Abstract S2-05: Event-free and overall survival following neoadjuvant weekly paclitaxel and dose-dense AC +/- carboplatin and/or bevacizumab in triple-negative breast cancer: outcomes from CALGB 40603 (Alliance)


Aurora St. Luke's Medical Center


BACKGROUND: CALGB (Alliance) 40603 measured the effects of adding carboplatin (Cb) and/or bevacizumab (Bev) to standard neoadjuvant chemotherapy (weekly paclitaxel x 12 then doxorubicin/cyclophosphamide every 2 weeks x 4) on pathologic complete response (pCR) rates in stage II-III triple-negative breast cancer (TNBC). As previously reported (Sikov et al, JCO 2015), pCR breast (ypT0/is) and pCR breast/axilla (pCR Br/Ax) (ypT0/isN0) rates increased from 46% to 60% and 41% to 54%, respectively, with Cb and from 48% to 59% and 44% to 52%, respectively, with Bev. Secondary endpoints included event-free survival (EFS) and overall survival (OS).

METHODS: EFS is measured from study entry to ipsilateral invasive breast or locoregional recurrence, distant recurrence or death from any cause and OS from study entry to death from any cause in all patients (pts) who started study treatment. Pts without an event were censored as of their last clinical assessment. Hazard ratios (HR) were calculated for pts who achieved pCR vs. not and for pts assigned to receive drug (Cb or Bev) vs. not. All p-values are 2-sided.

RESULTS: 443 pts started study treatment. Median follow-up was 39 months (range 28-66). 110 EFS events and 77 deaths have been reported. At 3 yrs, overall EFS was 74.1% and OS 83.2%. Pts who achieved pCR breast had 3-yr EFS of 84.8% vs. 61.8% for those who did not. Table 1 shows the association between pCR and pCR or minimal residual invasive disease (Residual Cancer Burden Class I (RCB I), Symmans et al, JCO 2007) and outcomes; p-values for all comparisons <0.0001:

Table 1 pCR Breast pCR Br/Ax pCR Br/Ax or RCB I Yes/No N (%) 231 (52%)/212 (48%) 207 (47%)/236 (53%) 266 (60%)/177 (40%) EFS-HR 0.33 (0.22-0.50) 0.30 (0.19-0.46) 0.29 (0.20-0.43) OS-HR 0.28 (0.17-0.46) 0.20 (0.11-0.36) 0.21 (0.13-0.34)

Pts assigned to Cb vs. not had 3-yr EFS 76.5% vs. 71.6% and OS 81.9% vs. 84.6%. Pts assigned to Bev vs. not had 3-yr EFS 75.5% vs. 72.9% and OS 85.5% vs. 80.9%. Table 2 shows HRs by assigned treatment:

Table 2 Cb Bev N (Yes/No) 225/218 222/221 EFS - HR 0.84 (0.58-1.22) p=0.36 0.80 (0.55-1.17) p=0.25 OS - HR 1.15 (0.74-1.79) p=0.53 0.76 (0.49-1.19) p=0.23

CONCLUSIONS: Pts with TNBC who achieved pCR with study treatment had significantly better EFS and OS than pts who did not, consistent with findings from a published meta-analysis (Cortazar et al, Lancet 2014); the addition of RCB I did not weaken this association. Our study was not powered to assess the impact of Cb or Bev on these endpoints. While our findings are consistent with predictions from the meta-analysis as to the impact of raising the pCR rate on EFS (Berry-Hudis, JAMA Oncology 2015), the wide confidence intervals illustrate the challenge of conclusively demonstrating a correlation between pCR increment and EFS benefit, especially as the control pCR rate rises. While the addition of Bev has failed to improve long-term outcomes in TNBC in large randomized adjuvant trials, our results support ongoing and planned neoadjuvant and adjuvant studies designed to further assess the value of Cb-containing regimens in stage II-III TNBC.

Support: U10s - CA180821, CA180882, CA180820, CA076001, CA025224, CA180868, CA180888.

Document Type


Link to Full Text