- Email: [email protected]
Comparison of HER-2 status between primary breast cancer and corresponding distant metastatic sites
Original article
Annals of Oncology 13: 1036–1043, 2002 DOI: 10.1093/annonc/mdf252
Comparison of HER-2 status between primary breast cancer and corresponding distant metastatic sites D. Gancberg1, A. Di Leo1, F. Cardoso1, G. Rouas 1, M. Pedrocchi2, M. Paesmans3, A. Verhest1, C. Bernard-Marty1, M. J. Piccart1 & D. Larsimont1* 1
Translational Research Unit and 3Data Centre, Jules Bordet Institute, Brussels, Belgium; 2 F. Hoffmann-La Roche Ltd, Basel, Switzerland
Received 22 January 2002; revised 20 March 2002; accepted 11 April 2002
Background: The humanized anti-HER-2 monoclonal antibody trastuzumab (Herceptin®) is a new treatment modality for metastatic breast cancer, the efficacy of which is directly correlated with the HER-2 status of the tumour, evaluated either by immunohistochemistry (IHC) and/or by fluorescence in situ hybridisation (FISH). This analysis is generally performed on the primary tumour. There are few data regarding the HER-2 status in the corresponding distant metastases. Methods: HER-2 status in 107 patients with a primary breast tumour and at least one distant metastatic lesion was analysed by IHC and FISH. Results: We found similar levels of amplification (25% and 24%) and overexpression (13% and 19%) of HER-2 in primary and metastatic samples, respectively. Among paired primary/metastatic tumours, six (6%) showed discordance by HercepTest™ (n = 100): all six cases showed greater Her-2 overexpression in the metastatic tissue. By FISH (n = 68), five (7%) cases were discordant: two cases were amplified in the primary tumour but not in the metastasis, and three samples showed amplification in the metastasis but not in the primary. Finally, we analysed HER-2 status in different metastatic lesions from 17 patients that had at least two distant metastatic sites. Discordance between different sites from the same patient was 18% by IHC and 19% by FISH. Conclusions: Between the paired primary tumour and distant metastatic lesions, 94% and 93% of samples had concordant HER-2 status when analysed by IHC or FISH, respectively. These results do not support routine determination of HER-2 on metastatic sites, particularly when FISH results from the primary tumour are available. Key words: breast cancer, HER-2, metastatic, primary
Introduction Breast cancer is a heterogeneous disease [1]. It has been postulated that the equilibrium of different clones within a tumour is eventually overcome by a biologically dominant one with enhanced metastatic potential [2, 3]. Instead of a linear progression from the primary tumour to the metastasis, an early stem-line clone might evolve independently in both sites [4]. The resulting heterogeneity of metastatic breast cancer (MBC) may explain why biomarkers of prognosis or therapy responsiveness measured exclusively from primary tumours may not entirely reveal all the biological properties of MBC. Overexpression or amplification of proto-oncogenes such as HER-2 may confer biological advantage to tumour cells. The Her-2 oncoprotein (for a review see [5, 6]) is a transmembrane
*Correspondence to: Dr D. Larisimont, Department of Pathology, Jules Bordet Institute, Boulevard de Waterloo 125, 1000 Brussels, Belgium. Tel: +32-2-5413115; Fax: +32-2-5413281; E-mail: [email protected] © 2002 European Society for Medical Oncology
receptor, belonging to the epidermal growth factor receptor (EGFR) family, with tyrosine kinase activity that has been associated with shortened survival [7], enhanced aggressiveness [7, 8] and other poor prognostic factors. However, the literature does not uniformly support a predictive role for HER-2 in breast tumours, mainly due to the lack of prospective studies and the loss of reproducibility of the assessment methods across different laboratories [5, 9, 10]. The only exception is the value of HER-2 as a predictive factor for responsiveness to trastuzumab (Herceptin®), the humanized anti-HER-2 monoclonal antibody [5]. In recent years, the efficacy of trastuzumab in the treatment of MBC that overexpresses HER-2 has been proven, and it is currently under evaluation in the adjuvant setting. Monotherapy with trastuzumab achieves overall response rates of 35% when given as first-line therapy, and 18% in patients previously treated with chemotherapy for MBC [11–13]. Combination treatment with chemotherapy as first-line treatment achieves response rates of 45% and a survival benefit of ∼5 months [14]. These results
1037 led to the approval of trastuzumab as monotherapy for patients with HER-2-positive metastatic breast tumours who received anthracyclines and taxanes for metastatic disease, and in combination with paclitaxel as first-line therapy for HER-2positive patients previously treated with anthracyclines. It has also been proven that the efficacy of trastuzumab is highly dependent on the HER-2 status of the tumour [11–15]. The main targets of any therapy in MBC are the metastases. However, in the great majority of cases, HER-2 status is determined on the primary tumour, and there are few published data regarding the comparison of HER-2 status between the primary and the metastatic sites. Studies have reported a high level of consistency, although not complete, in HER-2 status between primary tumours and locoregional metastases using immunohistochemistry (IHC) [16–20], fluorescence in situ hybridisation (FISH) or both [21–24]. Data referring to distant metastases are extremely scarce [18, 20, 22, 25, 26]. In this study, we compared HER-2 overexpression and amplification in primary tumours and their distant metastases from a series of 107 breast cancer patients with at least one distant metastatic site, using both IHC (HercepTest™) and FISH (triple probe LSI HER-2/topoIIα/CEP17) methods. Additionally, the evaluation of HER-2 status in different metastatic sites from the same patient was performed in 17 cases. This is, to our knowledge, the first study to compare HER-2 status, evaluated by both these techniques, between primary tumours and their distant metastases (locoregional metastases not included), in a relatively large number of patients.
Materials and methods Breast cancer specimens A series of 107 paraffin-embedded blocks of primary breast carcinomas, routinely examined in the Jules Bordet Institute (Brussels, Belgium) from 1981 to 1999, and corresponding distant metastases, were analysed. No selection criteria were used to identify these samples except that the metastatic lesion had to be distant and not locoregional. For each case, haematoxylin and eosin staining was performed to determine whether an invasive tumour was present in the sample. In 17 patients, at least two metastatic samples were available for HER-2 evaluation.
Immunohistochemistry IHC was performed using the HercepTest™ (Dako, Glostrup, Denmark) according to the manufacturer’s instructions. A score of 3+ was considered to be a true positive. Each run included positive and negative controls. The slides were reviewed and scored by one pathologist (D.L.), and only once. The observer was blinded with regard to the FISH data and to the HER-2 status in the paired samples.
Fluorescence in situ hybridisation Slides were deparaffinised in xylene for 5 min, three times, and dehydrated in two changes of 100% ethanol for 5 min. After incubation in a solution of 0.2 M HCl for 20 min at room temperature, slides were washed in deionised water for 1 min and incubated in the pre-treatment reagent (1 M NaSCN) for 30 min at 80°C, and rinsed in deionised water for 3 min.
Slides were then incubated in a solution of 0.2 M HCl/4 mg/ml protease (Paraffin Pre-treatment kit; Vysis, Inc., Downers Grove, IL, USA) at 37°C for 10–20 min, and rinsed in deionised water for 3 min. After controlling the digestion, slides were incubated in neutral-buffered formalin for 10 min, rinsed in deionised water and dehydrated through graded ethanol series (70, 85 and 100%), 1 min each. Slides were placed on a hot plate (73°C) for 5 min with a 10 µl probe (Vysis multicolor-probe Topo IIα Spectrum Green, HER-2 Spectrum Orange and CEP17 Spectrum Aqua), coverslipped and sealed. Denaturation was followed by incubation in humid atmosphere at 37°C for 16–18 h. The next day, the rubber cement was removed and coverslips floated off by soaking the slides at room temperature in 2× SSC/0.3% Nonidet P-40 for 2 min. Slides were then incubated in a solution of 2× SSC/0.3% Nonidet P-40 at 73°C for 2 min, dried in a dark place and counterstained with 10 µl of 0.2 µM 4,6-diamidino-2-phenylindole (DAPI) in antifade solution (Vectashield; Vector Laboratories, Inc., Burlingame, CA, USA). Signals were counted in 60 nuclei using a Leica DMRB or an Olympus BX51 epifluorescence microscope equipped with the following filter sets: DAPI, aqua, fluorescein isothiocyanate (FITC), Texas Red, DAPI/FITC/TRITC, and 40-, 63- and 100-fold magnification oil immersion objectives. Gene amplification was defined as a HER-2 to CEP17 ratio >2. The slides were analysed and classified by two observers (D.G. and G.R.) who were blinded with regard to the IHC data and to the HER-2 status in the paired sample. In only two cases was there discordance between both observers, and in both cases the FISH analysis was repeated and concordance obtained. Owing to problems with tissue preparation that prevent FISH analysis, such as Bouin fixation or decalcification treatment, FISH was successfully performed on 68 primary-metastasis pairs out of the 107 originally available.
Statistical analysis Theoretical proportions of HER-2 positivity were estimated on the basis of the observed proportions in the samples. These estimations were carried out using the samples obtained from the primary tumour and using the samples obtained for the first (chronologically) developed metastatic site. The assessments done on other metastatic sites from the same patient were not considered for estimation of the positivity rates in order to avoid dependency between the observed data (positivity on one metastatic site being most probably linked to positivity on the other sites). Inclusion of multiple sites from the same patient could consequently lead to a possible over- or underestimation of positivity rates. The 95% confidence intervals (CIs) were calculated for the theoretical proportion on the basis of asymptotic normal distributions. Rates of positivity for the primary tumours and the first metastatic site were compared using two-sided paired McNemar’s test at a usual 5% level. Discordance rates between the primary tumour/first metastatic site pairs were similarly analysed. Finally, a descriptive analysis was performed for the 17 patients for whom several metastatic sites were available, in order to assess the possible discordance between the sites. No inferential analysis was performed due to the small number of patients and the fact that the sites were not synchronous.
Results A total of 129 metastatic samples were available: 107 from a first metastatic site, 17 from a second and five from a third. All but four metastases were asynchronous. Metastases were located in bone (n = 38), soft tissue (not locoregional; n = 32), liver (n = 26), lung/bronchus/pleura (n = 13), other gastrointestinal (e.g. stomach/duodenum/biliary tract/peritoneum;
1038 Table 1. (A) HER-2 status assessed by IHC in primary and metastatic breast tumours IHC score (HercepTest™)
Primary tumours (n = 100)
Metastatic tumours (n = 100a)
Events (%)
Events (%)
0
57 (57)
52 (52)
1+
14 (14)
21 (21)
2+
16 (16)
8 (8)
3+
13 (13)
19 (19)
Not evaluable
7
7
a
In the case of multiple metastatic sites only the score observed in the first metastatic site was considered for this analysis.
(B) HER-2 status assessed by FISH in primary and metastatic breast tumours FISH score
Primary tumours (n = 85)
Metastatic tumours (n = 84)a
Events (%)
Events (%)
Amplification
21 (25)
20 (24)
No amplification
64 (75)
63 (76)
Not evaluable
22
24
a
In the case of multiple metastatic sites only the score observed in the first metastatic site was considered for this analysis.
n = 9), ovary (n = 6), brain (n = 2) or other sites (parotid, bladder, pericardium). The time period between surgical removal of the primary tumour and the surgery/biopsy of the metastatic lesion ranged from 1 month to 18 years.
HER-2 status in primary and metastatic breast tumours (Table 1) Using IHC (Table 1A), we found that 13% (95% CI 6% to 20%) of the primary tumours (n = 100) revealed Her-2 overexpression (score 3+). In metastatic lesions (n = 100), 19% (95% CI 11% to 27%) showed 3+ Her-2 overexpression. FISH analysis (Table 1B) of the primary (n = 85) and metastatic (n = 84) samples demonstrated 25% (95% CI 16% to 34%) and 24% (95% CI 15% to 33%) of HER-2/neu amplification, respectively.
Discordance in HER-2 status between IHC and FISH FISH and IHC methods were concomitantly successful for detection of HER-2 in 79 primary tumours (74% of the 107 available patients) and in 97 metastatic samples (75% of the 129 available samples). Regarding the primary tumours, there were nine (11%) discordant cases between the two techniques. All these nine cases were negative by IHC (score 0 to –2) but positive by FISH. Among the 76 evaluable samples of the first metastatic site (71% of the 107 available), there were seven (9%) cases of discordance between IHC and FISH results: five out of seven
of these cases were negative by IHC but amplified by FISH, while two out of seven were positive by IHC but not amplified by FISH. For the samples collected from the second metastatic site, 16 of 17 were evaluable by both techniques. Only three of 16 presented discordant results, two of which were negative by IHC and amplified by FISH, while the inverse was seen in the remaining one. Five patients had samples available from a third metastatic site. In all of these IHC and FISH could be performed and yielded concordant results. Overall, more samples were HER-2 positive when evaluated by FISH than by IHC (P = 0.01). While this occurred in both the primary tumours (P = 0.04) and the metastatic sites (P = 0.73), the difference reached statistical significance only for the primary tumours.
Discordance in HER-2 status between primary and metastases (Table 2) One hundred paired samples were available for IHC analysis (Table 2A). Seven cases were not evaluable due to detachment of the tissue during the pre-treatment. There were 13 and 19 cases with 3+ Her-2 overexpression in the primary tumours and in the metastatic lesions, respectively. Six of these 100 cases (6%) scored differently in the primary tumour and its respective first metastatic site. All cases showed greater Her-2 overexpression in the metastatic site as compared with
1039 Table 2. (A) Discordant data between primary and paired metastatic tumours by IHC (n = 100) Case no.
Primary tumour
Date
First metastasis
Metastatic site
Date
1
Negative
May 1989
Positive
Liver
July 1990
2
Negative
September 1990
Positive
Bronchus
May 1995
3
Negative
November 1991
Positive
Bone
November 1993
4
Negative
August 1995
Positive
Bone
March 1997
5
Negative
January 1994
Positive
Parotid
November 1999
6
Negative
January 1992
Positive
Pericardium
March 1998
(B) Discordant data between primary and paired metastatic tumours by FISH (n = 68) Case no.
Primary tumour
Date
Metastasis
Metastatic site
Date
1
Negative
May 1990
Positive
Bronchus
October 1992
2
Negative
April 1985
Positive
Bone
December 1989
3
Negative
February 1999
Positive
Node
November 2000
4
Positive
June 1989
Negative
Node
August 1992
5
Positive
March 1985
Negative
Pleura
December 1992
the primary tumour. This increase was statistically significant (P = 0.03). When FISH was used (Table 2B), only 68 paired primary/ metastatic tumours were evaluable, due to technical difficulties with Bouin-fixed or decalcified samples. Among these, the HER-2 gene was amplified in 16 primary tumours (24%) and 17 metastatic lesions (25%). Five of 68 cases (7%) were discordant for HER-2 status between the primary site and the metastases. Three cases showed HER-2 amplification in the metastasis but not in the primary sample, and two cases showed HER-2 amplification in the primary tumour but not in the metastasis. No correlation was observed between changes in the HER-2 status and the location of the metastases.
Comparison of HER-2 status in multiple metastatic sites from the same patient (Table 3) Seventeen cases with multiple metastatic sites were evaluable by IHC. Twelve patients had two distinct metastatic sites, and five patients had three metastases. We found discordant Her-2 overexpression in three of 17 (18%; 95% CI 0% to 26%) cases. These three cases, when evaluated by FISH, showed concordant results. When FISH was used, 16 of 17 cases were successfully hybridised. We found differences in HER-2 amplification between different metastatic sites in three of 16 (19%; CI 0% to 38%) cases. These cases were not the same as those that had been found discordant with IHC and, when evaluated by IHC, they showed concordant results. The discordant results among the different metastases were not associated with a particular location of the metastases.
Discussion The present study is, to our knowledge, the first to compare the HER-2 status of primary breast tumours exclusively with their distant metastases, using both IHC and FISH techniques, in a relatively large number of MBC patients. This issue has become important in recent years due to the development and approval of trastuzumab for the treatment of HER-2-overexpressing MBC. In the metastatic setting, where the disease is virtually incurable, few drugs have shown a significant advantage, not only in terms of response rates but particularly in terms of overall survival. Trastuzumab, when given in combination with paclitaxel, achieved a survival benefit of 5 months compared with single-agent treatment of paclitaxel [14]. On the other hand, although very well tolerated, trastuzumab has a non-negligible risk of cardiotoxicity. Since trastuzumab activity is directly dependent on the HER-2 status of the tumour, a careful selection of patients is crucial for increasing its clinical benefit and avoiding unnecessary treatment of patients who most likely will not benefit from it. To date, the HER-2 status is almost always evaluated on the primary tumour sample, since routine biopsy of metastatic sites is not a standard procedure. Thus, it becomes imperative to determine whether homogenous HER-2 expression exists between the primary tumour and its distant metastases. Most published studies have compared HER-2 status in primary tumours with their locoregional or concurrent lymph node metastases. Only five studies have presented data concerning distant metastases [18, 20, 22, 25, 26], and only one of these [25] focused exclusively on distant metastases. However, it should not be assumed that both types of metastases are biologically equivalent. Distant metastases often represent clonal outgrowths with genetic modifications not detected in
1040 Table 3. (A) Comparison of multiple metastatic (M) sites from the same patient by IHC (n = 17) Case
M1 (site)
M2 (site)
M3 (site)
Primary
1
Negative (node)
Negative (node)
Negative
2
Negative (node)
Negative (node)
Negative
3
Negative (stomach)
Negative (stomach)
4
Negative (node)
Positive (node)
5
Positive (node)
Negative (ovary)
6
Positive (pericardium)
Negative (pleura)
7
Negative (bone)
Negative (bone)
Negative
8
Negative (brain)
Negative (brain)
Negative
9
Negative Negative (pleura)
Negative Positive
Negative (diaphragm)
Negative
Negative (bronchus)
Negative (bronchus)
Negative
10
Negative (ovary)
Negative (ovary)
Negative
11
Negative (bone)
Negative (bone)
Negative
12
Negative (bone)
Negative (bone)
Negative
13
Negative (skin)
Negative (skin)
14
Negative (node)
Negative (skin)
15
Negative (duodenum)
Negative (stomach–antrum)
16
Positive (bone)
Positive (bone)
17
Positive (ovary)
Positive (epiploon)
Negative (skin)
Negative Negative
Negative (stomach–fundus)
Negative Positive
Positive (node)
Positive
Discordant cases between different metastatic sites are reported in bold type (B) Comparison of multiple metastatic (M) sites from the same patient by FISH (n = 16) Case
M1 (site)
M2 (site)
1
Negative (node)
Positive (node)
M3 (site)
Positive
2
Negative (node)
Positive (node)
Negative
3
Positive (stomach)
Negative (stomach)
Positive
4
Negative (node)
Negative (node)
5
Positive (node)
Positive (ovary)
6
Negative (pericardium)
Negative (pleura)
8
Negative (brain)
Negative (brain)
Negative
9
Negative (pleura)
Primary
Negative Positive
Negative (diaphragm)
Negative
Negative (bronchus)
Negative (bronchus)
NE
10
Negative (ovary)
Negative (ovary)
Negative
11
Negative (bone)
Negative (bone)
Negative
12
Negative (bone)
Negative (bone)
Negative
13
Negative (skin)
Negative (skin)
14
Negative (node)
Negative (skin)
15
Negative (duodenum)
Negative (stomach–antrum)
16
Positive (bone)
Positive (bone)
17
Positive (ovary)
Positive (epiploon)
Negative (skin)
Negative NE
Negative (stomach–fundus)
NE Positive
Positive (node)
Positive
In case 7, FISH results were not available for the second metastatic site. Discordant cases between different metastatic sites are reported in bold type. NE, not evaluable.
the primary tumour, and are generally identified years after resection (and therapy) of the primary cancer. In addition, cells that metastasise locally via the lymphatic system may possess biological properties distinct from those that have the
capacity to invade vasculature, travel to distant sites and establish clinically significant disease. Cytogenetic analyses have demonstrated extreme genetic heterogeneity (up to 70% of polyclonality) in breast cancer [27–31]. Kuukasjärvi et al. [4]
1041 reported that the genetic composition of 31% of metastases differed almost completely from that of the paired primary tumours. Our group has previously compared the Her-2 status, among other biological markers, between 370 primary breast cancer sites and their metastatic ipsilateral axillary lymph nodes [16]. Her-2, evaluated by IHC, was the marker with the best overall concordance (98%); no biological marker showed 100% concordant results. Similarly high levels of concordance have been reported by other authors [17, 19, 21, 23, 24]. However, as has been pointed out, all these studies were restricted to locoregional metastases. Three of these studies determined HER-2 status by both IHC and a DNA hybridisation technique [21, 23, 24], and two by IHC alone [17, 19]. Another common pitfall is the number of evaluated patients: only two of the six studies had >100 evaluable patients [23, 24], and none had >200; the number of paired samples (primary/locoregional metastasis) is even smaller. Only five studies compared the HER-2 status in breast primary tumours and distant metastases [18, 20, 22, 25, 26]. In 1993, Niehans et al. [25] retrospectively evaluated Her-2 expression in primary breast cancer and different metastatic sites in autopsy tumour samples from 30 patients who had died from the disease. Eight of these patients (27%) were Her-2 positive and, among those, there was a single case of discordant staining (3%) between the primary site and two of its metastatic lesions. The authors concluded that Her-2 expression is generally congruent at different metastatic sites. This is the only published study that evaluated exclusively distant metastases. Unfortunately, the actual number of evaluable patients was very low, and the uncontrolled fixation time of autopsy samples jeopardizes the IHC data that can be obtained from these samples. In 2000, Shimizu et al. [18] evaluated Her-2 protein levels, by IHC, in primary and metastatic breast cancer samples from 21 patients. Only seven of the metastatic samples were from distant sites, with the remaining 14 belonging to locoregional relapses. The authors found no significant differences in the Her-2 expression between the primary tumours and their metastases. Also in 2000, Masood et al. [20] evaluated Her-2 overexpression in 56 patients by IHC. Once again, only a minority (11 cases) of the metastases were from distant sites. The pattern and intensity of Her-2 overexpression in the primary tumours and their metastases were found to be almost identical, with heterogeneity present in only one case. Recently, Tanner et al. [22] analysed HER-2 amplification in 46 breast primary tumours and their metastases, using IHC and DNA in situ hybridisation techniques. The authors found complete concordance regarding HER-2 amplification between the primary tumours and their metastases. However, in only 12 cases were the analysed metastases from a distant site. In the 2000 San Antonio Breast Cancer Meeting, Edgerton et al. [26] presented some preliminary results of a study comparing HER-2 status, determined by IHC and FISH, on 193 primary breast cancers and corresponding local recur-
rences (68), lymph node metastases (32) and distant metastases (93). Overall, the authors found 25% discordant cases. Publication of detailed results and methodology of this study would be of interest. In the present study, we analysed both Her-2 protein overexpression by IHC and HER-2 oncogene amplification by FISH in 107 patients with at least one distant metastatic site. Owing to technical difficulties, the number of evaluable pairs (primary/distant metastasis) was 100 for IHC and 68 for FISH. The relatively large number of samples not evaluable is an important limitation, encountered in all studies performed in paraffin-embedded tumour samples, and was mainly due to bouin fixation or decalcification treatment (needed for bone metastases), which compromise hybridisation. Seventeen cases with multiple metastatic sites were also comparatively studied. The Vysis multicolour FISH probe, which includes the topoisomerase IIα gene analysis, was used since the original plan was the evaluate the status of both the HER-2 and the topoisomerase IIα gene in the primary tumour and its corresponding metastatic sites. Topoisomerase IIα gene results are not reported, since the probe was found to be inadequate to evaluate this gene. A new topoisomerase IIα probe has been developed by Vysis and is now being used to evaluate the status of this gene in the present series. Similar levels of amplification of the HER-2 oncogene were found in the primary tumours and in the metastatic samples (25% and 24%, respectively) using FISH (Table 1B). Overexpression of Her-2, assessed by HercepTest™, was found in 13% of the primary and 19% of the metastatic samples when only 3+ scores were considered to be positive. When 2+ scores were included, Her-2 overexpression was found in 29% of the primary tumours and 27% of the metastatic lesions (Table 1A). This means that using strong staining (3+) with the HercepTest™ as the sole indicator of Her-2 positivity was very stringent and that it underestimated the number of samples with HER-2 gene amplification (P = 0.01). Controversy related to the best IHC protocol, in particular the best antibody, to use for Her-2 determination is still ongoing. However, we have adopted an IHC protocol using the monoclonal antibody CB-11 with antigen retrieval, which allows us to decrease the rate of false negatives; all positive results are confirmed by FISH, therefore reducing the rate of false positives. This attitude seems well balanced in terms of efficiency and costs. HER-2 status in the primary and corresponding metastatic lesions revealed a high level of concordance (94% and 93% when analysed by IHC or FISH, respectively). Discordant results were observed in only six of 100 cases (6%) by IHC (Table 2A) and five of 68 cases (7%) by FISH (Table 2B). Regarding IHC, all six discordant cases presented an increase in staining in the metastatic site in comparison to the primary tumour. This increase in IHC staining when the tumour disseminates was statistically significant (P = 0.03). Four of these six cases had a 2+ score in the primary tumour. Among the five discordant cases by FISH, in three the HER-2 gene
1042 became amplified in the metastatic site while it was nonamplified in the primary tumour, and in two the opposite occurred (HER-2 gene amplification in the primary tumour but not in the metastasis). Including all metastatic sites in the analysis, our IHC data suggest that Her-2-positive metastatic lesions with negative primary tumours were more frequent than the converse situation (Table 2A). This is in line with the biological fact that Her-2 is correlated with enhanced tumour aggressiveness. Another possible explanation may be an underestimation of Her-2 protein overexpression in the primary tumour by HercepTest™; this could be due to the fact that most primary samples were collected many years before the present analysis (see Table 2 for specific dates), and therefore some protein degradation might have occurred. Previous studies have shown that the sensitivity of IHC when used in paraffinembedded fixed tumour samples varies widely, depending not only on the antibody used but also on the tissue processing, namely tissue fixation and the ‘age’ of the samples, which interferes greatly with their quality [32]. This explanation is supported by the FISH data, which show a very similar percentage of HER-2-positive samples between pairs and therefore seems not to support the hypothesis of enhanced tumour aggressiveness. In fact, when FISH was used, among the 16 of 68 primary samples with HER-2 gene amplification (Table 2B), two (13%) had corresponding distant metastases with a non-amplified HER-2 gene, while among the 52 of 68 HER-2-negative primary samples, three (6%) had corresponding metastatic sites with HER-2 gene amplification. The HER-2 status in samples from multiple distant metastatic sites belonging to the same patient were analysed in 17 cases (Table 3). Thirteen percent and 18% discordant results between samples from different metastases were found when assessed by FISH and IHC, respectively. Previously, only the study by Niehans et al. [25] had looked into this issue, but did so using solely the IHC technique: they found a single case among the 30 evaluated that presented different staining among the four metastatic sites (two were Her-2 positive and two were Her-2 negative). In conclusion, this study does not support the routine testing of metastases to confirm HER-2 positivity when detected in the primary tumour, particularly if results obtained by FISH are available. Assessment of HER-2 status in one of the metastatic sites may be worthwhile only in some patients with easily accessible metastases and for whom Her-2 evaluation by IHC, recently performed in a primary tumour sample collected many years before, shows a negative score. An alternative solution would be the determination of HER-2 amplification by FISH in the primary tumour sample. Translational research efforts to understand the mechanisms of response and resistance to trastuzumab should be a priority. Studies aimed at identify additional predictive markers of responsiveness to trastuzumab (e.g. phosphorylated Her-2, EGFR, HER-3, HER-4) and other possible molecular mechanisms involved are underway.
Acknowledgements The authors thank Carolyn Straehle, PhD, for editorial assistance. G.R. is supported by a grant from ‘Les Amis de l’Institut Bordet’, Brussels, Belgium. The costs related to the detection kit for HER-2 evaluation by FISH and IHC were funded by Hoffmann-La Roche, Basel, Switzerland.
References 1. Bièche I, Lidereau R. Genetic alterations in breast cancer. Genes Chromosom Cancer 1995; 14: 227–251. 2. Heppner GH. Tumor cell societies. J Natl Cancer Inst 1989; 81: 648–649. 3. Kerbel RS. Growth dominance of the metastatic cancer cell: cellular and molecular aspects. Adv Cancer Res 1990; 55: 87–132. 4. Kuukasjärvi T, Karhu R, Tanner M et al. Genetic heterogeneity and clonal evolution underlying development of asynchronous metastasis in human breast cancer. Cancer Res 1997; 57: 1597–1604. 5. Ross JS, Fletcher JA. HER-2/neu (c-erb-B2) gene and protein in breast cancer. Am J Clin Pathol 1999; 112: S53–S67. 6. Schnitt SJ. Breast cancer in the 21st century: neu opportunities and neu challenges. Mod Pathol 2001; 14: 213–218. 7. Slamon DJ, Clark GM, Wong SG et al. Human breast cancer: correlation of relapse and survival with amplification of the HER-2/neu oncogene. Science 1987; 235: 177–182. 8. Tsuda H, Hirohashi S, Shimosato Y et al. Correlation between longterm survival in breast cancer patients and amplification of twoputative oncogene co-amplification units: hst-1/int-2 and c-erbB-2/ ear-1. Cancer Res 1989; 49: 3104–3108. 9. Di Leo A, Larsimont D, Gancberg D et al. Predictive molecular markers: a new window of opportunity in the adjuvant therapy of breast cancer. In Naboltz JM, Tonkin K, Aapro MS, Buzdar AU (eds): Breast Cancer Management—Application of Evidence to Patient Care. London, UK: Martin Dunitz 2000; 429–439. 10. Gancberg D, Järvinen T, Di Leo A et al. Evaluation of Her-2/neu protein expression in breast cancer by immunohistochemistry: an interlaboratory study assessing the reproducibility of Her-2 testing. Breast Cancer Res Treat 2002; in press. 11. Vogel C, Cobleigh M, Tripathy D et al. First-line, non-hormonal, treatment of women with HER2 overexpressing metastatic breast cancer with Herceptin (trastuzumab, humanized anti-HER2 antibody). Proc Am Soc Clin Oncol 2000; 19: 71a (Abstr 275). 12. Vogel C, Cobleigh M, Tripathy D et al. First-line single-agent Herceptin (H) therapy for women with HER2-overexpressing metastatic breast cancer (MBC). Eur J Cancer 2000; 36 (Suppl 5): S53 (Abstr 40). 13. Cobleigh MA, Vogel CL, Tripathy D et al. Multinational study of the efficacy and safety of humanized anti-HER2 monoclonal antibody in women who have HER2-overexpressing metastatic breast cancer that has progressed after chemotherapy for metastatic disease. J Clin Oncol 1999; 17: 2639–2648. 14. Slamon DJ, Leyland-Jones B, Shak S et al. Use of chemotherapy plus a monoclonal antibody against HER2 for metastatic breast cancer that overexpresses HER2. N Engl J Med 2001; 344: 783–792. 15. Vogel C, Cobleigh M, Tripathy D et al. Superior outcomes with Herceptin (trastuzumab) in fluorescence in situ hybridization (FISH)selected patients. Proc Am Soc Clin Oncol 2001; 20: 23a (Abstr 86).
1043 16. Cardoso F, Di Leo A, Larsimont D et al. Evaluation of HER-2, p53, bcl-2, topoisomerase II-α, heat shock proteins 27 and 70 in primary breast cancer and metastatic ipsilateral axillary lymph nodes. Ann Oncol 2001; 12: 615–620. 17. Symmans WF, Liu J, Knowles DM et al. Breast cancer heterogeneity: evaluation of clonality in primary and metastatic lesions. Hum Pathol 1995; 26: 210–216. 18. Shimizu C, Fukutomi T, Tsuda H et al. c-erbB-2 protein overexpression and p53 immunoreaction in primary and recurrent breast cancer tissues. J Surg Oncol 2000; 73: 17–20. 19. Barnes DM, Lammie GA, Millis RR et al. An immunohistochemical evaluation of c-erbB-2 expression in human breast carcinoma. Br J Cancer 1988; 58: 448–452. 20. Masood S, Bui MM. Assessment of HER-2/neu overexpression in primary breast cancers and their metastatic lesions: an immunohistochemical study. Ann Clin Lab Sci 2000; 30: 259–265. 21. Lacroix H, Iglehart JD, Skinner MA et al. Overexpression of erbB-2 or EGF receptor proteins present in early stage mammary carcinoma is detected simultaneously in matched primary tumors and regional metastases. Oncogene 1989; 4: 145–151. 22. Tanner M, Järvinen P, Isola J. Amplification of HER-2/neu and topoisomerase II-α in primary and metastatic breast cancer. Cancer Res 2001; 61: 5345–5348. 23. Simon R, Nocito A, Hübscher T et al. Patterns of HER-2/neu amplification and overexpression in primary and metastatic breast cancer. J Natl Cancer Inst 2001; 93: 1141–1146. 24. Iglehart JD, Kraus MH, Langton BC et al. Increased erbB-2 gene copies and expression in multiple stages of breast cancer. Cancer Res 1990; 50: 6701–6707.
25. Niehans GA, Singleton TP, Dykoski D, Kiang DT. Stability of HER-2/neu expression over time and at multiple metastatic sites. J Natl Cancer Inst 1993; 85: 1230–1235. 26. Edgerton SE, Merkel D, Moore DH et al. HER-2/neu/erbB-2 status by immunohistochemistry and FISH: clonality and pogression with recurrence and metastases. Breast Cancer Res Treat 2000; 64: 55 (Abstr 180). 27. Teixeira MR, Teixeira MR, Bardi G et al. Clonal heterogeneity in breast cancer: karyotypic comparisons of multiple intra- and extratumorous samples from 3 patients. Int J Cancer 1995; 63: 63–68. 28. Teixeira MR, Pandis N, Berdi G et al. Karyotypic comparisons of multiple tumours and macroscopically normal surrounding tissue samples from patients with breast cancer. Cancer Res 1996; 56: 855–859. 29. Pandis N, Jin Y, Gorunova L et al. Chromosome analysis of 97 primary breast carcinomas: identification of eight karyotypic subgroups. Genes Chromosom Cancer 1995; 12: 173–185. 30. Trent J, Yang JM, Emerson J et al. Clonal chromosome abnormalities in human breast carcinoma II. Thirty-four cases with metastatic disease. Genes Chromosom Cancer 1993; 7: 194–203. 31. Kuukasjärvi T, Kononen J, Helin H et al. Loss of oestrogen receptor in recurrent breast cancer is associated with poor response to endocrine therapy. J Clin Oncol 1996; 14: 2584–2589. 32. Press MF, Hung G, Godolphin W et al. Sensitivity of HER-2/neu antibodies in archival tissue samples: Potential source of error in immunohistochemical studies of oncogene expression. Cancer Res 1994; 54: 2771–2777.
Annals of Oncology 13: 1036–1043, 2002 DOI: 10.1093/annonc/mdf252
Comparison of HER-2 status between primary breast cancer and corresponding distant metastatic sites D. Gancberg1, A. Di Leo1, F. Cardoso1, G. Rouas 1, M. Pedrocchi2, M. Paesmans3, A. Verhest1, C. Bernard-Marty1, M. J. Piccart1 & D. Larsimont1* 1
Translational Research Unit and 3Data Centre, Jules Bordet Institute, Brussels, Belgium; 2 F. Hoffmann-La Roche Ltd, Basel, Switzerland
Received 22 January 2002; revised 20 March 2002; accepted 11 April 2002
Background: The humanized anti-HER-2 monoclonal antibody trastuzumab (Herceptin®) is a new treatment modality for metastatic breast cancer, the efficacy of which is directly correlated with the HER-2 status of the tumour, evaluated either by immunohistochemistry (IHC) and/or by fluorescence in situ hybridisation (FISH). This analysis is generally performed on the primary tumour. There are few data regarding the HER-2 status in the corresponding distant metastases. Methods: HER-2 status in 107 patients with a primary breast tumour and at least one distant metastatic lesion was analysed by IHC and FISH. Results: We found similar levels of amplification (25% and 24%) and overexpression (13% and 19%) of HER-2 in primary and metastatic samples, respectively. Among paired primary/metastatic tumours, six (6%) showed discordance by HercepTest™ (n = 100): all six cases showed greater Her-2 overexpression in the metastatic tissue. By FISH (n = 68), five (7%) cases were discordant: two cases were amplified in the primary tumour but not in the metastasis, and three samples showed amplification in the metastasis but not in the primary. Finally, we analysed HER-2 status in different metastatic lesions from 17 patients that had at least two distant metastatic sites. Discordance between different sites from the same patient was 18% by IHC and 19% by FISH. Conclusions: Between the paired primary tumour and distant metastatic lesions, 94% and 93% of samples had concordant HER-2 status when analysed by IHC or FISH, respectively. These results do not support routine determination of HER-2 on metastatic sites, particularly when FISH results from the primary tumour are available. Key words: breast cancer, HER-2, metastatic, primary
Introduction Breast cancer is a heterogeneous disease [1]. It has been postulated that the equilibrium of different clones within a tumour is eventually overcome by a biologically dominant one with enhanced metastatic potential [2, 3]. Instead of a linear progression from the primary tumour to the metastasis, an early stem-line clone might evolve independently in both sites [4]. The resulting heterogeneity of metastatic breast cancer (MBC) may explain why biomarkers of prognosis or therapy responsiveness measured exclusively from primary tumours may not entirely reveal all the biological properties of MBC. Overexpression or amplification of proto-oncogenes such as HER-2 may confer biological advantage to tumour cells. The Her-2 oncoprotein (for a review see [5, 6]) is a transmembrane
*Correspondence to: Dr D. Larisimont, Department of Pathology, Jules Bordet Institute, Boulevard de Waterloo 125, 1000 Brussels, Belgium. Tel: +32-2-5413115; Fax: +32-2-5413281; E-mail: [email protected] © 2002 European Society for Medical Oncology
receptor, belonging to the epidermal growth factor receptor (EGFR) family, with tyrosine kinase activity that has been associated with shortened survival [7], enhanced aggressiveness [7, 8] and other poor prognostic factors. However, the literature does not uniformly support a predictive role for HER-2 in breast tumours, mainly due to the lack of prospective studies and the loss of reproducibility of the assessment methods across different laboratories [5, 9, 10]. The only exception is the value of HER-2 as a predictive factor for responsiveness to trastuzumab (Herceptin®), the humanized anti-HER-2 monoclonal antibody [5]. In recent years, the efficacy of trastuzumab in the treatment of MBC that overexpresses HER-2 has been proven, and it is currently under evaluation in the adjuvant setting. Monotherapy with trastuzumab achieves overall response rates of 35% when given as first-line therapy, and 18% in patients previously treated with chemotherapy for MBC [11–13]. Combination treatment with chemotherapy as first-line treatment achieves response rates of 45% and a survival benefit of ∼5 months [14]. These results
1037 led to the approval of trastuzumab as monotherapy for patients with HER-2-positive metastatic breast tumours who received anthracyclines and taxanes for metastatic disease, and in combination with paclitaxel as first-line therapy for HER-2positive patients previously treated with anthracyclines. It has also been proven that the efficacy of trastuzumab is highly dependent on the HER-2 status of the tumour [11–15]. The main targets of any therapy in MBC are the metastases. However, in the great majority of cases, HER-2 status is determined on the primary tumour, and there are few published data regarding the comparison of HER-2 status between the primary and the metastatic sites. Studies have reported a high level of consistency, although not complete, in HER-2 status between primary tumours and locoregional metastases using immunohistochemistry (IHC) [16–20], fluorescence in situ hybridisation (FISH) or both [21–24]. Data referring to distant metastases are extremely scarce [18, 20, 22, 25, 26]. In this study, we compared HER-2 overexpression and amplification in primary tumours and their distant metastases from a series of 107 breast cancer patients with at least one distant metastatic site, using both IHC (HercepTest™) and FISH (triple probe LSI HER-2/topoIIα/CEP17) methods. Additionally, the evaluation of HER-2 status in different metastatic sites from the same patient was performed in 17 cases. This is, to our knowledge, the first study to compare HER-2 status, evaluated by both these techniques, between primary tumours and their distant metastases (locoregional metastases not included), in a relatively large number of patients.
Materials and methods Breast cancer specimens A series of 107 paraffin-embedded blocks of primary breast carcinomas, routinely examined in the Jules Bordet Institute (Brussels, Belgium) from 1981 to 1999, and corresponding distant metastases, were analysed. No selection criteria were used to identify these samples except that the metastatic lesion had to be distant and not locoregional. For each case, haematoxylin and eosin staining was performed to determine whether an invasive tumour was present in the sample. In 17 patients, at least two metastatic samples were available for HER-2 evaluation.
Immunohistochemistry IHC was performed using the HercepTest™ (Dako, Glostrup, Denmark) according to the manufacturer’s instructions. A score of 3+ was considered to be a true positive. Each run included positive and negative controls. The slides were reviewed and scored by one pathologist (D.L.), and only once. The observer was blinded with regard to the FISH data and to the HER-2 status in the paired samples.
Fluorescence in situ hybridisation Slides were deparaffinised in xylene for 5 min, three times, and dehydrated in two changes of 100% ethanol for 5 min. After incubation in a solution of 0.2 M HCl for 20 min at room temperature, slides were washed in deionised water for 1 min and incubated in the pre-treatment reagent (1 M NaSCN) for 30 min at 80°C, and rinsed in deionised water for 3 min.
Slides were then incubated in a solution of 0.2 M HCl/4 mg/ml protease (Paraffin Pre-treatment kit; Vysis, Inc., Downers Grove, IL, USA) at 37°C for 10–20 min, and rinsed in deionised water for 3 min. After controlling the digestion, slides were incubated in neutral-buffered formalin for 10 min, rinsed in deionised water and dehydrated through graded ethanol series (70, 85 and 100%), 1 min each. Slides were placed on a hot plate (73°C) for 5 min with a 10 µl probe (Vysis multicolor-probe Topo IIα Spectrum Green, HER-2 Spectrum Orange and CEP17 Spectrum Aqua), coverslipped and sealed. Denaturation was followed by incubation in humid atmosphere at 37°C for 16–18 h. The next day, the rubber cement was removed and coverslips floated off by soaking the slides at room temperature in 2× SSC/0.3% Nonidet P-40 for 2 min. Slides were then incubated in a solution of 2× SSC/0.3% Nonidet P-40 at 73°C for 2 min, dried in a dark place and counterstained with 10 µl of 0.2 µM 4,6-diamidino-2-phenylindole (DAPI) in antifade solution (Vectashield; Vector Laboratories, Inc., Burlingame, CA, USA). Signals were counted in 60 nuclei using a Leica DMRB or an Olympus BX51 epifluorescence microscope equipped with the following filter sets: DAPI, aqua, fluorescein isothiocyanate (FITC), Texas Red, DAPI/FITC/TRITC, and 40-, 63- and 100-fold magnification oil immersion objectives. Gene amplification was defined as a HER-2 to CEP17 ratio >2. The slides were analysed and classified by two observers (D.G. and G.R.) who were blinded with regard to the IHC data and to the HER-2 status in the paired sample. In only two cases was there discordance between both observers, and in both cases the FISH analysis was repeated and concordance obtained. Owing to problems with tissue preparation that prevent FISH analysis, such as Bouin fixation or decalcification treatment, FISH was successfully performed on 68 primary-metastasis pairs out of the 107 originally available.
Statistical analysis Theoretical proportions of HER-2 positivity were estimated on the basis of the observed proportions in the samples. These estimations were carried out using the samples obtained from the primary tumour and using the samples obtained for the first (chronologically) developed metastatic site. The assessments done on other metastatic sites from the same patient were not considered for estimation of the positivity rates in order to avoid dependency between the observed data (positivity on one metastatic site being most probably linked to positivity on the other sites). Inclusion of multiple sites from the same patient could consequently lead to a possible over- or underestimation of positivity rates. The 95% confidence intervals (CIs) were calculated for the theoretical proportion on the basis of asymptotic normal distributions. Rates of positivity for the primary tumours and the first metastatic site were compared using two-sided paired McNemar’s test at a usual 5% level. Discordance rates between the primary tumour/first metastatic site pairs were similarly analysed. Finally, a descriptive analysis was performed for the 17 patients for whom several metastatic sites were available, in order to assess the possible discordance between the sites. No inferential analysis was performed due to the small number of patients and the fact that the sites were not synchronous.
Results A total of 129 metastatic samples were available: 107 from a first metastatic site, 17 from a second and five from a third. All but four metastases were asynchronous. Metastases were located in bone (n = 38), soft tissue (not locoregional; n = 32), liver (n = 26), lung/bronchus/pleura (n = 13), other gastrointestinal (e.g. stomach/duodenum/biliary tract/peritoneum;
1038 Table 1. (A) HER-2 status assessed by IHC in primary and metastatic breast tumours IHC score (HercepTest™)
Primary tumours (n = 100)
Metastatic tumours (n = 100a)
Events (%)
Events (%)
0
57 (57)
52 (52)
1+
14 (14)
21 (21)
2+
16 (16)
8 (8)
3+
13 (13)
19 (19)
Not evaluable
7
7
a
In the case of multiple metastatic sites only the score observed in the first metastatic site was considered for this analysis.
(B) HER-2 status assessed by FISH in primary and metastatic breast tumours FISH score
Primary tumours (n = 85)
Metastatic tumours (n = 84)a
Events (%)
Events (%)
Amplification
21 (25)
20 (24)
No amplification
64 (75)
63 (76)
Not evaluable
22
24
a
In the case of multiple metastatic sites only the score observed in the first metastatic site was considered for this analysis.
n = 9), ovary (n = 6), brain (n = 2) or other sites (parotid, bladder, pericardium). The time period between surgical removal of the primary tumour and the surgery/biopsy of the metastatic lesion ranged from 1 month to 18 years.
HER-2 status in primary and metastatic breast tumours (Table 1) Using IHC (Table 1A), we found that 13% (95% CI 6% to 20%) of the primary tumours (n = 100) revealed Her-2 overexpression (score 3+). In metastatic lesions (n = 100), 19% (95% CI 11% to 27%) showed 3+ Her-2 overexpression. FISH analysis (Table 1B) of the primary (n = 85) and metastatic (n = 84) samples demonstrated 25% (95% CI 16% to 34%) and 24% (95% CI 15% to 33%) of HER-2/neu amplification, respectively.
Discordance in HER-2 status between IHC and FISH FISH and IHC methods were concomitantly successful for detection of HER-2 in 79 primary tumours (74% of the 107 available patients) and in 97 metastatic samples (75% of the 129 available samples). Regarding the primary tumours, there were nine (11%) discordant cases between the two techniques. All these nine cases were negative by IHC (score 0 to –2) but positive by FISH. Among the 76 evaluable samples of the first metastatic site (71% of the 107 available), there were seven (9%) cases of discordance between IHC and FISH results: five out of seven
of these cases were negative by IHC but amplified by FISH, while two out of seven were positive by IHC but not amplified by FISH. For the samples collected from the second metastatic site, 16 of 17 were evaluable by both techniques. Only three of 16 presented discordant results, two of which were negative by IHC and amplified by FISH, while the inverse was seen in the remaining one. Five patients had samples available from a third metastatic site. In all of these IHC and FISH could be performed and yielded concordant results. Overall, more samples were HER-2 positive when evaluated by FISH than by IHC (P = 0.01). While this occurred in both the primary tumours (P = 0.04) and the metastatic sites (P = 0.73), the difference reached statistical significance only for the primary tumours.
Discordance in HER-2 status between primary and metastases (Table 2) One hundred paired samples were available for IHC analysis (Table 2A). Seven cases were not evaluable due to detachment of the tissue during the pre-treatment. There were 13 and 19 cases with 3+ Her-2 overexpression in the primary tumours and in the metastatic lesions, respectively. Six of these 100 cases (6%) scored differently in the primary tumour and its respective first metastatic site. All cases showed greater Her-2 overexpression in the metastatic site as compared with
1039 Table 2. (A) Discordant data between primary and paired metastatic tumours by IHC (n = 100) Case no.
Primary tumour
Date
First metastasis
Metastatic site
Date
1
Negative
May 1989
Positive
Liver
July 1990
2
Negative
September 1990
Positive
Bronchus
May 1995
3
Negative
November 1991
Positive
Bone
November 1993
4
Negative
August 1995
Positive
Bone
March 1997
5
Negative
January 1994
Positive
Parotid
November 1999
6
Negative
January 1992
Positive
Pericardium
March 1998
(B) Discordant data between primary and paired metastatic tumours by FISH (n = 68) Case no.
Primary tumour
Date
Metastasis
Metastatic site
Date
1
Negative
May 1990
Positive
Bronchus
October 1992
2
Negative
April 1985
Positive
Bone
December 1989
3
Negative
February 1999
Positive
Node
November 2000
4
Positive
June 1989
Negative
Node
August 1992
5
Positive
March 1985
Negative
Pleura
December 1992
the primary tumour. This increase was statistically significant (P = 0.03). When FISH was used (Table 2B), only 68 paired primary/ metastatic tumours were evaluable, due to technical difficulties with Bouin-fixed or decalcified samples. Among these, the HER-2 gene was amplified in 16 primary tumours (24%) and 17 metastatic lesions (25%). Five of 68 cases (7%) were discordant for HER-2 status between the primary site and the metastases. Three cases showed HER-2 amplification in the metastasis but not in the primary sample, and two cases showed HER-2 amplification in the primary tumour but not in the metastasis. No correlation was observed between changes in the HER-2 status and the location of the metastases.
Comparison of HER-2 status in multiple metastatic sites from the same patient (Table 3) Seventeen cases with multiple metastatic sites were evaluable by IHC. Twelve patients had two distinct metastatic sites, and five patients had three metastases. We found discordant Her-2 overexpression in three of 17 (18%; 95% CI 0% to 26%) cases. These three cases, when evaluated by FISH, showed concordant results. When FISH was used, 16 of 17 cases were successfully hybridised. We found differences in HER-2 amplification between different metastatic sites in three of 16 (19%; CI 0% to 38%) cases. These cases were not the same as those that had been found discordant with IHC and, when evaluated by IHC, they showed concordant results. The discordant results among the different metastases were not associated with a particular location of the metastases.
Discussion The present study is, to our knowledge, the first to compare the HER-2 status of primary breast tumours exclusively with their distant metastases, using both IHC and FISH techniques, in a relatively large number of MBC patients. This issue has become important in recent years due to the development and approval of trastuzumab for the treatment of HER-2-overexpressing MBC. In the metastatic setting, where the disease is virtually incurable, few drugs have shown a significant advantage, not only in terms of response rates but particularly in terms of overall survival. Trastuzumab, when given in combination with paclitaxel, achieved a survival benefit of 5 months compared with single-agent treatment of paclitaxel [14]. On the other hand, although very well tolerated, trastuzumab has a non-negligible risk of cardiotoxicity. Since trastuzumab activity is directly dependent on the HER-2 status of the tumour, a careful selection of patients is crucial for increasing its clinical benefit and avoiding unnecessary treatment of patients who most likely will not benefit from it. To date, the HER-2 status is almost always evaluated on the primary tumour sample, since routine biopsy of metastatic sites is not a standard procedure. Thus, it becomes imperative to determine whether homogenous HER-2 expression exists between the primary tumour and its distant metastases. Most published studies have compared HER-2 status in primary tumours with their locoregional or concurrent lymph node metastases. Only five studies have presented data concerning distant metastases [18, 20, 22, 25, 26], and only one of these [25] focused exclusively on distant metastases. However, it should not be assumed that both types of metastases are biologically equivalent. Distant metastases often represent clonal outgrowths with genetic modifications not detected in
1040 Table 3. (A) Comparison of multiple metastatic (M) sites from the same patient by IHC (n = 17) Case
M1 (site)
M2 (site)
M3 (site)
Primary
1
Negative (node)
Negative (node)
Negative
2
Negative (node)
Negative (node)
Negative
3
Negative (stomach)
Negative (stomach)
4
Negative (node)
Positive (node)
5
Positive (node)
Negative (ovary)
6
Positive (pericardium)
Negative (pleura)
7
Negative (bone)
Negative (bone)
Negative
8
Negative (brain)
Negative (brain)
Negative
9
Negative Negative (pleura)
Negative Positive
Negative (diaphragm)
Negative
Negative (bronchus)
Negative (bronchus)
Negative
10
Negative (ovary)
Negative (ovary)
Negative
11
Negative (bone)
Negative (bone)
Negative
12
Negative (bone)
Negative (bone)
Negative
13
Negative (skin)
Negative (skin)
14
Negative (node)
Negative (skin)
15
Negative (duodenum)
Negative (stomach–antrum)
16
Positive (bone)
Positive (bone)
17
Positive (ovary)
Positive (epiploon)
Negative (skin)
Negative Negative
Negative (stomach–fundus)
Negative Positive
Positive (node)
Positive
Discordant cases between different metastatic sites are reported in bold type (B) Comparison of multiple metastatic (M) sites from the same patient by FISH (n = 16) Case
M1 (site)
M2 (site)
1
Negative (node)
Positive (node)
M3 (site)
Positive
2
Negative (node)
Positive (node)
Negative
3
Positive (stomach)
Negative (stomach)
Positive
4
Negative (node)
Negative (node)
5
Positive (node)
Positive (ovary)
6
Negative (pericardium)
Negative (pleura)
8
Negative (brain)
Negative (brain)
Negative
9
Negative (pleura)
Primary
Negative Positive
Negative (diaphragm)
Negative
Negative (bronchus)
Negative (bronchus)
NE
10
Negative (ovary)
Negative (ovary)
Negative
11
Negative (bone)
Negative (bone)
Negative
12
Negative (bone)
Negative (bone)
Negative
13
Negative (skin)
Negative (skin)
14
Negative (node)
Negative (skin)
15
Negative (duodenum)
Negative (stomach–antrum)
16
Positive (bone)
Positive (bone)
17
Positive (ovary)
Positive (epiploon)
Negative (skin)
Negative NE
Negative (stomach–fundus)
NE Positive
Positive (node)
Positive
In case 7, FISH results were not available for the second metastatic site. Discordant cases between different metastatic sites are reported in bold type. NE, not evaluable.
the primary tumour, and are generally identified years after resection (and therapy) of the primary cancer. In addition, cells that metastasise locally via the lymphatic system may possess biological properties distinct from those that have the
capacity to invade vasculature, travel to distant sites and establish clinically significant disease. Cytogenetic analyses have demonstrated extreme genetic heterogeneity (up to 70% of polyclonality) in breast cancer [27–31]. Kuukasjärvi et al. [4]
1041 reported that the genetic composition of 31% of metastases differed almost completely from that of the paired primary tumours. Our group has previously compared the Her-2 status, among other biological markers, between 370 primary breast cancer sites and their metastatic ipsilateral axillary lymph nodes [16]. Her-2, evaluated by IHC, was the marker with the best overall concordance (98%); no biological marker showed 100% concordant results. Similarly high levels of concordance have been reported by other authors [17, 19, 21, 23, 24]. However, as has been pointed out, all these studies were restricted to locoregional metastases. Three of these studies determined HER-2 status by both IHC and a DNA hybridisation technique [21, 23, 24], and two by IHC alone [17, 19]. Another common pitfall is the number of evaluated patients: only two of the six studies had >100 evaluable patients [23, 24], and none had >200; the number of paired samples (primary/locoregional metastasis) is even smaller. Only five studies compared the HER-2 status in breast primary tumours and distant metastases [18, 20, 22, 25, 26]. In 1993, Niehans et al. [25] retrospectively evaluated Her-2 expression in primary breast cancer and different metastatic sites in autopsy tumour samples from 30 patients who had died from the disease. Eight of these patients (27%) were Her-2 positive and, among those, there was a single case of discordant staining (3%) between the primary site and two of its metastatic lesions. The authors concluded that Her-2 expression is generally congruent at different metastatic sites. This is the only published study that evaluated exclusively distant metastases. Unfortunately, the actual number of evaluable patients was very low, and the uncontrolled fixation time of autopsy samples jeopardizes the IHC data that can be obtained from these samples. In 2000, Shimizu et al. [18] evaluated Her-2 protein levels, by IHC, in primary and metastatic breast cancer samples from 21 patients. Only seven of the metastatic samples were from distant sites, with the remaining 14 belonging to locoregional relapses. The authors found no significant differences in the Her-2 expression between the primary tumours and their metastases. Also in 2000, Masood et al. [20] evaluated Her-2 overexpression in 56 patients by IHC. Once again, only a minority (11 cases) of the metastases were from distant sites. The pattern and intensity of Her-2 overexpression in the primary tumours and their metastases were found to be almost identical, with heterogeneity present in only one case. Recently, Tanner et al. [22] analysed HER-2 amplification in 46 breast primary tumours and their metastases, using IHC and DNA in situ hybridisation techniques. The authors found complete concordance regarding HER-2 amplification between the primary tumours and their metastases. However, in only 12 cases were the analysed metastases from a distant site. In the 2000 San Antonio Breast Cancer Meeting, Edgerton et al. [26] presented some preliminary results of a study comparing HER-2 status, determined by IHC and FISH, on 193 primary breast cancers and corresponding local recur-
rences (68), lymph node metastases (32) and distant metastases (93). Overall, the authors found 25% discordant cases. Publication of detailed results and methodology of this study would be of interest. In the present study, we analysed both Her-2 protein overexpression by IHC and HER-2 oncogene amplification by FISH in 107 patients with at least one distant metastatic site. Owing to technical difficulties, the number of evaluable pairs (primary/distant metastasis) was 100 for IHC and 68 for FISH. The relatively large number of samples not evaluable is an important limitation, encountered in all studies performed in paraffin-embedded tumour samples, and was mainly due to bouin fixation or decalcification treatment (needed for bone metastases), which compromise hybridisation. Seventeen cases with multiple metastatic sites were also comparatively studied. The Vysis multicolour FISH probe, which includes the topoisomerase IIα gene analysis, was used since the original plan was the evaluate the status of both the HER-2 and the topoisomerase IIα gene in the primary tumour and its corresponding metastatic sites. Topoisomerase IIα gene results are not reported, since the probe was found to be inadequate to evaluate this gene. A new topoisomerase IIα probe has been developed by Vysis and is now being used to evaluate the status of this gene in the present series. Similar levels of amplification of the HER-2 oncogene were found in the primary tumours and in the metastatic samples (25% and 24%, respectively) using FISH (Table 1B). Overexpression of Her-2, assessed by HercepTest™, was found in 13% of the primary and 19% of the metastatic samples when only 3+ scores were considered to be positive. When 2+ scores were included, Her-2 overexpression was found in 29% of the primary tumours and 27% of the metastatic lesions (Table 1A). This means that using strong staining (3+) with the HercepTest™ as the sole indicator of Her-2 positivity was very stringent and that it underestimated the number of samples with HER-2 gene amplification (P = 0.01). Controversy related to the best IHC protocol, in particular the best antibody, to use for Her-2 determination is still ongoing. However, we have adopted an IHC protocol using the monoclonal antibody CB-11 with antigen retrieval, which allows us to decrease the rate of false negatives; all positive results are confirmed by FISH, therefore reducing the rate of false positives. This attitude seems well balanced in terms of efficiency and costs. HER-2 status in the primary and corresponding metastatic lesions revealed a high level of concordance (94% and 93% when analysed by IHC or FISH, respectively). Discordant results were observed in only six of 100 cases (6%) by IHC (Table 2A) and five of 68 cases (7%) by FISH (Table 2B). Regarding IHC, all six discordant cases presented an increase in staining in the metastatic site in comparison to the primary tumour. This increase in IHC staining when the tumour disseminates was statistically significant (P = 0.03). Four of these six cases had a 2+ score in the primary tumour. Among the five discordant cases by FISH, in three the HER-2 gene
1042 became amplified in the metastatic site while it was nonamplified in the primary tumour, and in two the opposite occurred (HER-2 gene amplification in the primary tumour but not in the metastasis). Including all metastatic sites in the analysis, our IHC data suggest that Her-2-positive metastatic lesions with negative primary tumours were more frequent than the converse situation (Table 2A). This is in line with the biological fact that Her-2 is correlated with enhanced tumour aggressiveness. Another possible explanation may be an underestimation of Her-2 protein overexpression in the primary tumour by HercepTest™; this could be due to the fact that most primary samples were collected many years before the present analysis (see Table 2 for specific dates), and therefore some protein degradation might have occurred. Previous studies have shown that the sensitivity of IHC when used in paraffinembedded fixed tumour samples varies widely, depending not only on the antibody used but also on the tissue processing, namely tissue fixation and the ‘age’ of the samples, which interferes greatly with their quality [32]. This explanation is supported by the FISH data, which show a very similar percentage of HER-2-positive samples between pairs and therefore seems not to support the hypothesis of enhanced tumour aggressiveness. In fact, when FISH was used, among the 16 of 68 primary samples with HER-2 gene amplification (Table 2B), two (13%) had corresponding distant metastases with a non-amplified HER-2 gene, while among the 52 of 68 HER-2-negative primary samples, three (6%) had corresponding metastatic sites with HER-2 gene amplification. The HER-2 status in samples from multiple distant metastatic sites belonging to the same patient were analysed in 17 cases (Table 3). Thirteen percent and 18% discordant results between samples from different metastases were found when assessed by FISH and IHC, respectively. Previously, only the study by Niehans et al. [25] had looked into this issue, but did so using solely the IHC technique: they found a single case among the 30 evaluated that presented different staining among the four metastatic sites (two were Her-2 positive and two were Her-2 negative). In conclusion, this study does not support the routine testing of metastases to confirm HER-2 positivity when detected in the primary tumour, particularly if results obtained by FISH are available. Assessment of HER-2 status in one of the metastatic sites may be worthwhile only in some patients with easily accessible metastases and for whom Her-2 evaluation by IHC, recently performed in a primary tumour sample collected many years before, shows a negative score. An alternative solution would be the determination of HER-2 amplification by FISH in the primary tumour sample. Translational research efforts to understand the mechanisms of response and resistance to trastuzumab should be a priority. Studies aimed at identify additional predictive markers of responsiveness to trastuzumab (e.g. phosphorylated Her-2, EGFR, HER-3, HER-4) and other possible molecular mechanisms involved are underway.
Acknowledgements The authors thank Carolyn Straehle, PhD, for editorial assistance. G.R. is supported by a grant from ‘Les Amis de l’Institut Bordet’, Brussels, Belgium. The costs related to the detection kit for HER-2 evaluation by FISH and IHC were funded by Hoffmann-La Roche, Basel, Switzerland.
References 1. Bièche I, Lidereau R. Genetic alterations in breast cancer. Genes Chromosom Cancer 1995; 14: 227–251. 2. Heppner GH. Tumor cell societies. J Natl Cancer Inst 1989; 81: 648–649. 3. Kerbel RS. Growth dominance of the metastatic cancer cell: cellular and molecular aspects. Adv Cancer Res 1990; 55: 87–132. 4. Kuukasjärvi T, Karhu R, Tanner M et al. Genetic heterogeneity and clonal evolution underlying development of asynchronous metastasis in human breast cancer. Cancer Res 1997; 57: 1597–1604. 5. Ross JS, Fletcher JA. HER-2/neu (c-erb-B2) gene and protein in breast cancer. Am J Clin Pathol 1999; 112: S53–S67. 6. Schnitt SJ. Breast cancer in the 21st century: neu opportunities and neu challenges. Mod Pathol 2001; 14: 213–218. 7. Slamon DJ, Clark GM, Wong SG et al. Human breast cancer: correlation of relapse and survival with amplification of the HER-2/neu oncogene. Science 1987; 235: 177–182. 8. Tsuda H, Hirohashi S, Shimosato Y et al. Correlation between longterm survival in breast cancer patients and amplification of twoputative oncogene co-amplification units: hst-1/int-2 and c-erbB-2/ ear-1. Cancer Res 1989; 49: 3104–3108. 9. Di Leo A, Larsimont D, Gancberg D et al. Predictive molecular markers: a new window of opportunity in the adjuvant therapy of breast cancer. In Naboltz JM, Tonkin K, Aapro MS, Buzdar AU (eds): Breast Cancer Management—Application of Evidence to Patient Care. London, UK: Martin Dunitz 2000; 429–439. 10. Gancberg D, Järvinen T, Di Leo A et al. Evaluation of Her-2/neu protein expression in breast cancer by immunohistochemistry: an interlaboratory study assessing the reproducibility of Her-2 testing. Breast Cancer Res Treat 2002; in press. 11. Vogel C, Cobleigh M, Tripathy D et al. First-line, non-hormonal, treatment of women with HER2 overexpressing metastatic breast cancer with Herceptin (trastuzumab, humanized anti-HER2 antibody). Proc Am Soc Clin Oncol 2000; 19: 71a (Abstr 275). 12. Vogel C, Cobleigh M, Tripathy D et al. First-line single-agent Herceptin (H) therapy for women with HER2-overexpressing metastatic breast cancer (MBC). Eur J Cancer 2000; 36 (Suppl 5): S53 (Abstr 40). 13. Cobleigh MA, Vogel CL, Tripathy D et al. Multinational study of the efficacy and safety of humanized anti-HER2 monoclonal antibody in women who have HER2-overexpressing metastatic breast cancer that has progressed after chemotherapy for metastatic disease. J Clin Oncol 1999; 17: 2639–2648. 14. Slamon DJ, Leyland-Jones B, Shak S et al. Use of chemotherapy plus a monoclonal antibody against HER2 for metastatic breast cancer that overexpresses HER2. N Engl J Med 2001; 344: 783–792. 15. Vogel C, Cobleigh M, Tripathy D et al. Superior outcomes with Herceptin (trastuzumab) in fluorescence in situ hybridization (FISH)selected patients. Proc Am Soc Clin Oncol 2001; 20: 23a (Abstr 86).
1043 16. Cardoso F, Di Leo A, Larsimont D et al. Evaluation of HER-2, p53, bcl-2, topoisomerase II-α, heat shock proteins 27 and 70 in primary breast cancer and metastatic ipsilateral axillary lymph nodes. Ann Oncol 2001; 12: 615–620. 17. Symmans WF, Liu J, Knowles DM et al. Breast cancer heterogeneity: evaluation of clonality in primary and metastatic lesions. Hum Pathol 1995; 26: 210–216. 18. Shimizu C, Fukutomi T, Tsuda H et al. c-erbB-2 protein overexpression and p53 immunoreaction in primary and recurrent breast cancer tissues. J Surg Oncol 2000; 73: 17–20. 19. Barnes DM, Lammie GA, Millis RR et al. An immunohistochemical evaluation of c-erbB-2 expression in human breast carcinoma. Br J Cancer 1988; 58: 448–452. 20. Masood S, Bui MM. Assessment of HER-2/neu overexpression in primary breast cancers and their metastatic lesions: an immunohistochemical study. Ann Clin Lab Sci 2000; 30: 259–265. 21. Lacroix H, Iglehart JD, Skinner MA et al. Overexpression of erbB-2 or EGF receptor proteins present in early stage mammary carcinoma is detected simultaneously in matched primary tumors and regional metastases. Oncogene 1989; 4: 145–151. 22. Tanner M, Järvinen P, Isola J. Amplification of HER-2/neu and topoisomerase II-α in primary and metastatic breast cancer. Cancer Res 2001; 61: 5345–5348. 23. Simon R, Nocito A, Hübscher T et al. Patterns of HER-2/neu amplification and overexpression in primary and metastatic breast cancer. J Natl Cancer Inst 2001; 93: 1141–1146. 24. Iglehart JD, Kraus MH, Langton BC et al. Increased erbB-2 gene copies and expression in multiple stages of breast cancer. Cancer Res 1990; 50: 6701–6707.
25. Niehans GA, Singleton TP, Dykoski D, Kiang DT. Stability of HER-2/neu expression over time and at multiple metastatic sites. J Natl Cancer Inst 1993; 85: 1230–1235. 26. Edgerton SE, Merkel D, Moore DH et al. HER-2/neu/erbB-2 status by immunohistochemistry and FISH: clonality and pogression with recurrence and metastases. Breast Cancer Res Treat 2000; 64: 55 (Abstr 180). 27. Teixeira MR, Teixeira MR, Bardi G et al. Clonal heterogeneity in breast cancer: karyotypic comparisons of multiple intra- and extratumorous samples from 3 patients. Int J Cancer 1995; 63: 63–68. 28. Teixeira MR, Pandis N, Berdi G et al. Karyotypic comparisons of multiple tumours and macroscopically normal surrounding tissue samples from patients with breast cancer. Cancer Res 1996; 56: 855–859. 29. Pandis N, Jin Y, Gorunova L et al. Chromosome analysis of 97 primary breast carcinomas: identification of eight karyotypic subgroups. Genes Chromosom Cancer 1995; 12: 173–185. 30. Trent J, Yang JM, Emerson J et al. Clonal chromosome abnormalities in human breast carcinoma II. Thirty-four cases with metastatic disease. Genes Chromosom Cancer 1993; 7: 194–203. 31. Kuukasjärvi T, Kononen J, Helin H et al. Loss of oestrogen receptor in recurrent breast cancer is associated with poor response to endocrine therapy. J Clin Oncol 1996; 14: 2584–2589. 32. Press MF, Hung G, Godolphin W et al. Sensitivity of HER-2/neu antibodies in archival tissue samples: Potential source of error in immunohistochemical studies of oncogene expression. Cancer Res 1994; 54: 2771–2777.