Is it possible to predict the efficacy of a combination of Panitumumab plus FEC 100 followed by docetaxel (T) for patients with triple negative breast cancer (TNBC)? Final biomarker results from a phase II neoadjuvant trial.

Background: TNBC is an heterogeneous group of tumors for some of which the Epithelial Growth Factor Receptor pathway (EGFR) may play an important role. We evaluated the efficacy and toxicity of an anti-EGFR antibody (panitumumab) combined with a standard neoadjuvant chemotherapy in order to identify predictive biomarkers of efficacy and target biologically defined subpopulations for potential further development. Methods: Sixty patients with stage II-IIIA disease were prospectively included in this multicentre neoadjuvant study. Systemic therapy (ST) consisted of panitumumab (9 mg/kg q.3 weeks x8) combined with FEC 100 (500/100/500 mg/m² q.3 weeks x4) followed by 4 cycles of T (100 mg/m² q.3weeks x4). All patients underwent surgery at completion of ST. Paraffin-embedded and frozen samples were systematically collected before and after ST for biologic studies. Patients characteristics are as follows: mean age 50 [27-72]; median tumor size: 40 mm [20-120]; invasive ductal carcinoma: 96.7%; Scarff-Bloom-Richardson Grade III: 71.7%, grade II: 28.3%. Complete pathological response (pCR) rate was 52.3% [95% IC: 37.3-67.5] (Sataloff's classification) and 46.7% [95% IC: 31.6-61.4](Chevallier's classification). Conservative surgery was performed in 88% of cases. Skin toxicity was the main side-effect: Cutaneous toxicity grade IV: 5%, grade III: 30%, grade II: 20%. Neutropenia grade IV: 27%; febrile neutropenia: 5%. Infection: 0%. Hand- foot syndrome grade III: 3.3%. Ungueal toxicity grade III: 1.6%, grade II: 20%. Results: We performed a ROC curve to identify the best cut-off value for KI-67, EGFR, cytokeratin 5-6 and p53 in order to predict a pCR. Tumors with more than 40% of positive cells for KI-67 and tumors with a score for EGFR greater than 70 tend to be associated with pCR according to Chevallier's classification (p = 0.06). No predictive value was identified for Cytokeratin 5-6 and p53 (p = 0.61 and p = 0.27, respectively). Immunohistochemistry results show that two thirds of tumors have more than 40% of positive cells for KI-67 and that two thirds of tumors present a score for EGFR greater than 70. About half of the tumors express cytokeratin 5-6 and p53 (cut off: 1%). Chi-squared tests were performed to assess relations between cutaneous toxicities and pCR. The cutaneous toxicities were not predictive of pCR (p = 0.94) and no correlations were found with KI-67, EGFR, Cytokeratin 5-6 and p53. In terms of BRCA1 and BRCA2 status, 35 tumors were analysed so far: BRCA1: 6 mutations (17%); BRCA2 (30 patients): 1 mutation (3.3%). Conclusions: These results suggest the possibility to identify a subpopulation with high probability of pCR (KI-67 > 40%, EGFR score > 70). Further biological studies are ongoing and will be presented at the meeting, including EGFR polymorphisms, C-met, ALDH1, pCadherine and PTEN. This will help us further define subpopulations of TNBC patients, potential targets for antiEGFR development. [ABSTRACT FROM AUTHOR]