| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Molecular Oncology, Markers, Clinical Correlates |
1 Program in Cancer Genetics, Departments of Oncology and Human Genetics, and Cancer Prevention Centre, Sir M. B. Davis-Jewish General Hospital, McGill University, Montreal, Quebec, Canada; 2 Faculty of Nursing, University of Toronto, Toronto, Ontario, Canada; 3 Centre for Research on Women’s Health and Sunnybrook and Women’s College Health Sciences Centre, University of Toronto, Toronto, Ontario, Canada; 4 Department of Preventive Medicine and Public Health, Creighton University School of Medicine, Omaha, Nebraska; 5 Epidemiology Research Unit, Centre de Recherché du CHUM, Pavillon Masson de l’Hôtel-Dieu, Montreal, Quebec, Canada; 6 Beth Israel Deaconess Hospital, Boston, Massachusetts; 7 Department of Medicine, University of Chicago, Illinois; 8 Departments of Medicine and Genetics and Hematology/Oncology, University of Pennsylvania, Philadelphia, Pennsylvania; 9 Cancer Risk Program, University of California-San Francisco Comprehensive Cancer Center, San Francisco, California; 10 British Columbia Cancer Agency, Victoria, British Columbia, Canada; and 11 Department of Pathology, Sacré Coeur Hospital, Montreal, Quebec, Canada
 |
ABSTRACT |
|---|
 Top ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Experimental Design: ER status, grade, and histological tumor type were evaluated in 1131 women with invasive breast cancer, ascertained at 10 centers in North America. There were 208 BRCA1 mutation carriers, 88 BRCA2 carriers, and 804 women without a known mutation. We stratified the patients by mutation status, grade, age, and histological type and calculated the percentage of ER-positive tumors within each stratum.
Results: BRCA1 mutation carriers were more likely to have ER-negative breast cancers than were women in other groups, after adjustment for age, grade, and histological subtype (P < 0.001). Only 3.9% of BRCA1-related breast cancers were ER-positive cancers occurring in women in their postmenopausal years. The direction and magnitude of the change in ER status with increasing age at diagnosis in BRCA1 carriers was significantly different from in BRCA2 carriers (Pintercept = 0.0002, Pslope = 0.04). Notably, changes in ER status with age at diagnosis for BRCA1 carriers and noncarriers were almost identical (Pslope = 0.98).
Conclusions: The strong relationship between the presence of a BRCA1 mutation and the ER-negative status of the breast cancers is neither a consequence of the young age at onset nor the high grade but is an intrinsic property of BRCA1-related cancers. The ER-negative status of these cancers may reflect the cell of origin of BRCA1-related cancers.
|
INTRODUCTION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
We have recently shown that tamoxifen is effective at reducing both local recurrence and contralateral breast cancer among women with BRCA1 mutations (3) . However, in that study, much of the ER data were missing. One possible way to circumvent the missing ER data are to identify other features of the breast cancer that predict ER status and then use these tumor characteristics as surrogate markers of ER status. In a small series of cases from Montreal, Quebec, Canada, we observed that 16 of 18 (89%) grade 3, infiltrating ductal breast cancers, which occurred among BRCA1 carriers, were ER negative. Here, we attempted to confirm this initial observation in an expanded series of cases. We also wished to establish whether BRCA1-related breast cancer is more likely to be ER negative than non-BRCA1-related breast cancer, even after adjusting for the effects of age, grade, and histological type.
|
MATERIALS AND METHODS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
10% of the total number of tumor cells examined showed clear and intense staining with the anti-ER antibody used, with no background cytoplasmic staining, as described previously (1)
. The ascertainment criteria and mutation analysis performed for these women have been described elsewhere (10
, 11)
. Tumors 51 mm were excluded, so as to match the second series (see below). Among the 246 eligible women, 30 were BRCA1 carriers, 7 were BRCA2 carriers, and 239 were noncarriers.
The second series of women consists of a historical cohort of BRCA1 and BRCA2 carriers affected with breast cancer. Four hundred ninety-two BRCA1/2 carriers (328 BRCA1, 152 BRCA2, and 12 BRCA1+BRCA2) women were ascertained as part of a North American study of the effects of various treatments of hereditary breast cancer. There were 10 participating centers. The inclusion criteria were as follows: presence of a germ-line BRCA1 or BRCA2 mutation; invasive breast cancer; diagnosed in 1975 or thereafter; at age 
65 years; and residents of North America at the time of diagnosis. Exclusion criteria were as follows: women with a history of breast cancer or any other cancer; before 1975; carcinoma in situ with no invasive component; tumor size > 50 mm; and evidence for locoregional or distant metastases at time of diagnosis and those with fixed mass of axillary nodes. Histological type, histological grade, and ER status were extracted from the patient charts by local collaborators. ER status was measured both biochemically and by using immunohistochemistry, but we do not have details of individuals scores because the entry sheet used allowed for three responses: positive; negative; and unknown. There were 178 BRCA1 carriers and 81 BRCA2 carriers with complete information, and these were included in the present study. In total, there were 208 BRCA1 carriers and 88 BRCA2 carriers (both study sets combined).
The third study group consisted of consecutive cases of women diagnosed with invasive breast cancer at <65 years of age at Women’s College Hospital, Toronto, Ontario, Canada, from 1987 to 1997, for whom histological subtype, ER status, and histological grade were available. In all cases, ER was measured biochemically in all cases at the time of diagnosis, and a level of >10 fmol/mg protein was recorded as positive for ER. The range of values observed was 0–776 fmol/mg protein. No immunohistochemical results were used in the analyses reported here. Patients with primary tumors 51 mm and larger in size were excluded, so as to match set 2 (BRCA1/2 carriers). The 596 women in this data set were not tested for BRCA1 or BRCA2 mutations. This set is referred to as the untested set in the tables.
Statistical Analysis.
The percentages of ER-positive tumors were calculated by histological grade (Table 1)
, by mutation status (Table 2)
, and by age at diagnosis (Table 3
). The significance of differences was tested using the t test. The trend in percent ER positivity as histological grade increased was tested by the Cochran-Armitage test. A multiple regression model (12)
was created:
 |
|
|
|
. As a result of regression of model (1)
, the four regression equations of the four groups were as follows: Y = 
1 + ß1ageca (for group 1), and Y = 1 + j + (ß1+ ßj)ageca, for j = 2, 3, 4 (for groups 2, 3, 4), respectively.
|
.
|
+ ß1 BR+ ß2 ageca, where BR is Bloom and Richardson histological grade. All statistical analyses were done using SAS, version 8.2. |
RESULTS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. This is observed in patients with and without mutations in BRCA1/2 (Table 1)
. However, for tumors of grade 2 or 3, BRCA1-related breast cancers are significantly more likely to be ER negative than are BRCA2-related noncarriers or untested individuals (all P < 0.001). There are too few grade 1 BRCA1-related cancers to observe an effect, but the trend is in the same direction. The proportion of cancers that are ER positive decreases dramatically from grade 1 to grade 3 in noncarriers: (grade 1, 90.1% ER positive and grade 3, 45.7% ER positive; P < 0.0001) and in BRCA1 carriers (71.4 and 15.7%, respectively; P = 0.0024). In contrast to these groups, the difference in ER positivity between grade 1 and grade 3 tumors is not significant for BRCA2-related tumors (80.0 versus 61.0%, P = 0.64), although the sample size is small (Table 1)
. There were 209 nonductal cancers (18.5% of the total). The same relationship between grade, age, and ER status is observed in nonductal cancers as in ductal cancers, but because of small numbers, statistical significance is not reached in all comparisons for nonductal cancers (data not shown).
To study the effect of mutation status on ER status independent of grade and histological type, we stratified all ductal grade 3 cancers by mutation status (Table 2)
. BRCA1-related breast cancers are significantly more likely to be ER negative than tumors in BRCA2 carriers, noncarriers, or untested individuals (all P < 0.0001). Among grade 3 ductal cancers, BRCA1 tumors are 4.8 times (95% CI 2.4–9.36) more likely to be ER negative than were tumors in noncarriers. Among BRCA1 carriers who were <45 years when they were diagnosed with a grade 3 ductal breast cancer, 87.1% (88 of 101) were ER negative. This fell to 70% (7 of 10) in women 55–65 years old at diagnosis (P = 0.44). Interestingly, BRCA2 carriers with grade 2 or 3 cancers are more likely to have an ER-positive cancer than are noncarriers (80.8 versus 60.9%, P = 0.0015; Tables 1
and 2
).
For each age group (<45, 45–54, and 55–64 years), BRCA1-related breast cancers are more likely to be ER negative than cancers occurring in any other group (Table 3)
. For all groups other than BRCA2 carriers, a similar increase in the percentage of tumors that were ER positive with increasing age was observed. For BRCA2-related breast cancers, there is a much smaller change in ER status with increasing age at diagnosis (3.7%, P = 0.78; Table 3
) or grade (19%, P = 0.64; Table 1
) compared with the other subtypes studied.
The relationship between age at diagnosis and percentage of ER-positive cancers was studied further by a regression analysis (model 1). The percentage of ER-positive cancers (y axis) was plotted against age at diagnosis (x axis; Fig. 1
), with the parameters. Notwithstanding the different starting points of the slopes (noncarrier versus BRCA1, P = 0.0.13; noncarriers versus untested, P = 0.25), BRCA1 carriers, noncarriers, and untested women have slopes that are almost parallel from ages 24 to 64.9 years (noncarrier versus BRCA1, P = 0.98; noncarriers versus untested, P = 0.40; Table 4
). From Fig. 1
, it can be deduced that a 60-year-old BRCA1 carrier has the same probability of an ER-negative breast cancer as a 25-year-old noncarrier. At the youngest ages at diagnosis, the proportion of cancers that were ER positive was 0.42 (noncarriers, youngest age at diagnosis: 26.5 years), 0.06 (BRCA1 carriers, 23.9 years), 0.85 (BRCA2 carriers, 25.7 years), and 0.54 (untested individuals, 24.6 years). Despite these differences, the difference in the proportion of cancers that are ER positive between the youngest and oldest ages is almost identical in BRCA1 carriers (0.40) and noncarriers (0.41, P = 0.98; Fig. 1
). The relationship between age at diagnosis and ER status in BRCA2 carriers is significantly different from that of the noncarrier group in both intercept (P = 0.01) and slope (P = 0.034) and that from BRCA1 carriers (P = 0.0002 for intercept and P = 0.041 for slope; Table 4
). For the untested group (n = 596), we also plotted absolute level of ER in fmol/mg of protein against the age at diagnosis. This resulted in a slope with an R2 of 0.06 (data not shown).
|
in keeping with the results shown in Tables 1
and 3
. The intercept and slope for grade 3 cancers are statistically significantly different from that seen in grade 1 cancers. When comparing the intercepts and slopes between grade 2 and grade 3 cancers, the intercepts are very different, but the slopes (i.e., the change in ER status with increasing age at diagnosis) are similar (P for differences between intercepts = 0.002, P for slope = 0.34; Fig. 2
). Table 5
shows in more detail the parameters estimated using model (1)
. The intercept () and slope (ß) for grade 3 cancers are both significantly different from the grade 1 values, whereas the parameters for grade 2 cancers are not different from grade 1 cancers. This indicates that the relationship between the ER status of a breast cancer and the age at diagnosis is dependent on the grade of the cancer, insofar as the relationship is not uniform across all grades. As the age at breast cancer diagnosis increases, the difference between the three groups becomes less noticeable. This is in contrast to the results shown in Figure 1
. Grade 1 cancers are likely to be ER positive, no matter the age at diagnosis; this is not the case for grade 3 cancers. Notably, a woman who develops a grade 3 breast cancer at 40 years of age is five times more likely to have an ER-negative cancer than a woman diagnosed with a grade 1 cancer at the same age (ER percent negative = 60 versus 12%).
|
= 0.74, P < 0.0001; ß1 = -2.7, P < 0.0001 for grade; and ß2 = 0.009, P < 0.0001 for age), but histological grade is the most important variable in predicting ER status. |
DISCUSSION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
). Our findings suggest that ER-negative status is an intrinsic feature of BRCA1-related breast cancer. As suggested previously, it is possible that most, if not all ER-positive breast cancers occurring in BRCA1 carriers are, in fact, sporadic cancers. We could not address this question in this study, but BRCA1 sequencing or loss of heterozygosity analysis to compare the frequency of complete inactivation of BRCA1 in ER-negative and ER-positive breast cancers could help to resolve this issue. Even if this hypothesis were supported, the magnitude of the reduction in risk for contralateral breast cancer in BRCA1 carriers who received tamoxifen treatment appears to be too large to be attributable to the successful treatment of sporadic, BRCA1-unrelated breast cancers (3)
.
Although this hypothesis remains a possibility, it is notable that the slope of the change in ER positivity with increasing age is almost identical for BRCA1 carriers, noncarriers, and untested individuals (Fig. 1)
. This similarity in the slopes suggests that BRCA1 carriers are likely to be susceptible to the same hormonal and/or environmental factors that result in the change in ER positivity with age in the general population. The hormonal milieu of the woman at risk of breast cancer due to a BRCA1 mutation probably influences the ER status of the cancer, but the presence of the mutation sets the limits of the response in terms of the likely ER status of the breast cancer. BRCA1 carriers in the oldest age group were 2.6 times as likely as those in the youngest age group to have an ER-positive cancer (P = 0.08), but this phenotype was less than half as likely to occur in older BRCA1 carriers, when compared with older noncarriers (38.0 versus 78.7%, P = 0.0004; Table 3
). The lack of change with age in the slope for BRCA2 carriers (Fig. 1)
suggests that the ER status of breast cancer occurring in BRCA2 carriers is not under the same hormonal and environmental control as is nonhereditary breast cancer.
Among BRCA1 carriers, grade 3, ductal breast cancers are highly likely to be ER negative: only 21 of 138 of this type were ER positive (15.2%; Table 2
). This suggests that in any study of breast cancer outcomes after adjuvant hormonal treatment, it is reasonable to assume that the great majority of BRCA1-related, grade 3 invasive ductal cancers are ER negative. This finding is relevant to the debate as to the potential benefit of tamoxifen for the primary prevention of BRCA1-related breast cancer. The data presented here do not suggest that ER-positive cancers predominate in any subgroup of BRCA1-related breast cancers, whether defined by age at diagnosis, grade, or histological type. Recent case-control (7)
and retrospective (3)
studies of contralateral breast cancer in BRCA1 and BRCA2 carriers have shown that tamoxifen is capable of reducing the incidence of second primary breast cancers by 
50%. The reduction was equally large in women whose first cancer was a high-grade ductal cancer as it was for other groups. Moreover, as stated above, our results suggest that the percentage of all BRCA1 carriers who develop an ER-positive breast cancer in the postmenopausal years is too small (8 of 208, or 3.9%) to account for the observed risk reductions, even if ER-positive breast cancers in BRCA1 carriers are actually sporadic breast cancers.
There are two main weaknesses in our study that may limit the interpretation of our results. Because of the diverse nature of the three study groups and the partly retrospective collection of data, we were unable to arrange central pathology review. Similarly, we were unable to determine ER status using one standardized method. Both these weaknesses mean that some of our analyses may be subject to error. However, should grade or ER measurement errors exist, they are likely to bias the results toward the null. It would be more troubling if our lack of standardization were to lead to differential misclassification. However, in group 1, all tumors were analyzed by a single pathologist without knowledge of the BRCA1/2 status, and given the era, very few of the pathologists could have been aware of the BRCA1/2-status in group 2. It is therefore implausible that there the lack of standardization has introduced a systematic bias in the measuring or reporting of either grade or ER status. It has been suggested that tumor-associated infiltrates, which are common in BRCA1-related breast cancers, may express ER (21) , and ligand-binding assays could therefore overestimate tumor ER levels. If this observation were to contribute significantly to our findings, it would mean that tumor cells in BRCA1-related cancers actually contain less ER than reported here. In this case, BRCA1-related breast cancers are even more likely to be ER-negative than we observed. Finally, we do not know the BRCA1/2 status of the untested cases (group 3). It is difficult to be sure of the expected numbers of BRCA1/2 carriers, but it is likely to be substantially <5% in most outbred populations (22) , and therefore, these hidden carriers are likely to contribute very little to the overall results obtained from the untested group.
The observation here and elsewhere (23) that BRCA1-related breast cancer is likely to be ER negative, regardless of the grade or age at diagnosis, suggests that BRCA1-related breast cancers may arise from an ER-negative breast cell. This conjecture is supported by the observation that ductal carcinoma in situ occurring in BRCA1 carriers is usually ER negative (24) . Some studies suggest that the breast stem cell is an ER-negative cell but is surrounded by ER-positive cells (25 , 26) . If the regulation of ER-negative breast stem cells is partly controlled by paracrine factors released by the adjacent ER-positive cells (26) , then inhibiting estrogenic stimulation of these supporting cells either directly by estrogen deprivation, or indirectly by use of tamoxifen, could play a role in reducing the incidence of breast cancers in BRCA1 carriers.
Our observations led us to conclude that the beneficial effect of tamoxifen on both local recurrence and contralateral cancer rates in BRCA1 carriers is due to the prevention of breast cancers that arose from breast cells that either were always ER negative or those that became ER negative very soon after an initiating event. Research focused on the mechanisms by which tamoxifen reduces risk of breast cancer in BRCA1 carriers is to be encouraged.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
Note: O. Olopade is supported by the Falk Medical Research Trust and is a Doris Duke Distinguished Clinical Scientist.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Requests for reprints: William D. Foulkes, Division of Medical Genetics, Department of Medicine, Room L10–120, Montreal General Hospital, 1650 Cedar Avenue, Montreal, Quebec, H3G 1A4 Canada. Phone: (514) 934-1934, ext. 44121; Fax: (514) 934-8273; E-mail: william.foulkes{at}mcgill.ca
Received 7/17/03; revised 12/23/03; accepted 12/29/03.
|
REFERENCES |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
is a novel marker expressed by follicular dendritic cells in lymph nodes and tumor-associated lymphoid infiltrates. Am J Pathol, 163: 1313-20, 2003.This article has been cited by other articles:
|
|
 |
|
  S. Beetstra, G. Suthers, V. Dhillon, C. Salisbury, J. Turner, M. Altree, R. McKinnon, and M. Fenech Methionine-Dependence Phenotype in the de novo Pathway in BRCA1 and BRCA2 Mutation Carriers with and without Breast Cancer Cancer Epidemiol. Biomarkers Prev., October 1, 2008; 17(10): 2565 - 2571. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. T. Chlebowski and R. L. Prentice Menopausal Hormone Therapy in BRCA1 Mutation Carriers: Uncertainty and Caution J Natl Cancer Inst, October 1, 2008; 100(19): 1341 - 1343. [Full Text] [PDF]
|
|
|
|
 |
|
 D S P Tan, C Marchio, and J S Reis-Filho Hereditary breast cancer: from molecular pathology to tailored therapies J. Clin. Pathol., October 1, 2008; 61(10): 1073 - 1082. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. D. Kauff, S. M. Domchek, T. M. Friebel, M. E. Robson, J. Lee, J. E. Garber, C. Isaacs, D. G. Evans, H. Lynch, R. A. Eeles, et al. Risk-Reducing Salpingo-Oophorectomy for the Prevention of BRCA1- and BRCA2-Associated Breast and Gynecologic Cancer: A Multicenter, Prospective Study J. Clin. Oncol., March 10, 2008; 26(8): 1331 - 1337. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. Britt, A. Ashworth, and M. Smalley Pregnancy and the risk of breast cancer Endocr. Relat. Cancer, December 1, 2007; 14(4): 907 - 933. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. M. Hosey, J. J. Gorski, M. M. Murray, J. E. Quinn, W. Y. Chung, G. E. Stewart, C. R. James, S. M. Farragher, J. M. Mulligan, A. N. Scott, et al. Molecular Basis for Estrogen Receptor {alpha} Deficiency in BRCA1-Linked Breast Cancer J Natl Cancer Inst, November 21, 2007; 99(22): 1683 - 1694. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
N. D. Kauff and R. R. Barakat Risk-Reducing Salpingo-Oophorectomy in Patients With Germline Mutations in BRCA1 or BRCA2 J. Clin. Oncol., July 10, 2007; 25(20): 2921 - 2927. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. D. Kauff and K. Offit Modeling Genetic Risk of Breast Cancer JAMA, June 20, 2007; 297(23): 2637 - 2639. [Full Text] [PDF]
|
|
|
|
 |
|
 S. Colilla, P. W. Kantoff, S. L. Neuhausen, A. K. Godwin, M. B. Daly, S. A. Narod, J. E. Garber, H. T. Lynch, M. Brown, B. L. Weber, et al. The joint effect of smoking and AIB1 on breast cancer risk in BRCA1 mutation carriers Carcinogenesis, March 1, 2006; 27(3): 599 - 605. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. Dai, L. van't Veer, J. Lamb, Y. D. He, M. Mao, B. M. Fine, R. Bernards, M. van de Vijver, P. Deutsch, A. Sachs, et al. A Cell Proliferation Signature Is a Marker of Extremely Poor Outcome in a Subpopulation of Breast Cancer Patients Cancer Res., May 15, 2005; 65(10): 4059 - 4066. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A.-M. Simeone, C.-X. Deng, G. J. Kelloff, V. E. Steele, M. M. Johnson, and A. M. Tari N-(4-Hydroxyphenyl)retinamide is more potent than other phenylretinamides in inhibiting the growth of BRCA1-mutated breast cancer cells Carcinogenesis, May 1, 2005; 26(5): 1000 - 1007. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
W. D. Foulkes Re: Estrogen Receptor Status of Primary Breast Cancer Is Predictive of Estrogen Receptor Status of Contralateral Breast Cancer J Natl Cancer Inst, July 7, 2004; 96(13): 1040 - 1041. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Cancer Research | Clinical Cancer Research |
| Cancer Epidemiology Biomarkers & Prevention | Molecular Cancer Therapeutics |
| Molecular Cancer Research | Cancer Prevention Research |
| Cancer Prevention Journals Portal | Cancer Reviews Online |
| Annual Meeting Education Book | Meeting Abstracts Online |
