Claire E Murphy1, Kinsey A McCormick, Veena Shankaran, Deepti M Reddi, Paul E Swanson, Melissa P Upton, Antonios Papanicolau-Sengos, Sara Khor, Maria Westerhoff. 1. From the *Department of Pathology and †Division of Medical Oncology, University of Washington School of Medicine; and ‡Division of Public Health Sciences and §Clinical Research Division, Fred Hutchinson Cancer Research Center, Seattle, WA; ∥Department of Pathology, Roswell Park Cancer Institute, Buffalo, NY; and ¶Surgical Outcomes Research Center, Department of Surgery, University of Washington, Seattle, WA.
Abstract
OBJECTIVES: The aim of this study was to evaluate the concordance in grade assignment for gastroenteropancreatic neuroendocrine tumors using mitotic count (MC), Ki-67 proliferative index (KPI), and phosphohistone H3 count (PHH3C). METHODS: Resected gastroenteropancreatic neuroendocrine tumors were graded based on MC, KPI, and PHH3C. Concordance was determined using a weighted κ statistic. Median survival across each grade category was determined using Kaplan-Meier methods. RESULTS: Of the 110 patients, the majority had gastrointestinal primaries and grade 1 or 2 tumors. Rates of discordance in grade assignment were 29% of cases for KPI versus MC (κW = 0.26), 32% for PHH3C versus MC (κW = 0.34), and 32% for PHH3C versus KPI (κW = 0.37). There was fair agreement between grading by KPI and MC. Relative to grade by KPI and MC, PHH3C tended to upgrade tumors. The proportion alive at 3 and 5 years was not significantly different for patients with grade 1 versus grade 2 tumors. CONCLUSIONS: The concordance between KPI and MC was fair. Phosphohistone H3 count tended to upgrade tumors using the cutoffs established by MC. Grade 1 and grade 2 tumors were associated with similar survival regardless of grading method. The overall relevance of the current cutoff values used in grading neuroendocrine tumors may need to be revisited.
OBJECTIVES: The aim of this study was to evaluate the concordance in grade assignment for gastroenteropancreatic neuroendocrine tumors using mitotic count (MC), Ki-67 proliferative index (KPI), and phosphohistone H3 count (PHH3C). METHODS: Resected gastroenteropancreatic neuroendocrine tumors were graded based on MC, KPI, and PHH3C. Concordance was determined using a weighted κ statistic. Median survival across each grade category was determined using Kaplan-Meier methods. RESULTS: Of the 110 patients, the majority had gastrointestinal primaries and grade 1 or 2 tumors. Rates of discordance in grade assignment were 29% of cases for KPI versus MC (κW = 0.26), 32% for PHH3C versus MC (κW = 0.34), and 32% for PHH3C versus KPI (κW = 0.37). There was fair agreement between grading by KPI and MC. Relative to grade by KPI and MC, PHH3C tended to upgrade tumors. The proportion alive at 3 and 5 years was not significantly different for patients with grade 1 versus grade 2 tumors. CONCLUSIONS: The concordance between KPI and MC was fair. Phosphohistone H3 count tended to upgrade tumors using the cutoffs established by MC. Grade 1 and grade 2 tumors were associated with similar survival regardless of grading method. The overall relevance of the current cutoff values used in grading neuroendocrine tumors may need to be revisited.
Authors: Mohammad Y Zaidi; Alexandra G Lopez-Aguiar; Jeffrey M Switchenko; Joseph Lipscomb; Valentina Andreasi; Stefano Partelli; Adriana C Gamboa; Rachel M Lee; George A Poultsides; Mary Dillhoff; Flavio G Rocha; Kamran Idrees; Clifford S Cho; Sharon M Weber; Ryan C Fields; Charles A Staley; Massimo Falconi; Shishir K Maithel Journal: Ann Surg Date: 2019-09 Impact factor: 12.969