Outcome Prediction After Neoadjuvant Chemotherapy (NAC) for Breast Cancer, Using Tumor-infiltrating Lymphocytes Within Fibrotic Foci of Tumor Stroma (FF-TILs)

Background/Aim: Tumor-infiltrating lymphocytes (TILs), which are indicators of immune response monitoring, are generally mononuclear immunocytes that aggregate with tumors and are thought to have a close relationship with cancer cells. On the other hand, a fibrotic focus (FF) within the stroma of a tumor is a histological formation that plays an important role in the cancer microenvironment with regard to proliferation and development. Here, we focused on TILs that exist within the FF and performed pathological evaluations. Patients and Methods: Of the 320 patients treated with neoadjuvant chemotherapy (NAC), 239 subjects who were able to evaluate FF-TILs were targeted. Lymphocytes that infiltrate the FF are FF-TILs. Results: The disease-free survival (DFS) period after NAC for the high-FF-TIL group was found to be significantly longer than that for the low-FF-TIL group for all cases (p<0.001) and for all subtypes of triple-negative breast cancer (TNBC) (p=0.001), human epidermal growth factor receptor 2-enriched breast cancer (HER2BC) (p=0.010), and hormone receptor-positive breast cancer (HRBC) (p=0.003). In multivariable analysis as well, high-FF-TIL group classification was an independent factor for recurrence after NAC for all cases [p<0.001, hazard ratio (HR)=0.198] and all subtypes of TNBC (p=0.006, HR=0.172), HER2BC (p=0.025, HR=0.135), and HRBC (p=0.007, HR=0.228). Conclusion: FF-TILs are possibly a useful factor for predicting recurrence of breast cancer after NAC.

Cancer cells have various mutations that allow them to proliferate spontaneously and survive. However, the surrounding environment also influences cancer cells in their entirety and their intrinsic characteristics (1). Therefore, monitoring the host immune response to tumors in the cancer microenvironment helps predict treatment response and outcome (1,2). Tumor-infiltrating lymphocytes (TILs), which are indicators for monitoring an immune response, are generally mononuclear immunocytes that aggregate with tumors and are thought to have a close relationship with cancer cells. Recently, the usefulness of methods that histopathologically evaluate the TILs function in situ has been elucidated (3)(4)(5). Moreover, it has been shown that morphological evaluation of TILs is useful for outcome prediction in the treatment of breast cancer, as well as for predicting the therapeutic effects of drugs; therefore, TILs are considered novel biomarkers (6)(7)(8)(9).
Methods for in situ evaluation of TILs and their cutoff values have not assimilated previously, but rather have depended on the previous reports (7,10,11). Thus, in 2014, a recommendation concerning evaluation methods for TILs was created by the International Working Group (3). TILs can be classified on the basis of the region in which they exist as either stromal TILs (within the stroma of the tumor) or intratumoral TILs (existing within tumor foci and therefore in contact with cancer cells). According to this recommendation, it is desirable to evaluate TILs according to the quantity of immune cells within the cancer stroma; therefore, this recommendation employs evaluation with stromal TILs. However, it has been suggested that the heterogeneity of TILs deters clear cutoff values from being defined.
Recently, it has been reported that the biological state of the cancer stroma is closely related to cancer proliferation and development (12)(13)(14). Therefore, the evaluation of histomorphological images of tumor stroma has come to be considered one of the major keys to understanding the cancer microenvironment. A fibrotic focus (FF) within the stroma of a tumor is a histological formation that plays an important role in the cancer microenvironment with regard to proliferation and development (15,16). Infiltrating tumor cells surround the FF, which can be recognized as a converged focus of the tissue component that exists within the tumor. In breast cancer, it has been reported that tumors with wide-ranging FF have a high likelihood of malignancy and poor prognosis (15,17,18). That is, it is thought that morphological evaluation of FF is an important indicator in the cancer microenvironment for understanding the transformation to malignancy (19,20).
Here, we focused on TILs that exist within the FF and performed pathological evaluations. TILs within FF are called FF-TILs, and we hypothesized that the evaluation of TILs that exist within these regions would be a more precise indicator than previous methods of evaluating TILs. Among patients undergoing neoadjuvant (pre-surgical) chemotherapy (NAC) for breast cancer, we evaluated the prediction of treatment effects using FF-TILs.
All patients received a standardized protocol of NAC consisting of four courses of FEC100 (500 mg/m 2 fluorouracil, 100 mg/m 2 epirubicin, and 500 mg/m 2 cyclophosphamide) every 3 weeks, followed by 12 courses of 80 mg/m 2 paclitaxel administered weekly. The patients with HER2-positive breast cancer, were additionally administered weekly (2 mg/kg) or tri-weekly (6 mg/kg) trastuzumab during paclitaxel treatment (23)(24)(25). All patients underwent chemotherapy as outpatients. Therapeutic anti-tumor effects were assessed in accordance with the Response Evaluation Criteria in Solid Tumors (RECIST) criteria (26). Patients underwent mastectomy or breast-conserving surgery following NAC (27). The pathological effect of chemotherapy was assessed for resected primary tumors after NAC. Pathological complete response (pCR) was defined as the complete disappearance of the invasive components of the lesion with or without intraductal components, including that in the lymph nodes according to the National Surgical Adjuvant Breast and Bowel Project (NSABP) B-18 protocol (28). All patients who underwent breast-conserving surgery were administered postoperative radiotherapy to the remnant breast. The standard postoperative adjuvant therapy for the subtype concerned was administered.
Overall survival (OS) time was the period from the initiation of NAC to the time of death from any cause. Disease-free survival (DFS) was defined as freedom from all local, locoregional, and distant recurrences. All patients received follow-up with physical examination every 3 months, US every 6 months, and CT and bone scintigraphy annually. The median follow-up period was 3.7 years (range=0.2-6.0 years) for the assessment of OS and 3.4 years (range=0.1-6.0 years) for DFS.
Histopathological evaluation of TILs. Histopathological assessment of predictive factors was made for CNB or VAB specimens at the time of the breast cancer diagnosis. Histopathological parameters examined included nuclear grade, histological type, presence of TILs, and correlation of these parameters with intrinsic subtypes and pCR. Histopathological analysis of the percentage of TILs was evaluated on a single full-face hematoxylin and eosin (HE)-stained tumor section according to criteria described by Salgado et al. (3). TILs were defined as the infiltrating lymphocytes within tumor stroma and expressed by the proportion to the field investigated, and the number of TILs in stroma surrounding the stained cancer cells was quantitatively measured in each field under 400× magnification (11,29). The area of in situ carcinoma and crush artifacts were not included. Proportional scores were defined as 3, 2, 1, and 0 if the area of stroma with lymphoplasmacytic infiltration around invasive tumor cell nests was >50%, >10-50%, ≤10%, and absent, respectively ( Figure 2). TILs were considered high when scores were ≥2 and low when scores were 1 and 0. Histopathological evaluation of FF-TILs. An FF is a scar-like lesion consisting of an area of mainly collagen and fibroblasts, often located near the center of a carcinoma. In addition, an FF is a converged focus of the tissue component of the stroma of a tumor, and is surrounded by infiltrating tumor cells. An FF is defined as "FF often consisted of fibrous bands radially expanding to the surrounding area, and FF was located within the tumor, was surrounded by a more cellular zone of infiltrating ductal carcinoma cells, and occupied various percentage of the tumor area" (15,16). The fibroblasts and collagen fiber that form the FF show a storiform pattern, and propagate intertwined with each other. Lymphocytes that infiltrate the FF are FF TILs. We refer to the high TIL group (TIL score ≥2) within an FF as the high-FF-TIL group, and the low TIL group (TIL score: 1 or 0) within a FF as the low-FF-TIL group. We refer to the prediction of treatment effect using previous TIL evaluation methods as the training set (TS), and that using FF-TILs as the validation set (VS). Histopathologic evaluation of TILs and FF was jointly performed by two breast pathologists who were blinded to clinical information, including treatment allocation and outcomes.
Sectional analysis. Statistical analysis was performed using the SPSS version 19.0 statistical software package (IBM Corp., Armonk, NY, USA). The associations between FF-TILs, and clinicopathological variables were examined using the chi-square test (or Fisher's exact test when necessary). Multivariable analysis of pCR was carried out using a binary logistic regression model. The Kaplan-Meier method was used to estimate DFS and OS, and the results were compared between groups with log-rank tests. The Cox proportional hazards model was used to compute univariable and multivariable hazard ratios (HR) for the study parameters with 95% confidence intervals (CIs), and used in a backward stepwise method for variable selection in multivariable analyses. A p-value <0.05 was considered significant. Cutoff values for different biomarkers included in this study were chosen before statistical analysis.

FF-TILs and clinicopathological investigation.
Among the 239 patients who underwent NAC, 123 (51.5%) were in the high-FF-TIL group, and 116 (48.5%) were in the low-FF-TIL group. In the high-FF-TIL group, the Ki67 value was significantly high (p=0.049), the proportion of HRBC was significantly low (p=0.013), and the pCR rate was significantly high (p=0.004) ( Table I). With regard to subtypes, there were 83 cases of TNBC (34.7%), 46 cases of HER2BC (19.2%), and 110 cases of HRBC (46.1%). In an investigation that considered the clinicopathological background, regarding HRBC, in the high-FF-TIL group the Ki67 values were significantly high (p=0.001) (Table II). However, for all the subtypes, no correlation with the pCR rate was found (TNBC: p=0.154, HER2BC: p=0.489, HRBC: p=0.083).

Discussion
Recently, in many large-scale prospective clinical trials, TILs have been shown to be useful as a prognostic factor (4) (6,8,9,30). These reports have suggested that TILs can be useful as clinical biomarkers, and today attempts to unify their evaluation methods are underway (3,4). According to the International Working Group, in the evaluation of TILs, the percentage of immunocytes within a tumor stroma should be measured at the border of the cancer infiltration region (3). However, there is not yet a standardized opinion regarding determination of the region within the tumor stroma. Therefore, we also paid attention to FF in regions that were within the tumor stroma and surrounded by infiltrating tumor cells. Wide-ranging FF contribute to the tumor's acquisition of a malignant condition due to the hypoxic environment within the tumor, and are related to prognosis and drug resistance (31)(32)(33). In the present study, when we investigated TILs that exist within FF as FF-TILs in patients undergoing NAC, the high-FF-TIL group had a longer DFS period after NAC than did the low-FF-TIL group.
When studied by subtype, TILs have been considered useful as prediction markers for treatment effect in subtypes with high immunological action, such as TNBC and HER2BC (4,8,9,30). However, in HRBC, which has the highest frequency, there are few reports that show a clinical relationship with TILs. In the present study, FF-TILs were found to be useful as prognostic factors following NAC not only for TNBC and HER2BC, but also for HRBC. Furthermore, for non-HRBC, such as TNBC and HER2BC, it is thought that wide-ranging FF exist (33), and in the present study as well, for non-HRBC, we observed that FF-TILs were significantly high.
In the current study, a reason that outcome prediction was possible even for HRBC is the heterogeneity of TILs in the cancer microenvironment. ER, which is important for the occurrence and development of HRBC, is activated not just by estrogen but also by the signal cascade of a pathway via phosphorylation due to various growth factors, and the control of ER depends on the cancer microenvironment (34). However, there are reports that, in a hypoxic environment, expression of hormone receptors is weakened (35,36). That is, it is thought that, in an FF region, hormone receptors are weakened owing to the hypoxic environment; therefore, biomarkers for ER-negative cancers are useful. It is possible to dynamically understand the changes in the cancer microenvironment more precisely with FF-TILs than with the previous evaluation range of TILs; therefore, we believe it may be a better evaluation method. In the present study, patients in the low-FF-TIL group experienced a high recurrence rate, and perhaps, depending on the subtype, adjuvant therapy should have been added. That is, in the selection of post-surgical adjuvant therapy for patients undergoing NAC, it is suggested that determination using FF-TILs can contribute to the choice of a proper treatment strategy. However, the retrospective nature of this study is a limitation, and regarding adjuvant therapy after NAC, since there are differences among cases, and since it is difficult to evaluate FF with CNB specimens, it is necessary to obtain more tissue through the use of methods such as VAB during diagnoses.

Conclusion
It is suggested that FF-TILs are a useful factor for predicting recurrence of breast cancer after NAC, and they may be a more precise indicator than previous evaluation methods with TILs.

Conflicts of Interest
All of the Authors have no conflicts of interest to disclose regarding this study.