Abstract No. 
San Antonio Breast Cancer Symposium
December 6-10

Lobular Breast Cancer and NAC: Combined Results from the NKI and I-SPY 1 Trial

Mukhtar RA, Lips E, Wesseling J, Livasy C, Yau C, Berry D, Veer Lvan't, Carey LA, Esserman LJ, Rodenhuis S, Hwang ES

Background The benefits of neoadjuvant chemotherapy (NAC) in breast cancer are twofold: allowing breast conservation surgery, and assessment of response. Having pathologic complete response (pCR) predicts improved outcomes. Many have suggested that invasive lobular cancers (ILC) do not respond to NAC as well as invasive ductal cancers (IDC), and recommend against offering NAC in such cases. We previously investigated differences in the response to NAC in ILC and IDC in the I-SPY 1 Trial, and now present findings from a joint analysis with the Netherlands Cancer Institute (NKI). 

Methods : We combined datasets comprising 676 patients (221 from I-SPY, 455 from NKI) enrolled in NAC trials. Eligible patients had palpable tumors ≥ 3 cm, and underwent serial biopsies, microarray analysis (Agilent 44K for I-SPY and Illumina 6v3 for NKI), and MRI imaging. ILC versus IDC histology was assigned by pathologic appearance, and ILCs were tested for e-cadherin expression and centrally reviewed by a breast pathologist at each site. We performed multivariate logistic regression analyses to assess differences in pCR. We compared intrinsic subtypes, prognostic gene signature expression, and MRI phenotypes. 

Results : There were 75 ILC and 601 IDC cases. ILCs had lower risk features and were more likely hormone receptor + (Table 1) but there was substantial heterogeneity and many ILC had high risk features. There were fewer T2 (33% versus 48%) but more T3 tumors (59% versus 37%) among the ILCs. Nearly 10% of ILC cases had the basal intrinsic molecular subtype; 22% expressed the activated wound healing signature, and 56% had a high risk 70-gene prognostic signature. While pCR rate was lower in ILCs, adjustment for low risk features showed lobular histology was not an independent predictor of pCR. ILC was less likely to present as a solid mass. 85% of ILC were e-cadherin negative. ![Graphic][1]</img> 

Conclusions : This combined analysis suggests that the decreased response to NAC in ILC is due to the association between ILC and low risk features. ILC in a clinically high risk setting has a higher proportion of high risk biology in spite of lobular histology. Lobular histology alone should not be used to exclude patients from NAC trials or treatment. Citation Information: Cancer Res 2011;71(24 Suppl):Abstract nr P1-06-10. [1]: /embed/inline-graphic-1.gif

View original