Departments of 1 Human Genetics, 2 Oncology, 3 Surgery, 4 Pathology and 5 Medicine, and 6 Research Institute of the McGill University Health Centre, 7 Program in Cancer Genetics, 8 Cancer Prevention Centre, Sir M. B. Davis-Jewish General Hospital, McGill University, Montreal, QC, Canada; 9 Divisions of Human Biology and Public Health Sciences, Fred Hutchinson Cancer Research Center and Department of Pathology, University of Washington, Seattle, WA, USA; 10 Department of Pathology, Sunnybrook and Women's College Health Sciences Centre, University of Toronto, Toronto, ON; 11 Algorithme Pharma, Montreal, QC, Canada
*Correspondence to: Dr P. O. Chappuis, Service of Oncology, University Hospitals of Geneva, 24 rue Micheli-du-Crest, 1211 Geneva 14, Switzerland. Tel: +41-22-372-98-78; Fax: +41-22-372-98-58; Email: pierre.chappuis{at}hcuge.ch
![]() |
Abstract |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Patients and methods:: We analyzed a historical cohort of 288 Ashkenazi Jewish women who were diagnosed with breast cancer between 1980 and 1995 and were previously tested for BRCA1/2 mutations. Protein levels of cyclin E and p27Kip1 were assessed by immunohistochemistry. Breast cancer-specific survival (BCSS) was the main outcome measured.
Results:: The median follow-up was 8 years. Thirty tumors carried germline BRCA1 mutations. These tumors were more likely to have high cyclin E protein levels [odds ratio (OR) 9.5; P <0.001] and low p27Kip1 protein levels (OR 2.8; P=0.03) than tumors from patients without BRCA1/2 mutations. High cyclin E expression level was the strongest predictor of BRCA1 germline mutations (multivariate OR 4.7; P=0.004). On univariate analysis, high cyclin E protein levels [relative risk (RR) 2.6; P <0.001] and low p27Kip1 protein levels (RR 2.3; P=0.006) were significant prognostic factors for a poorer BCSS. In Cox multivariate models, high cyclin E levels remained an independent indicator of poor outcome only in the subgroup of patients who did not receive chemotherapy (P=0.002).
Conclusions:: In this ethnically restricted cohort, a high level of cyclin E is a characteristic of BRCA1-related breast cancer, and is a marker of poor prognosis following breast cancer, particularly in the absence of adjuvant chemotherapy.
Key words: BRCA1, breast cancer, chemotherapy, cyclin E, KIP1, prognosis
![]() |
Introduction |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Germline mutations in the tumor suppressor genes BRCA1 and BRCA2 predispose individuals to early onset breast and ovarian cancers [2]. The biological activities of the BRCA1 gene product are not completely understood; nevertheless, roles in DNA damage repair, cell-cycle control and transcriptional regulation have been identified. From a clinical standpoint, BRCA1-related tumors demonstrate distinct features in histopathological [3
], immunohistochemical (IHC) [4
], cytogenetic [5
] and gene expression profile [6
, 7
] studies, when compared with either non-familial breast cancer cases or BRCA2-related breast cancer. Other studies have suggested an inferior survival rate for women developing a BRCA1-related breast cancer [8
11
].
Cyclins bind to and activate cyclin-dependent kinases to form serine/threonine kinase complexes that regulate eukaryotic cell cycle. The cyclin E/CCNE1 gene is located on chromosome 19q. In normal dividing cells, cyclin E, in collaboration with cyclins A and D, regulates the transition through the G1 phase to enter the S phase by activating CDK2, and high level of cyclin E protein accelerates this transition [12]. In a mouse model, conditional expression of cyclin E in mammary tissue resulted in the induction of mammary gland hyperplasia and carcinomas [13
]. In several studies, high levels of cyclin E expression have been associated in multivariate analyses with a worse outcome after breast cancer [14
, 15
].
The cyclin-dependent kinase inhibitor 1B (CDKN1B) gene is located on chromosome 12p. The major function of the gene product, known as p27Kip1, is to inhibit cyclin ECDK2 complexes and thus control cell proliferation [16]. In some breast cancer survival studies, low levels of p27Kip1, as detected by IHC, have been associated with a poor prognosis in univariate and multivariate analyses [16
], but other studies did not confirm these results [17
, 18
].
Several previous studies have shown that cyclin D1 levels are lower in BRCA1-related breast cancer than in cancers occurring in either BRCA2 carriers or non-carriers [6]. Recently, it has been suggested that the perturbation of the cyclin D1 and cyclin E pathways lead to separate and alternative breast cancer subtypes [19
]. Using a subset of the cases presented here, we previously showed that BRCA1/2-related breast cancer was associated with low levels of p27Kip1 [10
]. Furthermore, using the same dataset as that presented here, we have recently shown that the basal-like phenotype of breast is characterized by high levels of expression of cyclin E, expression of p53, the presence of glomerular microvascular proliferation and low levels of p27Kip1 [20
]. The latter three factors have previously been shown by our group and others to be associated with BRCA1-related breast cancer [10
, 21
, 22
]. Here, we wished to extend our analysis of the two functionally interacting cell cycle proteins, cyclin E and p27Kip1, both of which had a strong prognostic influence in the previous analysis, and to study their relationship with BRCA1-related breast cancer. These two areas were not covered in our previous publication [20
].
We compared the clinicopathological features of BRCA1-related and BRCA1/2-unrelated breast cancer, evaluated cyclin E and p27Kip1 expression levels as predictor of the BRCA1 status, and then performed survival analyses including standard prognostic variables such as tumor size, nuclear grade and axillary node status. The main end point studied was breast cancer-specific survival (BCSS). In addition, we studied the effect of adjuvant treatment on outcome in those whose tumors featured differing levels of cyclin E/p27Kip1 proteins.
![]() |
Patients and methods |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
BRCA1 and BRCA2 mutation status
Mutation analysis was carried out as described previously [8, 10
], looking specifically for the recurrent mutations in the Ashkenazi Jewish population (BRCA1: 185delAG, 5382insC; BRCA2: 6174delT). Haplotype analysis was also used to confirm 5382insC mutation. The BRCA2 mutation 6174delT was sought using single-strand conformation analysis, by a mutation-specific PCR-RFLP endonuclease digestion analysis and by direct sequencing. We also used a size assay as a second assay for all three mutations. BRCA2 mutation carriers (n=10) were excluded from further analysis because it is not established that they can be combined with BRCA1 carriers in survival analyses, and they constituted a group too small to be analyzed alone.
IHC analyses
All interpretation of IHC assays were made without knowledge of clinical outcome or any other clinical variables. ER protein levels were detected using a standard streptavidinbiotinperoxidase complex IHC technique on paraffin-embedded tissue. Positivity was recorded when >10% of tumor cell nuclei showed immunoreactivity.
Cyclin E protein levels by IHC were determined using affinity-purified polyclonal cyclin E antibody as described previously [14]. Comparison of the polyclonal antibody with HE12 cyclin E monoclonal antibody (Oncogene Research Products, Cambridge, MA, USA) by western blot showed similar patterns of staining, including identification of multiple bands of smaller molecular weight products with both antibodies (data not shown). Scoring of cyclin E IHC was a subjective interpretation of staining intensity and the percentage of tumor cells positive, as described previously [14
]. p27Kip1 expression was also evaluated as described previously [10
]. Briefly, streptavidinbiotinperoxidase complex IHC technique on paraffin-embedded tissue with an anti-p27 monoclonal antibody (Transduction Laboratories, Lexington, KY, USA) was used. Low levels of protein were recorded when
50% of tumor cell nuclei were stained, as previously indicated [10
].
Statistical analysis
Clinical, pathological and molecular data were collected in a mutually blinded fashion. Patient characteristics were compared using non-parametric Wilcoxon's test and Fisher's exact test. Trends in increasing odds ratios (ORs) were assessed by the CochranArmitage test. Exact confidence intervals (CI) were calculated for the proportions of BRCA1/2 mutation carriers, and tumors with high levels of cyclin E and low levels of p27Kip1 present in the cohort. A logistic regression model was developed to predict BRCA1 status where the following parameters were considered: age at diagnosis, tumor size, nuclear grade, axillary lymph nodes status, ER status, and cyclin E and p27Kip1 expression levels. This model was adjusted for missing values wherever possible.
Survival rates were calculated from the date of primary surgery until death from breast cancer for BCSS, which was the main outcome measure. Ten-year survival curves were estimated using the KaplanMeier method and significance was assessed using the log-rank test. To estimate the relative risk (RR) of death from breast cancer, Cox's proportional hazards model was used where the following prognostic factors were examined: tumor size, axillary lymph node status, nuclear grade, age at diagnosis, ER status, cyclin E, p27Kip, and BRCA1 mutation status; multivariate models were adjusted for missing variables wherever possible. All data were censored at 10 years time, significance was assessed at the 5% level, 95% CI were constructed and all statistical tests were two-sided. A competing risk analysis was not performed.
![]() |
Results |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
|
|
|
|
|
When the BCSS analysis was performed according to the type of adjuvant treatment, high cyclin E protein level was an independent prognostic factor (RR 5.7; P=0.002) among 142 patients evaluable who did not receive adjuvant chemotherapy and among 128 patients with adjuvant hormonal therapy (RR 4.7; P=0.03). Cyclin E protein level was not an independent prognostic factor in other groups. In contrast, low levels of p27Kip1 were associated with a worse survival only in Cox univariate models and when the analyses were performed among the subgroup of patients (n=128) who received adjuvant chemotherapy (RR 2.9; P=0.02) and the subgroup of patients (n=130) who did not receive hormonotherapy (RR 2.2; P=0.03).
Combined effect of cyclin E and p27Kip1 protein levels
We compared BCSS among four groups of patients according to the combination of various levels of cyclin E and p27Kip1 protein level (Figure 3). An overall difference was found between the four curves (P <0.0001). Patients whose breast cancers featured low cyclin E and high p27Kip1 protein levels had the highest BCSS, particularly when compared with patients with high cyclin E and low p27Kip1 tumor protein levels (85% versus 46% respectively; P <0.0001, at 10 years follow-up).
|
![]() |
Discussion |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
A recent study showed that a high level of cyclin E as measured by western blotting was the strongest independent factor in predicting survival following breast cancer [15]. It was pointed out by several commentators that grade was not included in the final multivariable model, thus diluting the impact of their finding. The authors argued in a response letter that grade is not universally recognized as a prognostic factor [24
]. This viewpoint is somewhat debatable. In the last international consensus established by the Early Breast Cancer Trialists' Collaborative Group, both histological and nuclear grade were considered as prognostic factors and thus are used to select patients who will receive adjuvant treatment, particularly in the subgroup of node-negative patients [1
]. These recommendations were reinforced during the last meeting that took place in 2003 in St Gallen, Switzerland. Interestingly, when we excluded grade from the Cox multivariate analysis, high cyclin E expression levels (RR 1.9; 95% CI 13.3; P=0.04), but not low p27Kip1 levels, was an independent prognostic factor in the entire cohort.
When we evaluated BCSS according to the type of adjuvant treatment, high levels of cyclin E protein expression was the strongest independent prognostic factor (RR 5.7; P=0.002) among the group of patients who did not receive adjuvant chemotherapy. When the analysis was adjusted for the type of chemotherapy received (anthracyclin-based versus other regimens), high cyclin E levels remained a prognostic factor, but of borderline significance in the multivariate analysis (RR 1.8; P=0.07). As expected, high cyclin E expression was also an independent prognostic factor (RR 4.7; P=0.03) among patients who received adjuvant hormonotherapy, reflecting the fact that most of these patients did not receive chemotherapy. This result suggests that chemotherapy may ameliorate the poor prognosis conferred by high levels of cyclin E protein. The role of differential protein levels of cyclin E and p27Kip1 as potential predictive markers of response to chemo- or hormonotherapy is unclear [18, 25
27
]. Available data are based on retrospective studies that employed non-standardized IHC methods, various adjuvant treatments and different stages of the disease, precluding any definitive conclusion. Nevertheless, this important issue deserves further well-designed prospective studies.
Simultaneous analysis of the expression of thousand of genes by DNA microarrays techniques have opened a new area in the search of markers that can predict the outcome of an individual patient. A recent study using this new high-throughput molecular technique has defined gene expression profiles associated with either a poor prognosis or a good prognosis after breast cancer [7]. In an extended study, the same research group recently showed that among patients associated with a genetically defined poor prognosis, 50% develop a distant progression of the disease and 45% died within 10 years. In the good prognosis group, 85% of patients stayed disease-free and 95% survived [28
]. The genetic profile was the strongest independent prognostic indicator in this series. These molecular signatures consisted of a selection of a subset of 70 genes the expression of which defined the outcome. Notably, cyclin E2, a recently discovered protein with activity similar to cyclin E [29
], was one of the genes identified in the 70-gene profile. In the van de Vijver et al. [28
] study, DDFS of 151 patients with lymph node-negative disease was evaluated according to the prognosis signature. At 10 years, distant disease free survival (DDFS) for 60 patients (40% of node-negative patients) with tumor harboring a good prognosis profile was 87%, and 44% for 91 patients (60%) with a poor prognosis profile. These data can be compared with the survival results in association with cyclin E protein levels in our cohort. The DDFS was 85.6% at 10 years follow-up for 90 (70%) node-negative patients with low levels of cyclin E expression compared with 59.4% for 38 (30%) patients with high levels of cyclin E expression (P <0.001). Therefore, in our study, low levels of cyclin E staining identified 70% of all node-negative patients, amongst whom the DDFS was >85% at 10 years.
This study has a robust design in that, our historical cohort approach, results in an unbiased ascertainment. The restriction of our study population to the Ashkenazim limits the genetic heterogeneity that could obscure associations between putative disease markers and germline mutations. In particular, we are confident that the restriction of the BRCA1/BRCA2 mutation analysis to the three common Ashkenazi Jewish founder mutations has not resulted in many women being misclassified as non-carriers [30]. The number of individuals with BRCA1 mutations is small, so larger, confirmatory studies are required. However, the overall cohort is now fairly mature with a median follow-up of survivors of nearly 10 years; one of the longest periods of follow-up for a study evaluating cyclin E as a prognostic factor in breast cancer. The studied population is ethnically restricted. As such, the expansion of these results to other populations should be evaluated.
This study has confirmed that cyclin E is an important prognostic marker, particularly when evaluated along with p27Kip1 levels. It has been shown that overexpression of cyclin E can promote chromosomal instability [31]. The observations that: (i) BRCA1-related breast cancers tend to be highly aneuploid [5
]; (ii) these cancers commonly overexpress cyclin E and underexpress p27Kip1; (iii) cyclin E overexpression and p27Kip1 underexpression are both associated with a poor outcome following breast cancer; and (iv) cyclin E overexpression mimicks the effect of germline BRCA1 mutations on response to adjuvant chemotherapy [23
] suggests strongly that, directly or indirectly, cyclin E overexpression is a major determinant of the BRCA1 breast cancer phenotype. While the data remain preliminary, it will be important to determine, in a prospective study, whether adjuvant chemotherapy can significantly ameliorate the poor survival associated with high protein levels of cyclin E.
![]() |
Acknowledgements |
---|
![]() |
Notes |
---|
Present address: Division of Hematology/Oncology, Department of Medicine, Tufts University, TuftsNew England Medical Center, Boston, MA 02115, USA
# Present address: Hôpital du Sacré-Cur de Montréal, Montreal, QC H4J 1C5, Canada
¶ Present address: Credit Valley Hospital, Mississauga, ON L5M 2N1, Canada
Received for publication September 8, 2004. Revision received December 17, 2004. Accepted for publication December 22, 2004.
![]() |
References |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
2. Narod SA, Foulkes WD. BRCA1 and BRCA2: 1994 and beyond. Nat Rev Cancer 2004; 4: 665676.[CrossRef][ISI][Medline]
3. Breast Cancer Linkage Consortium. Pathology of familial breast cancer: differences between breast cancers in carriers of BRCA1 or BRCA2 mutations and sporadic cases. Breast Cancer Linkage Consortium. Lancet 1997; 349: 15051510.[CrossRef][ISI][Medline]
4. Lakhani SR, Van De Vijver MJ, Jacquemier J et al. The pathology of familial breast cancer: predictive value of immunohistochemical markers estrogen receptor, progesterone receptor, HER-2, and p53 in patients with mutations in BRCA1 and BRCA2. J Clin Oncol 2002; 20: 23102318.
5. Tirkkonen M, Johannsson O, Agnarsson BA et al. Distinct somatic genetic changes associated with tumor progression in carriers of BRCA1 and BRCA2 germ-line mutations. Cancer Res 1997; 57: 12221227.[Abstract]
6. Hedenfalk I, Duggan D, Chen Y et al. Gene-expression profiles in hereditary breast cancer. N Engl J Med 2001; 344: 539548.
7. van't Veer LJ, Dai H, van de Vijver MJ et al. Gene expression profiling predicts clinical outcome of breast cancer. Nature 2002; 415: 530536.[CrossRef][ISI][Medline]
8. Foulkes WD, Wong N, Brunet JS et al. Germ-line BRCA1 mutation is an adverse prognostic factor in Ashkenazi Jewish women with breast cancer. Clin Cancer Res 1997; 3: 24652469.[Abstract]
9. Ansquer Y, Gautier C, Fourquet A et al. Survival in early-onset BRCA1 breast-cancer patients. Institut Curie Breast Cancer Group. Lancet 1998; 352: 541.[CrossRef]
10. Chappuis PO, Kapusta L, Begin LR et al. Germline BRCA1/2 mutations and p27(Kip1) protein levels independently predict outcome after breast cancer. J Clin Oncol 2000; 18: 40454052.
11. Stoppa-Lyonnet D, Ansquer Y, Dreyfus H et al. Familial invasive breast cancers: worse outcome related to BRCA1 mutations. J Clin Oncol 2000; 18: 40534059.
12. Moroy T, Geisen C. Cyclin E. Int J Biochem Cell Biol 2004; 36: 14241439.[CrossRef][ISI][Medline]
13. Bortner DM, Rosenberg MP. Induction of mammary gland hyperplasia and carcinomas in transgenic mice expressing human cyclin E. Mol Cell Biol 1997; 17: 453459.[Abstract]
14. Porter PL, Malone KE, Heagerty PJ et al. Expression of cell-cycle regulators p27Kip1 and cyclin E, alone and in combination, correlate with survival in young breast cancer patients. Nat Med 1997; 3: 222225.[CrossRef][ISI][Medline]
15. Keyomarsi K, Tucker SL, Buchholz TA et al. Cyclin E and survival in patients with breast cancer. N Engl J Med 2002; 347: 15661575.
16. Alkarain A, Jordan R, Slingerland J. p27 deregulation in breast cancer: prognostic significance and implications for therapy. J Mammary Gland Biol Neoplasia 2004; 9: 6780.[CrossRef][ISI][Medline]
17. Leivonen M, Nordling S, Lundin J et al. p27 expression correlates with short-term, but not with long-term prognosis in breast cancer. Breast Cancer Res Treat 2001; 67: 1522.[CrossRef][ISI][Medline]
18. Spataro VJ, Litman H, Viale G et al. Decreased immunoreactivity for p27 protein in patients with early-stage breast carcinoma is correlated with HER-2/neu overexpression and with benefit from one course of perioperative chemotherapy in patients with negative lymph node status: results from International Breast Cancer Study Group Trial V. Cancer 2003; 97: 15911600.[CrossRef][ISI][Medline]
19. Loden M, Stighall M, Nielsen NH et al. The cyclin D1 high and cyclin E high subgroups of breast cancer: separate pathways in tumorogenesis based on pattern of genetic aberrations and inactivation of the pRb node. Oncogene 2002; 21: 46804690.[CrossRef][ISI][Medline]
20. Foulkes WD, Brunet JS, Stefansson IM et al. The prognostic implication of the basal-like (cyclin E high/p27 low/p53+/glomeruloid-microvascular-proliferation+) phenotype of BRCA1-related breast cancer. Cancer Res 2004; 64: 830835.
21. Crook T, Crossland S, Crompton MR et al. p53 mutations in BRCA1-associated familial breast cancer. Lancet 1997; 350: 638639.[ISI][Medline]
22. Goffin JR, Straume O, Chappuis PO et al. Glomeruloid microvascular proliferation is associated with p53 expression, germline BRCA1 mutations and an adverse outcome following breast cancer. Br J Cancer 2003; 89: 10311034.[CrossRef][ISI][Medline]
23. Goffin JR, Chappuis PO, Begin LR et al. Impact of germline BRCA1 mutations and overexpression of p53 on prognosis and response to treatment following breast carcinoma: 10-year follow up data. Cancer 2003; 97: 527536.[CrossRef][ISI][Medline]
24. Keyomarsi K, Tucker SL, Bedrosian I. Cyclin E is a more powerful predictor of breast cancer outcome than proliferation. Nat Med 2003; 9: 152.[CrossRef][ISI][Medline]
25. St Croix B, Florenes VA, Rak JW et al. Impact of the cyclin-dependent kinase inhibitor p27Kip1 on resistance of tumor cells to anticancer agents. Nat Med 1996; 2: 12041210.[CrossRef][ISI][Medline]
26. Han S, Park K, Kim HY et al. Reduced expression of p27Kip1 protein is associated with poor clinical outcome of breast cancer patients treated with systemic chemotherapy and is linked to cell proliferation and differentiation. Breast Cancer Res Treat 1999; 55: 161167.[CrossRef][ISI][Medline]
27. Akli S, Zheng PJ, Multani AS et al. Tumor-specific low molecular weight forms of cyclin E induce genomic instability and resistance to p21, p27, and antiestrogens in breast cancer. Cancer Res 2004; 64: 31983208.
28. van de Vijver MJ, He YD, van't Veer LJ et al. A gene-expression signature as a predictor of survival in breast cancer. N Engl J Med 2002; 347: 19992009.
29. Gudas JM, Payton M, Thukral S et al. Cyclin E2, a novel G1 cyclin that binds Cdk2 and is aberrantly expressed in human cancers. Mol Cell Biol 1999; 19: 612622.
30. Phelan CM, Kwan E, Jack E et al. A low frequency of non-founder BRCA1 mutations in Ashkenazi Jewish breast-ovarian cancer families. Hum Mutat 2002; 20: 352357.[CrossRef][ISI][Medline]
31. Spruck CH, Won KA, Reed SI. Deregulated cyclin E induces chromosome instability. Nature 1999; 401: 297300.[CrossRef][ISI][Medline]