Background
Breast cancer is the most frequent non-skin malignancy and the second leading cause of cancer death in American and European women [
1,
2]. Adjuvant chemotherapy is administered to most patients with high-risk operable breast cancer, since it prolongs disease-free survival (DFS) and overall survival (OS) [
3]. Anthracyclines and taxanes are considered to be two of the most efficient drugs in this setting [
4,
5]. Despite intensive clinical research devoted to the role of adjuvant chemotherapy, the majority of patients do not benefit from its use and a small but considerable percentage of them suffer from long-term life-threatening toxicities, such as acute leukemia, myelodysplasia or irreversible congestive heart failure [
6,
7]. To select candidate patients for such aggressive treatments, robust prognostic markers in human breast cancer are needed. Investigators intensively evaluate well-established oncogenes or chromosome aberrations, using large tumor repositories, in an effort to widen their knowledge on the molecular mechanisms, gene interrelationships or gene function underlying breast cancer.
It has long been established that breast cancer is often characterized by gains or losses of specific chromosomes, leading to activation of oncogenes or inactivation of tumor suppressor genes [
8]. Chromosome 17 is the second most gene-dense chromosome in the human genome, housing important genes for breast cancer pathophysiology, such as
BRCA1, HER2, RAD51C, RARA, TOP2A and
TP53[
9]. Changes of chromosome 17 copy number (aneusomy) are extremely frequent in breast cancer [
10]. These chromosome aberrations (reviewed in ref. [
11]) are tightly linked to important cell functions, such as proliferation, apoptosis, angiogenesis and motility. Increased numbers of centromere 17 copies are seen in 10% to 50% of breast tumors [
12‐
14], depending on the criteria used, and this is more common in tumors with
HER2 gene amplification. However, it has to be stated that an increase in chromosome 17 signals seen with fluorescence in situ hybridization (FISH) does not always correspond to true polysomy of the whole chromosome, but may rather represent a focal pericentromeric gain or partial polysomy [
15]. Other abnormalities of chromosome 17 include losses and gains of genetic material in both the p and q arms, focal copy number gains and losses and other structural rearrangements [
15,
16]. Indeed, recent studies using different techniques, such as comparative genomic hybridization (CGH) [
17,
18], multiplex ligation-dependent probe amplification (MLPA) [
19], single nucleotide polymorphism arrays (SNP arrays) [
15], or FISH using alternative chromosome 17 reference genes (
RARA, TP53, SMS) [
20] suggested that true chromosome 17 polysomy is a rare event in breast cancer. In fact, in most of the cases, polysomy, as detected by FISH or chromogen in situ hybridization (CISH), actually reflects a gain or amplification in the pericentromeric region of the chromosome [
21]. For these reasons the term “CEP17 gain” instead of “chromosome 17 polysomy”, is used here, referring to its detection by the centromere 17 enumeration probe (CEP17 probe).
CEP17 gain has been incriminated for the inconsistencies seen in cases with HER2 gene amplification defined by absolute gene copy numbers, versus gene amplification defined by the ratio of HER2 gene copy number to CEP17. Misclassification of HER2 gene status based on dual-color FISH assays, due to CEP17 gain, may have important therapeutic implications since a number of patients considered being HER2-negative by the second definition could be denied trastuzumab.
Importantly, recently published data from retrospectively assessed (although prospectively collected) tumors, by triple color FISH, from 1762 patients who participated in the National Epirubicin Adjuvant Trial (NEAT/BR9601) suggested that CEP17 duplication was associated with increased relapse-free and overall survival in patients treated with an anthracycline compared to CMF [
22].
The
HER2 oncogene is located on the long arm of chromosome 17 (17q12) [
23].
HER2 amplification and/or protein overexpression has been identified in 15% to 25% of invasive breast tumors [
24] and is associated with worse prognosis [
25].
HER2 gene amplification has been shown to predict benefit from the use of several chemotherapeutic agents, including anthracyclines and paclitaxel [
26,
27]. Notably, a meta-analysis provided compelling evidence that the use of anthracyclines benefits exclusively those patients with
HER2 amplification [
28]. However, other investigators could not confirm these data [
29,
30], suggesting that other genes, also located on chromosome 17, may regulate anthracycline responsiveness [
26].
One such gene is the topoisomerase II alpha gene (
TOP2A), which is located ~700 kb telomerically to
HER2 and encodes the alpha isozyme of the human topoisomerase II. In general, topoisomerases are responsible for transcription, replication and chromosome condensation and segregation during cell division [
31,
32].
TOP2A in particular is considered a molecular target for anthracyclines and other chemotherapeutic agents [
33,
34]. The
TOP2A gene is amplified in 30%-40% of the tumors with
HER2 gene amplification, while deletions are frequently observed [
35]. TOP2A gene amplification [
36] and, perhaps, topoisomerase II alpha (TopoIIa) protein overexpression [
37] may benefit high-risk breast cancer patients treated with anthracyclines.
Information regarding the interaction of
HER2/TOP2A gene status in the presence of CEP17 gain with the outcome of breast cancer patients is limited. This urged us to investigate the prognostic role of HER2 and TopoIIa protein expression, as well as
HER2 and
TOP2A gene status along with CEP17 gain in a large cohort of breast cancer patients. This is a prospective-retrospective study as described by Simon [
38], performed in the context of two randomized, consecutively conducted, phase III trials (HE10/97 and HE10/00) with epirubicin-based adjuvant chemotherapy with or without paclitaxel [
39‐
42].
Discussion
In the present study we investigated the prognostic role of CEP17 gain in relation to
HER2 and
TOP2A gene status and protein expression in 1031 patients with operable breast cancer. All these patients were treated with epirubicin-based adjuvant chemotherapy in the context of two consecutively conducted phase III trials [
39‐
41]. In a previous study published by our group for the HE10/00 and HE10/97 cohorts [
42], patients with either luminal B, luminal-HER2 or HER2-enriched tumors performed worse than those with luminal A tumors, while patients with triple-negative tumors had the worst outcome. In addition, it was observed that the HER2-enriched subtype was predictive of response to paclitaxel-containing treatments. These prognostic and predictive HER2-related effects were breast cancer subtype specific and were not maintained in the present study. An earlier observation reported for this cohort, of HER2 amplification being predictive for OS benefit from adjuvant treatment with paclitaxel [
56], was not confirmed in the current analysis with updated follow-up. In both analyses however, the ability to detect any predictive impact of
HER2/
TOP2A amplification or CEP17 gain in the presence of taxanes was limited (only 1 of the 4 trial arms did not include taxanes). The present results concerning HER2 are in line with reports on the prognostic value of this marker [
59,
60]. A recent meta-analysis suggests that patients with both
HER2 amplified and non-amplified tumors may benefit from anthracyclines [
61,
62]. This could not be investigated in the current study, since all patients had been treated with anthracyclines.
Among
HER2 amplified tumors, 42% exhibited
TOP2A co-amplification, which is within the reported range of 35%-50% for this genomic alteration [
26,
29,
63,
64].
TOP2A deletions were more common in
HER2 amplified tumors, comprising approximately 10% of the
HER2 amplified cases.
TOP2A gene pathology (amplification, deletion and combinations of both) has been reported as a favorable prognostic and predictive marker in adjuvant-treated breast cancer patients [
36,
65]. However, in the present study we did not observe any association between patient outcome and
TOP2A amplification, deletion, or both, in accordance with the recent meta-analysis mentioned above [
62].
The clinical importance of CEP17 gain, as detected by FISH, in human breast cancer remains a controversial issue. From the biological perspective, CEP17 gain and chromosome 17 polysomy do not represent the same situation, since the first one corresponds to the fluorescent signals of a 5.6 kb region, while the second reflects aberrant numbers of the whole chromosome, which should be demonstrated with spectral karyotyping (SKY) or other cytogenetic approaches. The CEP17 FISH probe detects the alpha-satellite repeat region at the centromere of chromosome 17, at 17p11.1-17q11.1. The specificity of CEP17 remains undetermined, while this probe and a centromeric probe detecting additional neighboring regions on 17p11.2-12 yield different results concerning chromosome 17 status and, therefore, different
HER2 gene amplification status, when the latter is assessed as
HER2/CEP17 ratio of >2.2 [
66]. This may reflect the presence of the probed satellite repeats outside the centromeric region, which happen during evolution [
67] and probably during cancer clone evolution, as well. Another problem for assessing low copy gains is that during DNA synthesis and in the G2/M phases, the targeted regions will appear double (three to four copies instead of two, taking into account the nuclear truncation effect during paraffin block sectioning). The cut-off used in the present study for the classification of CEP17 gains has been shown to correct for the maximally four centromeric signals that would be expected in this situation [
55]. With this cut-off, we detected CEP17 gain in approximately 40% of all carcinomas examined, which is in the range of published results when using the FISH method in all-type breast carcinoma series (10-50%) [
12‐
14,
54,
68,
69].
Measurement of CEP17 probe signals reflects the condition of the corresponding centromeric area and can by no means reflect gains of the entire chromosome 17 or “chromosome 17 polysomy”, as often reported in the literature. In the same treatment settings, depending on how CEP17 signals are classified and interpreted, and also depending on the drugs administered, the effect of CEP17 status on patient outcome may vary. Thus, in the adjuvant setting, by using the same FISH probe, duplication of the CEP17 region of chromosome 17 seems to be predictive of benefit from anthracyclines [
22,
70] or of borderline association with clinical response to the same drugs [
71]. Furthermore, CEP17 gain in the absence of
HER2/
TOP2A amplification has been reported by one recent study to be an unfavorable prognostic marker [
72]; however, no prognostic value was identified for CEP17 gain in other studies [
55,
64,
69], which is in line with our present findings.
CEP17 appears to be related with disease prognosis when this marker is combined with
HER2 status. Whether “polysomy” 17 drives
HER2 amplification or the opposite is true, as was recently suggested [
54,
73], remains unanswered; it is, however, noteworthy that “polysomy” 17 is rarely observed in circulating tumor cells from patients with metastatic breast cancer and when present, it corresponds to HER2-negative primary tumors [
74]. In the absence of
HER2 amplification, CEP17 “polysomy” has been reported to confer a more favorable prognosis [
75] or to be associated with aggravating prognostic markers [
76]. In addition to these contradictory results, herein we did not observe any interaction between CEP17 and
HER2 status. Methodological differences in the assessment of these parameters and cohort-fitted results should account for the diversity of data regarding the role of CEP17/
HER2 status on adjuvant-treated breast cancer patients.
Although more than half of the tumors with CEP17 gain were not
HER2 or
TOP2A amplified, we did observe a higher than double incidence of CEP17 gain in tumors with aberrant
HER2 and
TOP2A genes (amplification, deletion, and, especially, high copy gains) in comparison to tumors with a normal status of these genes. These data are in line with “polysomy” 17 correlating with multiple copies of
HER2 but not with
HER2 amplification [
77], while they further justify the higher incidence of CEP17 gain in the luminal-HER2 and HER2-enriched subtypes, as described in this study.
Equivocal HER2 IHC findings were observed in cases with chromosome 17 “polysomy” and correspondingly increased
HER2 gene copy numbers [
54,
78]. In the present study, the incidence of CEP17 gains was strongly related to HER2 IHC grades, but it did not contribute to the further assessment of HER2 IHC 2+ cases. With respect to FISH equivocal cases, herein we used very stringent criteria involving both gene/CEP17 ratios and gene copy numbers. Increased CEP17 ratios might result in false negative
HER2/CEP17 ratios of ≤2.2; on the other hand, their coincidence with increased
HER2 copies is not equivalent to HER2-positive disease [
79]. The few (n = 10, <0.01% of the entire tumor series) FISH equivocal tumors for
HER2 gene status were HER2 IHC 0 to 2+. CEP17 gain was observed in only one such case. Hence, at least in the present series, CEP17 gain did not aid further in the classification of equivocal
HER2 gene status cases. In addition, we observed that most of the tumors with low-gain of
HER2 and/or
TOP2A copies indeed had CEP17 gains as well. However, in order to evaluate the impact of these concomitant alterations on patient outcome, larger patient series with this sub-category of tumors would be needed.
Acknowledgements
Presented in part at the 34th Annual San Antonio Breast Cancer Symposium, December 6–10, 2011.
The authors are indebted to all patients and their families for their trust and participation in the HE10/97 and HE10/00 trials and for the provision of biological material for research purposes.
The authors also wish to thank all HeCOG personnel (data managers, research assistants and monitors) for their dedication, M. Moschoni for data coordination, T. Spinari for collection of FFPE tissue blocks and S. Dallidou for secretarial assistance.
Supported by an internal Hellenic Cooperative Oncology Group (HeCOG) translational research grant (HE TRANS_BR).
Competing interests
On behalf of the Hellenic Foundation for Cancer Research, Athens, Greece, the senior author (GF) has pending patent applications with Siemens Healthcare Diagnostics, Tarrytown, NY. The rest of the authors declare that they have no competing interests.
Authors’ contributions
GF conceived of the study, participated in its design and coordination, contributed to the analysis and interpretation of data and drafted the manuscript. UD conceived of the study, participated in its design, contributed to the analysis and interpretation of data and drafted the manuscript. MB carried out the IHC and FISH analysis, contributed to the analysis and interpretation of data and drafted the manuscript. VK conceived of the study, participated in its design, carried out the molecular studies, contributed to the analysis and interpretation of data and drafted the manuscript. AB carried out the immunoassays and contributed to the analysis and interpretation of data. IX participated in the acquisition of data and contributed to the collection of the tumor tissue samples analyzed in the study. CP participated in the acquisition of data and contributed to the collection of the tumor tissue samples analyzed in the study. TK carried out the molecular studies and contributed to the analysis and interpretation of data. ET carried out the FISH assays and contributed to the analysis and interpretation of data. DT carried out the immunoassays and contributed to the analysis and interpretation of data. ET participated in the acquisition of data and contributed to the collection of the tumor tissue samples analyzed in the study. AK participated in the acquisition of data and contributed to the collection of the tumor tissue samples analyzed in the study. GK participated in the acquisition of data. ES participated in the acquisition of data and contributed to the collection of the tumor tissue samples analyzed in the study. NP participated in the acquisition of data and contributed to the collection of the tumor tissue samples analyzed in the study. CK participated in the acquisition of data and contributed to the collection of the tumor tissue samples analyzed in the study. IS participated in the acquisition of data. NP participated in the acquisition of data and contributed to the collection of the tumor tissue samples analyzed in the study. HG participated in the acquisition of data and contributed to the collection of the tumor tissue samples analyzed in the study. HL participated in the acquisition of data and contributed to the collection of the tumor tissue samples analyzed in the study. KTK conceived of the study, participated in its design, contributed to the analysis and interpretation of data and drafted the manuscript. DP conceived of the study, participated in its design and contributed to the analysis and interpretation of data. MAD conceived of the study, participated in its design, contributed to the analysis and interpretation of data and drafted the manuscript. All authors read and approved the final manuscript.