Abstract
BACKGROUND: Established measurements of proliferation in breast cancer are Ki67 and mitotic-activity-index (MAI), with problems in reproducibility and prognostic accuracy. Phosphohistone H3 (PHH3), a relatively novel IHC marker is specific for mitosis with good reproducibility. We hypothesized that PHH3 would be more reproducible and better represent proliferation than Ki67. RESULTS: PHH3 identified easily-missed mitosis by MAI, as demonstrated by upgrading M grade at diagnosis (n = 29/218, evenly distributed). PHH3 accurately found hot-spots, supported by mitotic count agreement between low-power and 10HPFs (R(2) = 0.999; P = 0.001). PHH3 was more reproducible than Ki67, measured by five-rater inter-class correlation coefficient (0.904 > 0.712; P = 0.008). Finally, despite a relatively short follow-up (median 46 months; 7 recurrences) PHH3 was the only variable correlated with disease-free survival (P = 0.043), while all other conventional clinicopathologic variables, including Ki67 (P = 0.356), did not. MATERIALS AND METHODS: We compared Ki67 and PHH3 for 218 breast cancer surgical cases diagnosed from 2012 to 2013 at Severance hospital. The most representative invasive breast cancer surgical slides were immunohistochemically stained for Ki67 and PHH3. CONCLUSIONS: Poor reproducibility and inadequate representation of proliferation of Ki67 and MAI may be improved by PHH3, allowing better accuracy in breast cancer diagnostics.