Transcription factor GATA-3 binds to the DNA consensus sequence (A/T) GATA (A/G) 1 and plays a crucial role at multiple stages of breast development including luminal cell differentiation. 2 Therefore, GATA-3 dysregulation has been implicated in the pathogenesis of breast cancer. GATA-3 expression has been detected in the luminal breast cancer subtypes 345 , and somatic mutations of GATA-3 have been identified in ER-positive breast carcinomas. 67 GATA-3 shows promise as a sensitive and relatively specific marker for breast cancer as its expression is retained in "estrogen/progesterone receptor loss" metastases of breast carcinoma. 8 Several studies have also shown that higher levels of GATA-3 are associated with better survival in breast cancer 9 , making it a biomarker with potential independent prognostic value. Considering the relative paucity of studies aimed at understanding the role of GATA-3 in breast carcinoma in India, this study was conducted to detect GATA-3 expression using immunohistochemistry and determine its association with the different established prognostic parameters of breast carcinoma such as the histologic grade, morphological and molecular subtypes of tumor.

This cross-sectional, observational study was conducted in the Department of Pathology, NRSMCH, Kolkata on 110 breast cancer patients who underwent surgery between February 2019 and July 2020 and gave consent for the study. Relevant history and clinical findings were noted. Formalin-fixed and paraffin-embedded tissue blocks were cut at 4μm thickness, stained with haematoxylin and eosin and examined for routine microscopic findings such as tumor morphological type 10 (Figure 1).

Out of 97 cases positive for GATA-3 immunostaining, 35% cases were weak positive (score 2), 35% moderately positive (score 3) and 30% strong positive (score 4) for GATA-3.

GATA-3 expression was not significantly associated with age (P=0.539), sex (P=0.358), tumor laterality (P=0.797), history of menopause (P=0.740), history of breast feeding (P=0.216) and history of chemotherapy (P=0.143). Interestingly, a significant association was found between GATA-3 expression and family history of breast carcinoma with lower expression of GATA-3 in cases with positive family history (P=0.019), although this analysis might have been compromised by the small number of family history positive cases (Table 2).

GATA-3 expression was significantly lower with larger tumor size (P=0.002), number of axillary lymph nodes involved (P=0.014), higher histological grade (P=0.014), presence of lymphovascular invasion (P=0.007), in-situ carcinoma component (P=0.017) and positive margins (P=0.007) (Table 3). However, GATA-3 expression was not significantly associated with the tumor morphological type (P=0.063).

GATA 3 expression was significantly associated with ER/PR/HER2neu and ki67 expression with P value<0.001 for each variable. Higher GATA-3 expression was noted in positive ER/PR and low ki67 cases while lower expression was seen in Her-2-neu positive and triple negative cases (Table 4).

In this study, 97 out of 110 breast carcinoma cases (88.1%) expressed GATA-3. No statistically significant association of GATA-3 was found with age and sex of patient, tumor laterality, history of menopause, breast-feeding or chemotherapy. These findings corroborate the findings of the study of Ismail et al. 11 where nuclear expression of GATA-3 was detected in 36 (72%) cases and no association was found between GATA-3 expression and the age of the patients or tumor laterality. However, in our study, there was statistically significant lower expression of GATA-3 in the six cases (5.5%) with a positive family history (P=0.019).

In our study, 92 cases (83.6%) were diagnosed as invasive ductal carcinoma, no special type (IDCNST) and no significant association was found between GATA-3 expression and the morphological subtype of breast cancer (P=0.063). This is also similar to the study of Ismail et al. 11 where GATA-3 expression between cases of invasive ductal and lobular carcinoma breast was statistically insignificant (t=0.625).

Similar to the study by Albergaria et al. 18 and Voduc et al. 12 in our study, reduced GATA-3 expression was found in higher histological grade 3. GATA-3 expression was significantly associated with tumor grades (P<0.001), CIS (P=0.017) and LVI (P=0.007). All the cases with higher GATA-3 expression had uninvolved margin status while the cases with lower expression of GATA-3 had a statistically significant positive margin involvement (P=0.007). This may be related to the fact that GATA-3 is associated with reversal of epithelialmesenchymal transition and inhibits tumor metastasis. 13 It was seen that with an increase in the tumor size, there was reduced expression of GATA-3 (P=0.002). This finding corroborates the findings of the study of Mehra et al. 5 who found low GATA-3 expression was associated with larger tumor size (P=0.03). A significant association was found between GATA-3 expression and pN in our study (P=0.014) unlike the study of Voduc et al. 12 where there was no significant association between GATA-3 and lymph node status (Pearson chi-square test, P=0.47).

In our study, a significant association was found between expression of GATA-3 and the molecular types of breast cancer (P<0.001) with higher GATA-3 expression in the luminal subtypes and lower GATA-3 expression in the triple negative cases. A study by Ismail et al. 17 through the analysis of KW test demonstrated a significant association between GATA-3 expression in both luminal A and B-HER2-negative tumors when compared with GATA-3 expression in the triple-negative subtype. The same finding was reported by Jiang et al. 14 These observations agree with the hypothesis that GATA-3 mutations might be important in the aetiology of luminal-like breast cancers.

Individually, GATA-3 expression showed a significant association with ER, PR and Her-2-neu expression with P-value<0.001 in all cases. More cases of triple negative breast cancer were in the group that expressed low levels of GATA-3 and a significant association was found between the two in this study (P=<0.001). A high Ki-67value was found in all cases with low GATA-3 expression and a low ki67 value was observed for the majority of GATA-3-Grade 3 tumors. This finding is consistent with the study of Yildirim et al. 15 which revealed that GATA-3 has an inverse relationship with Ki-67. There was a significant association of GATA 3 expression with the ki67 status in the Grade 2 tumors (P<0.001).

We have inferred statistical significance on the basis of the chi-square test. However, the trends are to be interpreted as only preliminary in view of the limited numbers in some of the cross-tabulation categories. Moreover, as no data on follow-up was available, the prognostic significance of GATA-3 in breast carcinoma could not be assessed.

GATA-3 is expressed in most cases of breast carcinoma. GATA-3 expression is significantly associated with different known prognostic factors in breast cancer such as the size and histology grade of tumor, axillary lymph node status, LVI, CIS and the molecular subtypes of breast cancer including ki67 index. GATA-3 expression was higher in the luminal subtypes and lower in the triple negative cases. Ki-67 was high in all low GATA-3 expressing tumors. Further studies on GATA-3 could be undertaken to assess its potential to predict prognosis in breast carcinoma or develop novel therapeutic approaches in the future.

None.

Institutional Ethics Committee approval no/NMC/10018; dated: 08-01-19.

None.