Prediction of Response to Neoadjuvant Chemotherapy in Egyptian Patients with Locally Advanced Breast Cancer: The Evolving Role of Tumour Infiltrating Lymphocytes (TILs)

Background & objectives: Presence of TILs in breast cancer indicates better therapeutic responses to neoadjuvant chemotherapy (NAC), increased pCR and improved outcome. We aim at evaluation of TILs in breast cancer biopsies in correlation with pathological response after NAC in locally advanced breast cancer Methods. This study was conducted at Ain Shams university hospital and Maadi Military hospital in Egypt. 45 Female patients with locally advanced breast cancer were treated with NAC; pathological response was assessed. Tumours were categorized into: luminal A, luminal B, Her2 enriched and TNBC. Assessment of TILs (CD4 and CD8 lymphocytes) was based on Immuno-Oncology Biomarker Working Group guidelines. Results: Tumours were classified into high and low TIL groups using the interquartile range cutoff (29%). High CD4 group showed increased pCR (p = 0.003) and smaller residual tumours (p = 0.04). High CD8 group showed a significant association with smaller residual tumours (p = 0.003). At follow-up of 24-months, CD4 and CD8 high groups showed significantly higher 2-years DFS. The difference between CD4 phological or genomics-based parameters helps to stratify patients of breast cancer.


Introduction
Neoadjuvant chemotherapy is increasingly used for the treatment of breast cancer aiming at reduction of the tumor size, allowing conservative surgical resection, eliminating clinically silent metastatic foci, and providing prognostic data based on the pathologic response of the tumor [1].
Evidence is emerging that pathological response after neoadjuvant chemotherapy, represented by total or near total destruction of the tumor in the surgical specimen, is believed to be a strong predictor for survival and an indicator of favourable overall prognosis [2].
The tumorigenesis of breast cancer is determined by genetic alterations, phenotypic variations of cancer cells and interaction between the tumor cells and the surrounding microenvironment. As important components of tumor microenvironment, tumor-infiltrating lymphocytes (TILs) are a group of mononuclear cells representing the immunological response to tumors [3]. Although breast cancer was not considered as a typical immunogenic disease, the morphological evaluation of TILs in breast cancer and their clinical relevance has become a focus of interest over the past few years. TILs are mostly comprised of T cells, followed by macrophages, B cells, natural killer and other immune cells. Immunohistochemistry (IHC) allows the identification of the different populations of TILs, by using specific markers of leukocytes [4] [5] [6].
Studies have indicated that presence of TILs is associated with improved therapeutic responses to chemotherapy in both neoadjuvant and adjuvant settings [7] [8]. Nevertheless, many trials have shown that TILs are associated with improved outcome of breast cancer patients; this could be reflected by the higher pathological complete response (pCR) rates in immune-rich cancers [9] [10]. These evidences support the fact that the efficacy of many cancer therapies relies upon establishing effective immunologic biomarkers within the microenvironment [11] [12] [13].
In this study we aim at evaluation of TILs (CD4 and CD8 T-lymphocytes) in breast cancer tissue biopsies and their correlation with the pathological response after neoadjuvant chemotherapy in Egyptian female patients with locally advanced breast cancer. We included pre-or post-menopausal female patients with locally advanced breast cancer showing no evidence of metastatic disease. Male patients and patients with bilateral breast cancer were excluded. Cases subjected to primary surgical treatment were not included in this cohort.

Material and Methods
Clinical evaluation, histopathologic assessment and treatment: Breast cancer diagnosis was confirmed by histopathologic assessment of core needle biopsy and staged by systemic imaging studies using Computed Tomography (CT), ultrasonography (US) and bone scan. Tumors were evaluated for the immunohistochemical expression of estrogen receptors (ER), progesterone receptors (PR), HER2-neu and Ki-67 index. For cases of HER2-neu score 2, silver in-situ hybridization (SISH) was used for confirmation of the final score. Based on their immunohistochemical expression profiles, all tumors were categorized into four immunophenotypes: luminal A (ER+ and/or PR+, HER2−, Ki67-low); luminal B (ER+ and/or PR+, HER2+) (ER+ and/or PR+, Her2−, Ki67-high), Her2 enriched (ER−, PR−, Her2+); and TNBC (negative for ER, PR, and Her2). Luminal A and luminal B cases were collectively categorized as hormone receptor-positive breast cancers.
All patients received a neoadjuvant chemotherapy protocol comprised of Flurouracil (500 mg/m 2 /21days), Epirubicin (100 mg/m 2 /21days) and Cyclophosphamide (500 mg/m 2 /21days), for three or four cycles, followed by Docetaxel (75 -100 mg/m 2 /21days) for three or four cycles. This was followed by radical surgery according to response, radiotherapy and hormonal treatment in hormone-positive cases. The pathologic effect of the neoadjuvant chemotherapy was assessed in all resected tumors. pCR was defined as the complete disappearance of all invasive tumor cells from breast tissue and regional lymph nodes regardless of the presence of residual ductal carcinoma in situ, according to Miller-Payne grading system [14]. Disease free survival (DFS) and overall survival (OS) were estimated over a follow-up period of two years.
Evaluation of TILs: Formalin fixed paraffin embedded sections from all tumors were examined to identify the areas with dense lymphocytic infiltration. Immunohistochemical staining was performed using the T-cell markers CD4 (EP204) rabbit monoclonal antibody (RabMAb), and CD8 (C8/144B) mouse monoclonal antibody (Sigma-Aldrich) through an avidin-biotin-peroxidase complex method with diaminobenzidine as chromogen according to the manufacturer's instructions.
Assessment of stromal TILs was done using criteria described by Salgado et al. [15]. TILs were defined as the infiltrating lymphocytes within tumor stroma and expressed as the percentage of tumor area that was occupied by inflammatory (CD4 and CD8) cells. TILs were scored as continuous variables with a positivity cut off set at 1% [16].
The stromal lymphocytes were counted in three high power fields (magnification, ×400). The counts were performed in areas of maximum lymphocytic infiltration. Foci demonstrating haemorrhage, necrosis, in situ carcinoma and crush artefacts were avoided. CD4 and CD8 lymphocytes expression levels were classified as either high and low based on values ≥ or < the inter-quartile range (IQR) of each, respectively. The TILs evaluation was performed by two pathologists who were blinded to clinical data [4] [17] [18].

Statistical Analysis
SPSS package (version 22.0) was used for data analysis. The association between clinical and pathologic parameters were tested with χ 2 test. Numerical data were expressed as mean and standard deviation or median and range as appropriate.
Qualitative data were expressed as frequency and percentage. Chi-square test (Fisher's exact test) was used to examine the relation between qualitative variables.
For quantitative data, comparison between two groups was done using either student t-test or Mann-Whitney test (non-parametric t-test). Disease free survival (DFS) and overall survival (OS) will be estimated using Kaplan-Meier method.

Results
Patient demographics and clinical characteristics: Patient age ranged between 33 and 66 with median age 53 years. Most of patients were older than 35 years old (93%). Premenopausal females were 21 patients (46%) while postmenopausal were 24 (53%). Tumor characteristics are summarized in (Table 1).
All tumors were classified into high and low TIL groups using the interquartile range (IQR). Cases showing levels equal to or higher than the IQR were included in the TILs high group, while TILs low group included cases expressing TILs levels lower than the IQR value.
In the overall cohort, the interquartile range (IQR) for both CD4 and CD8 was 30% (0% -60.0%) (Figure 2). The CD4 low and high groups included 36 and 9 cases respectively, while CD8 low and high groups included 38 and 7 cases.   TILs and clinico-pathological parameters: A significant difference in estrogen receptor status and molecular subtype was evident between both CD4 high and low groups. In CD4 groups, the ER positive cases represented 69.7% and 33.3% of cases included in CD4 low and CD4 high groups respectively (p = 0.029). Patients with Her2 enriched subtype showed higher CD4 expression when compared with other molecular subtypes (p = 0.007). On the other hand, the molecular subtype expressed a significant correlation with CD8 among other parameters (p = 0.018); with Her2 subtype showing the highest CD8 expression (Table 2) (Figure 3).  Prognostic value of TILs: High CD4 group showed significantly increased pCR in comparison to low CD4 (17.7% and 11.1% respectively) (p = 0.003). All patients in high CD4 group developed either pCR or partial remission but no patient experienced stationary disease or progression after neoadjuvant chemotherapy. With regards to pCR among different molecular subtypes, 10 tumors (76.9%) were Her2 positive, 2 (15.3%) were TNBC and 1 (7.6%) was luminal B.
In this analysis, we found that the correlation between pCR and presence of CD4 was nonlinear. Rates of pCR increased when levels of CD4 were greater than 3% (Figure 4).    By the end of the study 38 patients were still alive with 7 deaths occurring in low CD4 group while no deaths were seen in high CD4 group (p = 0.09).
Data regarding CD8 showed some differences as compared to CD4. High CD8 significantly improved the DFS (95% CI, 18.6 -22.9; p = 0.01). Also, High CD8 group showed a significant association with smaller residual tumors (p = 0.003), but there was no significant difference regarding pCR or OS between CD8 high and low groups (Table 3).
TILs and pattern of treatment failure: After neoadjuvant chemotherapy and radical surgery, 40% patients developed events in the form of local recurrence or distant metastasis. In CD4 low group, 2 patients developed local recurrence and 14 patients had distant metastasis while in CD4 high group only one patient had local recurrence and distant metastasis (p = 0.19). Similarly, in CD8 low group 2 patients had local recurrence and 15 patients developed distant metastasis while in CD8 high group, one patient had local recurrence only (p = 0.025) ( Table 3).

Discussion
The understanding of tumor immune microenvironment influences the development of immunotherapy, it also affects the tumor response to other anticancer drugs and treatment outcomes [19] [20].
Recently, TILs proved to play a promising role as predictors of response to neoadjuvant chemotherapy in breast cancer. This opens new opportunities for different treatment plans, based on modulation of immune responses [21] [22] [23].
In the current trial we investigated the correlation between TILs and the response to neoadjuvant chemotherapy in locally advanced breast cancer cases in Egyptian females. We classified TILs into two distinct groups (high and low) based on the interquartile range of stromal TILs levels. The IQR was equal in both CD4 and CD8 with significant results regarding its prognostic and predictive value.
In our study, the cut off between TILs high and low groups was 30%, this value allowed for a significant correlation between TILs and different prognostic indicators. Other reports employed different cut off points for this purpose such as 35% cut off in a study by Kotoula et al., in 2016, while Jang and colleagues reported 10% cut off value to define TILs levels that were associated with better outcomes for a subset of hormone positive breast cancer patients [24] [25].  [26]. Similar data were reported by the NeoALLTO trial, as a nonlinear relationship was seen between pCR and presence of TILs, rates of pCR increased sharply when levels of stromal TILs were greater than 5%, regardless of treatment [27]. These reports are comparable to our results as the high CD4 group in this cohort showed increased rates of pCR in contrast to low CD4  [26]. Similarly, we reported higher recurrence rate in TIL low group as compared to TIL high groups for both CD4 and CD8, however, statistical significance was evident in CD8 groups only (p = 0.02).
Recently, the guidelines of the Immuno-Oncology Biomarker Working Group have contributed to reproducibility of TILs evaluation in residual disease as TILs in the post-neoadjuvant residual disease setting are obtaining rising interest [15].
One of the limitations of our cohort was the small number of cases with residual disease. However, we elucidated a significant correlation between high TILs levels and smaller residual tumor. Another limitation of the study was the lack of TILs evaluation within in-situ carcinoma settings.

Conclusion
We concluded that presence of high TILs levels may identify a group of patients with excellent prognosis (achieved pCR after NAC). Nevertheless, we raise the concerns to find mutual agreement regarding a standard cut off point to define high and low TIL levels as the heterogeneity in different studies may be due to interobserver variability in TILs assessment. Further large-scale clinical studies may address such elusive issues and reach a consensus regarding the utility of TILs in combination with other morphological or genomics-based parameters to stratify patients of breast cancer after the use of NAC.

Funding
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.