1. Background

Breast cancer (BC) is the most common malignancy in females in Saudi Arabia, with evidence of a surge in annual incidence from 23.5 /100,000 in 2000 to 34.5 cases/100,000 in 2010 [1, 2]. Neoadjuvant chemotherapy (NAC) followed by definitive surgical resection is the usual therapeutic approach for locally advanced breast cancer (LABC). Besides improving the operability of these patients by downstaging their primary tumors, NAC is a pragmatic approach to preoperatively test chemosensitivity on an individual basis [3, 4].

Performing core needle biopsy (CNB) is a common practice to confirm the diagnosis and determine the presence of immunohistochemical (IHC) markers such as hormone (estrogen and progesterone) receptor (HR) and human epidermal growth factor receptor 2 (HER2), which are important prognostic indicators as well as key factors in the decision-making process regarding further adjuvant therapy [5, 6].

Changes in HR and HER2 receptor status following NAC have been interesting, yet conflicting, research areas for the past couple of years. While several studies suggested that the expression of these receptors is altered after NAC [7–12], others indicated that they remained stable [13, 14]. Moreover, it has been shown that BC patients demonstrating a conversion from HR (+) to HR (−) tended to benefit less from NAC, compared to those with no change or the opposite conversion; from HR (−) to HR (+) [7]. Breast cancer is the most common type of female malignancies in Saudi Arabia with little is known about its pathobiology [15]. Moreover, to the best of our knowledge, there is little information on the prognostic impact of receptor conversion caused post-NAC in patients with LABC. Therefore, we aimed in the current to assess any discordance in the expression profiles of HR and HER2 in patients with residual tumors after NAC, the rate of such discordance (if any), and the prognostic significance of such receptor changes in terms of survival outcomes.

2. Materials and methods

3. Results

3.1. Patient characteristics

The clinical characteristics of 91 patients with FFPE of pre-NAC and matched residual disease at the surgical specimens are detailed in Table 1. The median age of our cohort was 47 years (range of 23 to 74 years). The initial (pre-NAC) expression of ER, PR, and HER2 receptors was negative in 22 (24%), 28 (31%), and 59 (65%) of patients, respectively. There was a statistically significant difference in the patient’s clinical characteristics (age, menopausal status), tumor characteristics (stage, histopathology, and grade), pre-NAC expressions of ER, PR, and HER2 receptors, as well as the NAC regimens.

Table 1

Clinical characteristics of the cohort (n = 91)*.

Characteristic	Number	%	P-value
Age, years
Median (range)	47.0 (23.0–74.0)
Age groups			<0.001¶
20–40 years	24	26.4
40–60 years	51	56
≥ 60 years	16	17.6
Menopausal status
Premenopausal	57	63	0.021
Postmenopausal	34	37
Stage (pre-NAC)
II	31	34	0.003
III	60	66
Histopathology
Invasive ductal	86	95	<0.001¶
Invasive lobular	5	5
Tumor grade
Low	1	1	<0.001¶
Intermediate	43	47
High	47	52
Molecular subtypes			<0.001¶
Luminal	47	51.6
Triple-negative	12	13.2
ER+PR+HER2+	22	24.2
ER-PR-HER2+	10	11.0
Initial ER status
Negative	22	24	<0.001¶
Positive	69	76
Initial PR status
Negative	28	31	<0.001¶
Positive	63	69
Initial HER2 status
Negative	59	65	0.006
Positive	32	35
LVI
Negative	42	46	0.529
Positive	49	54
NAC regimens
FEC-D	53	58	<0.001¶
AC-T	23	25
TAC	15	17

NAC; neoadjuvant chemotherapy, ER; estrogen receptors, PR; progesterone receptors, HER2; Human epidermal growth factor receptor-2, LVI; lymphovascular invasion, FEC-D; 5- fluorouracil, epirubicin, cyclophosphamide-Taxotere, AC-T; cyclophosphamide, doxorubicin-paclitaxel, TAC; Taxotere, Adriamycin-cyclophosphamide

* Non-parametric One-sample Chi-square test

^¶ Non-parametric Binominal test

3.2. Conversion in receptor status

The statuses of hormone receptors in the 91-matched initial pre-NAC biopsies and post-NAC surgical specimens were analyzed using the McNemar test. The expression of ER remained the same in 74 (81.3%) specimens. One initially ER-negative tumor turned ER-positive and 16 (17.5%) initially ER-positive ones turned into ER-negative post NAC. The difference in ER expression pre and post-NAC was highly significant (p<0.001). PR expression remained the same in 76 (83.5%) of the specimens. Three (3.3%) initially PR negative tumors turned PR positive and 12 (13.2%) initially PR positive tumors turned PR negative post-NAC. The difference in PR expression pre and post-NAC was statistically significant (p = 0.035). In sum, the expression of both ER and PR remained the same in 70 (77%) specimens, while there was a conversion of both ER and PR in 21 (23%) ones. Four (4.4%) specimens were ER stable PR converted, while 6 (6.6%) were PR stable ER converted. Thus, 21 (23.1%) of 91 patients had ER and/or PR conversion after NAC; with a highly significant difference between patients with stable receptors and those with any receptor (ER and/or PR) conversion; p< 0.001. For HER2 expression, it remained the same in 76 (83.5%) specimens. Five (5.5%) initially HER2 negative tumors became HER2 positive and 10 (11%) initially HER2 positive tumors became HER2 negative after NAC. The difference in HER2 expression level before and after NAC was not statistically significant (p = 0.302). Table 2 shows the details of receptor status conversion among the study cohorts.

Table 2

Baseline and Post-NAC hormonal receptors’ status among the study cohorts (n = 91).

Changes in estrogen receptor (ER) status post-NAC
Pre-NAC/Post-NAC	No	%	P-value
ER-/ER-	21	23.0	p<0.001¶
ER-/ER+	1	1.0
ER+/ER-	16	17.5
ER+/ER+	53	48.5
Changes in progesterone receptor (PR) status post-NAC
Pre-NAC/Post-NAC	No	%	P-value
PR-/PR-	25	27.4	0.035¶
PR-/PR+	3	3.3
PR+/PR-	12	13.2
PR+/PR+	51	56.1
Any receptor changes post-NAC
Receptors stable vs conversion	No	%	P-value
Both ER&PR stable	70	77.0	P<0.001¶
Both ER&PR converted	11	12.0
ER stable PR converted	4	4.4
PR stable ER converted	6	6.6
Changes in HER2 receptor (HER2) status post-NAC
Pre-NAC/Post-NAC	No	%	P-value
HER2-/HER2-	54	59.3	0.302¶
HER2-/HER2+	5	5.5
HER2+/HER2-	10	11.0
HER2+/HER2+	22	24.2

^¶McNemar test

3.3. Survival analysis and prognostic impact of HR conversion

During the follow-up time (range, 16–149 months), 23 (25.3%) of 91 patients had died, and 13 (14.3%) of 91 patients had experienced disease recurrence. The median overall survival was 77.8 months, while the median DFS was 67.4 months. The relationship between the patients clinicopathological and post-NAC receptor changes on one hand and both 10-years DFS and OS, on the other hand, is shown in Tables 3 and 4. In Table 3, univariate analysis between the patient characteristics and DFS showed significant relations with the post-NAC expressions of ER, PR, and any receptor conversion and molecular subtypes, log-rank test; 12.338, p <0.001, 14.296, p <0.001, 10.547, p <0.001 and 19.934, p <0.001; respectively. However, in multivariate Cox regression analysis, none of these variables were independently related to 10-years DFS. Kaplan-Meier curves for DFS in the patients’ groups of stable and converted ER, PR, and any receptor are shown in Figs 1–3, respectively. The differences among the curves were statistically significant as determined by the log-rank test (P <0.001, P <0.001, and p<0.001), respectively. In Table 4, univariate analysis between the patient characteristics and the 10-years OS showed that only molecular subtypes had a significant relationship with overall survival, log-rank test 21.197, p = <0.001.

Fig 1

Kaplan-Meier curves of Disease-Free Survival (DFS) in patients with stable and converted estrogen receptors (ER) following neoadjuvant chemotherapy.

Log-rank test was significant (P < 0.001).

Fig 2

Kaplan-Meier curves of Disease-Free Survival (DFS) in patients with stable and converted Progesterone Receptors (PR) following neoadjuvant chemotherapy.

Log-rank test was significant (P < 0.001).

Fig 3

Kaplan-Meier curves of Disease-Free Survival (DFS) in patients with stable and converted any receptor (ER and/or PR) statuses following neoadjuvant chemotherapy.

Log-rank test was significant (P< 0.001).

Table 3

The effects of clinicopathological features on 10-years Disease-Free Survival (DFS); univariate and multivariate analyses.

Characteristic	N (%)	10-years Disease-free survival (DFS)
		Univariate		Multivariate
		HR (95%) CI)	P	HR (95% CI)	P
Age groups
20–40 y	24 (26)	0.278 (0.396–2.934)	0.840
40–60 y	51 (56)
≥ 60 y	16 (18)
Menopausal status
Premenopausal	57 (63)	1 1	0.143
Postmenopausal	34 (37)	. 0.734 (0.263–2.053)
Stage (before NAC)
II	31 (34)	0 0.664 (0.291–2.453)	0.293
III	60 (66)
Histopathology
Invasive ductal	86 (95)	1 0.452 (0.482–1.882)	0.787
Invasive lobular	5 (5)
Tumor grade
Low	1 (1)	1 0.631 (0.682–2.221)	0.652
Intermediate	43 (47)
High	47 (52)
Molecular subtypes
Luminal	47 (52)	2.775 (1.104–4.900)	<0.001	1.835 (0.997–3.532)	0.051
Triple negative	12(13)
ER+PR+HER2+	22 (24)
ER-PR-HER2+	10 (11)
Post-NAC ER status
Stable	74 (82)	1 1	<0.001	1	0.196
Converted	17 (18)	3.088 (1.815–5.253)		1.726 (0.632–4.363)
Post-NAC PR status
Stable	76 (84)	1 1	<0.001	1	0.053
Converted	15 (16)	1 3.855 (1.275–11.661)		2.410 (0.997–5.827)
Any receptor change
Stable	70 (77)	1	<0.001	1	0.618
Converted	21 (23)	2.189 (0.052–1.690)		0.684 (0.154–3.044)
Post-NAC HER2 status
Stable	76 (84)	1 1	0.143
Converted	15 (16)	0.974 (0.468–2.030)
LVI
Negative	42 (46)	1 1	0.441
Positive	49 (54)	0.723 (0.443–1.527)
NAC regimens
FEC-D	53 (58)	1 0.620 (0.582–1.981)	0.629
AC-T	23 (25)
TAC	15 (17)

Table 4

The effects of clinicopathological features on Overall Survival (OS); univariate analysis.

Characteristic	N (%)	10-years Overall survival (OS)
		Univariate analysis
		HR (95% CI)	P
Age groups	24 (26)	0.878 (0.396–2.934)	0.433
20–40 y	51 (56)
40–60 y	16 (18)
≥ 60 y	24 (26)
Menopausal status
Premenopausal	57 (63)	1	0.551
Postmenopausal	34 (37)	0.640 (0.515–1.714)
Stage (before NAC)
II	31 (34)	1.578 (0.948–5.565)	0.092
III	60 (66)
Histopathology
Invasive ductal	86 (95)	0.927 (0.296–2.234)	0.161
Invasive lobular	5 (5)
Tumour grade
Low	1 (1)	0.667 (0.496–1.924)	0.272
Intermediate	43 (47)
High	47 (52)
Molecular subtypes		2.678 (0.948–7.565)	<0.001
Luminal	47 (52)
Triple negative	12(13)
ER+PR+HER2+	22 (24)
ER-PR-HER2+	10(11)
Post-NAC ER status
Stable	74 (82)	1	0.062
Converted	17 (18)	2.075 (1.104–3.900)
Post-NAC PR status
Stable	76 (84)	1	0.065
Converted	15 (16)	1.620 (0.423–4.636)
Any receptor change
Stable	70 (77)	1	0.053
Converted	21 (23)	1.987 (0.083–1.810)
Post-NAC HER2 status
Stable	76 (84)	1	0.070
Converted	15 (16)	1.630 (0.631–2.805)
LVI
Negative	42 (46)	1	0.413
Positive	49 (54)	0.887 (0.396–2.544)
NAC regimens
FEC-D	53 (58)	0.351(0.353–1.197)	0.925
AC-T	23 (25)
TAC	15 (17)

4. Discussion

To the best of the authors’ knowledge, this is the first study that addresses the conversion in the hormonal status (pre-and post-NAC) and its prognostic significance in a well-characterized cohort of Saudi patients with LABC. Our results showed a conversion rate of 23.1% with a highly significant difference between patients with stable receptors and those with any receptor conversion. There was a prognostic influence of these hormonal changes on the 10-years DFS but none of them were independently related. There is a significant surge in the incidence of BC among our cohort, which occurs at an earlier age than in western countries [2].

Neoadjuvant chemotherapy (NAC) or primary systemic therapy is considered the standard treatment for LABC aiming towards a reduction in tumor size, increasing the chances of conservative surgery, and improvement of the cosmetic outcome. [3]. Besides, beyond initiating an early systemic treatment for clinically undetectable micro-metastases, NAC also provides an opportunity to evaluate the tumor sensitivity to different chemotherapeutic regimens [4].

It has been reported that discordance of hormonal expression before and after NAC ranges from 2.5 to 51.7% and the conversion of positive Her 2 neu expression to negative one was reported in 43% of patients with BC [18–22]. Several molecular mechanisms have been proposed to mediate primary or acquired resistance to ETs in HR+ BC. These mechanisms include mutations in the ESR1 gene, which encodes ERα, or alterations in the genes of the mitogen-activated protein kinase (MAPK) pathway (e.g., activating ERBB2 mutations and NF1 loss-of-function mutations) or ESR1 transcriptional regulators [23, 24].

In the current study, we had observed a conversion rate of 23.1% which is in concordance with those previously reported figures, particularly those of Tacca et al [18] and Yang et al [19]. Several studies have suggested that the expression of these receptors is altered after NAC [7–12], others indicated that they remained stable [13, 14]. Two studies [9, 25] have summarized data and concluded that NAC seemed able to change ER and PR receptors expression and status, but HER2 amplification appeared to be more stable. Discordance of the HR status was reported in 8–33% of the patients, and the authors recommended that retesting the receptor status of the residual tumor after NAC should be considered to improve the future tailored adjuvant therapies [8]. In a pooled meta-analysis of 14 eligible studies, Zhang et al [9] had demonstrated that HR status (ER and PR) can be significantly changed with the use of NAC, with significant differences in the reported ER and PR status; however, the effect of NAC on HER2 and Ki67 were not statistically significant. In the current retrospective analysis of 91 patients, our results confirmed the previously reported results [7–12], we observed conversion of HR expression status after NAC in 23.1% of our cohorts. More importantly, these hormonal changes had a prognostic significance as they were correlated significantly with the 10-years DFS. Interestingly, 17.5% and 13.2% of ER-positive and PR-positive tumors changed to ER-negative and PR-negative, while only 1% and 3.3% of ER-negative and PR-negative tumors changed to ER-positive and PR-positive respectively. Moreover, these results are similar to what was reported in the previous studies [10, 20]. Except for random changes due to small-sized samples, laboratory procedures, observer variability, the possible mechanisms for a change in receptor status or expression in BC cells after chemotherapy are complicated [20]. Intra-tumoral heterogeneity might lead to several different clones with varied phenotypes within individual tumors [22]. Even within the same tumor, some clones are HR (+) while others are HR (−). Likewise, HER2 (+) cells are not evenly distributed within individual tumors. The sensitivity to chemotherapy also differs between different clones, with HR (−) tumor cells being more sensitive to chemotherapy than HR (+) tumors cells, which are aptly named insensitive tumor cells [26], and are left behind as part of the residual disease after NAC [8, 18]. Another established mechanism of HR status conversion is the ER downregulation caused by NAC itself. It is reported that chemotherapy can suppress ovarian function and adrenal glands [27, 28], and the decrease in the circulating levels of hormone resulting from this suppression might alter the HR status of residual tumors from (+) to (−) post NAC [8]. This mechanism is considered to be the main reason for the switch of HR (+) to HR (−) post NAC, which was observed frequently in our cohorts. Other explanations for the conversion of receptor status include genetic mutations [29], staining techniques, and statistical errors [11].

The 10-years survival analysis of our cohort revealed worse DFS in patients with converted receptor status (ER, PR, and any receptor change) than those with stable statuses, respectively. Moreover, the DFS was significantly correlated with the molecular subtypes but in Cox Regression analysis none of these factors were independently correlated.

Except for molecular subtypes, no significant correlation was found for the OS. There is a dearth of literature on the prognostic value of changed receptor status. Few studies have demonstrated the relations between HR conversion and treatment response, but discordant conclusions were observed. Chen et al [7] reported that patients with an HR (+) to (-) switch, benefit less from endocrine therapy compared to those whose HR status remains stable. In contrast, Tacca et al [18] and Buchholz et al [30] observed that there were no significant differences in PFS and OS rates between endocrine therapy-administered to patients with HR-negative switch lesions, and those with HR-positive lesions, both pre and post-NAC. Regardless, the authors demonstrated that a positive switch of the HR-status could be an indicator for a better outcome. Of note, despite the differences in the patients’ numbers and epidemiologic characteristics, our finding of prognostic significance between HR conversion and only DFS are similar to those reported by the recent large retrospective trial came from MD Anderson Cancer Center [31] which reported that is a total of 398 women-any receptor change was correlated with increased relapse-free survival, but no correlation with OS. Still, however, our findings are different from those observed by Chen et al [7] and Ozmen et al [32] who reported that change in HR status was an independent predictor for a poorer DFS as well as OS.

The strengths of our study are being in agreement with some published data, significant changes in IHC expression of ER, PR in LABC post neoadjuvant chemotherapy has been detected. More importantly, post-NAC expressions of ER and PR were independently related to DFS. Inevitably, the current study has some limitations of being a retrospective study, including a relatively small number of patients and the existence of tumor heterogeneity has led to concerns that core biopsies may not be representing the whole tumor tissue as they are often restricted to the superficial aspects of the tumor.

5. Conclusion

Patients with LABC had significant changes in the ER and PR receptor expression status after neoadjuvant chemotherapy. Post-NAC expressions of ER and PR were correlated with DFS. Until more comparable and/or prospective studies are carried out, retesting of the hormone receptors should be considered after NAC in patients with LABC. Further prospective studies enrolling a larger number of patients are warranted to understand the relationship between NAC and hormonal pathways and to explore the strategies of more therapeutic benefit.