Dynamic evaluation of the immune infiltrate and immune function genes as predictive markers for neoadjuvant chemotherapy in hormone receptor positive, HER2 negative breast cancer

• Main Authors: Alexios Matikas, John Lövrot, Anna Ramberg, Margareta Eriksson, Therese Lindsten, Tobias Lekberg, Ingrid Hedenfalk, Niklas Loman, Jonas Bergh, Thomas Hatschek, Ann Erlandsson, Theodoros Foukakis
• Format: Article
• Language: English
• Published: Taylor & Francis Group 2018-09-01
• Series: OncoImmunology
• Subjects: biomarker; breast cancer; gene expression; predictive; tumor infiltrating lymphocytes
• Online Access: http://dx.doi.org/10.1080/2162402X.2018.1466017

Gene expression (GE) signatures and Tumor Infiltrating Lymphocytes (TIL) enumeration are predictive for response to neoadjuvant chemotherapy in HR- and in HER2+ breast cancer, but data are conflicting in HR+/HER2- disease. This study aimed to explore their predictive value in this subset, measured both at baseline and after short exposure to chemotherapy. Specifically, the PROMIX phase 2 trial enrolled patients with locally advanced HER2- BC to receive six cycles of epirubicin and docetaxel, plus bevacizumab during cycles 3–6. Patients underwent tumor biopsies at baseline and after cycle 2 for GE profiling and enumeration of TIL, FOXP3+ T-cells and CD163+ macrophages. An immune related gene module and the quantification of the immune infiltrate were analyzed for association with pathologic complete response (pCR), decrease in tumor size and disease-free survival (DFS). Of the 150 patients enrolled in PROMIX, 113 were HR+/HER2-. Baseline GE and immune cell enumeration data were available from 71 patients, while data after 2 cycles of chemotherapy were available from 41. At baseline, only GE was statistically significantly associated with higher pCR rates (OR 2.29, 95% CI 1.05 – 5.38, p = 0.037) and decrease in tumor size (r = 0.25, p = 0.047). In contrast, longitudinal data indicate that both GE (r = 0.54, p<0.001) and TIL abundance (p = 0.009) are stronger predictors for the reduction of tumor size, while low FOXP3+ was statistically significantly associated with an improved DFS (p = 0.027). In conclusion, GE analysis, TIL and FOXP3+ enumeration after short-term exposure to chemotherapy carry important predictive information in HR+/HER2- breast cancer at the neoadjuvant setting.