Feed

Ki-67 index and mitotic count form the basis of grading of small intestinal neuroendocrine tumours (siNET). We hypothesized that the mitosis-specific marker phosphohistone H3 (PHH3) might better correlate with cancer-specific survival (CSS) and with response to treatment. We evaluated the association between Ki-67 index, PHH3-estimated mitotic count, and survival outcomes in a retrospective cohort of 73 consecutive patients with metastatic siNET. Additionally, we estimated the optimal cut-off for PHH3 and cross-validated the outcome. Both markers adequately distinguished CCS when comparing lower and higher proliferation groups (Ki-67: 128 vs. 95 m; PHH3: 149 vs. 88 m). They were strongly associated with CSS as continuous (HR 1.18 [1.08-1.28] and 1.16 [1.09-1.25]), and dichotomous variables (HR 2.96 [1.31-6.67] and 3.11 [1.50-6.46]). The Cox model based on PHH3 displayed slightly better optimism-corrected Harrell's c-index (0.71 vs. 0.68) and Akaike information criterion (219 vs. 223). Additionally, PHH3 showed significant association with PFS after treatment with somatostatin analogues (HR 1.12 [1.03-1.21]), and borderline significant association with PFS after treatment with peptide receptor radionuclide therapy (HR 1.11 [1.00-1.24]). A cut-off of >2 mitoses per 10 high-power fields estimated by PHH3 seemed to have better discrimination power compared to the standard WHO cut-off (<2). Mitotic count based on PHH3 is associated with CSS and with PFS after treatment with first-line SSA and possibly with PRRT for metastatic siNET. It may be an alternative to Ki-67 for estimation of proliferation and grading. A cut-off of >2 mitoses per 10 HPF might better distinguish G1 and G2 tumours.