In our analysis of 1396 cases of female luminal breast cancer patients received standardized treatment from China, high TILs level were related to poor prognostic clinicopathological characteristics (such as age at diagnosis, lymph node status, Ki-67 index and advanced pathological stage. Although the incidence of LPBC in LBC was only 5.7%, it was significantly associated with Ki-67 and hormone receptor status, and the incidence was significantly higher in luminal B subtype, reaching 12.7% of luminal B/HER2 + and 10.6% of luminal B/HER2-. For the clinical outcome analysis, the high level of TILs (> 10%) was associated with better prognosis in Luminal /HERb2- patients, at the same time TIL levels and clinical outcome might show different directions for luminal/Her2 + and luminal/Her2- breast cancers.
Honda et al. rencently study in general consistent with our conclusion ,they found that in Luminal/HER2- breast cancer high-grade sTIL level was significant correlation with high tumor cell proliferation rate, as well as high levels of Ki-67 expression and patients with an intermediate or high sTIL level had a better prognosis compared with a low sTIL [19]. In another study, high TILs were associated with better DDFS in a cohort of 987 patients with ER+/HER2- breast tumor which treated with adjuvant chemotherapy, higher Ki-67 expression retained significant association with TILs in multivariable regression analysis[20]. A cohort study including 148 ER + breast cancer patients showed FOXP3 + lymphocytes were negative prognostic factors and increased FOXP3 + infiltrate was associated with worse DFS and OS; however the enriched of FOXP3 + cells in ER + breast cancer identified high-risk patients and predicted the risk of relapse after 5 years[21]. In contrast, Zhang et al. found that high TILs levels tended to have a worse DFS for luminal A breast cancer[18]. The inconsistent results may be due to the different study populations, which were composed of premenopausal women. Moreover, the cut-off values for the risk stratification of TILs are quite different. We believe that these factors may be responsible for the different prognostic significance of our study from other related studies. In another study of 563 ER + tumors, while CD4 and FOXP3 were not significantly associated with prognosis, patients with tumors classified as CD8-high had increased risk of recurrence. Lymphocytic infiltrate (most pronounced for FOXP3 + cells) was more often present in tumors with increased PI3K downstream phosphorylation, suggesting PI3K pathway alterations might be associated with the composition of the tumor microenvironment[22]. The limitation of our study was that we did not investigate tumor-infiltrating lymphocyte subpopulations in luminal breast cancer, which may further explore prognostic implications.
Breast cancer has long been considered as less immunogenic solid tumor, LBC has fewer immune mutations and presence of immune heterogeneity compared to TNBC and HER2 + subtypes, and the effect of endocrine therapy is remarkable leading to immunotherapy has not received enough attention, but there are still some subtypes of LBC with endocrine resistance and recurrent metastasis at an early stage[23]. In the tumor microenvironment, tumor cells interact with tumor-infltrating lymphocytes, mammary epithelial cells themselves are manipulators of immune regulation, and their immune evasion mechanisms are not yet clear[24]. Atashgaran V et al. concluded that mammary gland development depends on dynamic interactions between hormonally regulated mammary epithelia and the immune microenvironment. Alternating roles of macrophages and regulatory T cells (Tregs) may affected menstrual cycle-associated breast cancer risk, particularly during the process of mammary gland regression[25]. Interestingly, in our study found that TILs level was significantly associated with age and Ki-67 index, postmenopausal women were significantly more numerous in the group with high TIL level compared to the low level group which suggested changes in hormone levels affect the composition of the tumor microenvironment, which may be one reason that high levels of TILs were associated with improved DFS in luminal breast cancer.
The latest research shows that adoptive cell therapy (ACT) holds promise in solid tumor immunotherapy, which was using of patient's own highly heterogeneous TILs for precision targeting of tumor cells[26]. And responses have been achieved in patients with metastatic epithelial cancers, including breast cancer, by selecting TIL reactive to the products of nonsynonymous tumor mutations[27, 28]. Immunotherapy using checkpoint inhibitors has been successfully applied to triple-negative breast cancer, however, the lack of detailed understanding of the mechanisms of immune escape from breast tumors and the treatment of patients with highly heterogeneous metastatic disease has led to suboptimal results[29]. Combination of TIL-ACT and anti-PD-1/PD-L1 therapy is a good solution, it may bring important implications for immunotherapy in breast cancer[30]. Data from a phase II clinical study have been reported recently, the study investigates the immunogenicity of somatic mutations in breast cancer and explores the therapeutic efficacy in a pilot trial of mutation-reactive tumor-infiltrating lymphocytes (TILs) with a short course of pembrolizumab in patients with metastatic breast cancer. Ultimately, of the six patients who participated in the treatment, objective tumor regression was noted in three LBC patients, including one complete response (ongoing over 5.5 years) and two partial responses (6 and 10 months)[31]. Adoptive transfer of TIL is a highly personalized experimental option for BC patients shown to be capable of mediating objective responses in current trial and deserves further study.
The mainly advantages of our study including multivariate analysis was used to control the confounding factors affecting outcomes, it is a retrospective cohort study with a large sample size including a comprehensive collection of clinicopathologic factors, treatment, and clinical outcome data from the Asian population. And considering the accuracy of TILs assessment we strictly evaluated stromal TILs according to the guidelines recommended by the International Immuno-Oncology Biomarkers Working Group. However, due to the better prognosis of LBC, extended follow-up end points may be more conducive to accurate and comprehensive analysis of the prognostic status of the population. Although multifactor analysis is used to control for confounding factors, there are still inevitable potential confounding factors.
In conclusion, we found that stromal TILs level showed significant prognostic value for luminal breast cancer in Asian populations with standard treatment. Specifically, high levels of TILs were associated with improved DFS in luminal breast cancer, the luminal B/HER2- subgroup showed significant association. Interestingly, the associations indicated different outcome directions in luminal/HER2 + and luminal/HER2- subtypes, indicating that high level of TILs may associated with worse outcome in luminal B/HER2 + patients, although the association was not statistically significant. There are significant differences in the composition of tumor immune microenvironment of specific LBC tumors, which also have different implications for prognosis. Therefore, further investiging on the relationship between subset of TILs and specific biomarker of immunogenicity could better reveal prognosis and provide more information on LBC immunotherapy.