HER-2 status discrepancy between primary breast cancer and metastatic sites. Impact on target therapy

Authors

  • Alfredo Santinelli,

    Corresponding author
    1. Pathological Anatomy and Histopathology, Department of Neuroscience, Polytechnic University of Marche Region, Azienda Ospedaliero-Universitaria “Ospedali Riuniti Umberto I–G.M. Lancisi–G. Salesi,” Via Conca 71, I-60020 Torrette di Ancona, Ancona, Italy
    • Pathological Anatomy and Histopathology, Department of Neuroscience, Polytechnic University of Marche Region, Azienda Ospedaliero-Universitaria “Ospedali Riuniti Umberto I–G.M. Lancisi–G. Salesi,” Via Conca 71, I-60020 Torrette di Ancona, Ancona, Italy
    Search for more papers by this author
    • Fax: +39-071-889985.

  • Eleonora Pisa,

    1. Pathological Anatomy and Histopathology, Department of Neuroscience, Polytechnic University of Marche Region, Azienda Ospedaliero-Universitaria “Ospedali Riuniti Umberto I–G.M. Lancisi–G. Salesi,” Via Conca 71, I-60020 Torrette di Ancona, Ancona, Italy
    Search for more papers by this author
  • Daniela Stramazzotti,

    1. Pathological Anatomy and Histopathology, Department of Neuroscience, Polytechnic University of Marche Region, Azienda Ospedaliero-Universitaria “Ospedali Riuniti Umberto I–G.M. Lancisi–G. Salesi,” Via Conca 71, I-60020 Torrette di Ancona, Ancona, Italy
    Search for more papers by this author
  • Guidalberto Fabris

    1. Pathological Anatomy and Histopathology, Department of Neuroscience, Polytechnic University of Marche Region, Azienda Ospedaliero-Universitaria “Ospedali Riuniti Umberto I–G.M. Lancisi–G. Salesi,” Via Conca 71, I-60020 Torrette di Ancona, Ancona, Italy
    Search for more papers by this author

Abstract

In this prospective study, we determined HER-2 status in primary breast invasive carcinomas and in the paired lymph node metastases (synchronous and metachronous), local recurrence and metachronous distant metastases, to verify the percentage of discordant cases. HercepTest™ and Fluorescence in situ hybridization (FISH) were used to determine HER-2 status on 119 cases of primary infiltrating breast carcinoma and paired metastases (45 cases with synchronous lymph node metastases, 9 cases with metachronous lymph node metastases, 30 cases with local recurrence, and 35 cases with metachronous distant metastases). A therapeutically significant HER-2 status discordance was demonstrated between primary carcinoma and synchronous lymph node metastases (6.7%), local recurrence (13.3%) and metachronous distant metastases (28.6%). In the first comparison, there was a normal HER-2 status in primary tumours and HER-2 amplification in paired metastases, in the second the opposite phenomenon was present, and both types of discordance were evident in the third comparison. Considering the cases of local recurrences and metachronous distant metastases all together, 14 out of 65 cases (21.5%) showed a therapeutically significant discordance of HER-2 status between the primary tumour and the paired metachronous recurrence or metastasis (p < 0.001), the 15.4% of cases showing normal HER-2 status in the primary tumour and HER-2 amplification in the neoplastic relapse. For the treatment of metastatic patients, the evaluation of HER-2 status should be performed in neoplastic tissue from metastatic site, whenever possible. This procedure could be also suggested in the patients that are metastatic at the time of diagnosis. © 2007 Wiley-Liss, Inc.

The HER-2 proto-oncogene is located on chromosome 17q21 and encodes a transmembrane tyrosine kinase protein, which belongs to the epithelial growth factor receptors (EGFRs) family. Overexpression of the receptor or amplification of the HER-2 gene has been identified in 15–25% of invasive breast carcinomas and is very important both for the prognosis,1 and for the therapy.2, 3 Trastuzumab is a humanized monoclonal antibody, which is used in the therapy of invasive breast carcinomas overexpressing HER-2 protein and/or showing HER-2 gene amplification. Trastuzumab treatment, in combination with chemotherapy, has given very good results, in terms of disease free survival and overall survival times, in patients with metastatic breast cancer2; moreover, the results of recent trials on the efficacy of the drug in an adjuvant regimen have showed excellent results in terms of drastic reduction of the precocious relapses, when combined with chemotherapy.4, 5

Immunohistochemistry (IHC) and Fluorescence in situ hybridization (FISH) are the techniques recommended for HER-2 determination, according to the algorithm showed in Figure 1. Chromogenic in situ hybridization (CISH) has furnished promising results but is not yet considered an alternative choice in the selection of the patients to treat with trastuzumab. In some European countries, precise recommendations for HER-2 testing have been given and accepted all over the country.6 These recommendations identify reference laboratories, which should have an annual caseload volume of at least 250 immunohistochemical detections and of at least 100 FISH assays. HER-2 testing should be assessed only in these laboratories, because there is evidence of higher consistency of assay quality when tests are performed by high volume reference laboratories.7, 8

Figure 1.

Algorithm showing the use of immunohistochemistry and FISH technique in the determination of HER-2 status for clinical purpose. CEP17 = Chromosome 17 centromere.

Nevertheless, even if there is a worldwide agreement about the assays and the way to use them for the determination, to our knowledge there is disagreement on whether HER-2 status should be assessed in primary tumour or metastasis, for the selection of metastatic patients to treat with trastuzumab. Controversial opinions exist both about the stability of HER-2 status in breast carcinoma throughout the course of the disease, and about whether chemotherapy (neoadjuvant or adjuvant) can modify HER2 status. Most data have shown good overall concordance between primary and metastatic lesions,9, 10, 11, 12, 13, 14, 15 but some data have demonstrated discordance in up to 20% of cases.9, 12, 16, 17, 18 These differences could be due to a possible genetic drift or clonal selection for HER-2, which may happen during tumour progression,16 or to the presence of intratumoural heterogeneity of HER-2 status,19, 20, 21 with a clone having enhanced metastatic potential, leading to metastases with different HER-2 status from that of the primary lesion.

This phenomenon could have a great impact on the clinical management of the metastatic patient. In fact, if HER-2 status is assessed on the primary tumour, to avoid the biopsy of the metastasis that sometimes is clinically very difficult, it could be a HER-2 positive primitive tumour with a negative metastasis or a HER-2 negative primitive tumour with a positive metastasis. In the former situation, the treatment with trastuzumab could be unsuccessful, whereas in the latter one, a patient could be deprived of this drug that give excellent results in terms of prolonged disease free survival time.

The aim of this prospective study, in which HercepTest™ and FISH were used to determine HER-2 status in primary breast invasive carcinomas and in the paired lymph node metastases (synchronous and metachronous), local recurrence and metachronous distant metastases, was to verify the stability of HER-2 status in the following steps of neoplastic progression of invasive breast carcinoma, which is fundamental for an appropriate therapeutic approach.

Material and methods

Sample collection

In this prospective study we considered all the consecutive metachronous breast cancer metastases (in axillary lymph nodes and other organs) and local recurrences diagnosed in the Institute of Pathological Anatomy and Histopathology of the Polytechnic University of Marche Region in the period from January 2001 to December 2006. In this way, we collected 9 metachronous lymph node metastases, 35 metachronous distant metastases and 30 local recurrences. For each of these cases, the corresponding primary infiltrating breast carcinomas were retrieved from the files of the Institute of Pathological Anatomy and Histopathology of the Polytechnic University of Marche Region. Moreover we also considered all the consecutive infiltrating breast carcinomas with synchronous lymph node metastases (45 cases) diagnosed in the same Institute in the period from January to December 2006.

All the enrolled cases were subdivided in the following 4 groups: 45 cases with synchronous lymph node metastases (group A), 9 cases with metachronous lymph node metastases (group B), 30 cases with local recurrences (group C) and 35 cases with metachronous distant metastases (group D).

All the neoplastic tissue samples were fixed in 10% buffered formalin for a maximum of 48 hr and embedded in paraffin.

Immunohistochemistry

Immunohistochemistry to detect HER-2 protein was performed using HercepTest™ Kit (DAKO Corporation, Glostrup, Denmark and Carpinteria, CA) and applying the method illustrated in our previous study.22 For each of the embedded breast tissue older than 10 years, normal embedded breast tissue and HER-2 positive embedded breast carcinomas of the same period were used as negative and positive control, respectively.

Two independent expert pathologists (A.S. and E.P.) semiquantitatively evaluated the whole breast infiltrating carcinoma present in each of the immunostained slides, utilizing a light microscope and applying a double blind method. The eventually discordant results were then discussed at a conference microscope. The results were expressed according the Score System proposed by the manufacturer and illustrated in our previous study.22

Fluorescence in situ hybridization

FISH analysis was performed, using PathVysion HER-2 DNA Probe Kit™ (Vysis, Woodcreek Drive, IL), in all the cases.

Briefly, formalin fixed, paraffin wax embedded neoplastic tissues were cut into 4–5-μm-thick sections, dried overnight in a 56°C and 30 min in a 65°C oven. Sections were deparaffinized by 3 changes in xilene followed by 2 changes in 100% ethanol. Pretreatments were performed as following: sections were placed in 400 ml antigen retrieval citrate buffer, pH 6.0 and microwaved for 10 min at 98°C, allowed to cool for 20 min and subsequently rinsed in 2X SSC pH 7.0. Sections were then digested in 0.05% pepsin solution, pH 2.0 at 37°C for 40 min. After a buffer wash in 2X SSC, sections were dehydrated in 70, 85 and 100% ethanol, then a 10 μl mixture of 2 directly labelled probes—HER-2/neu specific sequence probe (LSI Her-2/neu spectrum orange) and α-satellite sequence at chromosome 17 (CEP 17 spectrum green) was added to tissue sections. Codenaturation at 85°C for 4 min and hybridization at 37°C for 14–18 hr were carried out in HYBrite (Vysis). Slides were then washed in a post hybridization buffer (0.4% SSC/ 0.3% NP40) at 73°C for 3 min, subsequently air dried, counterstained with DAPI II (125 ng/ml) and covered with coverslips. ProbeChek™ HER-2 Normal and Amplified Ratio Control Slides were used as FISH assay quality controls for appropriate hybridization conditions and as controls for proper reagents preparation.

For each specimen, 100 neoplastic infiltrating cells were counted both for HER-2 gene and chromosome 17 centromere signals, with an ECLIPSE E800 epi-fluorescence microscope (Nikon Inc. Instrument Group, Melville, NY, USA) under oil immersion at 100× magnification, using the recommended optic filters. The evaluation was performed by 2 expert pathologists (A.S. and E.P.) with a double blind method; the 100 cells were counted in 5 different fields of infiltrating tumour. The result was reported as the ratio of the average copy number of the HER-2 genes to the chromosome 17 centromeres (HER-2/CEP17 ratio). Tumours were scored as “non amplified” when the signal ratio was <1.8, and as “amplified” when the signal ratio was more than 2.2. Specimens with a ratio between 1.8 and 2.2 were submitted to a new evaluation. The cases in which more than 50% of the nuclei contained more than 2 green signals were also classified as polysomic for chromosome 17. In all the cases the 2 pathologists achieved concordant results in terms of amplification and polysomy.

According to the literature6 a case was considered good for trastuzumab therapy when it was classified as 3+ by immunohistochemistry or 2+ and 1+ by immunohistochemistry with HER-2 gene amplification by FISH.

Statistical analysis

A Fisher's exact test was used to evaluate statistical significant differences between the primary tumours and the recurrences or metastases in all the 4 groups. A p-value lower than 0.05 was considered as statistically significant.

Results

Clinical and pathological features concerning the primary tumours of the 4 groups are shown in Table I. Table II highlights the data concerning synchronous and metachronous metastases of all the groups.

Table I. Age and Pathological Features in all the Groups1
FeaturesNo. of patients (%) group ANo. of patients (%) group BNo. of patients(%) group CNo. of patients(%) group D
  • 1

    Cases with synchronous lymph node metastasis (group A), cases with metachronous lymph node metastasis (group B), cases with local recurrence (group C), cases with metachronous distant metastasis (group D).

Age (yrs)
 Mean53.352.153.450.4
 Range32–7138–6426–7531–76
Tumour size
 T1a2 (5)01 (3)4 (11)
 T1b6 (13)1 (12)5 (17)7 (20)
 T1c21 (47)7 (76)13 (43)13 (37)
 T214 (30)1 (12)8 (27)8 (23)
 T32 (5)03 (10)3 (9)
Histologic type
 Invasive ductal carcinoma40 (89)8 (88)27 (90)33 (94)
 Invasive lobular carcinoma5 (11)1 (12)3 (10)2 (6)
Histologic grade (according to Elston-Ellis) 
 G14 (10)1 (12)4 (13)3 (9)
 G215 (33)4 (44)16 (53)14 (40)
 G326 (57)4 (44)10 (34)18 (51)
Estrogen receptor status
 Positive12 (27)3 (33)4 (13)9 (26)
 Negative26 (57)5 (55)19 (63)16 (46)
 Not determined7 (16)1 (12)7 (23)10 (28)
Progesterone receptor status
 Positive17 (37)4 (44)11 (37)11 (32)
 Negative21 (47)4 (44)12 (40)14 (40)
 Not determined7 (16)1 (12)7 (23)10 (28)
Table II. Times and Sites of Synchronous and Metachronous Metastases in all the Groups
Metastatic siteInterval (months)Range (months)No. of patients (%)
Synchronous lymph node metastases (group A)45
 Axillary lymph node45 (100)
Metachronous lymph node metastases (group B)72.912–2649
 Axillary lymph node4912–1088 (88)
 Supraclavicular lymph node2641 (12)
Local recurrence (group C)62.69–25230
 Breast44.812–795 (17)
 Thorax skin and chest wall66.29–25225 (83)
Distant metastases (group D)66.712–22835
 Liver4812–964 (11)
 Lung9624–1923 (9)
 Pleura8812–2289 (26)
 Bone18.512–364 (11)
 CNS81.612–1565 (14)
 Skin2515–363 (9)
 Colon4824–722 (6)
 Ovary361 (3)
 Peritoneum1441 (3)
 Stomach841 (3)
 Retroperitoneum361 (3)
 Cervical node601 (3)

In the 45 cases of group A, an HER-2 therapeutically significant discordance between primitive tumour and paired synchronous lymph node metastasis was evident in 1 out of 7 HercepTest™ negative (1+) primitive cases (concordance = 85.7%), in 1 out of 9 HercepTest™ positive (2+) primitive cases (concordance = 88.9%) and in 1 out of 14 HercepTest™ positive (3+) primitive cases (concordance = 92.9%). In summary, considering all the 45 cases of the group A, in 3 cases (6.7%) there was a therapeutically significant discordance of HER-2 status between the primary tumours and the paired synchronous lymph node metastases (p = not calculable), with HER-2 gene amplification in the primary tumour and normal HER-2 gene in the paired metastases.

In the group B, 5 primitive tumours were classified as negative (0) by HercepTest™, 2 as positive (2+) and 2 as positive (3+). The paired metachronous lymph node metastases had the same protein expression in the positive cases (2 cases 2+ and 2 cases 3+), whereas the metastases corresponding to the 5 negative primitive tumours were classified as negative (0) in 2 cases, negative (1+) in 2 cases and positive (2+) in 1 case (concordance = 66.6%). 2 out of 9 primitive tumours had HER-2 gene amplification by FISH and the same results were obtained in the paired lymph node metastases (concordance = 100%). Considering all the 9 cases of group B, there wasn't therapeutically significant discordance of HER-2 status between the primary tumours and the paired metachronous lymph node metastases.

The data regarding each of the 30 cases of group C are showed in Table III. A therapeutically significant discordance between primitive tumour and paired local recurrence was evident in 3 out of 20 HercepTest™ negative (0) primitive cases (concordance = 85%) and in 1 out of 3 HercepTest™ positive (2+) primitive cases (concordance = 66.6%). In summary, 4 cases (13.3%) showed a therapeutically significant discordance of HER-2 status (p = not calculable), with HER-2 gene amplification in the local recurrences and normal HER-2 gene in the corresponding primitive tumours.

Table III. HER-2 Status in the 30 Primary Breast Infiltrating Carcinomas and the Paired Local Recurrences (Group C)
CaseHER-2 (IHC) primary cancerHER-2 (FISH) primary cancerLocal recurrence siteHER-2 (IHC) local recurrenceHER-2 (FISH) local recurrenceInterval (months)
  1. IHC, immunohistochemistry; FISH, fluorescence in situ hybridisation.

10Not amplifiedchest0Not amplified22
20Not amplifiedchest0Not amplified108
30Not amplifiedchest0Not amplified22
40Not amplifiedchest0Not amplified17
50Not amplifiedchest0Not amplified90
60Not amplifiedchest0Not amplified168
70Not amplifiedchest0Not amplified180
80Not amplifiedbreast0Not amplified72
90Not amplifiedthorax skin0Not amplified84
100Not amplifiedbreast0Not amplified12
110Not amplifiedchest0Not amplified60
120Not amplifiedchest1+Not amplified108
130Not amplifiedchest1+Not amplified24
140Not amplifiedchest1+Not amplified12
150Not amplifiedthorax skin1+Not amplified66
160Not amplifiedchest2+Amplified101
170Not amplifiedthorax skin2+Not amplified96
180Not amplifiedchest2+Not amplified252
190Not amplifiedchest3+Amplified23
200Not amplifiedchest3+Amplified23
211+Not amplifiedchest0Not amplified14
221+Amplifiedchest1+Amplified24
231+Amplifiedchest1+Amplified36
241+Not amplifiedchest1+Not amplified55
252+Not amplifiedbreast2+Amplified79
262+Not amplifiedchest2+Not amplified9
272+Not amplifiedbreast2+Not amplified42
283+Amplifiedthorax skin2+Amplified12
293+Amplifiedbreast3+Amplified19
303+Amplifiedchest3+Amplified48

Table IV highlights the data concerning the 35 cases of the group D. A therapeutically significant discordance between primitive tumour and paired distant metastasis was evident in 1 out of 14 HercepTest™ negative (0) primitive cases (concordance = 92.9%), in 2 out of 3 HercepTest™ negative (1+) primitive cases (concordance = 33.3%), in 6 out of 12 HercepTest™ positive (2+) primitive cases (concordance = 50%) and in 1 out of 6 HercepTest™ positive (3+) primitive cases (concordance = 83.3%). In summary, considering all the 35 cases of the group D, 10 cases (28.6%) showed a therapeutically significant discordance of HER-2 status (p = 0.084): 6 cases (17.2%) with HER-2 gene amplification in the distant metastases and normal HER-2 gene in the corresponding primitive tumours, and 4 cases (11.4%) with HER-2 gene amplification in the primary tumours and normal HER-2 gene in the paired distant metastases.

Table IV. HER-2 Status in the 35 Primary Breast Infiltrating Carcinomas and the Paired Metachronous Distant Metastases (Group D)
CaseHER-2 (IHC) primary cancerHER-2 (FISH) primary cancerDistant metastatic siteHER-2 (IHC)distant metastasisHER-2 (FISH) distant metastasisInterval (months)
  1. IHC, immunohistochemistry; FISH, fluorescence in situ hybridisation.

10Not amplifiedCNS0Not amplified96
20Not amplifiedpleura0Not amplified228
30Not amplifiedpleura0Not amplified24
40Not amplifiedbone0Not amplified36
50Not amplifiedpleura0Not amplified120
60Not amplifiedpleura0Not amplified60
70Not amplifiedbone1+Not amplified12
80Not amplifiedperitoneum1+Not amplified144
90Not amplifiedCNS1+Not amplified120
100Not amplifiedskin1+Not amplified24
110Not amplifiedCNS2+Not amplified156
120Not amplifiedstomach2+Not amplified84
130Not amplifiedliver2+Not amplified12
140Not amplifiedliver3+Amplified48
151+Not amplifiedbone1+Amplified12
161+Not amplifiedbone1+Not amplified12
171+Not amplifiedovary2+Amplified36
182+Amplifiedpleura2+Not amplified84
192+Amplifiedretroperitoneum2+Not amplified36
202+Amplifiedpleura2+Not amplified120
212+Not amplifiedliver1+Not amplified84
222+Not amplifiedlung2+Not amplified36
232+Not amplifiedskin2+Not amplified12
242+Not amplifiedpleura2+Not amplified48
252+Not amplifiedlung2+Not amplified192
262+Not amplifiedcolon2+Not amplified84
272+Not amplifiedlung3+Amplified72
282+Not amplifiedcervical node3+Amplified60
292+Not amplifiedpleura3+Amplified132
303+Amplifiedpleura1+Not amplified12
313+AmplifiedCNS3+Amplified24
323+Amplifiedcolon3+Amplified22
333+Amplifiedskin3+Amplified40
343+AmplifiedCNS3+Amplified12
353+Amplifiedliver3+Amplified36

Considering the cases of group C and group D all together, 14 out of 65 cases (21.5%) showed a therapeutically significant discordance of HER-2 status between the primary tumours and the paired metachronous recurrences or metastases (p < 0.001): in 10 cases (15.4%) there was a HER-2 gene amplification in the neoplastic relapses with normal HER-2 gene in the primary tumours, and in 4 cases (6.1%) the opposite phenomenon was described.

Discussion

Nowadays, the evaluation of HER-2 status is a fundamental biological feature that the pathologist must furnish to the clinician for the management of a patient suffering from infiltrating breast carcinoma. This feature is important both for the prognosis,1 and for the treatment of the patients in adjuvant and metastatic setting.2, 3, 4, 5, 23

For this last purpose, in the past 5 years the evaluation of HER-2 status was generally performed on the primary neoplasm, avoiding the biopsy of metastatic site, and in the assumption that the HER-2 status should remain stable in the years, as demonstrated in numerous previous studies, mostly based only on immunohistochemistry.9, 10, 11, 13, 16, 17 For this reason, to our knowledge, the necessity of HER-2 determination in metastatic lesion is not well established, at present time.15

In this prospective study we have considered 119 pairs of primary tumours and their metastases subdivided in lymph node metastases, synchronous and metachronous, local recurrences and metachronous distant metastases; we have used immunohistochemistry (HercepTest™) and FISH to evaluate whether HER-2 status changed during disease progression and after adjuvant hormonal therapy and/or chemotherapy. Our results showed a HER-2 status concordance only between primitive tumour and metachronous lymph node metastasis (group B). On the contrary, a discordance about HER-2 status was evident in the other 3 groups, with variable percentages of involved cases and different way: from positive HER-2 status in the primary tumour to negative in the paired metastasis, and vice versa. In particular, in the group A the 6.7% of the cases, in the group C the 13.3% of the cases, and in the group D the 28.6% of the cases showed a significant discordance. In all the discordant cases of the group A, HER-2 status was positive in the primary tumour and negative in the synchronous lymph node metastasis; on the contrary, in all the discordant cases of the group C, HER-2 status was negative in the primary tumour and positive in the local recurrence. Both types of discordance were present in the group D: 17.2% of cases had a HER-2 negative tumour and HER-2 positive distant metastasis, and 11.4% of cases showed the opposite HER-2 status. In the scientific literature of the last 3 years, we can appreciate different results about this issue. Regitnig et al.24 have studied 31 breast carcinomas and their corresponding lymph node (10 cases) and distant (31 cases) metastases, by applying immunohistochemistry and FISH. They demonstrated that HER-2 status remained the same in the lymph node metastases of the 10 cases considered, whereas it changed in 48.4% of the 31 distant metastases, with a therapeutically significant change in 16.1% of the cases. The authors recommended future studies with larger number of cases to apply these results to the clinical practice. Similar results were obtained by Zidan et al.18 in primary breast carcinomas and paired metastases of 58 patients. They found a discordant HER-2 status in 14% of patients; in 2% of patients, HER-2 was positive in primitive tumour and negative in paired metastasis, and in 12% of patients, HER-2 had the opposite status. On the basis of this discordance rate they suggested taking HER-2 evaluation into consideration in metastatic sites, especially when HER-2 is negative in the primary tumour; nevertheless they also stressed the necessity to verify their results with larger numbers of patients. Different results were obtained from Gong et al.15 in a study concerning 60 patients with breast carcinomas (43 with loco-regional recurrence and 17 with distant metastases). 42 out of 43 (98%) and 16 out of 17 (94%) patients had a concordance of HER-2 status in primary tumours and paired loco-regional recurrences, and in primary tumours and paired distant metastases, respectively. Moreover, among the 32 patients who received chemotherapy, the HER-2 status remained unchanged in all patients (concordance rate: 100%). They concluded that HER-2 status can be reliably evaluated in material from either primary or metastatic tumours in most patients, and that other studies with larger series was warranted to clarify the significance of rare discordant results. Finally, Varga et al.25 studied prospectively the effect of chemotherapy on HER-2 status in 23 patients with stage-II, -III and -IV breast cancer and found that during the therapy 3 patients (13%) changed HER-2 status. For this reason they suggested to analyze HER-2 status more than once during the disease course.

Despite the discordant results obtained in the different above cited studies, our data appear to be concordant, in all the studied groups of patients, towards a precise idea: the HER-2 status can be different in the metastasis with respect to the primitive tumour. Only the group B showed the same HER-2 status in the primitive tumour and metachronous lymph node metastasis, but in this group only 9 cases were studied and the results can be influenced by this low number. Nevertheless, the possible explanations of HER-2 discordance in each of the 3 groups can be different, as well as the comments and the conclusions that we can draw. The discordance of HER-2 status in the 3 cases of the group A can be principally explained by the biologic phenomenon of tumour heterogeneity that, even if quite rare (<5% of primary breast carcinomas),10, 11, 21 could determine the presence, in the primitive tumour, of a small neoplastic clone that not only may have a different HER-2 status, but also may acquire the genotypic and phenotypic capability to metastasize. This small neoplastic clone could be negligible in the tumour nodule and could not be detected in the immunohistochemical staining of the larger primary carcinoma26; in this situation, the metastasis has a HER-2 status different from the primary carcinoma. The same biological phenomenon could be hypothesized for the HER-2 status discordance highlighted in the groups C and D; nevertheless, in these cases of local recurrences or metachronous distant metastases, it's more probable that HER-2 status alteration may not be present in the primitive tumour, but it may happen in the neoplastic cells eventually remained in the patient after surgical and adjuvant therapy. Breast cancer cells that survive adjuvant treatment may undergo genetic changes resulting in either a loss or gain of expression of some biological markers.27 The greater genetic instability of these cells during the disease progression and as consequence of adjuvant endocrine therapy28, 29 and/or chemotherapy30 could also explain the greater frequency of discordance in HER-2 status between primary tumour and local recurrence or metastasis highlighted in groups C and D with respect to group A.

In conclusion, the present study demonstrates that HER-2 status in the metastases can be discordant from that observed in the primitive tumour in about the 21% of the patients. From a clinically point of view, these data suggest that the evaluation of HER-2 status, principally for the use of trastuzumab in metastatic breast cancer patients, should be performed in neoplastic tissue from metastatic site, whenever possible. This procedure could be also suggested in the patients with metastatic breast cancer at the time of diagnosis.

Acknowledgements

The content of this study is solely the responsibility of the authors and does not necessarily represent the official views of the Polytechnic University of Marche Region (Ancona, Italy). We gratefully thank Mrs. Donatella Diotallevi, Mrs. Graziella Marchetti and Miss. Aurora Baldini for their excellent technical contribution.

Ancillary