Impact of Tumor-Infiltrating Lymphocytes in Breast Cancer Patients with Neoadjuvant Chemotherapy Depends on Subtypes

Background: The aim of this study is to evaluate the relationship between TILs and response to neoadjuvant chemotherapy (NAC) as well as long-term outcome in breast cancer including luminal, triple negative (TN) and HER2-positive subtypes. Methods: Thirty-two patients receiving NAC from January 2010 to November 2015 were analyzed. Intratumoral lymphocytes (ItuLy) and stromal lymphocytes (StrLy) were evaluated in hematoxylin and eosin-stained sections from pretherapeutic core needle biopsies. Lymphocyte-predominant breast cancer (LPBC) was defined as tumors with >50% of ItuLy or StrLy. Results: A total of 32 patients was divided into 15 luminal (46.9%), 5 TN (15.6%) and 12 HER2-positive (37.5%) subtypes. A significant correlation between ItuLy (the median percentage: 5%) and StrLy (25%) was observed (Spearman's correlation coefficient= 0.647, P<0.01). LPBC comprised 28.1% (9/32) of the total population, with frequencies of 50.0%, 20.0% and 13.3% in HER2-positive, TN and luminal subtypes, respectively (P=0.112). LPBC group had a significantly higher pathological complete response (pCR) rate of 66.6% (6/9), compared with 4.3% (1/23) for non-LPBC group (P=0.001). LPBC showed the modest correlation with pCR in luminal (100% (2/2) vs. 0% (0/13), P=0.01) and HER2-positive subtypes (66.6% (4/6) vs. 16.7% (1/6), P=0.121). In TN subtype, neither LPBC nor non-LPBC showed pCR. After a median follow-up of 29.9 months, patients with any TILs were significantly associated with better disease-free survival (DFS) than those without TILs (24.3 vs. 23.6 months, P= 0.014). The most significant impact of TILs on DFS was observed in HER2-positive subtype (23.2 vs. 15.2 months, P=0.025), whereas TILs did not associate with DFS in luminal and TN subtypes. Conclusions: The presence of TILs tended to predict a response to NAC except in TN subtype and would provide valid information of prognosis in HER2-positive breast cancer.