Breast cancer is one of the leading causes of cancer-related mortality and morbidity in females. Various advancements have been made in the diagnosis and treatment of breast cancer in recent decades. Due to these advancements in breast cancer therapy, the current focus is on identifying the optimal treatment strategy for individual breast cancer patients. For this purpose, breast cancer patients can be classified into various subgroups based on their clinicopathological features and the ideal treatment for each subgroup is being studied [1]. These clinicopathological variables, otherwise called the “predictive markers” indicate the sensitivity or resistance to various forms of therapies in a particular patient thereby helping to choose the ideal therapy for them.

The molecular subtype is one such predictive marker that has got prognostic significance as well [2] There are 5 molecular subtypes of breast cancer Luminal A, Luminal B, HER-2-neu enriched, Basal-like and Normal breast-like. Although these subtypes indicate the response to various forms of therapies, identifying the molecular subtype in each patient needs gene assays which are costly and needs advanced equipment as well. Such facilities might not be available routinely across all the centres, particularly in the developing and under-developed countries.

Hence surrogate subtyping of breast cancer has been proposed based on immunohistochemical features which give results closer to the molecular subtyping. However, the criteria for this surrogate subtyping has constantly undergone modifications based on various studies. The recent 15th St. Gallen International Expert Consensus Conference on Breast Cancer held in the year 2017 has further modified the criteria for this surrogate subtyping [1]^. Hence this study was done to assess the frequency of different intrinsic subtypes of breast cancer and their clinicopathological features based on these recent St. Gallen’s modifications.

Place of study: Our study was done at a tertiary care health centre in South India from April 2019 to June 2020.

Inclusion criteria: All cases of invasive breast carcinomas in which modified radical mastectomy had been done and received in the Department of

College of American Pathologists (ASCO/CAP) guidelines ER/PR were considered positive when 1% or more cells showed nuclear staining. For HER -2, if there were no membranous staining or faint incomplete membranous staining in less than 10% of cells, the score was given as zero.

Faint incomplete membranous staining in >10 % of cells was scored 1+. Weak to moderate complete membranous staining in >10% cells was scored 2+. Strong complete membranous staining in >10% cells was given a score of 3+. Cases with score 0 and 1 were considered HER-2 negative. Score 3+ was taken as HER-2 positive. Score 2 was taken as an equivocal result and Fluorescence in situ hybridization (FISH) was done in these cases to confirm HER-2 status.

Based on the expression of ER, PR and HER-2, all the cases were subtyped into four categories according to St. Gallen 2017 guidelines. These four categories were triple-negative (hormone receptor (HR) negative, HER-2 negative) tumours, HR negative HER-2-positive tumours, HR-positive HER-2 positive tumours and HR-positive HER-2 negative tumours. The last group was further classified into “Luminal A-like” (low grade) and “Luminal B-like” (high grade) tumours.

For grade II cases, Ki-67 immunohistochemical staining (PathnSitu) was done and when <20% cells were positive they were categorised as “Luminal Alike” tumours. When >20% cells were Ki-67 positive they were categorised as “Luminal B- like tumours”. Clinicopathological parameters like age, T (tumour size) staging, N (lymph node) staging and histopathological grading were studied in all these groups and compared.

Statistical analysis: Statistical analysis of the data was done using SPSS software 24.0 by Chi-square test. p-value less than 0.05 was considered statistically significant.

Percentage of larger tumours (T3) was maximum in the triple-negative subtype (30.8%) and HR negative HER-2 positive subtype (28.6%) when compared to other subtypes (Table IV). On the other hand, smaller tumours (T1) were more frequent in the HR-positive HER-2 negative group (29.1%).

In this subtype also smaller tumours were more common in “ luminal Alike” cases (42.1%) compared to “luminal B- like “ cases. Triple-negative and HR negative HER -2 positive rarely presented with T1 tumours. The association between various subtypes and tumour size was also found to be statistically significant (p-value < 0.001).

With regards to age, triple-negative (69.2%) subtype often presented in women of the age group of fewer than 50 years (Table V). All the other subtypes were more common in women of more than 50 years. On the statistical analysis of data, the association between various subtypes and age was not significant (p value=0.401).

Breast cancer is a heterogeneous disease with varied clinicopathological features. As a result all the cases do not respond uniformly to the different forms of therapy.

They further recommended this classification to be used in planning treatment strategy in individual breast carcinoma patients. HR-positive HER-2 negative was the most common subtype in our study. This in turn was due to the high frequency of “luminal Alike” cases. In the earlier classificatio, “ luminal A” used to be a separate subtype and was the most common subtype in many previous studies [3,5,6,7,8,9]. Studies on the Indian population done by Munjal et al [10] and Ashima et al [9] had also shown “luminal Alike” as the predominant subtype. Howeve few other Indian studies by Shabnam et al [11] and Kunheri et al [12] observed triple-negative and “luminal B -like” to be the most common subtypes respectively. It needs to be studied further whether this is a geographic variation in different parts of India due to some genetic and geographic factors.

Subtypes and tumour size: Our study found a significant association between the subtypes and the tumour size. The The proportion of tumours larger than 5cms was maximum intriple-negative subtype in our study. Spitale et al [3] had observed the same in their study. After thetriple-negative subtype, the frequency of such larger tumours was maximum in the HR negative HER-2 positive subtype. Studies by Onitillo et al [13] and Engstrom et al [14] had also observed that tumours larger than 5cm were frequent in this subtype. At the other end, tumours less than 2cms were frequent in theHR-positive HER-2 negative subtype particularly in the “Luminal Alike” cases in our study. Such an association between small tumour size and “Luminal Alike” cases was noted in many previous studies as well [3, 5, 7, 13, 14]. Although few studies have noted no significant association between subtypes and tumour size they seem infrequent [10,15 ].

Subtypes and lymph node metastasis: The The triple-negative type was frequently associated with the the highest stage of lymph node metastasis in our study. In contrast to our results nodal metastasis was frequently absent in triple-negative subtype in studies by Spitale et al and Vasconcelos et al [3, 5]. However many other study results by Zaha et al, Munjal et al, Kunheri et al and Blows et al were similar to our study strengthening our results [6,10,12,16]. Lymph node metastasis was often absent in “Luminal Alike” tumours in our study. This was similar to the results of Tang et al [17] and Onitillo et al [13].

Subtypes and tumour grade: Triple-negative subtype cases were often of the highest histological

While “Luminal A- like” tumours were often of smaller tumour size without nodal metastasis in many cases, “Luminal-B” like tumours were often of a larger tumours size with nodal metastasis.

One of the limitations of our study was we were not able to do a long term follow up of these patients to assess the difference in treatment response and survival rates among the different subtypes. This was because many patients were lost to follow-up. Another limitation was Ki-67 was not done in all cases due to economic constraints.

However this was not necessary in all cases as according to the St. Gallen guidelines, the grade of the tumour can be used instead of Ki-67 for classifying the “Luminal A- like” and “Luminal B- like” tumours. We followed the same guidelines. In cases where the grade was intermediate alone, we used Ki-67 for classification. Such an approach would be cost-effective for subtyping breast carcinomas in developing countries.

Breast cancer is a heterogeneous disease with different subtypes differing in their clinicopathological features. Since these subtypes are likely to differ in their treatment response as well, it is essential to identify the subtype in every individual patient.

Since molecular subtyping might not be affordable in developing and undeveloped countries, surrogate immunohistochemical subtyping can be used. We adopted the recent St. Gallen guidelines for this immunohistochemical subtyping and observed that the subtypes differed significantly in their clinicopathological features further validating the guidelines.

Further long term studies with a large sample size and adequate follow up will help in assessing the treatment response and the survival rates in these different subtypes. Such large scale studies will help in planning the treatment guidelines for each subtype particularly in the Indian population.

We thank all our Department faculties, Postgraduates and Technicians for their kind help during the study. I thank Mrs. Anitha (statistician) for helping me in statistical analysis.