|Home | About | Journals | Submit | Contact Us | Français|
To fully characterize the risk of contralateral breast cancer (CBC) in patients with breast cancer with a family history who test negative for BRCA1 and BRCA2 mutations.
From our population-based case-control study comparing women with CBC to women with unilateral breast cancer (UBC), we selected women who tested negative for BRCA1 and BRCA2 mutations (594 patients with CBC/1,119 control patients with UBC). Rate ratios (RRs) and 95% CIs were estimated to examine the association between family history of breast cancer and risk of asynchronous CBC. Age- and family history–specific 10-year cumulative absolute risks of CBC were estimated.
Family history of breast cancer was associated with increased CBC risk; risk was highest among young women (< 45 years) with first-degree relatives affected at young ages (< 45 years; RR, 2.5; 95% CI, 1.1 to 5.3) or women with first-degree relatives with bilateral disease (RR, 3.6; 95% CI, 2.0 to 6.4). Women diagnosed with UBC before age 55 years with a first-degree family history of CBC had a 10-year risk of CBC of 15.6%.
Young women with breast cancer who have a family history of breast cancer and who test negative for deleterious mutations in BRCA1 and BRCA2 are at significantly greater risk of CBC than other breast cancer survivors. This risk varies with diagnosis age, family history of CBC, and degree of relationship to an affected relative. Women with a first-degree family history of bilateral disease have risks of CBC similar to mutation carriers. This has important implications for the clinical management of patients with breast cancer with family history of the disease.
A family history of breast cancer is a well-established and significant risk factor for breast cancer.1–3 Relative risks vary depending on the age at diagnosis of the affected relative(s) and the number of affected relatives, ranging from two-fold for one affected first-degree relative to three-fold for two affected first-degree relatives and four-fold for three affected first-degree relatives.1–3 Young ages at diagnosis and bilaterality in first-degree relatives further increase risk.2
After diagnosis of a first primary breast cancer, women with an intact contralateral breast are at risk of developing contralateral breast cancer (CBC). The majority of studies,4–15 but not all,16,17 have identified family history as a risk factor for CBC. From the few studies presenting risk by detailed family history, there is evidence that a family history of bilateral breast cancer, multiple affected relatives, and early-onset disease significantly increase risk of CBC.5,13,18
We19 and others20–22 have shown that CBC risk is greatly increased in women carrying deleterious mutations in BRCA1 or BRCA2 and that the magnitude of this association is modified by age at diagnosis of the first breast cancer. Graeser et al20 found that women who were carrying a deleterious mutation in BRCA1 or BRCA2 and who were older than 50 years at first breast cancer diagnosis had a 10-year CBC risk of 8.4%, whereas carriers younger than 40 years had a 10-year CBC risk of 28.3%. In the Women's Environmental Cancer and Radiation Epidemiology (WECARE) Study, we previously reported that the youngest women (25 to 29 years) who were carrying a deleterious BRCA1 or BRCA2 mutation had a 10-year CBC risk of 28.2% and those older than 50 years had an analogous risk of 10.8%.19 No studies to date have examined the predictive contribution of various components of family history in relation to the risk of CBC for women in whom no mutation in BRCA1 or BRCA2 has been identified. Our goal in this study is to more clearly define risk prediction for women with a first primary breast cancer who test negatively for deleterious BRCA1 and BRCA2 mutations.
The WECARE Study is a population-based case-control study comparing women with CBC (cases) to controls with unilateral breast cancer (UBC).23 In the current analyses, we examined the relationship between family history of breast cancer and the risk of developing asynchronous CBC in women testing negatively for deleterious mutations in BRCA1 and BRCA2. Age- and family history–specific 10-year cumulative absolute risks of CBC were also estimated.
The WECARE Study is a population-based case-control study that identified patients with CBC and UBC through four cancer registries in the United States (Los Angeles County Cancer Surveillance Program, the Cancer Surveillance System of the Fred Hutchinson Cancer Research Center in Seattle, the State Health Registry of Iowa, and the Cancer Surveillance Program of Orange Country/San Diego-Imperial Organization for Cancer Control) and through the Danish Breast Cancer Cooperative Group Registry, supplemented by data from the Danish Cancer Registry. The study has been described in detail previously.23 Women with CBC (n = 708 patient cases) and women with UBC (n = 1,399 controls) completed the study interview and provided a blood sample. Two controls were individually matched to each patient case on year of birth (5-year strata), year of diagnosis (4-year strata), cancer registry, and race. Additionally, to improve statistical efficiency for the original study aims, case-control triplets were counter-matched on cancer registry–reported radiation exposure such that two members of each triplet received radiation therapy and one member did not.23 All participants were diagnosed with their first primary breast cancer between January 1985 and December 2000 at age 54 years or younger. Eligible CBC patient cases had at least 1 year between their two breast cancer diagnoses, resided in the same reporting area for both diagnoses, and had not been diagnosed with any prior or intervening cancers (except for nonmelanoma skin cancer or in situ cervical cancer). Eligible controls had no subsequent cancer diagnoses and had no prophylactic mastectomy of the contralateral breast. All eligible participants were alive at the time of contact and able to provide written informed consent, complete the interview, and provide a blood sample.
All WECARE Study participants were interviewed by telephone using a structured questionnaire on known or suspected breast cancer risk factors, including age at menarche, age at menopause, number of full-term pregnancies, treatment for breast cancer, and detailed family history of breast cancer in female first- and second-degree relatives. A first-degree family history of breast cancer was defined as a diagnosis of breast cancer in at least one first-degree relative (ie, mother, sister, or daughter). A second-degree family history involved a diagnosis of breast cancer in at least one grandmother, aunt, or half-sister. In the event of multiple first-degree relatives affected with breast cancer, the age at diagnosis of the youngest relative was used in the analyses. Similarly, if there was at least one first-degree relative with bilateral breast cancer, then the woman was categorized as having a bilaterally affected first-degree relative.
As previously described,19 mutation screening of BRCA1 and BRCA2 coding exons and flanking introns in the WECARE Study population identified 113 unique deleterious mutations, including 73 frameshift deletions/insertions, 26 nonsense mutations, seven splice sites, and seven missense mutations. One hundred nine women were BRCA1 mutation carriers, and 72 women were BRCA2 mutation carriers. Women carrying these known deleterious mutations were excluded from the analyses, as were 14 adopted women without information on family history of breast cancer. In the analyses, if a patient case was excluded as a result of carrying a deleterious mutation, both of her matched controls were also excluded. In the two instances where both controls in a triplet were carriers, their matched patient case was also excluded. Women who were carriers of BRCA1 and BRCA2 variants of unknown significance were not excluded because most variants of unknown significance are likely to be neutral in effect. The final analyses were based on 594 patient cases and 1,119 matched controls.
Multivariable-adjusted rate ratios (RRs) and corresponding 95% CIs were calculated using conditional logistic regression adjusting for variables known to affect breast cancer risk, including age at first primary breast cancer diagnosis, age at menarche, age at menopause, number of full-term pregnancies, histology, estrogen receptor (ER) and progesterone receptor status of the first primary cancer, and chemotherapy and hormonal treatments. Models included an offset term to take into account the counter-matched study design.23 Cumulative risks of CBC by age at first breast cancer diagnosis and family history status were estimated using previously described methodology.19 Briefly, prevalences and relative risks were estimated directly from WECARE Study data and combined with population-based SEER*Stat software cancer incidence data for women age 18 to 54 years diagnosed between 1985 and 2000 for comparability to women in the WECARE Study.24 Additionally, in a post hoc analysis, models from Table 2 were stratified by ER status. This was done where sample size allowed (any affected first- or second-degree relative, any affected first-degree relative, any affected second-degree relative, and affected second-degree relative only). Heterogeneity was tested using a likelihood ratio test of nested models.
Patient demographics and the distribution of risk factors and treatment by case-control status are listed in Table 1. By design, patient cases and controls were comparable with respect to matching factors but differed somewhat on chemotherapy and/or hormonal therapy treatment. Whereas 55% of controls and 51% of patient cases were ER positive, only 34% of controls and 30% of patient cases received hormonal therapy. Table 2 presents the relationship between family history and CBC risk for women in whom no mutation in BRCA1 or BRCA2 was identified. Compared with women who reported no affected relatives, those with a first-degree family history of breast cancer had a nearly two-fold greater risk of developing CBC (RR, 1.9; 95% CI, 1.4 to 2.6). Increases in CBC risks were similar (about two-fold), regardless of the relationship of the affected relative to the index patient (mother, sister, or grandmother; data not shown).
Early age at diagnosis (before age 45 years) in the proband and early age at diagnosis in the affected relative were both associated with increased CBC risk (RR, 2.2; 95% CI, 1.4 to 3.6; and RR, 2.2; 95% CI, 1.3 to 3.9, respectively). When the proband or relative was diagnosed at age 45 years or older, the RRs remained statistically significantly elevated but were lower (1.7 and 1.8, respectively). The highest risk was seen when both the proband and the affected relative were diagnosed before age 45 years (RR, 2.5; 95% CI, 1.1 to 5.3). The RR for CBC was 3.6 (95% CI, 2.0 to 6.4) in women whose first-degree relative had experienced CBC. A history of bilateral breast cancer in an affected relative was observed to be a more important predictor of risk than the age at diagnosis of the first primary cancer in the relative (Table 2). Where sample size allowed, results from Table 2 were stratified by ER status (results not shown). There was no effect modification by ER status on the association between family history and CBC risk.
Table 3 lists the 10-year cumulative risks of CBC for women in whom no deleterious mutation in BRCA1 or BRCA2 was identified. Risks are shown as a function of age at diagnosis and the nature of the family history. For comparison, analogous previously published cumulative risks of CBC are also provided for BRCA1 and BRCA2 mutation carriers.19 Among women age 25 to 54 years, the 10-year cumulative risks of CBC for those with no family history, second-degree family history only, a first-degree family history, or a history of bilateral breast cancer in a first-degree relative were 4.6% (95% CI, 4.0% to 5.1%), 5.9% (95% CI, 4.6% to 8.6%), 8.6% (95% CI, 6.1% to 11.5%), and 15.6% (95% CI, 8.5% to 28.5%), respectively. For comparison, the previously published 10-year cumulative risk of CBC among women age 25 to 54 years carrying mutations in BRCA1 or BRCA2 was 18.4% (95% CI, 16.0% to 21.3%).19 Figure 1 provides the cumulative absolute risks by time since diagnosis for women diagnosed before age 55 years with varying degrees of family history and bilaterality; for comparison, previously published risks for BRCA1 and BRCA2 mutation carriers19 are also shown.
Family history of breast cancer is a well-established risk factor for CBC. In this study, we demonstrated that women with a family history of breast cancer have a high risk of CBC, even after testing negative for deleterious mutations in the known breast cancer susceptibility genes BRCA1 and BRCA2.
The relative risks for the association of family history of CBC that we observed in non–BRCA1 and BRCA2 mutation carriers are comparable to those of prior studies that evaluated the relative risk of first primary incident breast cancer and used unaffected controls as the comparison group. A meta-analysis of case-control studies of first primary breast cancer reported a two-fold increased risk associated with having a first-degree family history of breast cancer (meta-analysis odds ratio, 2.1; 95% CI, 2.0 to 2.2; WECARE Study RR, 1.9; 95% CI, 1.4 to 2.6) and a three-fold increased risk when both the proband and relative were both diagnosed at younger ages (meta-analysis odds ratio, 3.3; 95% CI, 2.8 to 3.9; WECARE Study RR, 2.5; 95% CI, 1.1 to 5.3).3
Our results also confirm those from prior studies that reported associations of CBC risk with any family history of breast cancer, family history of bilateral breast cancer, and family history of early-onset breast cancer.4–15,18 In a study with 136 CBCs and 4,524 UBCs where BRCA1 and BRCA2 mutation status was unknown, Bernstein et al5 found that compared with women with no first-degree family history of breast cancer, women with a first-degree relative diagnosed at age 45 years or younger had 2.7-fold greater risk of CBC (RR, 2.7; 95% CI, 1.5 to 5.0)5 and also noted, as we did, a significantly increased risk of CBC for women whose first-degree relative was affected bilaterally.
In a recent study by Nichols et al,25 temporal trends of CBC incidence in the United States were found to diverge starting in 1985, depending on ER status of the first primary breast cancer. A favorable decrease of 3% per year was found in women with ER-positive first primary breast cancer. No significant decrease was found in women diagnosed with ER-negative first primary breast cancer. In another study, Bedrosian et al26 found that women with ER-positive tumors had a significantly lower risk of CBC when compared with women with ER-negative tumors. Because of these differential incidence patterns by hormone receptor status and because of the discrepancy in the current study between percentage of women who were ER positive and percentage of women who received hormonal therapy, we included both ER and progesterone receptor status of the first primary breast cancer as confounders in the current study. The addition of these variables did not substantively change estimates from models without these indicators. To further investigate the interplay between ER status, family history, and risk of CBC, where sample size allowed, we stratified family history of breast cancer models by ER status while adjusting for hormonal therapy. No significant heterogeneity was established, suggesting family history is an important risk factor of CBC regardless of a woman's ER status.
On the basis of data from the SEER cancer registry program, it is estimated that women with a first primary invasive breast cancer have an approximately 7% risk of developing CBC 25 years after the first diagnosis.27 Ten-year cumulative incidence rates of CBC among BRCA1 and BRCA2 mutation carriers with breast cancer vary widely.20,22,28,29 In this WECARE Study population, we found that 10-year rates ranged from 10% to 32%, depending on age at first diagnosis,19 with women younger than 55 years having a 10-year cumulative CBC rate of 18%. In the present analysis, we estimated the age- and family history–specific absolute risks of CBC and found that noncarriers diagnosed with UBC before age 55 years with a first-degree family history of bilateral breast cancer had a cumulative risk of CBC of 15.6% 10 years after their first primary breast cancer diagnosis. Thus, noncarriers with a bilaterally affected first-degree relative have a 10-year cumulative risk of CBC that is nearly as high as that of women who are BRCA1 or BRCA2 mutation carriers (15.6% v 18.4%, respectively). It is important to note that other studies have found higher 10-year rates of CBC for BRCA carriers in the 30% to 40% range.22,29–32 However, these studies often differ from the current study by using even younger women or including just BRCA1 carriers. We elected to compare results produced from the same population to take advantage of all similarities including study design and patient case ascertainment.
Previously, we reversed the time frame to estimate the cumulative risk to relatives of women with CBC or UBC. In that analysis,33 we performed a kin-cohort analysis of WECARE data to estimate the breast cancer risks to relatives of patients with CBC and UBC by the proband's BRCA1 and BRCA2 carrier status and ages of the relatives. We showed that relatives of patients with CBC were at a higher risk than relatives of patients with UBC, among both carriers and noncarriers. Although this provides evidence of the effect of family history of bilateral breast cancer, we did not specifically address the risk of CBC in the relatives either among carriers or noncarriers in the prior study.
The strengths of the present study include its population-based study design unselected for family history, the large number of CBCs, detailed family history data, and complete BRCA1 and BRCA2 mutation screening of all study participants. Additionally, all women were younger than age 55 years at the time of first breast cancer diagnosis, thereby focusing this study on an enriched population of young women at risk of CBC. Further, because controls were also affected with breast cancer, it is highly unlikely that differential recall of reported family history biased our results. Although our total sample size is large, a limitation of our study is the small number of women in several categories, which precludes further stratified assessments of the associations of CBC risk with family history.
We have demonstrated that after excluding carriers of pathogenic mutations in BRCA1 and BRCA2, family history remains a substantial risk factor for CBC. In particular, young noncarriers who have a family history of breast cancer have a considerably greater risk of CBC than those without a family history. Further, women with a first-degree history of bilateral breast cancer have a 10-year cumulative risk of CBC that is comparable to risks previously reported for BRCA1 and BRCA2 mutation carriers. These results underscore the critical importance of obtaining detailed family histories from all women diagnosed with breast cancer, regardless of BRCA1 and BRCA2 mutation carrier status. Because women with a family history of bilateral breast cancer have risks of CBC similar to those of mutation carriers, these women should receive counseling for preventive measures.
The WECARE Study Collaborative Group includes the following members: Memorial Sloan-Kettering Cancer Center (New York, NY): Jonine L. Bernstein, PhD (WECARE Study primary investigator [PI]); Colin Begg, PhD; Jennifer D. Brooks, PhD; Marinela Capanu, PhD; Xiaolin Liang, MD; Anne S. Reiner, MPH; Meghan Woods, MPH; Beckman Research Institute, City of Hope National Medical Center (Duarte, CA): Leslie Bernstein, PhD (subcontract PI); Cancer Prevention Institute of California (Fremont, CA): Esther M. John, PhD (subcontract PI); Danish Cancer Society (Copenhagen, Denmark): Jørgen H. Olsen, MD, DMSc (subcontract PI); Lene Mellemkjær, PhD; Fred Hutchinson Cancer Research Center (Seattle, WA): Kathleen E. Malone, PhD (subcontract PI); International Epidemiology Institute (Rockville, MD) and Vanderbilt University (Nashville, TN): John D. Boice Jr, ScD (subcontract PI); National Cancer Institute (Bethesda, MD): Daniela Seminara, PhD, MPH; New York University (New York, NY): Roy E. Shore, PhD, DrPH (subcontract PI); Samuel Lunenfeld Research Institute, Mount Sinai Hospital (Toronto, Ontario, Canada): Julia Knight, PhD (subcontract PI); Anna Chiarelli, PhD (co-investigator); Translational Genomics Research Institute (TGen) (Phoenix, AZ): David Duggan PhD (subcontract PI); University of Iowa (Iowa City, IA): Charles F. Lynch, MD, PhD (subcontract PI); Michele West, PhD; University of Southern California (Los Angeles, CA): Robert W. Haile, DrPH (subcontract PI); Daniel Stram, PhD (co-investigator); Duncan C. Thomas, PhD (co-investigator); Anh T. Diep (co-investigator); Shanyan Xue, MD; Nianmin Zhou, MD; Evgenia Ter-Karapetova; The University of Texas MD Anderson Cancer Center (Houston, TX): Marilyn Stovall, PhD (subcontract PI); Susan Smith, MPH (co-investigator); University of Virginia (Charlottesville, VA): Patrick Concannon, PhD (subcontract PI); Sharon Teraoka, PhD (co-investigator).
Author affiliations appear at the end of this article.
Listen to the podcast by Dr Narod at www.jco.org/podcasts
Written on behalf of the Women's Environmental Cancer and Radiation Epidemiology (WECARE) Study Collaborative Group (see Appendix, online only).
Supported by National Cancer Institute Grants No. R01CA129639, R01CA114236, R01CA097397, and U01CA083178.
Terms in blue are defined in the glossary, found at the end of this article and online at www.jco.org.
Authors' disclosures of potential conflicts of interest and author contributions are found at the end of this article.
The author(s) indicated no potential conflicts of interest.
Conception and design: Anne S. Reiner, Jennifer D. Brooks, Leslie Bernstein, Lene Mellemkjær, Kathleen E. Malone, Marinela Capanu, Duncan C. Thomas, Colin B. Begg, Jonine L. Bernstein
Financial support: Leslie Bernstein, Kathleen E. Malone, Jonine L. Bernstein
Administrative support: Jonine L. Bernstein
Provision of study materials or patients: Charles F. Lynch, Leslie Bernstein, Kathleen E. Malone, Marilyn Stovall, Jonine L. Bernstein
Collection and assembly of data: Charles F. Lynch, Leslie Bernstein, Lene Mellemkjær, Kathleen E. Malone, Sharon N. Teraoka, Patrick Concannon, Xiaolin Liang, Susan A. Smith, Marilyn Stovall, Jonine L. Bernstein
Data analysis and interpretation: Anne S. Reiner, Esther M. John, Jennifer D. Brooks, Lene Mellemkjær, Kathleen E. Malone, Julia A. Knight, Marinela Capanu, Jane C. Figueiredo, Malcolm C. Pike, Robert W. Haile, Duncan C. Thomas, Colin B. Begg, Jonine L. Bernstein
Manuscript writing: All authors
Final approval of manuscript: All authors