Prognostic impact of a tumor-infiltrating lymphocyte subtype in triple negative cancer of the breast

Background Tumor-infiltrating lymphocytes (TILs) have recently been reported as an important factor in the tumor microenvironment and influence the growth and progression of cancer. However, the relationship between immune cell subpopulations, such as CD4+, CD8+, and FOXP3+, in breast cancer, especially in triple negative carcinoma (TNC), remains unclear. Methods The subjects were 107 patients with TNC that were surgically resected at Dokkyo Medical University Hospital between 2006 and 2018. The expression of CD4+, CD8+, and FOXP3+ was evaluated in TILs and expressed as the numbers of positive cells. Results Univariate analysis revealed that the TILs were not prognostically significant. In multivariate analyses, increased infiltration of intratumoral (i) CD4+ TILs was found to have a good prognosis in relapse-free survival (RFS). In contrast, a high stromal CD8+ TILs level was found to be a favorable prognostic factor in RFS (p = 0.038) and overall survival (OS) (p = 0.046). A low sFOXP3 + TILs level was significantly associated with favorable RFS (p < 0.001) and OS (p = 0.029). Conclusions The present study demonstrated no difference in TILs and survival in TNC. However, there was a significant correlation in prognosis with levels of iCD4+, sCD8+, and sFOXP3 + TILs in TNC. The difference in TNC clinical outcome may be due to the subtype of the infiltrating TILs.


Introduction
Gene expression profiling studies have divided invasive breast cancer into several major subtypes [1]. The so-called 'triple negative carcinoma' (TNC) is characterized by a lack of expression of the estrogen receptor (ER) and progesterone receptor (PgR), and absence of human epidermal growth factor receptor 2 (HER2) protein overexpression; this type is known to have a poor prognosis [2,3]. Tumor-infiltrating lymphocytes (TILs) have recently been reported as an important factor in the tumor microenvironment and influence the growth and progression of cancer.
The majority of TILs in cancer are of the T-cell phenotype, which includes CD4+ (helper cells) and CD8+ (cytotoxic cells) lymphocytes. CD4 + T lymphocytes are important for priming tumor-specific CD8 + TILs as well as for the secondary expansion and memory of CD8 + TILs [4]. Furthermore, many immunohistochemical studies have concluded that CD8 + TILs have antitumor activity as evaluated by the favorable prognosis in colorectal [5], ovarian [6], esophageal [7], renal [8], lung [9], and pancreatic [10] tumors. However, the impact of CD8 + TILs in breast cancer is controversial. Previous breast cancer studies have reported that marked infiltration of CD8 + TILs is associated with good prognosis, while several studies have found a negative correlation or no correlation with prognosis [11][12][13][14].

3
Therefore, the assumption that lymphocyte infiltration promotes or prevents cancer cannot be confirmed without clarifying which immune cell phenotype is involved.
The role of CD4 + TILs in immune activity has been reported in many cancer patients. However, the discovery of regulatory T cells (Treg) has markedly changed conventional speculation regarding the role of CD4 + T lymphocytes in anti-tumor immunity.
Forkhead box protein 3 (FOXP3) plays a critical role in the generation of immune-suppressive CD4 + Tregs, and this leads to immune tolerance of CD8+ killer cells [15]. Excess FOXP3 expression leads to Treg proliferation and severe immunodeficiency, whereas lack of FOXP3 results in immune system activation and aggressive lymphoproliferation. Furthermore, FOXP3 is involved in immune escape mechanisms and both poor survival and improved survival in breast cancer have been reported [15][16][17]. There have been several reports on TILs in breast cancer [18,19]. However, the relationship between immune cells subpopulations, such as CD4+, CD8+ and FOXP3+, in breast cancer, especially in TNC, remains unclear. We herein discuss the clinicopathological features and possible roles of immune cells in TNC.

Patients
The subjects were 107 patients with TNC, which was surgically resected at Dokkyo Medical University Hospital between 2006 and 2018. Patients' clinical information was retrieved from institutional medical records. Clinical outcome was also documented. For each case, all available hematoxylin and eosin-stained whole-tissue sections were reviewed to confirm the diagnosis of mammary disease with no knowledge of prior histological results or clinical outcomes. The present study was approved by the Ethics Committees of Dokkyo Medical University (Tochigi, Japan; registration number 28009).

Immunohistochemistry (IHC)
Surgical sections were immunostained for ER (clone SP1, Novocastra (Leica), prediluted, nuclear), PgR (clone 1E2, Novocastra (Leica), prediluted, nuclear), HER2 (clone 4B5, Roche (VENTANA), prediluted, membranous), CD4 (CD4, clone 1F6, Novocastra (Leica), 1:40), CD8 (CD8, clone 4B11, Novocastra (Leica), prediluted) and FOXP3 (FOXP3, clone 236A/E7, abcam, 1:50). Counterstaining was performed with hematoxylin. ER and PgR status were considered positive if any positive cells were detected within the tumor. HER2 status was assessed according to the guidelines defined by the American Society of Clinical Oncology/College of American Pathologists [20]. We estimated the TILs on hematoxylin and eosin (H&E) stained sections according to the criteria proposed by the International Immuno-Oncology Biomarkers Working Group [21]. Lymphocytes in contact with or within the tumor epithelium were defined as intratumoral (i), whereas lymphocytes in the interstitial space or in the stromal areas were defined as stromal (s). TILs were defined as all mononuclear cells, including lymphocytes, within the stromal area, and excluded necrosis, crush artifacts, regressive hyalinization, as well as granulocytes and other polymorphonuclear leukocytes (Fig. 1). TILs levels were categorized as high (≥ 30%) and low (< 30%) adopting previously validated cut-offs [22]. The expressions of CD4, CD8, and FOXP3 were evaluated in TILs and expressed as the numbers of positive cells counted in each case at ×400 magnification (×40 objective) (Fig. 1). For statistical analyses, the number of positive cells was divided into lower and higher groups based on cut-off points according to the median. As a result, the cut-off for iCD4+ was 3, sCD4+ was 54, iCD8+ was 7, sCD8+ was 43, iFOXP3+ was 3, and sFOXP3+ was 32. All sections were evaluated by two pathologists (TJ and HK) who had no previous knowledge of the patients' clinical information, and the results were averaged.

Statistical analysis
The associations between CD4+, CD8+, and FOXP3 + TILs and clinicopathological variables were examined by x 2 -test. Relapse-free survival (RFS) was defined as the time from surgery to recurrence, including metastatic disease. Overall survival (OS) was determined from the date of surgery to the date of death by cancer or to the date of the last follow-up. For significance testing in Kaplan-Meier survival analysis, we used the log-rank test. The hazard ratios (HRs) and 95% confident intervals (CIs) were calculated by Cox proportional hazard models. Multivariate Cox regression analysis including all potential variables that were significantly associated with survival in each univariate analysis was performed. All statistical tests were considered significant at the p < 0.05 level. Statistical analysis was performed using IBM SPSS Statistics 25 (IBM, Armonk, NY, United States).

Results
Clinicopathological findings and expression of immune markers (CD4 + , CD8 + , FOXP3 +) are summarized in Tables 1 and 2, respectively. The patient ages ranged from 28 to 89 years, with a mean of 58.9 years. Tumor size ranged from 0.3 to 10.0 cm; 64.5% were ≤ 2.0 cm and 35.5% were > 2.0 cm in diameter. They presented with histological 1 3 high grade (76/107) and lymph node metastasis (30/107). Recurrence occurred in 19 (19.2%) of 99 patients, and cancer-associated death occurred in 12 (12.1%) of 99 patients. Median follow-up for the assessment of RFS was 42.13 (0-120) months and that for overall survival was 43.86 (0-120) months. The high expression of iCD4 + TILs was associated with larger tumor size (p = 0.034) and higher histological grade (p = 0.006). Lymph node status (p = 0.036, p = 0.05) and expression of CD8 + TILs in both i and s areas had a significant correlation in the x 2 -test.

Univariate and multivariate Cox regression analysis correlation with CD4, CD8, and FOXP3 levels with RFS and OS
Univariate and multivariate Cox regression analysis of RFS and OS were performed using clinicopathological findings and expression of TILs (Tables 2, 3). Univariate analysis revealed that the conventional clinicohistological tumor parameters, including age, tumor size, histological grade, lymph node status, and TILs were not prognostically significant. Furthermore, the iCD4 + TILs demonstrated a significant association with RFS (p = 0.044), but no significant difference in terms of OS (p = 0.074). Both in the RFS and OS, sCD4 + TILs patients showed no significance in univariate analysis (p = 0.261; p = 0.254). However, the expressions of sCD8 + and sFOXP3 + TILs were associated with RFS and OS (p = 0.020; p = 0.032). The iCD4 + , sCD8 + and sFOXP3 + TILs found to have significant prognostic value in univariate analysis were selected for Cox proportional hazard analyses and the significance of their prognostic association was confirmed by multivariate assessment. In multivariate analyses, patients with high expression of iCD4 + TILs had a significantly longer RFS (HR 0.172, 95% CI 0.037-0.792, p = 0.024). Increased infiltration of sCD8 + TILs was found to be a favorable prognostic factor in RFS (HR 0.225, 95% CI 0.061-0.836, p = 0.026) and OS (HR 0.263, 95% CI 0.071-0.975, p = 0.046). In contrast, a low sFOXP3 + TILs level was found to be significantly associated with favorable RFS (HR 7.426, 95% CI 1.596-34.552, p < 0.011) and OS (HR 5.467, 95% CI 1.192-25.07, p = 0.029) ( Table 4).

Discussion
The characteristic features of TNC are large anaplastic cells and poor prognosis. Invasive carcinoma including medullary features with massive TILs has a better prognosis than the typical types of invasive mammary carcinomas [3,23]. Furthermore, a recent report suggested the prognostic importance of TILs in high-grade breast cancers. Kurozumi et al. recently investigated the relationship between TILs and prognosis in 294 cases and reported that high stromal TILs expression was a good prognostic marker in ER-negative cancers [24]. Ibrahim et al. also demonstrated that TILs were significantly correlated with a favorable breast cancer outcome in ER-negative tumors using meta-analysis including data on 2,987 patients [25]. However, we could not find significant differences in either iTILs or sTILs regarding the prognosis in TNC. Therefore, it is difficult to reach a conclusion regarding the prognosis of breast cancer based only on the TILs in TNC. Our results suggest that this prognosis in TNC is due, at least in part, to the presence of immune cell types that are closely associated with the tumor. We found that TNC patients with a good prognosis had a predominance of sCD8 + TILs in both RFS and OS. Lymphocytes infiltrating a tumor indicate a local immune response and they play an important role in tumor progression [11,12,26]. The majority of infiltrating lymphocytes in tumors are CD8 + TILs and these have a cytotoxic effect [12,27]. In several organs, high levels of CD8 + TILs infiltration were associated with better prognosis [5][6][7][8][9][10]. In breast cancer, Ali et al. reported that iCD8 + and sCD8 + T cell infiltration was also associated with a significant reduction in the relative risk of death [11]. Furthermore, Liu et al. reported that iCD8 + and sCD8 + tumor-infiltrating lymphocytes are an independent prognostic factor associated with better survival in TNC [28]. Therefore, the greater predominance of CD8 + lymphocyte infiltration in TNC suggests that a strong immune response is occurring. However, we found increased infiltration of iCD4 + TILs was significantly associated with good prognosis only in RFS by multivariate analysis. In contrast, there have been a few reports examining the role of CD4 + TILs in breast cancers; they were associated with more aggressive behavior. Huang et al. reported that iCD4 + TILs negatively correlated with RFS in breast cancer [26]. Rubbert et al. reported a predominance of sCD4 + TILs among tumor-infiltrating lymphocytes in patients with larger tumors [29]. Furthermore, Macchetti et al. observed that in patients with lymph node metastasis, there was increased infiltration of sCD4 + TILs with a corresponding reduction in CD8 + cells [30]. Since CD4 + TILs are expressed in many T cell subsets including T helper 1 (Th1) cells, T helper 2 (Th2) cells and Tregs, each of these may have a different impact on prognosis. Th1 cells secrete several cytokines such as interferon gamma (IFNg), transforming growth factor beta (TGFβ), tumor necrosis factor alpha (TNF), and interleukin 2 (IL-2) [31]. These cytokines are involved in the function of CD8 + TILs and protect against tumor development and progression. In contrast, Th2 cells express several types of interleukin and induce loss of cytotoxicity [32]. Thus, CD4 + TILs that include many T cell subsets may explain why iCD4 + TILs was different in TNC.
In the present study, TNC patients with a good prognosis showed significantly lower expression of sFOXP3 + Tregs. Tregs are important mediators of immune tolerance that Table 4 Hazards for triple-negative cancer (TNC) overall survival (OS) in the entire cohort with univariate and multivariate analyses Multivariate cox regression analyses were performed for all potential variables that were significantly associated with survival in univariate analysis OS overall survival. TNC triple-negative cancer, HR hazard ratio, CI confidence interval, i intratumoral, s stromal, TILs tumor-infiltrating lymphocytes, FOXP3 Forkhead box P3 * The P value is significant  good survival of breast cancer [27]. A recent study by Miyashita et al. demonstrated that a high sCD8/FOXP3 ratio was associated with improved prognosis in TNC [39]. Furthermore, our study confirmed that not only sCD8/FOXP3, but also the sFOXP3/CD4 ratio, were significantly associated with both RFS and OS. It seemed that both FOXP3 and CD4 were associated with tumor progression, but FOXP3 was a stronger indicator. These results suggest that activation of cytotoxic TILs and Tregs may affect the clinical outcome.

Conclusion
The present study demonstrated no difference in either iTILs or sTILs and survival in TNC. However, we found higher numbers of iCD4 + TILs were significantly associated with good prognosis in RFS. Further, decreased sFOXP3 + TILs infiltrate and higher numbers of sCD8 + TILs in TNC were associated with a significantly good prognosis in both RFS and OS. Therefore,  included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creat iveco mmons .org/licen ses/by/4.0/.