2. |
- Matikas, A., et al.
(author)
-
Immune function and response to neoadjuvant chemotherapy in hormone receptor positive, HER2-negative breast cancer
- 2017
-
In: Annals of Oncology. - : Oxford University Press. - 0923-7534 .- 1569-8041. ; 28
-
Journal article (other academic/artistic)abstract
- BackgroundGene expression (GE) signatures and Tumor Infiltrating Lymphocyte (TILs) enumeration have shown promise as predictors of response to neoadjuvant chemotherapy in Hormone Receptor negative (HR-) and HER2+, but not in HR+/HER2- breast cancer (BC). This study aimed to explore their predictive value in HR+/HER2- BC, based on previous work from our group on the association of immune function and chemosensitivity in advanced HR+ BC.MethodsThe 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 using DNA microarrays and TIL enumeration according to standard guidelines. Since pathologic complete remission (pCR) is relatively rare in HR+ BC, we also associated an immune gene module score (IMS) and TIL counts with the non-dichotomous variable of decrease in tumor size.ResultsOf the 150 enrolled patients, n = 113 were HR+. For n = 71, both TIL and GE data were available at baseline, while for n = 78 and n = 49 patients longitudinal TIL and GE data at baseline and cycle 2 were available, respectively. At baseline, on both univariate (OR = 2.29, P = 0.037) and multivariate analysis (OR = 2.35, P = 0.044) IMS was associated with pCR, while its association with tumor shrinkage was only apparent on univariate (P = 0.047) and not multivariate analysis (P = 0.061). TIL infiltration >50% (n = 9) was associated with neither pCR (OR = 1.812, P = 0.61) nor tumor shrinkage (P = 0.99). However, decreases in TIL counts in cycle 2 compared with baseline were associated with lesser decreases in tumor size (P = 0.043 for univariate and P = 0.044 for multivariate analysis).
|
|