PROLIFERATION AND APOPTOSIS BEFORE AND AFTER PRIMARY SYSTEMIC THERAPY FOR OPERABLE BREAST CANCER

Background: The identification of biological markers able to predict tumor sensitivity to primary systemic therapy (PST) may have an important clinical output. Moreover, the effect of treatment on these biomarkers might have prognostic relevance. Aims of this analysis: to evaluate the predictive role of proliferation (Ki-67), apoptotic index (AI) and growth index (GI) before PST, the effect of PST on these parameters and their relation with tumor response and patients outcome. Methods: Ki-67, AI (Tunel Test), and GI were assessed at baseline and at surgery in 90 stage II-IIIB breast cancer patients enrolled in a phase II randomized trial of PST with 4 cycles of epirubicin-paclitaxel +/- gefitinib. Pathologic response was defined as tumor regression grade (TRG) (criteria of Miller and Payne). Results: patients characteristics were as follows: median age, 51 yrs (range 29-69); stage IIA 41%, IIB-IIIA 59%. A significant correlation between Ki-67 and AI was found at baseline (p=0.001). A significantly higher Ki-67 and AI were observed in ER- vs ER+ tumors (median Ki-67: 62% vs 25%, p<0.0001; median AI: 1.33 vs 0.85, p=0.009). A significantly higher Ki-67 and GI were observed in EGFR+ vs EGFR- tumors (median Ki-67: 60% vs 25%, p=0.0006; median GI: 48 vs 30.7, p=0.03). No significant differences were observed according to HER2 expression. After PST, 84% of pts was classified as having a TRG 4 or 5, and 16% as having a TRG 1-3. No significant correlation among baseline Ki-67, AI and GI and TRG was observed. After PST, Ki 67 and AI were significantly reduced from baseline. Lower Ki 67 and lower AI were significantly correlated with higher TRG (p=0.008 and 0.01 respectively). We observed 1 cancer death and 5 relapses only, therefore correlation with patient outcome has been planned for the time of the meeting Conclusions: ER- and EGFR+ tumors show more aggressive behaviour. PST significantly reduces Ki-67 and AI, and this reduction is significantly correlated with TRG. In particular, the reduced AI at surgery was not found to be a marker of resistance, more probably reflecting the reduced Ki-67 of responding tumors.