Modern pathology : an official journal of the United States and Canadian Academy of Pathology, Inc
-
Comparative Study
Calculation of the Ki67 index in pancreatic neuroendocrine tumors: a comparative analysis of four counting methodologies.
Ki67 index is now an essential part of classification of pancreatic neuroendocrine tumors. However, its adaptation into daily practice has been fraught with challenges related to counting methodology. In this study, three reviewers used four counting methodologies to calculate Ki67 index in 68 well-differentiated pancreatic neuroendocrine tumors: (1) 'eye-ball' estimation, which has been advocated as reliable and is widely used; (2) automated counting by image analyzer; (3) manual eye-counting (eye under a microscope without a grid); and (4) manual count of camera-captured/printed image. ⋯ The results showed that: (1) 'eye-balling' was least expensive and fastest (average time <1 min) but had poor reliability and reproducibility. (2) Automated count was the most expensive and least practical with major impact on turnaround time (limited by machine and personnel accessibility), and, more importantly, had inaccuracies in overcounting unwanted material. (3) Manual eye count had no additional cost, averaged 6 min, but proved impractical and poorly reproducible. (4) Camera-captured/printed image was most reliable, had highest reproducibility, but took longer than 'eye-balling'. In conclusion, based on its comparatively low cost/benefit ratio and reproducibility, camera-captured/printed image appears to be the most practical for calculating Ki67 index. Although automated counting is generally advertised as the gold standard for index calculation, in this study it was not as accurate or cost-effective as camera-captured/printed image and was highly operator-dependent. 'Eye-balling' produces highly inaccurate and unreliable results, and is not recommended for routine use.
-
We assessed the frequency and clinicopathologic significance of 19 genes currently identified as significantly mutated in myeloid neoplasms, RUNX1, ASXL1, TET2, CEBPA, IDH1, IDH2, DNMT3A, FLT3, NPM1, TP53, NRAS, EZH2, CBL, U2AF1, SF3B1, SRSF2, JAK2, CSF3R, and SETBP1, across 93 cases of acute myeloid leukemia (AML) using capture target enrichment and next-generation sequencing. Of these cases, 79% showed at least one nonsynonymous mutation, and cases of AML with recurrent genetic abnormalities showed a lower frequency of mutations versus AML with myelodysplasia-related changes (P<0.001). Mutational analysis further demonstrated that TP53 mutations are associated with complex karyotype AML, whereas ASXL1 and U2AF1 mutations are associated with AML with myelodysplasia-related changes. ⋯ Univariate analysis demonstrated that U2AF1 and TP53 mutations are associated with absence of clinical remission, poor overall survival (OS), and poor disease-free survival (DFS; P<0.0001), whereas TET2 and ASXL1 mutations are associated with poor OS (P<0.03). In multivariate analysis, U2AF1 and TP53 mutations retained independent prognostic significance in OS and DFS, respectively. Our results demonstrate unique relationships between mutations in AML, clinicopathologic prognosis, subtype categorization, and morphologic dysplasia.