well-differentiated neuroendocrine tumors.

Assessment of proliferative activity is required for grading well-differentiated pancreatic neuroendocrine tumors. However, a standardized method for obtaining the Ki-67 proliferative index is lacking. This study compared proliferative activity obtained by 3 methods: single-field hot spot (Ki-67 HS) and 10 consecutive field average (Ki-67 CFA) using the Ventana image analysis system (Ventana Medical Systems, Tucson, AZ) and mitotic index (MI). These methods resulted in discrepant grades in 30 (67%) of our 45 cases. With the current Ki-67 cutoff of more than 2% for intermediate-grade tumors, MI, CFA, and HS resulted in specificities of 91%, 94%, and 31%, respectively, for detecting metastasis, with positive predictive values (PPVs) of 25%, 67%, and 31%, respectively. At a higher Ki-67 cutoff of 7.5%, HS analysis resulted in a specificity of 94% and PPV of 71% for predicting metastasis. While single-field HS analysis may be practical and reliable at a higher cutoff, this study emphasizes the variability that can exist when different methods of assessment are used.