CNS metastases of breast cancer show discordant immunohistochemical phenotype compared to primary

Abstract

Purpose

A challenge in management of breast cancer is the development of brain metastases (BM). Because of improvements in systemic therapy with longer survival of patients with advanced cancer, BM can appear at a time when extra-BM disease is under control. Development of potential preventive strategies are considered, and new developments in systemic approaches to treatment of BM, (cytotoxic/targeted therapy), are explored. In primary breast cancer, ER/PR, HER2 are important biological markers for predicting prognosis and making effective treatment decisions. Known are changes in markers due to metastases, but clinical significance is still unclear. Aim of this retrospective study is to detect changes in immunohistochemical markers of primary and BM, to recognize concordance and impact on prognosis.

Methods

Twenty-one consecutive primary breast cancer patients who developed BM and got surgical resection of BM were enrolled in this study. Matched-pair analyses of primary and BM were done with evaluation by immunostaining (ER, PR, HER2).

Results

Loss of ER/PR receptor positivity was seen in BM compared to primary (ER: 47.6 %/9.0 %; PR: 42.9 %/0 %), respectively. High concordance exists for HER2 status in primary and BM (>80 %). HER2-positive breast cancer had a shorter median interval until appearance of metastases than HER2-negative patients (32.1/39 months; p = n.s.).

Conclusion

With loss of receptor positivity (ER/PR) in BM treatment, decisions are very difficult. High concordance of HER2 status was seen in matched-pair analysis. Further studies had to investigate whether HER3/4 is a possible target for further therapy.

Introduction

Development of brain metastases (BM) in patients with breast cancer is an increasing clinical challenge and is associated with poor prognosis (Gabos et al. 2006; Leyland-Jones 2009). Brain metastases usually develop within 2 years after first diagnosis of breast cancer (Gabos et al. 2006; Pestalozzi et al. 2006). Data from an autopsy study showed that the incidence of BM in patients with breast cancer is approximately 30 % (Tsukada et al. 1983). Risk factors for the development of brain metastases include young age, poorly differentiated tumors (grade 3), estrogen receptor (ER)-negative or triple-negative (ER, PR, HER2) status and human epidermal growth factor-2 (HER2)-positive status (Anders et al. 2011; Gabos et al. 2006; Kallioniemi et al. 1991; Pestalozzi et al. 2006; Dawood et al. 2010). Risk of BM in patients with HER2-positive breast cancer has been highlighted by several authors (Leyland-Jones 2009; Musolino et al. 2011; Slimane et al. 2004), but only a few studies have examined the HER2 status of both the primary tumor and the BM (Fuchs et al. 2002). Previous studies have shown that the immunophenotype of distant breast cancer metastases (e.g., liver) may be different from that of primary, leading to inappropriate choice of systemic treatment, rarely BM are investigated in these studies. When patients with breast cancer develop distant metastases, choice of systemic treatment is usually based on tissue characteristics of primary tumor as determined by immunohistochemistry (IHC) (Taucher et al. 2003). Discordance in immunohistochemical phenotype between primary and metastatic sites may have therapeutic implications, but the knowledge for receptor conversion (ER/PR/HER2) for BM is still small, and clinical implications remain unknown (Pusztai et al. 2010). The aim is to investigate whether there is receptor conversion for ER, PR and HER2 in BM of breast cancer and its impact on prognosis.

Methods

Therefore, in a single center analysis, 107 consecutive patients who have developed BM following primary treatment for primary non-metastatic operable breast cancer were enrolled in this retrospective study. All patients were treated between January 1, 2001 and June 30, 2009 at the Department of Gynecology at the University of Tübingen, Germany. All 107 patients were eligible for inclusion in the present study.

At initial diagnosis of primary breast cancer, the patients underwent mastectomy or breast-conserving surgery with axillary lymph node dissection according to guidelines (AGO-Mamma). Median age of patients at diagnosis of breast cancer was 53.9 years (range: 35–73 years). Treatment decisions for each patient and choice of systemic adjuvant therapy were at the decision of the treating physician according to TNM stage and guidelines. In patients with HER2-positive primary tumors, trastuzumab could be administered as adjuvant therapy or for metastatic disease. Patients got a routine follow-up. Patients with primary metastatic breast cancer were excluded. Twenty-eight of 107 consecutive patients with BM underwent neurosurgical resection for BM and were further evaluated. In our collective, neurosurgical resection of BM was done by the known criteria for a neurosurgical intervention with possibility of R0 resection, solitary BM. All patients got a whole brain radiotherapy (WBRT) after neurosurgical intervention.

Twenty-one of 28 patients got breast cancer surgery and neurosurgical intervention at University Hospital Tübingen, and all findings were available; these patients were further evaluated. Samples from primary and metastatic lesions (BM) were obtained from all 21 patients for matched-pair analyses to investigate immunohistochemical status of primary and BM. Patients' characteristics are shown in Table 1. All findings were available from 18 patients and were further evaluated. All patients provided written informed consent for data acquisition and use of biopsy material. For this study, there exists a votum of the ethics committee of University of Tübingen (No. 171/2012R).

Table 1.

Patients’ characteristics: primary breast cancer (n = 21): hormone receptor (ER, PR) and HER2 status of the primary breast cancer tissue

Parameter	Number of patients (n)
Histology
Ductal carcinoma	18
Lobular carcinoma	2
Medullary carcinoma	1
T classification
T1	15
T2	5
T3	1
T4	0
Histologic grade
1 and 2	14
3	7
Axillary lymph node involvement
No	8
Yes	13
ER status (primary)
Positive	10
Negative	11
PR status (primary)
Positive	9
Negative	12
HER2 status (primary)
Positive	11
Negative	10

HER2 status was determined by IHC or in situ hybridization; ER and PR status were determined by IHC

Data were extracted for stage at initial diagnosis, age, histology, grading, node involvement, ER−/PR−, HER2 status of primary. Additionally, histology, ER−/PR−/HER2 status of BM of each patient were extracted. Time to BM was calculated from initial diagnosis of breast cancer to resection of BM. We examined conventional clinicopathologic factors including the biological factors (ER, PR, HER2) by immunohistochemistry (IHC) and their association with prognosis. ER and PR status was considered positive if >10 % of the cells stained positive by IHC. HER2 score was assessed by means of intensity and percentage of immunoreactive tumor cells. Intensity was recorded as 0 (no reaction) to 3 (strong reaction). Fraction of immunoreactive tumor cells was recorded as 0 (no positive cells), 1 (<10 % positive cells), 2 (10–50 % positive cells), or 3 (>50 % positive cells). HER2 scores of 0, 1 and 2 were considered negative, and a score of IHC 3+ was considered positive (Edgerton et al. 2003). HER2 in situ hybridization was performed in all cases of IHC conversion or when an equivocal (IHC 2+) result was obtained for the primary tumor or BM. Triple-negative subtype was defined as ER negative (<10 % of the cells stained positive by IHC), PR negative (<10 % of the cells stained positive by IHC) and HER2 negative (score 0/1/FISH negative). For all receptors, results were considered “concordant” if primary and metastatic lesions were both positive and negative using the above-mentioned criteria, and different combinations were considered “discordant” (Table 3).

Table 3.

Example for receptor conversion and non-conversion of ER/PR/HER2; this patient shows conversion for ER and PR and no conversion for HER 2

Patient	Primary tumor			CNS metastases
	ER	PR	HER2	ER	PR	HER2
1	+	+	–	–	–	–

Follow-up

Follow-up data were collected when patients presented in our department for checkup.

Mean follow-up time was 30.3 months, and data of all 21 patients were evaluated.

Statistical analysis

Data were stored in a database and analyzed using PASW (Version 19 SPSS Inc., Chicago, IL, USA). Multivariate analyses were performed with PASW (Version 19 SPSS Inc., Chicago, IL, USA). Results are expressed as means, standard deviations, minimums, maximums and percentages. Survival curves were calculated using Kaplan−Meier analyses for OS to calculate hazard ratio (HR) and 95 % CI. The log-rank test was used to test for significant differences between the groups. Student’s t test and Mann−Whitney test were used to test for significant differences between the different factors of interest. P values less than 0.05 were considered statistically significant.

Results

Matched-pair analyses of all 21 consecutive patients were done. Most frequent pathologic type was invasive ductal carcinoma (86 %); 33 % were grade 3 (Table 1). ER, PR and HER2 in primary were positive in 47.6, 42.8 and 52.4 %, respectively (Table 1).

Rarely, BM are the only location of metastases: Of 21 patients with neurosurgical intervention, 4 patients had isolated BM; 6 patients had metastases in another location and 11 patients had more than 1 additional location of metastases (Table 1). Median interval between first diagnosis of primary breast cancer and detection of BM was 35.4 months (7–85 months). Patients with HER2-positive breast cancer had a shorter median interval until appearence of metastases than HER2-negative breast cancer patients (32.1 months vs. 39 months; p: n.s.).

Conversion of ER/PR/HER2 status mostly was loss of receptor expression in BM (Table 2). ER, PR and HER2 in BM were positive in 11.1, 0 and 55.5 %, respectively (Table 2).

Table 2.

Eighteen patients; paired analysis of receptor expression in breast cancer and brain metastases

Patient (n = 18)	Primary n (%)	CNS metastases n (%)	Concordance n (%)
ER
Positive	9 (50)	2 (11.1)	2/9 (22.2)
Negative	9 (50)	16 (88.8)	9/9 (100)
PR
Positive	7 (38.8)	0 (0)	0 (0)
Negative	11 (61.1)	18 (100)	11/12 (91.6)
HER2
Positive	8 (44.4)	10 (55.5)	7/8 (87.5)
Negative	10 (55.5)	8 (44.4)	7/10 (70)
Triple-negative	3 (16.6)	7 (38.8)	3/3 (100)

Conversion in ER occurred in 7 of 18 patients (38.8 %), in 7 of 9 (77.7 %) patients from positive to negative of originally positive breast cancer (Table 2) without conversion from negative to positive (not significant). PR conversion occurred in 7 of 18 patients (38.8 %); BM showed PR receptor loss in all cases (Table 2). More often there is PR than ER receptor loss in BM.

HER2 status in primary was mostly concordant in BM (Table 2). Conversion in HER2 occurred in 4 of 18 patients (22.2 %); 10 patients had negative HER2 status in primary compared to 10 HER2 positive patients in BM (Table 2); 3 patients changed from HER2 positive to negative of originally positive breast cancer; 1 patient with HER2-negative primary changed to positive in BM (Table 2). This patient developed triple-negative BM (Table 2). HER status is high concordant in primary and BM (87.5 %/70 %; Table 2). Overall, the proportion of patients with triple-negative disease was higher in BM (38.8 %) than in primary (16.6 %). All three patients with primary triple-negative tumors had triple-negative BM (concordance: 100 %). Additional four patients with triple-negative BM were associated with loss of ER or PR positivity in 3 patients and loss of positive HER2 status in one patient (Table 2). Highest concordance was seen for HER2 status between primary and BM (about 80 %; Table 2). Changes in subtypes were seen in about 25 %; however, the lowest rate of change was seen in triple-negative cases. Regarding receptor conversion, there was a trend to shorter interval to BM with receptor conversion compared to patients without receptor conversion. Data not demonstrated (Table 3).

Discussion

Overall, 10–13 % of patients with primary breast cancer show HER2-positive tumors, and this is a known risk factor for BM (11). Results of our study obtained from 18 patients showed that receptor conversion does occur, but is relatively uncommon for HER2 compared to ER/PR. These data are consistent with reports from literature showing receptor discordance between primary and metastatic sites, most frequently with PR status, particularly in brain, liver and gastrointestinal metastases (Broom et al. 2009; Idirisinghe et al. 2010). Negativity of ER/PR is probably explained by a dedifferentiation of metastatic tissue. If this phenomenon can be explained due to a performed endocrine therapy is unclear. Although discordance between HER2 status of primary and metastatic breast cancer has been reported by some authors (Edgerton et al. 2003; Yonemori et al. 2007, 2008), evidence from our analysis and other studies (Fuchs et al. 2002; Lear-Kaul et al. 2003; Broom et al. 2009; Pusztai et al. 2010) show that HER2 status of BM and primary tumors is highly concordant, indicating that HER2 testing of primary breast cancer is predictive of HER2 status of BM. Fuchs investigated whether HER2 testing of BM tissue rather than primary breast cancer is warranted (Fuchs et al. 2002). Possibility of a change in receptor status is an important consideration because it may require modification of treatment strategy, particularly in the era of targeted therapy (Simmons et al. 2009). Until now knowledge for receptor conversion (ER/PR/HER2) for BM in breast cancer and their impact on prognosis is still small and clinical implications remain unknown.

HER2 status of primary breast cancer appears to be the most important factor determining development of BM. Multivariate analysis has shown that HER2-positive status, trastuzumab treatment and hormone receptor negativity are independent risk factors for the development of BM (Bria et al. 2008). In addition, patients with HER2-positive primary breast cancer were shown to have a higher risk for and shorter time to the development of BM than HER2-negative tumors. Despite achieving excellent control of extracranial disease, trastuzumab has no known direct central nervous action because it is not able to pass the intact BBB. In patients with HER2-positive breast cancer and BM, continuation of trastuzumab may significantly prolong OS (Bartsch et al. 2007; Church et al. 2008; Dawood et al. 2008; Lower et al. 2003; Park et al. 2009). In a multivariate model (Dawood et al. 2008), risk of death was increased in patients with HER2-negative disease (hazard ratio [HR] 1.66, 95 % CI 1.31–2.12, P < 0.0001) and those with HER2-positive disease who had never received trastuzumab (HR 1.34, 95 % CI 0.78–2.30, P = 0.28) compared to patients with HER2-positive disease who had received trastuzumab before or at the time BM were diagnosed (Romond et al. 2005; Smith et al. 2007). Phase III studies with HER2-targeted therapy in combination with chemotherapy (lapatinib + capecitabine) have shown positive impact on prognosis in BM of breast cancer. So targeted therapy plays an important role in metastatic breast cancer.

Consequently, current standard of care for patients with BM is WBRT, with or without surgery, or stereotactic radiosurgery. In the future, novel therapies or combinations of therapies may provide further benefit in these patients, and perhaps in special cases, radiotherapy could be done delayed after neurosurgical intervention (HER2 positive BM). Possibly further targets of HER group (e.g., HER3) could be used as a new therapeutic intervention (Koutras et al. 2010; Da Silva et al. 2010). Until now there are only few data on this topic.

The subgroup of patients with triple-negative primary breast cancer had poor prognosis (Dawood et al. 2009, 2011). High concordance between primary and BM was seen in triple-negative disease in our study (Table 2). In study population, 3 of 10 patients with HER2-negative primary breast cancer had primary triple-negative tumors, whereas the rate of triple-negative BM was notably higher (38.8 %, Table 2). Development of triple-negative breast cancer may worsen prognosis and makes further treatment challenging because there are no specifically identified therapeutic targets and presents additional challenges for further therapy, even in case of CNS lesion. In literature deregulated HER family receptors, particularly HER3, their downstream pathways are implicated in colonization of brain metastasis. If other HER-targeted molecules will play a role in therapy of metastatic breast cancer to CNS is still unknown and had to be investigated in upcoming studies.

With a small number of patients in this study, interesting results were found, but clinical implications or definitive conclusions cannot be made at present. The increased risk of BM in patients with HER2-positive primary breast cancer should be considered in the overall management strategy from the outset. There is no need to determine the HER2 status of BM if the status of the primary tumor is known. Continuation of trastuzumab in patients with HER2-positive primary or metastatic breast cancer (BM) is associated with improved survival, but WBRT (with or without surgery) or stereotactic radiosurgery remains the standard of care (Dawood et al. 2010).

However, natural history of HER2-positive metastatic breast cancer has significantly improved following the introduction of anti-HER2 monoclonal antibody trastuzumab, which significantly improves progression-free survival (PFS) and overall survival (OS) when combined with chemotherapy (Slamon et al. 2001; Marty et al. 2005).

Conclusion

HER2 status of primary breast cancer appears to be one of the most important factors for the development of BM. The incidence of brain metastases in breast cancer is very low, so that a general screening is not required. Studies have shown that a HER2-targeted therapy can improve prognosis. Although receptor conversion assessed by IHC does occur between primary breast tumors and BM, this phenomenon is relatively uncommon, particularly for HER2. Therefore, concordance between HER2 status of primary and BM can be expected. Additionally in our study, discordance for ER and PR was seen in BM with negative receptor status in most cases; this is an important consideration because it may require modification of treatment strategy, particularly in the era of targeted therapy. Further studies are required to investigate the validity of this assumption on treatment strategies.

Possibly HER3/4 status is an important factor in the development of BM and could provide a potential target. If other factors, like HER3/4 specific agents, can improve prognosis is still unclear. Nevertheless, further studies must use HER2 status as a predictive factor for selection of targeted therapies, to avoid or delay development of brain metastases in high-risk patients. Currently, extended therapy with trastuzumab and/or prophylactic brain irradiation may improve prognosis, but additional studies are required to show the impact of this treatment strategy on survival.