HER2 status in operable breast cancers from Vietnamese women: Analysis by immunohistochemistry (IHC) and automated silver enhanced in situ hybridization (SISH)

Abstract

Background. HER2 amplification has been reported to occur in 19–30% of breast cancers in women from Western countries. Little is known about the HER2 status in breast cancers from Vietnamese women. The aim of this study was to assess the HER2 status in Vietnamese women with operable breast cancer using immunohistochemistry and silver in situ hybridization techniques. Material and method. Tissue blocks from 242 Vietnamese women with invasive primary breast carcinoma were stained by immunohistochemistry (IHC) and a dual silver in situ hybridization (SISH) for assessing HER2 status. The analysis followed international recommendations with a semi-quantitative grading of the reaction in four levels, “0”, “1+”, “2+” and “3+”. The HER2 gene amplification was assessed by calculating the ratio of HER2/chromosome 17 in 20 tumor cell nuclei. A ratio of <1.8 was classified as non-amplification and a ratio >2.2 indicated tumors with gene amplification. A ratio between 1.8 and 2.2 was equivocal. Results. Using IHC, 39% of the tumors strongly expressed “3+” the HER2 protein. An intermediate level “2+” of the protein was found in 11% while 50% showed no or low “0/1+” levels of protein expression. Gene amplification was found in 95% and 46% of the tumors with high “3+” and intermediate “2+” levels of protein expression, respectively. Thus, 41% of the patients had tumors with HER2 amplification. The concordance between IHC and SISH was 87%. Gene amplification was more frequent in ER negative than in ER positive tumors, 57% versus 33%, respectively. Additionally, tumors from postmenopausal women were amplified in 55% as compared to 36% in premenopausal women. Conclusions. HER2 gene amplification was found in 41% of Vietnamese breast cancers; this level is considerably higher than that previously found in a series of Swedish breast cancer women. The HER2-positive tumors were more often found in post-menopausal women than in ER negative tumors.

The incidence rate of breast cancer in Vietnam is low, 17.5/100 000 women, while the incidence rate for Vietnamese women living in USA is double that value [1,2]. These figures are much lower than equivalent ones for incidence rates of women originating from Western countries.

Previous studies have reported that 19–30% primary breast cancers in Western women showed overexpression of the HER2 protein or gene amplification [3–5]. The gene for human epidermal growth receptor 2 (ERBB2 or HER2) is an oncogene that encodes for a transmembrance glycoprotein with tyrosine kinase activity. Overexpression of HER2 protein and/or HER2 oncogene amplification is associated to a significant level with poor prognoses and has been shown to be a prerequisite for responses to the monoclonal antibody raised against HER2, i.e. trastuzumab. Moreover, HER2-positive tumors respond poorly to endocrine therapy [6]. There are several procedures available with which to assess the HER2 status in tumors such as protein expression by immunohistochemistry (IHC) or gene expression using either fluorescence in situ hybridization (FISH), chromogenic in situ hybridization (CISH), real time polymerase chain reaction (RT-PCR) or mRNA by Northern blotting [7,8]. Today, FISH testing is considered as a gold standard [3]. However, FISH employs fluorescence microscopy, which is a rather laborious technique and is often not available in pathology laboratories.

In recent years, a new method, silver in situ hybridization (SISH), has been reported to be a reliable technique with which to assess gene copies of HER2. Most previous studies have found that the concordance between SISH and FISH is high [7,9,10]. In addition, HER2 testing using the SISH method showed a good correlation with breast cancer-specific survival which therefore confirmed the clinical validity of this technique [8]. There are only a few reports on the HER2 status in breast cancers from Asian women. One study showed that 26% premenopausal Vietnamese and Chinese women with operable breast cancers had tumors that overexpressed the HER2 protein [11]. To date there is no study reporting HER2 gene amplification in breast cancer from Vietnamese women. One study of Chinese women reported that 65% the patients had tumors with HER2 gene amplification as analyzed by FISH [12]. Moreover, in another related study, it was found that 32% breast cancers from Asian women living in Shanghai showed HER2 gene amplification [13]. The aim of the current study was to assess the HER2 status in pre- and postmenopausal Vietnamese women with operable breast cancer using both immunohistochemistry and the SISH technique.

Material and methods

Clinical materials

Primary breast tumor specimens were randomly selected from a total of 242 Vietnamese women with histologically proven, invasive, operable breast carcinomas. These patients were operated on between 2002 and 2003 at the National Cancer Hospital in Hanoi, Vietnam, with either modified radical mastectomy or breast conserving surgery plus axillary dissection. Clinical information recorded of these women included age at diagnosis, clinical TNM stage including the stages I, II and III according to the UICC (Union for International Cancer Control) criteria. Patients with stage IV and inflammatory cancers were excluded. The women in this study were defined as pre-menopausal if they had had at least one menstrual period in 12 months preceding breast cancer surgery. Histopathological findings were classified according to the current World Health Organization criteria. Patients with in situ carcinomas were excluded from this study. All original breast tumor tissues were sectioned to confirm the diagnosis as invasive carcinomas and tumor grading was assessed according to the Scarff-Bloom-Richardson system.

A series of 2225 Swedish women operated on between 2007 and 2008 for invasive breast cancer in the Stockholm region, Sweden, was used to compare their HER2 status with that of Vietnamese patients.

Immunohistochemistry analysis

Methods for tissue handling, including fixation, were as the previously described [11]. The 4μm-thick sections for IHC and SISH were paralleled to sections for ER (Estrogen receptor) and PR (Progesterone receptor) analyses. Immunohistochemistry was performed using the Ventana HX automatic system BenchMark (Ventana Medical Systems, SA, lllkirch Cedex, France), with an anti-HER2 rabbit monoclonal antibody (clone 4B5) which identifies the internal domain of the c-erbB-2 oncoprotein (HER2). After antigen retrieval and blocking of endogenous peroxidase activity, the tissue sections were incubated with the primary antibody for 32 min at 42°C. Immunoperoxidase staining was performed using the LSAB system, NeuVision, according to the instructions of manufacturer (Ventana), and sections were then counterstained with haematoxylin. Positive tissue controls were run with every set of stained slides. In the cases of tumors of the invasive carcinoma type with a ductal carcinoma in situ component, scoring was done in the invasive area. Positive staining, i.e. “3+”, was defined as strong, complete, homogeneous membrane staining (“chicken-wire” pattern) in >30% of cells. Equivocal staining, i.e. “2+”, was assessed as moderate/strong, incomplete membrane staining in less than 30% of cells. No staining, i.e. “0”, or weak, incomplete membrane staining, i.e. “1+”, in any percentage of cells were categorized as negative. Two investigators VHT and LS evaluated independently the stained tissue samples.

SISH analysis

Automated SISH staining was performed using the Ventana Benchmark XT (Ventana Medical System), according to the manufacturer's protocols as well as methodology used in a previous study [9] for the INFORM HER2 DNA and Chromosome 17 probes. Both probes were labelled with dinitrophenol (DNP) and visualized using the rabbit anti-DNP primary antibody and Ultra view SISH Detection kit. The sections were denatured at 95°C for four minutes and hybridization with the chromosome 17 probe was carried out at 44°C for two hours. A second denaturation at 95°C for eight minutes was performed and hybridization with the HER2 probe was done at 52°C for two hours. The final reaction product was metallic silver, which was driven by the sequential addition of silver acetate, hydroquinone and hydrogen peroxide to the peroxidase-conjugated goat anti-rabbit antibody of the detection kit whereby silver was deposited in the nuclei of cancer cells. A single copy of the HER2 gene and chromosome 17 was visualized as one black and one pink dot, respectively. The slides were then counterstained for interpretation by light microscopy according to the guidelines given at http://www.her2sish.com.

SISH scoring was carried out independently following the manufacturer's recommendations. The adequacy of staining was assessed by examining each slide for HER2 and chromosome 17 signals identified as black and pink dots, respectively, in the nuclei of normal cells, which include fibroblasts, endothelial cells or normal epithelial cells. The entire stained section was then scanned and an area of the invasive tumor with the most easily identifiable signal deposits was chosen for counting. A total of 20 cancer cells was counted for dots representing both HER2, and chromosome 17. A single dot was counted as one, a small cluster as six and a large cluster as 12. Overlapping nuclei or cells in which black dots were found in the cytoplasm were not counted. Finally, the ratio of HER2 gene/ chromosome 17 was calculated by dividing the total score for HER2 by the total score for chromosome 17. A ratio of <1.8 indicated that the HER2 gene was not amplified, whereas a ratio of >2.2 indicated amplification of the gene. In tumors with a ratio between 1.8 and 2.2, a final decision on the degree of amplification was made after counting an additional 20 nuclei.

FISH analysis

In the series of 2225 Swedish women, FISH analysis was carried out as a part of histological routine using PatVision (Vysis/Abbott), including a probe to the gene and a centromere chromosome 17. Amplification status of the HER2 gene was as classified according to the manufacturer's instructions. The cut-off point was set at 2.0 copies per cell. More than 2.0 gene copies per cell was interpreted as amplification [14].

Statistical analysis

The SPSS statistical program, version 15.0 (SPSS, Los Angeles, CA), was used for statistical analysis. The χ² test was used to examine the categorical variables and the association between HER2 status and other clinicopathological variables in a univariate analysis with the odds ratio (OR) on a 95% confidence interval (CI). A value <0.05 was taken as statically significant. Kappa statistics with values less than 0.4 represented poor agreement, values between 0.4 to 0.6 moderate agreement, values between 0.6 to 0.8 good agreement and values of more than 0.8 excellent agreement.

Results

The clinical and pathologic parameters of 242 primary breast cancers from the Vietnamese women included in this study are presented in Table I. The median age was 47 years (range 26–72). The menopausal status of 239 tumors was defined with 156 (65%) premenopausal patients, 83 postmenopausal patients (34%) and three patients with unknown menstrual status (due to previous hysterectomies). Most patients (81%) had tumors larger than 2 cm, and pathology with low-intermediate grade tumors. Two hundred of the tumors (83%) were ductal invasive carcinomas and 16 (7%) had an in situ component with ductal invasion.

Table I. Clinicopathology features.

Variables	n	%
Menopausal status
Pre	156	65
Post	83	34
unknown	3	1
Tumor size
≤2 cm	38	16
>2 cm	197	81
unknown	7	3
Lymph node metastasis
Negative	136	56
Positive	105	44
unknown	1
Pathology type
Ductal invasive	217	90
Lobular invasive	11	4
Others	14	6
Tumor grade
I + II	184	85
III	27	12
unknown	6	3
ER status
Positive	151	62
Negative	91	38
PgR status
Positive	112	47
Negative	129	53
unknown	1
HER2 protein expression
0	84	35
1+	36	15
2+	28	11
3+	94	39

HER2 staining by IHC was scored as “0/ + 1” in 120 tumors (50%), “+2” in 28 tumors (11%), and “+3” in 94 tumors (39.0%). One hundred twenty two tumors with “2+” /or “3+” of protein expression were analyzed by SISH, three tumors failed because of technical reasons. It can be seen from Table II that 100 tumors (84%) were found to have gene amplification. Of 93 tumors with “3+” HER2 protein overexpression, 88 tumors (95%) were assessed as amplified (HER2/Chr17 ratio >2.2), and four tumors were not amplified and one tumor was equivocal (5%). Of 26 tumors with “2+” by IHC, amplification by SISH was found in 12 (46%) tumors. The concordance between the two techniques was 87%.

Table II. Concordance between HER2 status as assessed by SISH and by IHC.

		IHC
		2+	3+
SISH testing	Non-amplification	14	4
	Equivocal	0	1
	Amplification	12	88
	Total	26	93

The overall agreement between two methods in this study is 87%.

The correlation between clincopathologic factors and HER2 gene amplification, as determined by SISH, is summarized in Table III. It can be seen from the data shown in this table that HER2 was more frequently amplified in postmenopausal patients than in premenopausal patients. In accordance with this, there was a significant association with age of patients and HER2 gene amplification. Sixty two percent of the patients had ER positive tumors, of which 33% had amplified tumors, while ER negative tumors were HER2 amplified in 57% cases. An almost identical rate of amplification, 30% and 53%, was observed respectively for PgR positive and PgR negative tumors. This difference is statistically significant (p < 0.001). It can also be seen from Table III that tumors of low to intermediate grade (I and II) were more often amplified than those of high grade (III), but this difference was not statistically significant. HER2 gene amplification was not significantly correlated with histopathological subtypes (p = 0.07) and axillary lymph node metastasis (p = 0.85).

Table III. Association of HER2 gene amplification with clinicopathological parameters.

	SISH
Parameters	Negative n (%)	Positive n (%)	OR (p-value)
Age
≤50	97 (65)	53 (35)	2.1 (0.006)
>50	41 (47)	47 (53)
Menopause
Pre	98 (64)	55 (36)	2.2 (0.005)
Post	37 (45)	45 (55)
Tumor stage
≤T1	21 (55)	17 (45)	0.93 (0.84)
>T1	110 (57)	83 (43)
Histo-pathology
Ductal carcinoma	120 (56)	94 (44)	2.3 (0.07)
Others	18 (75)	6 (25)
Tumor grade
III	17 (63)	10 (37)	0.68 (0.36)
I, II	97 (54)	84 (46)
Lymph node
(+)	59 (57)	44 (43)	1.0 (0.85)
(−)	79 (59)	56 (41)
ER status
ER (−)	38 (43)	50 (57)	2.6 (0.0004)
ER (+)	100 (67)	50 (33)
PR status
PgR (−)	60 (47)	67 (53)	2.6 (0.0004)
PgR (+)	77 (70)	33 (30)

As can be seen from Table IV, HER2 gene amplification in Swedish patients was observed in 13% patients as analyzed by FISH. The rate of amplification was 16% for premenopausal and 12% for postmenopausal women. There was strong association of HER2 amplification with ER negativity. Thus, ER negative tumors had HER2 amplification in 31% as compared to 9% for ER positive tumors.

Table IV. HER2 amplification in breast cancers from Vietnamese and Swedish women (%).

	Vietnamese (n = 242)	Swedish (n = 2225)	P-value
All patients	41	13	<0.001
Premenopausal	36	16	<0.001
Postmenopausal	55	12	<0.001
ER positive	33	9	<0.001
ER negative	57	31	<0.001

Discussion

Breast cancers containing cells in which the HER2 gene is amplified respond to targeting therapy either with the humanized monoclonal antibody trastuzumab or with the dual tyrosine kinase inhibitor lapatinib [15]. Immunohistochemistry and fluorescence in situ hybridization have been used for HER2 testing. Currently, FISH is generally accepted as the gold standard. However, recently chromogenic in situ hybridization (CISH) has been introduced as a reliable alternative to FISH testing. CISH generates stable chromogenic signals which are visible by ordinary microscopy. In addition, several studies have reported an excellent correlation between FISH and SISH [9,16,17]. The main disadvantage of CISH is that it uses a single probe for the HER2 gene without allowing detection the control probe (i.e. chromosome 17). This disadvantage, however, was overcome when dual-color SISH was introduced and this technique is based on staining both the HER2 oncogene and the chromosome 17 centromere [18]. This has proven to be a highly reliable technique for HER2 testing with an excellent concordance rate compared to FISH. In the current study, an excellent correlation was also found between HER2 testing using IHC and dual color SISH. A total of 120 tumors were scored as “0/1+” on IHC and all were non-amplified. An IHC score of “3+” was found in 93 tumors, of which 88 (95%) were amplified as scored by SISH. Intermediate protein levels, score “2+”, were found in 11% tumors and of these 12 (46%) were amplified. The overall concordance between IHC and SISH was 87%, which is in good agreement with several earlier reports [9,17]. Once more, in the current study, it was confirmed that gene amplification is found in the majority of tumors characterized by HER2 protein overexpression.

Previous studies have suggested that Asian patients more often have HER2 positive breast cancers as compared to Western women. In one comparative cohort study on women younger than 45 years with breast cancer, it was found that 47.5% Korean patients had tumors which overexpressed HER2 as compared to 15.8% Caucasian women [19]. Similarly, high rates (32–65%) of HER2 gene amplification were reported for a large series of Chinese patients with breast cancer [12,13] and these figures are higher than the frequencies reported (i.e. 15–25%) from Western patients [4,5,20]. In another study on Chinese women, 34.6% breast cancer tissues showed HER2 amplification [21]. The results of the current study are therefore consistent with previous findings for Asian women. In the current study, HER2 gene amplification was present in 41% Vietnamese breast cancers which is significantly higher than the 14% recorded from a series of breast cancers surgically removed between 2007 and 2008 from 2225 Swedish women in Stockholm (p < 0.01) (Table IV). Thus, HER2 protein overexpression or gene amplification in breast cancers is more frequent in Asian women than in Caucasian women.

A strong correlation was found in the current study between age and HER2 amplification in Vietnamese cancer patients over 50 years who had 53% tumors with gene amplification as compared to those under 50 years who had tumors in which amplification was detectable in 35%, p = 0.006 (Table III). In previous studies, HER2 amplification rate in Western women was 14% for those over 41 years as compared to 18% in women 40 years and younger [20], but this difference was not statistically significant. A correlation between HER2 gene amplification and age was not found in two earlier studies on Asian women although the number of patients studied was relatively low [21,22]. However, HER2 amplification showed no correlation to age in patients from Sri Lanka [23]. By contrast, HER2 positivity showed a significant correlation to age in a large study on Danish breast cancer patients [4]. Therefore, correlation of HER2 amplification with age remains uncertain and further investigations are required to confirm or eliminate this possibility.

Conflicting results both for Asian and Western women have been published concerning HER2 amplification and menopausal status. Thus, one previous study from Vietnam showed that 26% breast cancers from premenopausal women overexpressed HER2 protein [11]. Our finding showed that cancerous breast tissue samples from postmenopausal Vietnamese women were HER2-amplified in 55% as compared to 36% for premenopausal patients. This distribution contrasts with the findings for samples taken from Swedish pre- and postmenopausal patients in the current study which were amplified at the frequencies of 16% and 12%, respectively. In one other recent study on Chinese women living in Shanghai, there was no significant difference of HER2 gene amplification between pre- and postmenopausal women [13]. Similarly, there was no association between HER2 amplification and menopausal status in similar comparative studies carried out on tissue samples from Norwegian and Chinese women, respectively [21,24]. However, it must be recognized that the sample sizes in these reports were modest. In conclusion, the findings of our current study, together with previous reports by others, indicate clearly that there is a difference in HER2 status between patients of different ethnicities, the reasons for which remain obscure.

In the present study, there was an inverse correlation between ER content and HER2 amplification. Thus, 57% and 33% of ER negative tumors were HER2 amplified in tumors from Vietnamese and Swedish women, respectively. A similar correlation between ER negativity and HER2 amplification has been reported for both Western and Asian women by others. For example, HER2 gene amplification was found in 10% ER positive tumors versus 28% ER negative tumors and this represented a significant association [20]. In addition, Kammori reported that the frequency of HER2 amplification was much higher in ER-negative tumors compared to ER-positive tumors from Japanese women [22]. Similarly, the ER positivity was significantly correlated to HER2 amplification or protein expression [23,25]. It appears therefore that HER2 amplification is correlated to ER negativity as has been reported in several studies on breast cancer patients belonging to different ethnic groups.

There are several reports on the association between HER2 status and tumor grade. In the current study, it was observed that HER2 amplification was lower in higher grade tumors as compared to those of low-intermediate grade ones, but this difference was not statistically significant. Our results contrast those obtained by several other workers which showed that high tumor grade tumors more often than not were HER2 positive as compared to low and intermediate grade tumors [25]. For instance, HER2 positivity was also more frequent in high grade tumors than in low grade tumors from Danish patients [4]. In a later study from USA, it was shown that gene amplification in breast cancers from Western women was more frequent in higher grade than in lower grade tumors [20]. However, no such correlation between HER2 amplification and tumor grade was reported in other studies on Chinese and Sri Lankan women [21,23].

More than 95% Vietnamese breast cancer patients are operated with total mastectomy plus axillary dissection. In patients with lymph node metastasis, adjuvant chemotherapy is also administered. Premenopausal patients with ER and/or PgR positive tumors will also be treated with oophorectomy plus tamoxifen and tamoxifen alone for postmenopausal patients with hormone receptor positive tumors. This type of adjuvant endocrine treatment is, however, only given in the National Cancer Hospital in Hanoi and HoChiMinh City Cancer Hospital where hormone receptor analysis is performed routinely. Other Vietnamese hospitals very rarely perform analysis of these markers. To the best of our knowledge, the current study is the first report of the HER2 status in Vietnamese breast cancers and our findings suggest that 41% patients could have benefited from anti-HER2 therapy, which up to now has not been used. It should also be noted that 43% patients with lymph node metastasis and 57% those with ER negative tumors had HER2 amplification. It is very likely that these high risk patients would have benefited from a more aggressive therapy such as anthracycline and trastuzumab. Our findings which showed that Vietnamese breast cancer patients often have HER2 amplified tumors suggests that HER2 status should be tested routinely in a series of patients and that the results should be used to improve the therapeutic strategies for a substantial portion of breast cancer patients. Unfortunately, it is unlikely that the current economic situation in Vietnam will allow wide spread use of agents such as trastuzumab.

Conclusions

HER2 protein overexpression and gene amplification rate as analyzed by IHC and SISH showed that breast cancers in Vietnamese women are more often positive than what is found for Swedish women. The findings suggest that a new treatment strategy might be used for Vietnamese women after analysis of HER2 gene amplification.

Acknowledgements

This study was supported by a grant from Sida/SAREC within the Karolinska International Research and Training (KIRT) Programme. Access to clinical information and biological results for the Swedish patients was granted by Stockholm Breast Cancer Study Group. We would like to thank Cham Nguyen Trong, Torsten Hägerström and Marianne Lunding for providing excellent technical assistance and Sinclair H. Mantell for efficiently revising and correcting the English manuscript. The authors report no conflict of interest. The authors are responsible for the content and writing of the paper. This study was approved by the committee of the Karolinska Institute (Sweden) and the Hanoi Medical University (Vietnam), respectively.