Translating Breast Cancer Immune Features into Biomarkers and Therapies

Abstract

Breast cancer remains one of the main causes of cancer-related deaths in women worldwide. In the past years, advances in breast cancer treatment have been made, namely with the introduction of preoperative neoadjuvant chemotherapy (NACT) in selected cases of inflammatory/inoperable or advanced tumors (size larger than 2 cm and/or disease extension to the axillary lymph node). This treatment is effective in reducing the size of the primary tumor, allowing breast conservation. However, less than half of the patients achieve a pathological complete response and residual disease after NACT is a strong predictor of relapse. Hence, it is essential to find a suitable marker of response to this treatment, to promptly direct NACT non-responder patients to alternative therapies. Tumor infiltrating lymphocytes (TILs), specifically CD8+ cytotoxic T cells (CTLs) have been appointed as biomarkers of response. However, the clinical usefulness of this biomarker is still controversial and their evaluation is not yet implemented. This controversy could be explained by the fact that tumor cells can escape the immune system by releasing cytokines or expressing immune checkpoint inhibitors, dampening CTLs activity. HLA-DR, a T cell activation marker, could be a more reliable biomarker of response to NACT, than the presence of CTLs per se, since it may reflect the overall immune status of the tumor microenvironment and their functionality. To get more insights into the immune component of the tumor microenvironment, fresh biopsies, surgical specimens and blood were collected from breast cancer patients. In order to assess the differences between breast cancer aggressiveness, the patients were divided in two groups: the ones that have metastasis in the axillary lymph node and the ones that don’t have. Although the immunophenotype was similar in both groups of patients, a significant difference was observed for the expression level of HLA-DR in CTLs and regulatory T cells (Tregs). Indeed, patients without axillary lymph node metastasis had higher level of HLA-DR in CTLs and lower in Tregs when compared to patients with axillary lymph node metastasis. Given this result, we wondered if this immune trait could be used to predict response to NACT. In biopsies of two independent cohorts of breast cancer patients selected for NACT, we observed that high HLA-DR expression level in CTLs was strongly correlated with good response to NACT, with a high sensitivity (94.12% and 80% in cohort 1 and 2, respectively) and specificity (100% and 85.71% in cohort 1 and 2, respectively). Therefore we propose that HLADR expression level in CTLs above a threshold value, calculated by a ROC curve, would identify patients that would be responders to NACT. Additionally, by multivariable analysis, we noted that HLA-DR expression in CTLs was an independent predictor of response to NACT. Interestingly, HLA-DR expression in CTLs also have the likelihood of being a prognostic marker of breast cancer patients’ outcome, since a progression-free survival analysis revealed that patients with low levels of HLA-DR in CTLs tend to relapse sooner.Moreover, we observed that this immune feature was systemically reflected. Indeed, with the assessment of HLA-DR expression level in circulating CTLs we could differentiate breast cancer aggressiveness and even response to NACT, although in a less striking manner when compared to the analysis of HLA-DR expression level in intratumor CTLs. To characterize HLA-DR+ CTLs, we performed a gene expression as well as a broad surface markers’ expression analysis. When comparing to HLA-DR negative CTLs, HLA-DR+ CTLs had higher expression of cytotoxicity-related molecules Granzyme B, IFN-, Perforin, TNF-, Eomes and lower expression of Tbet. Additionally, they had a high proliferative capacity (high Ki67 levels) and an intermediate level of exhaustion markers, namely PD-1, Tim3 and CD127. These results suggest that HLA-DR+ CTLs had a phenotype closer to effector memory T cells (TEM), which have the capacity to home to tissues and to recirculate, rapidly release effector molecules and differentiate in effector CTLs. Then, we developed a 3D culture platform to shed some light on the functionality of these HLA-DR+ CTLs and their contribution for NACT success. Namely, we used two different breast cancer cell lines (MCF-7 and MDA-MB-231) and allowed them to spontaneously form spheroids. Peripheral blood mononuclear cells (PBMCs) isolated from NACT-responders and non-responders were added to the culture, infiltrating the 3D tumor-like structure. Interestingly, the PBMCs from NACT-responders were able by themselves to reduce the viability of the breast cancer cell line; whereas the PBMCs from NACT non-responders showed no effect. This emphasizes that the immune cells from NACT-responders are activated and have cytotoxic capacities; on the opposite, the immune cells from NACT non-responders are immunosuppressed and cannot exert their cytotoxic function. Furthermore, we added doxorubicin (a NACT agent) to the 3D co-culture and observed that NACT-responders’ PBMCs had a synergistic effect with this drug in decreasing the tumor cells’ viability. On the contrary, the addition of doxorubicin and NACT non-responders’ PBMCs did not alter the viability of the tumor cells. With these in vitro assays we validated the clinical observations. To confirm that the anti-tumor activity of NACT-responders’ PBMCs were indeed due to HLADR+ CTLs, which are more abundant in the blood of these patients than in the blood of NACT non-responders, we took advantage of the 3D system implemented. In fact, sorted HLA-DR+ CTLs, but not HLA-DR negative CTLs, reduce the viability of MCF-7, attesting the cytotoxic capacity and anti-tumor properties of HLA-DR+ CTLs. To investigate if it would be possible to revert the immunosuppressed phenotype of NACT non-responders’ CTLs we stimulated their PBMCs with PMA/ionomycin or by T cell receptor (TCR) engagement. Notably, both stimuli raised the HLA-DR levels in CTLs in NACT nonresponders’ PBMCs allowing these cells to reduce the viability of the breast cancer cells, similarly to NACT-responders’ PBMCs. The observed effect was even more striking with the addition of doxorubicin. This result opens the possibility to develop new treatments for breast cancer patients, whose tumors are not susceptible to NACT regimens alone.