Evaluation of tumour infiltrating lymphocytes in core needle biopsies and resected specimens of breast carcinoma

Results: The mean value of tumor-infiltrating lymphocytes in 34 cases was 27.03% in core needle biopsies and 34.6% in surgically resected specimens. High tumor-infiltrating lymphocytes score was seen in 8 (23.6%) and 11 (32.4%) cases and low to intermediate score was seen in 26 (76.4%) cases and 23 (67.6%) in core needle biopsies and surgically resected specimens respectively. Intermediate and high tumor-infiltrating lymphocytes categories in core needle biopsies accounted for good agreement with surgically resected specimens (60%-intermediate, 87.5 %-high). More than three cores had a better agreement.


INTRODUCTION
The immune system has a crucial role to play in the disease outcome of patients with cancer. 1 Several factors like age, histological type, grade, tumour size, lymph node status, hormone receptor, and HER2neu status and vascular invasion provide a variable prognosis on the survival of breast carcinoma patients. 2 Research studies have proven that a relationship lies between tumour and host inflammatory response and its influence on the prognosis and metastasis of cancer. 3,4 Tumour infiltrating lymphocytes (TILs) has been recognized as a biomarker of an antitumor immune high (40-90%). 24 For simplifying the results and for better understanding, the TILs score were also assessed by twotier classification low/intermediate (<40%) and high (40%-90%). In the CNB, all cores containing invasive tumour cells were examined. Immunohistochemistry was performed on 4 micron thick sections. The primary antibodies used were ER, PR, HER2neu, and Ki67. Allred score was used to quantify the hormonal status. HER2neu staining was assessed as per the American Society of Clinical Oncology/ College of American Pathologists guidelines. Ki67 was assessed as an average number of positive cells for 1000 tumour cells and then reported as a percentage.

Statistical methods
Data were analyzed using SPS (v.20.0; SPSS, IBM Corporation, Bangalore, India). P-values of <0.05 were considered to be significant. The TIL scores of CNB and SRS were compared using paired t-tests and Pearson's correlation. Pearson's r-correlation was used to compare scores between CNB and SRS. Other relevant statistical tests like Fisher's test and ANOVA (one way) were applied to compare the results. The distribution and agreement of TIL score between CNB and SRS were done using the kappa statistical test. Values of κ> 0.6 were correlated with good agreement, values between 0.4 and 0.6 considered moderate agreement, values < 0.4 corresponded to fair, and values < 0.2 reflected poor agreement. Univariate analysis was applied for risk factors affecting the TIL score based on 2 tier classification.
The mean value of TILs in CNB was 27.03% and 34.6% for the surgically resected specimen. The different parameters like age, subtype, grade, immunohistochemistry expression, vascular invasion were analyzed with respect to TILs between CNB and SRS. The mean value of TILs for the response. [5][6][7][8] Recently it is found that TILs and microvessel density have a role in modifying the prognosis and survival of breast cancer patients. 9,10 However, the literature survey suggests inconsistency about the results in the literature due to the absence of methodical validation concerning the outcome of TILs in breast carcinoma. 11 In order to streamline this, international guidelines for evaluation of TILs in breast cancer was prepared by the International working group for TILs 2014. The aim was to standardize the evaluation of TILs based on Haematoxylins and eosin (H&E) stained sections of core needle biopsies (CNB) and surgical resection specimen (SRS). After the guidelines were laid, many studies were conducted and observed that increase TILs are predictive of a better response to neoadjuvant therapy (NAT) and overall survival. [12][13][14][15][16][17][18][19] In an era of NAT, where NAT has shown to have better clinical response and survival, the importance of CNB has routinely come into practice as it determines the specific NAT required. 20 Studies have shown that CD8+ T lymphocytes and FOXp3+ cell comprise the breast cancer TILs. CD8+ cell activation causes autoregression of larger tumours. FOXp3+ cells are known to inactivate the CD 8+ T cells. Studies show that the immune modulator used in NAT determines the survival of breast cancer patients. 21 Thus, assessment of TILs and type of TILs helps in the introduction of an immune modulator in NAT. When CNB and SRS were studied concerning estrogen receptor (ER), progesterone receptor (PR), HER2neu receptor, and Ki67 disagreement were confounded. 22,23 There are limited studies for TILs correlation between CNB and SRS. The present study was designed to compare the TILs status between CNB and SRS of breast carcinoma and to assess the reliability of the TILs score in CNB.

MATERIALS AND METHODS
It is a retrospective study from January 2016 to March 2019 conducted in the department of pathology at a tertiary care centre in southern India. 34 cases of breast carcinoma cases for which both CNB and SRS were obtained were included. Patients who received neoadjuvant chemotherapy, metastatic breast carcinoma, and recurrent carcinoma were excluded from the study group. Clinical, pathological, and treatment details with follow-up were obtained from department archives and computerized medical records.
Routine tissue processing on formalin-fixed tissues of CNB and SRS was performed and stained with H&E stain. H&E slides were double-blinded and TIL score was done by two pathologists according to the international guidelines for the assessment of TIL 2014. 24 Areas outside the tumour border, around the intraductal component and normal lobules, TILs with crush artifacts and necrosis were not considered. TILs included only lymphocytes and plasma cells. The average score was represented as a percentage. TIL scores were sub-classified as low (≤10%), intermediate (11-39%), and Tumour infiltrating lymphocytes in breast carcinoma ER-positive and PR-positive tumour (HR phenotype tumour) had low to intermediate TILs score than ERnegative tumours in both CNB and SRS. HER 2 phenotypic tumours had a very low TIL score in CNB (26.2%) than SRS (43.2%). TNBC( 40.33%) and HER2(26.27%) phenotypic tumour showed a high TIL score against very low TIL values in HR phenotypic tumours (13.27%). When low and high levels Ki67 percentage was compared, a TIL score of 24.21% (low Ki67), 32.4% (high Ki67) was encountered in CNB and 30.6% (low Ki67), 37.5% (high Ki67) TIL score in SRS.
On performing paired T-test upon the clinicopathological parameters between CNB TIL scores and SRS TIL scores, young age group was 24.4% which was much lower than the scores in SRS (37.5%). In elderly adults, the mean value of TILs for both age groups was quite close, which is 29.6% in CNB and 31.7% in SRS.
Infiltrating ductal carcinoma had a significant difference between TIL score of CNB (26.4%) and SRS (34.4%) in the present study. TIL score was higher in grade 3 (33.23%) followed by grade 2 (23.5%) tumours and grade 1      were categorised from fair to near perfect levels(0.3-0.81) in present study.

Figure. 2: Case 2-HR phenotypic tumour showed TILs score on (a) CNB-9% (b) SRS-15.5% (HE-stain; X40)
After univariate analysis, the parameters which had confounding factors that affected the TIL score were age, PR status, Her2 phenotypic tumour, and the number of cores. Our study showed that more number of cores (≥ 3) in CNB was significantly associated with a smaller difference in the TIL score. (Table 4) On comparing the TIL score with respect to phenotype classification of breast carcinoma by ANOVA, the mean TIL score in HER2, HR, TNBC was lower in CNB than the SRS and the difference was statistically significant. The difference in TIL score between HR and HER2 is not significant in CNB but is significant in SRS. Similarly, the difference in TILs score between HR and TNBC is significant in both CNB and SRS. The difference of TIL score in HER2 versus TNBC is not significant in both CNB and SRS since both types of tumours have a high TIL score. The microphotographs of the TIL distribution in various phenotypic tumours are explained ( fig. 1-3).

DISCUSSION
The present study examined the clinicopathological profiles of TILs in patients with primary breast carcinoma. We found that there was a considerable disparity between TIL score of CNB and SRS with younger age (<50 years), PR status, and HER2 phenotype tumours. CNB is expected to be composed of tissue from the periphery of the tumour and less from the centre. 20 Few studies hold an opinion that tissue with high microvessel density sets off more immune activation which brings about more hypoxia and necrotic change in the malignant tumours. 26 These are the rationale behind the difference in CNB and SRS TIL score. The present study and the study done by Cha et al support this hypothesis as we have got a lower mean TIL score for CNBs than the mean TIL score in SRS.
On analyzing the scores at a young age, the antecedent cause was firstly CNB procedure elicits an immune response in breast cancer which justifies the reason for the difference in TIL score between the two. [26][27][28] The TIL change after CNB between young and elderly is significant because the immune activation in the elderly is subtle when compared to younger patients because of a decrease in immune function, degeneration of thymus gland, less recruitment of T cell. 29 Thus, TIL is more consistent and uniformly distributed in the elderly than young patients. Cha et al also showed that there was less change in elderly patients between CNB and SRS. 30 TIL score when compared between different grades of the tumour showed a difference between CNB and SRS. The reason may be while performing histological grading of the tumour on CNBs there is a propensity to under categorize the tumour by virtue of underestimating the mitotic counts during biopsy and overestimation of overall nuclear grade on SRS. 20 Our study showed TIL score was low in the majority of cases and intermediate category for ER/PR positive tumours. Loi et al showed that there is lower activation of TILs in ERpositive tumours than triple-negative tumours. 31 Cha et al showed that ER/PR positive tumour had an intermediate mean score for TIL. 30 PR positivity is heterogeneous in the tumour when compared to ER positivity which is more homogenously stained. 32 Therefore PR positivity acts as a confounding factor for TIL scoring (p<0.05).
Our study unveils that there is no correlation in the TIL score between CNB and SRS in tumours with HER2 phenotype. HER2 subtypes tumours show a higher level of intratumoral heterogeneity in TIL. But a study done by Cha et al showed there is near perfect reliability between biopsy and SRS in HER2 phenotypic tumours against fair reliability in our study. 30 Our study showed that the variables responsible for discordance were younger age, HER2 status, and PR status.
The present study showed that the correlation between CNB and SRS improved for ≥3 cores. Cha et al proved that ≥5 cores are sufficient to get good agreement between CNB and SRS.
The limitations of the study are firstly the sample size is very small. In the phenotypic sub-classification of breast cancers where the essence of the study lies in the era of chemotherapy, there is an uneven distribution of cases. This can lead to uncertain results.
More studies are required to evaluate spatial and temporal heterogeneity in breast cancer by tissue microarrays and immunohistochemistry (CD8, FOXP3). More studies with complete follow-up along with the determination of overall survival and disease-free survival with neoadjuvant chemotherapy, immunomodulation therapies with regard to TIL is required. Recommendations by an International TILs Working Group 2014 while dealing with tumour with high TIL heterogeneity should be reconsidered.

CONCLUSIONS
TIL score in CNB is of reliable value in breast cancer. Discrepancies may occur in tumours with HER2 phenotype, PR status and younger age. More than three cores may accurately predict the TIL score.