Abstract Introduction: Management of triple negative breast cancer (TNBC) is still a major challenge: a significant proportion patients relapse despite adjuvant chemotherapy, and the median survival from relapse rarely exceeds 1 year. Angiogenesis activation as well as impairment in mechanisms of DNA repair are frequently described in sporadic TNBCs, but the role of angiogenesis inhibitors and platinum salts is still controversial. This is a phase II trial of preoperative carboplatin-paclitaxel in combination with bevacizumab in TNBC patients with previously untreated, stage II-III disease. The primary aim of the study is the pathologic complete response (pCR) rate. Among the secondary aims: safety, breast conserving surgery rate, early response assessment with dynamic contrast-enhanced Magnetic Resonance Imaging (DCE-MRI), biomarker analyses. Methods: Patients with hormone receptor negative (ER and PgR <1%), HER2 negative stage II-III breast cancer are eligible. At baseline, patients undergo breast DCE-MRI, followed by a single dose of bevacizumab 5 mg/kg (day -7). DCE-MRI is repeated prior to start chemotherapy (CT). On day 1, patients start CT with paclitaxel 80 mg/sqm+carboplatin AUC2 on days 1,8, 15 q 28, combined with bevacizumab 10 mg/kg on days 1 and 15. Patients received 5 preoperative cycles and undergo surgery within 6 weeks from the last CT course. Bevacizumab is omitted on cycle 5, to minimize the risk of surgical complications. Tissue samples are centralized to evaluate treatment effect on several tissue markers, as well as to identify potential predictors of treatment benefit. The sample size is estimated with the two-stage Simon's design, on the basis of a 40% expected pCR rate. A total of 43 patients are required, with the combination deemed worthwhile in case of at least 13 pCRs. Results: the accrual of the 44 planned patients is completed. Patients characteristics were as follows: median age 46 yrs (29-74); clinical stage IIA: 38.2%, IIB: 35.3%, III: 26.5%; ductal histology: 89%; histologic grade 3: 95%. Notably, 59% of the patients had clinically involved axillary nodes. At present, 33 patients are assessable for toxicity. Bevacizumab-associated adverse events (AEs) were mild: grade 1-2 hypertension occurred in 3 patients (9%) and grade 1 bleeding in 9 patients (27%). No grade 4 non-hematologic AEs were recorded; Grade 3 AEs were: liver function tests abnormalities in 3 patients, diarrhea in 3 patients, and neuropathy in 1 patient. Thirty-one patients underwent surgery, 16 patients (51%) received breast conserving surgery. A pCR in breast and axillary lymph-nodes was achieved in 16 patients (51%); 25 patients (81%) had negative axillary nodes (yN0). Conclusions: No unexpected toxicity was observed by combining bevacizumab to neoadjuvant platinum-taxane based CT. This combination is active in TNBCs, and the rate of pCR is in the expected range. Moreover, 80% of the patients have nodal negativity at surgery, and these results are of particular interest taking into account the high proportion of patients with clinically involved nodes at the time of diagnosis. The final results along with biomarkers data and early response assessment with DCE-MRI will be available at the time of the meeting. Citation Information: Cancer Res 2013;73(24 Suppl): Abstract nr P4-14-03.