1393 Frequency of BRCA1/BRCA2 Mutations in a Population-Based Sample of Young Breast Carcinoma Cases Kathleen E. Malone, Ph.D.1,2 Janet R. Daling, Ph.D.1,2 Cassandra Neal, B.S.3 Nicola M. Suter, Ph.D.3 Cecilia O’Brien, B.A.1 Kara Cushing-Haugen, M.S.1 Thora J. Jonasdottir, D.V.M.3 Jennifer D. Thompson, B.A.3 Elaine A. Ostrander, Ph.D.3 1 Public Health Sciences Division, Fred Hutchinson Cancer Research Center, Seattle, Washington. 2 School of Public, Health and Community Medicine, Department of Epidemiology, University of Washington, Seattle, Washington. 3 Clinical Research Division, Fred Hutchinson Cancer Research Center, Seattle, Washington. Supported in part by grants and contracts R01CA-63705, R01-CA-41416, N01-CP-95671, R01CA-59736, and R21-CA-66186 from the National Cancer Institute. The authors thank the study participants and area physicians for their generous contributions to this study, Kay Byron for programming assistance, Julie Bittner for project coordination efforts, and Neil Wiegand for technical assistance. This article repesents a multidisciplinary effort. Dr. Ostrander led the molecular analysis and Dr. Malone led the epidemiologic and statistical component. Both authors contributed equally to this work. Address for reprints: Elaine A. Ostrander, Ph.D., Clinical Research Division, Fred Hutchinson Cancer Research Center, 1100 Fairview Ave. N., MP-381, P.O. Box 19024, Seattle, WA 98109-1024 Received June 28, 1999; revision received December 13, 1999; accepted December 13, 1999. © 2000 American Cancer Society BACKGROUND. There is a clear and growing need for data regarding BRCA1 and BRCA2 mutation frequencies among breast carcinoma cases not specifically ascertained on the basis of extreme family history profiles. Toward this end, the authors previously reported results with regard to BRCA1 in breast carcinoma patients drawn from a population-based study. In the current study the authors present new findings concerning BRCA2 mutation frequency in this same population, as well as summary data regarding the combined contribution of these two genes. METHODS. Subjects were drawn from two population-based, case-control studies of breast carcinoma in young women conducted in western Washington State and focused on 1) women diagnosed with breast carcinoma before age 35 years (n ⫽ 203); and 2) women with a first-degree family history of breast carcinoma who were diagnosed before age 45 years (n ⫽ 225). Similarities and differences between BRCA2 carriers and BRCA1 carriers were analyzed in terms of age at diagnosis, family history status, and disease features. RESULTS. Of cases diagnosed before age 35 years, all of whom were unselected for family history, 9.4% carried germline mutations (3.4% for BRCA2 and 5.9% for BRCA1). Of cases diagnosed before age 45 years who had a first-degree family history of breast carcinoma, 12.0% carried germline mutations (4.9% for BRCA2 and 7.1% for BRCA1). Increased frequencies of mutations were observed in cases with a personal or family history of early age at diagnosis and in those with four or more family members affected with breast carcinoma. BRCA2 mutations were less common than BRCA1 mutations in families with any history of ovarian carcinoma. CONCLUSIONS. Overall, given current constraints on health care resources, these data suggest that screening for germline mutations in these breast carcinoma susceptibility genes may have the greatest impact on overall health care if it is prioritized toward high and moderate risk populations. Cancer 2000;88: 1393–1402. © 2000 American Cancer Society. KEYWORDS: breast carcinoma, BRCA1, BRCA2, population. R apid advances in the development and availability of BRCA1 and BRCA2 genetic testing have generated the need to better understand the contributions of these genes to the overall incidence of breast carcinoma in the general population. Most insights to date regarding the frequency of mutations in the breast carcinoma susceptibility genes have been drawn from unusual cancer-rich families that are not representative of the family history profiles of women in the general population who develop breast carcinoma. Chromosomal mapping of the BRCA1 and BRCA2 genes relied exclusively on the analysis of rare, deliberately selected, “high-risk” families, each of which had multiple generations with multiple mem- 1394 CANCER March 15, 2000 / Volume 88 / Number 6 bers affected with breast and/or ovarian carcinoma.1-3 Mutation carrier frequency subsequently has been assessed in several distinct collections of families for both BRCA1 and BRCA2.4-10 These studies provided important insights regarding the type and frequency of mutations found in very high risk families and the likely gene penetrance in the context of a pronounced family history. However, it is unclear the extent to which findings in high risk families extrapolate to breast carcinoma patients who lack such pronounced high risk profiles but who nevertheless may be at risk as BRCA1 and/or BRCA2 mutation carriers. Previously we have shown that of women drawn from a population-based cohort, 6.2% of breast carcinoma cases diagnosed before age 35 years and 7.2% of breast carcinoma cases diagnosed before age 45 years who had a first-degree family history of breast carcinoma were BRCA1 mutation carriers.11 Similar results have been reported in a study of African-American and white women.12 Recently, several studies have assessed the frequency of single founder mutations in BRCA2 within populations believed to carry only one or two mutations, namely Ashkenazi Jewish13,14 and Icelandic breast carcinoma patients.15 However, to our knowledge, to date no studies have assessed comprehensively the role of BRCA2 mutations in the entire BRCA2 coding region in breast carcinoma cases from the general population, nor have studies profiled the differences between BRCA1 and BRCA2 carriers drawn from the general population in terms of family history and features of disease, or documented the combined contributions of BRCA1 and BRCA2. The current study addresses these needs by 1) assessing the type and frequency of BRCA2 germline mutations in two potentially “at-risk” subgroups of women with breast carcinoma who were drawn from the general population; and 2) comparing these findings with our previously reported results of BRCA1 mutation analysis in the same group of women. This approach allows us to describe how features of proband or family history characterize the comparatively lower or higher proportions of BRCA1 and BRCA2 carriers among the subsets of women studied. Due to the population-based study design, the results derived offer a more generalizable approximation of the actual BRCA1 and/or BRCA2 carrier frequency in these two potentially “at risk for carrying a mutation” groups, an assessment that to date has been possible only through statistical extrapolation. MATERIALS AND METHODS The subjects for this study were drawn from two previously completed, population-based, case– control studies of breast carcinoma conducted in western Washington State.16,17 Cases (women with breast carcinoma) were identified through the Cancer Surveillance System, a population-based cancer registry, and controls (similarly aged women without breast carcinoma) were identified through random digit dialing. The first study ascertained all incident cases of breast carcinoma diagnosed between January 1, 1983 and April 30, 1990 in white women born after 1944 who were residents of King, Pierce, and Snohomish counties at the time of diagnosis. The second study targeted all incident cases diagnosed before age 45 years between May 1, 1990 and December 31, l992 among women of all races residing in the 3 county areas. Interviews were completed on 845 cases (83.6% of those eligible) and 961 controls (75.5% of those eligible) in the first study and 648 cases (87% of those eligible) and 610 controls (78.7% of those eligible) in the second study. Supplemental funding to the first study allowed for subsequent collection of blood specimens from 592 cases and a subset of control women (n ⫽ 165). In the second study, blood was collected from 545 of the cases and 473 of the controls. Information regarding potential risk factors for breast carcinoma, including family history, was obtained through a structured interview. A subsequent follow-up study allowed us to update the pedigree information that was provided originally. This study was approved by the Institutional Review Board at the Fred Hutchinson Cancer Research Center and informed consent was obtained from all subjects prior to participation. The molecular analyses described herein focused on 3 subgroups of women drawn from the 2 case– control studies: 1) cases with a first-degree family history of breast carcinoma (i.e., an affected mother and/or sister; 2) cases diagnosed before age 35 years; and 3) controls with a first-degree family history of breast carcinoma. These particular groups were targeted because the presence of family history and/or a young age at diagnosis are believed to be indicative of an increased probability of carrying a BRCA1/BRCA2 mutation. Molecular Analyses Genomic DNA was purified from either frozen buffy coats or immortalized lymphoblastoid cell lines. Polymerase chain reaction was performed under conditions described previously11 using primers that had been described previously by others.6,18,19 Novel primers designed to provide redundant coverage of the gene in some regions also were constructed. All variant bands occurring in ⬍ 5% of samples were sequenced as described previously.11,20 BRCA1/BRCA2 Mutations in Young Women/Malone et al. Statistical Analyses The Fisher exact test was used to evaluate among tested women whether the proportions with mutation varied by age, disease stage, vital status, and features of family history and to assess differences in age, disease stage, vital status, and family history between women who were and were not tested. Binomial exact 95% confidence intervals (95% CI) were calculated for the proportions with a mutation. P values and 95% CIs were calculated using Stata statistical software (Stata Corporation, College Station, TX). RESULTS Study Population BRCA2 mutation status was assessed in 2 groups of women with breast carcinoma designated as “at-risk”: 225 women diagnosed with breast carcinoma before age 45 years who also had a family history of a firstdegree relative with breast carcinoma and 203 women diagnosed before age 35 years. These groups overlapped; altogether, 386 cases were tested. The racial distribution of the tested women was as follows: 96.6% were white, 1% were African-American, 1.6% were Asian/Pacific Islander, 0.3% were American Indian/ Aleutian, and 0.6% were classified as “other”. No other specific information regarding ethnic background or the country of origin of the subject’s parents and/or grandparents was collected. The tested cases included 23 cases not available at the time of our initial BRCA1 analysis in this group.11 In both case groups studied, the proportions tested among women who were eligible did not vary substantively by age at diagnosis or family history as described elsewhere in more detail.11 Tested cases were more likely to be alive at last follow-up and were less likely to have had advanced stage disease at diagnosis than those unavailable for testing, largely due to the lag between the interview and blood collection in the first case– control study. In addition, 71 controls age ⬍ 45 years who had a family history of a firstdegree relative with breast carcinoma were assessed for mutations in both BRCA1 and BRCA2. Distribution and Frequency of Disease-Associated Mutations Among the 225 cases ages 21– 44 years with a family history of a first-degree relative with breast carcinoma who were tested, 11 (4.9%; 95% CI 2.5– 8.6) had germline BRCA2 mutations (Table 1) compared with 16 (7.1%; 95% CI 4.1–11.3) who had germline BRCA1 mutations. BRCA2 mutation frequency was highest among women age ⬍ 30 years and decreased by age at diagnosis. Similar results were observed for BRCA1. 1395 Four BRCA2 mutations (21.0%) and 2 BRCA1 mutations (10.5%) were observed in the 19 women who had both a mother and sister with breast carcinoma. Cases who had a first-degree relative diagnosed with breast carcinoma before age 45 years had a higher frequency of mutations for both BRCA2 (6.6%) and BRCA1 (11.8%) than did cases whose only first-degree affected relative(s) had breast carcinoma at or after age 45 years (4.2%, respectively, for both genes). In addition, the mutation frequency for both genes was higher in cases from families with ⱖ 4 affected members including the proband (10.7%, respectively, for both genes) versus those with fewer affected relatives (BRCA2: 3.0% and BRCA1: 5.9%). Among the 203 women diagnosed with breast carcinoma before age 35 years who were tested, 7 (3.4%; 95% CI 1.4 –7.0) had a BRCA2 mutation (Table 2) compared with 12 (5.9%; 95% CI 3.1–10.1) who had a BRCA1 mutation. BRCA2 mutations were slightly less frequent in women ages 30 –34 years (3.2%) versus women age ⬍ 30 years (4.4%). This is distinct from BRCA1, in which mutation frequency decreased noticeably with age (11.1% in women age ⬍ 30 years vs. 4.4% in women ages 30 –34 years). Among women under 35, mutation frequency in both BRCA1 and BRCA2 was related to family history of breast carcinoma. Of the 96 cases diagnosed with breast carcinoma before age 35 years who had no family history of breast carcinoma, none carried a BRCA2 mutation and only 1 carried a BRCA1 mutation. Of the 42 cases diagnosed with breast carcinoma before age 35 years who had a mother and/or sister with breast carcinoma, 3 (7.1%) had a BRCA2 mutation. By comparison, 5 women (11.9%) had a BRCA1 mutation. The mutation frequencies were lower for both genes in those cases age ⬍ 35 years who only had an aunt and/or grandmother with breast carcinoma (2.2% for BRCA2 and 6.7% for BRCA1). Of the 4 cases diagnosed before age 35 years who had both an affected mother and sister, 2 had a BRCA2 mutation and 1 carried a BRCA1 mutation. For both genes, mutation frequency was higher in cases with at least 1 relative with breast carcinoma diagnosed before age 45 years versus those cases with ⱖ 1 relatives with breast carcinoma, none of whom were diagnosed before age 45 years. Of the 11 BRCA2 mutations that occurred in women with a family history of a first-degree relative with breast carcinoma, none involved women with an additional family history of ovarian carcinoma (data not shown). In contrast, BRCA1 mutations were more common in women with a family history of a firstdegree relative with breast carcinoma who also had a relative with ovarian carcinoma (20% in those with a family history of ovarian carcinoma in a first-degree 1396 CANCER March 15, 2000 / Volume 88 / Number 6 TABLE 1 Distribution of BRCA1 and BRCA2 Mutations According to Disease Characteristics and Family History Features among Breast Carcinoma Cases with a First-Degree Family History of Breast Carcinoma All tested cases BRCA1 diseaseassociated mutations BRCA2 diseaseassociated mutations Characteristics No. No. (%) 95% CIa P valueb No. (%) 95% CIa P valueb All cases with first-degree family history 225 16 (7.1) (4.1–11.3) —c 11 (4.9) (2.5–8.6) —c Age at diagnosis (yrs) 21–29 30–34 35–39 40–44 13 29 87 96 3 2 8 3 (23.1) (6.9) (9.2) (3.1) (5.0–53.8) (0.8–22.8) (4.0–17.3) (0.6–8.9) 2 1 6 2 (15.4) (3.4) (6.9) (2.1) (1.9–45.4) (0.1–17.8) (2.6–14.4) (0.2–7.3) Family history of breast ca Mother and/or sister had breast ca Both mother and sister had breast ca Mother had breast ca Sister had breast ca Aunt and/or grandmother also had breast ca 225 19 188 56 97 16 2 13 5 8 (7.1) (10.5) (6.9) (8.9) (8.2) (4.1–11.3) (1.3–33.1) (3.7–11.5) (3.0–19.6) (3.6–15.6) —c —c —c —c —c 11 4 11 4 6 (4.9) (21.0) (5.8) (7.1) (6.2) (2.5–8.6) (6.0–45.6) (3.0–10.2) (2.0–17.3) (2.3–13.0) —c —c —c —c —c Mother/sister had breast ca ⬍ age 45 yrs Mother/sister had breast ca ⱖ age 45 yrs 76 143 9 6 (11.8) (4.2) (5.6–21.3) (1.6–8.9) 0.05 5 6 (6.6) (4.2) (2.2–14.7) (1.6–8.9) 0.52 At least 1 relative had breast ca ⬍ age 45 yrs Affected relatives all had breast ca ⱖ age 45 yrs 89 132 10 6 (11.2) (4.5) (5.5–19.7) (1.7–9.6) 0.07 5 6 (5.6) (4.5) (1.8–12.6) (1.7–9.6) 0.76 1–3 relatives with breast cad ⬎ 3 relatives with breast cad 169 56 10 6 (5.9) (10.7) (2.9–10.6) (4.0–21.9) 0.24 5 6 (3.0) (10.7) (1.0–6.8) (4.0–21.9) 0.03 0.04 0.10 95% CI: 95% confidence interval; ca: carcinoma. a 95% confidence interval for the proportion with mutations. b P value for differences in mutation frequency between groups. c P value not presented because there is no comparison or categories are not mutually exclusive. d Proband’s breast carcinoma is included. relative and 33.3% in those with only a second-degree relative with ovarian carcinoma; P ⬍ 0.001) than in those with no family history of ovarian carcinoma (4.1%). In the 7 cases diagnosed before age 35 years who had a mother and/or sister with ovarian carcinoma, we observed 1 BRCA2 mutation (14.3%) and no BRCA1 mutations. One BRCA2 mutation (9.1%) was detected among the cases diagnosed before age 35 years who had only a second-degree relative with ovarian carcinoma. In this same group, 3 BRCA1 mutations (27.3%) were found. To maximize precision, several characteristics were examined among all 386 cases tested (Table 3). Family history of bilateral breast carcinoma did not appear to be an indicator of either BRCA2 or BRCA1 status, and combined BRCA2 and BRCA1 mutation frequency was identical across family histories that included bilateral breast carcinoma (9.4%) versus those known not to include any bilateral breast carcinoma (9.5%). Of the 33 cases with a family history of both breast and ovarian carcinoma, none carried a BRCA2 mutation, but 8 (24.2%) were BRCA1 mutation carriers (data not shown). Of the 7 cases with a family history of ovarian but not breast carcinoma, there were no BRCA2 mutations and only 1 BRCA1 mutation (14.3%). Finally, of the 245 cases with a family history of breast but not ovarian carcinoma, there were 13 BRCA2 mutation carriers (5.3%) and 12 BRCA1 mutation carriers (4.9%). Lastly, we examined mutation frequency jointly according to the number of family members with breast carcinoma and the presence/absence of ovarian carcinoma in the entire series of 386 cases tested. The majority of cases (n ⫽ 286) in this study fell into the category of ⬍ 4 family members affected with breast carcinoma with no family history of ovarian carcinoma; 7 (2.4%) carried BRCA2 mutations and 11 (3.8%) had BRCA1 mutations. Of the 48 cases with ⱖ 4 family members affected with breast carcinoma but no family members with ovarian carcinoma, there was only 1 BRCA1 mutation (2.1%) observed compared with 6 BRCA2 mutations (12.5%). Of 28 cases with ⬍ 4 BRCA1/BRCA2 Mutations in Young Women/Malone et al. 1397 TABLE 2 Distribution of BRCA1 and BRCA2 Mutations According to Disease Characteristics and Family History among Breast Carcinoma Cases Diagnosed before Age 35 Years All tested cases BRCA1 diseaseassociated mutations Characteristics No. No. (%) 95% CIa P valueb No. (%) 95% CIa P valueb All cases diagnosed before age 35 yrs 203 12 (5.9) (3.1–10.1) —c 7 (3.4) (1.4–7.0) —c Age at diagnosis (yrs) 21–29 30–34 45 158 5 7 (11.1) (4.4) (3.7–24.0) (1.8–8.9) 0.14 2 5 (4.4) (3.2) (0.5–15.1) (1.0–7.2) 0.65 96 42 1 5 (1.0) (11.9) (0.0–5.7) (4.0–25.6) 0 3 — (7.1) (0.0–3.8)d (1.5–19.5) 45 8 12 3 1 2 (6.7) (12.5) (16.7) (1.4–18.3) (0.3–52.6) (2.1–48.4) 1 1 2 (2.2) (12.5) (16.7) (0.1–11.8) (0.3–52.6) (2.1–48.4) Both mother and sister had breast ca Mother had breast ca Sister had breast ca 4 33 13 1 4 2 (25.0) (12.1) (15.4) (0.6–80.6) (3.4–28.2) (1.9–45.4) —c —c —c 2 3 2 (50.0) (9.1) (15.4) (6.8–93.2) (1.9–24.3) (1.9–45.4) —c —c —c Mother/sister had breast ca ⬍ age 45 yrs Mother/sister had breast ca ⱖ age 45 yrs 20 20 3 2 (15.0) (10.0) (3.2–37.9) (1.2–31.7) 1.00 3 0 (15.0) — (3.2–37.9) (0.0–16.8)d 0.23 at least 1 relative had breast ca ⬍ age 45 yrs Relative(s) had breast ca ⱖ age 45 yrs 32 48 5 3 (15.6) (6.2) (5.3–32.8) (1.3–17.2) 0.26 3 1 (9.4) (2.1) (2.0–25.0) (0.0–11.1) 0.30 1–3 relatives with breast cae ⬎ 3 relatives with breast cae 177 14 9 1 (5.1) (7.1) (2.4–9.4) (0.2–33.9) 0.54 4 1 (2.3) (7.1) (0.6–5.7) (0.2–33.9) 0.32 Family history of breast ca None Mother and/or sister had breast ca Aunt and/or grandmother only had breast ca Other relative Adopted, family history unknown BRCA2 diseaseassociated mutations 0.02 0.01 95% CI: 95% confidence interval; ca: carcinoma. a 95% confidence interval for the proportion of cases with mutations. b P value for differences in the mutation frequency between groups. c P value not presented because there is no comparison or categories are not mutually exclusive. d One-sided 97.5% confidence interval. e Proband’s breast carcinoma is included. family members with breast carcinoma but at least 1 relative with ovarian carcinoma, no BRCA2 mutations were found, although BRCA1 mutations were observed in 4 cases (14.3%). Of the 12 cases from families with ⱖ 4 family members with breast carcinoma plus at least 1 family member with ovarian carcinoma, 5 (41.7%) had mutations in BRCA1 whereas none were observed to have a BRCA2 mutation. There were no women who carried mutations in both genes. In addition, no disease-associated mutations were observed in the 71 controls with a firstdegree family history who were tested (data not shown). Frameshift Mutations The majority of mutations observed for both genes in our studies are small insertions or deletions resulting in premature stop codons (Fig. 1). Nearly all BRCA1 mutations observed in these women, except for the stop codons at 1728 and 1854, have been noted previously in various studies of high risk families and are listed in the Breast Cancer Information Core (BIC). However, in the case of BRCA2, we noted that 4 of 14 frameshifts had not been reported previously in the BIC. These included small deletions of 1–7 base pairs that resulted in premature stop codons at 3019, 1739, and 2845, respectively. We also noted a deletion of a single amino acid resulting from a 3-base pair deletion at nucleotide 4081, resulting in the loss of a lysine at codon 1285. Although we believe too little is known regarding the functional domains of the protein to classify this change with absolute certainty, we provisionally have classified it as a disease-associated mutation. 1398 CANCER March 15, 2000 / Volume 88 / Number 6 TABLE 3 Distribution of BRCA1 and BRCA2 Mutations among all 386 Breast Carcinoma Cases Tested All tested cases BRCA1 diseaseassociated mutations BRCA2 diseaseassociated mutations BRCA1 and BRCA2 mutations combined Characteristics No. No. (%) 95% CIa P valueb No. (%) 95% CIa P valueb No. (%) 95% CIa P valueb All cases tested 386 23 (6.0) (3.8–8.8) —c 15 (3.9) (2.2–6.3) —c 38 (9.8) (7.1–13.3) —c Family history of breast ca At least one relative had breast ca ⬍ age 45 yrs Affected relatives all had breast ca ⱖ age 45 yrs 99 12 (12.1) (6.4–20.2) 0.03 5 (5.0) (1.7–11.4) 0.77 17 (17.2) (10.3–26.1) 0.05 162 7 (4.3) (1.8–8.7) 7 (4.3) (1.8–8.7) 14 (8.6) (4.8–14.1) 314 60 15 6 (4.8) (10.0) (2.7–7.8) (3.8–20.5) 7 6 (2.2) (10.0) (0.9–4.5) (3.8–20.5) 22 12 (7.0) (20.0) (4.4–10.4) (10.8–32.3) 137 64 69 9 3 7 (6.6) (4.7) (10.1) (3.0–12.1) (1.0–13.1) (4.2–19.8) 4 3 5 (2.9) (4.7) (7.2) (0.8–7.3) (1.0–13.1) (2.4–16.1) 13 6 12 (9.5) (9.4) (17.4) (5.1–15.7) (3.5–19.3) (9.3–28.4) 334 14 12 2 (3.6) (14.3) (1.9–6.2) (1.8–42.8) 13 1 (3.9) (7.1) (2.1–6.6) (0.2–33.9) 25 3 (7.5) (21.4) (4.9–10.8) (4.7–50.8 25 3 7 0 (28.0) — (12.1–49.4) (0.0–70.8)e 1 0 (4.0) — (0.1–20.4) (0.0–70.8)e 8 0 (32.0) — (14.9–53.5) (0.0–70.8)e 286 28 48 12 11 4 1 5 (3.8) (14.3) (2.1) (41.7) (1.9–6.8) (4.0–32.7) (0.0–11.1) (15.2–72.3) 7 0 6 0 (2.4) — (12.5) — (1.0–5.0) (0.0–12.3)e 0.02 (4.7–25.2) (0.0–26.5)e 18 4 7 5 (6.3) (14.3) (14.6) (41.7) (3.8–9.8) (4.0–32.7) (6.1–27.8) (15.2–72.3) No. of family members with breast ca 1–3 4⫹ d Family history of bilateral breast ca Unilateral only Bilateral Bilateral status unknown Family history of ovarian ca None Mother and/or sister had ovarian ca Aunt and/or grandmother only had ovarian ca Other relative only had ovarian ca Number with breast ca & presence/absence of ovarian ca in relativesd ⬍ 4 breast ca, no ovarian ca ⬍ 4 breast ca, yes ovarian ca 4⫹ breast ca, no ovarian ca 4⫹ breast ca, yes ovarian ca 0.12 0.47 ⬍ 0.001 ⬍ 0.001 0.009 0.32 0.65 0.005 0.24 0.001 0.001 95% CI: 95% confidence interval; ca: carcinoma. a 95% confidence interval for the proportion of cases with mutations. b P value for differences in the mutation frequency between groups. c P value not presented because there is no comparison or categories are not mutually exclusive. d Proband’s breast carcinoma is included. e One-sided 97.5% confidence interval. Missense Changes Missense changes for both genes are noted in Figure 1. Although several of these changes were found within domains predicted to have functional significance (i.e., the isoleucine to methionine at amino acid 379 in the p53-binding domain21), only four also are observed in conserved positions in the protein when the sequence of human, mouse, and either dog or rat is considered; for BRCA1 these are a methionine to threonine change at amino acid 18, which abuts the transcriptional RING finger, and a glycine to valine change at amino acid 1788 within the BRCT domain. For BRCA2 these include an asparagine to glycine change at amino acid 2811 and the glutamic acid to alanine change at amino acid 2856, both with the BRCT domains. No unique missense changes were noted in the control (women without breast carcinoma) samples for BRCA2. Polymorphisms Single stranded confirmation polymorphism (SSCP) analysis variants that occurred in ⬍ 5% of individuals were sequenced. For both genes a number of changes that are believed to be common polymorphisms not associated with disease predisposition were noted. These included changes in DNA sequence that did not change the resulting amino acid or that were within intron sequences and are unlikely to affect splicing of BRCA1/BRCA2 Mutations in Young Women/Malone et al. 1399 FIGURE 1. The diagram includes designation of areas of potential functional significance together with all missense, nonsense, and frameshift changes noted in the cases and controls tested. Findings in controls are illustrated by circles and cases are shown as squares. Deletions of a single repeat within a reiterated sequence are counted as a deletion of the first repeat unit. (a) BRCA1 (1863AA) functional domains and mutations. (b) BRCA2 (3418) functional domains and mutations. 1400 CANCER March 15, 2000 / Volume 88 / Number 6 exons. In addition, we noted a nonsense change in two cases and three controls, at base 10204, which generated a stop signal in BRCA2 at codon 3326. This previously has been described22 as being a polymorphic stop that is not associated with disease. DISCUSSION It is clear that a comprehensive understanding of BRCA1 and BRCA2 carrier frequency will require a variety of study designs. Estimates of gene carrier frequency from studies of women with pronounced family histories involving multiple generations of affected relatives vary noticeably, from 20 –50%.5-10,23-27 Such studies provide critical data relevant to women from these types of families, but these studies may vary by family size, degree of relatedness among relatives, and the overall prevalence of breast carcinoma in the families and thus are not easily manipulated to provide risk assessment information regarding women with more modest family histories. Similarly, studies derived from high risk breast carcinoma screening clinics also have been extremely useful, but estimates derived from such studies often reflect gene frequencies of select groups of women seeking specialized services rather than frequencies likely to be encountered in the population at large.28,29 In this report, we have expanded our earlier study of BRCA1 mutations in two subgroups of cases drawn from a population-based case– control study of early onset breast carcinoma in western Washington to now include an analysis of BRCA2 mutation frequency.11 In addition, we provide a comparative analysis of BRCA2 versus BRCA1 carriers with regard to features of disease and family history. The results of the current study show that overall, BRCA2 mutations were less frequent than were BRCA1 mutations. Of women diagnosed with breast carcinoma before age 35 years who were tested, 3.4% were BRCA2 carriers and 5.9% were BRCA1 carriers. Of cases diagnosed before age 45 who were selected because of their family history of one or more affected first-degree relatives, 4.9% were BRCA2 carriers and 7.1% carried a BRCA1 mutation. For both groups of cases there were variations in the BRCA1/BRCA2 mutation frequency that appeared to increase with decreasing age at diagnosis, the presence of more than three relatives with breast carcinoma, and, in the case of BRCA1 mutation frequency, a family history of ovarian carcinoma. Among all 386 cases studied overall, 9.8% carried a mutation in either gene. The mutation frequencies observed in this investigation are compatible with those derived from statistical analyses of interview data collected from a number of population-based studies.30-32 Such studies suggested that mutations in autosomal dominant, high penetrance genes could account for up to 36% of breast carcinomas in women age ⬍ 30 years30 and approximately 5–11% in women age ⬍ 40 years.31,32 Mutation frequency is expected to decrease dramatically in older women, with the proportion of breast carcinoma cases in the general population due to autosomal dominant breast carcinoma susceptibility genes being 2.2% between ages 40 – 49 years and 1.1% between ages 50 –70 years.33 Some findings in the current study were particular to one or the other gene. For instance, although the likelihood of being a BRCA1 versus BRCA2 carrier was similar for women with no family history of ovarian carcinoma who had ⬍ 4 family members affected with breast carcinoma (3.8% and 2.4%, respectively), it was much higher for BRCA2 in women with ⱖ 4 family members with breast carcinoma (2.1% vs. 12.5% for BRCA1 and BRCA2, respectively). In addition, in the current data set, a family history of ovarian carcinoma was more predictive of BRCA1 than BRCA2 mutations. BRCA1 mutations were significantly more common in women with a family history of ovarian carcinoma (21.4% or 9 of 42 women with a family history of ovarian carcinoma) than in women without such a history (3.6%) (P ⬍ 0.001) (data not shown). BRCA2 mutations were found in only 2 of 42 women with a family history of ovarian carcinoma (4.8%) versus 13 of 334 women without a family history of ovarian carcinoma (3.8%) (P ⫽ 0.67) (data not shown), suggesting that, in this data set, ovarian carcinoma is not a strong predictor of BRCA2 mutations. In addition, prior investigations have suggested that male breast carcinoma may be another hallmark for the presence of BRCA2 mutations.18,23 Five of the 386 cases tested reported a family history of male breast carcinoma, none of whom had a BRCA1 mutation and 1 of whom (20%) had a BRCA2 mutation. Although the rarity of male breast carcinoma makes it a relatively infrequently occurring indicator for consideration of genetic testing, our finding of a BRCA2 mutation within the five cases with a male relative with breast carcinoma is consistent with prior studies. Finally, one additional factor that appeared important when considering the likelihood of being a BRCA1 versus BRCA2 carrier was the age of onset in relatives. In cases with at least 1 affected family member diagnosed before age 45 years, we noted that 12.1% and 5% of cases were BRCA1 and BRCA2 mutation carriers, respectively. But in cases whose affected relatives were all diagnosed after age 45 years, BRCA1 and BRCA2 mutations occurred with similar frequency (4.3% each). It is interesting to note that in the single largest investigation of high risk breast carcinoma (or BRCA1/BRCA2 Mutations in Young Women/Malone et al. breast-ovarian carcinoma) families published to date, a study of 237 breast carcinoma families originally identified for mapping purposes, Ford et al.33 showed that the overall lifetime risk of breast carcinoma in BRCA1 and BRCA2 carriers was similar. However, there was some suggestion of lower risk in BRCA2 carriers age ⬍ 50 years.33 Similar findings were reported by Oddoux et al.13 and Schubert et al.26 Therefore, the results presented in the current study simply may represent the delayed penetrance for BRCA2 carriers over BRCA1 in younger women. However, because we have not specifically studied at women age ⬎ 45 years, we cannot address this question directly. Ford et al. also have shown that in the same cohort breast carcinoma was linked to BRCA1 in an estimated 52% of families, to BRCA2 in 32% of families, and to neither gene in 16% of families.33 In the data presented in the current study we observed that the ratio of BRCA1 to BRCA2 carriers among the women tested was 1.5 to 1, suggesting that in the general population, as in high risk families, BRCA2 mutations occur approximately half as frequently as BRCA1 mutations. The above conclusions must be considered in light of potential limitations in our study design and the sensitivity of our mutational screening strategy. A study such as the current one, of women drawn from the general population, offers the advantage of a broad generalizability of results for the many women who have modest to minimal family histories of cancer. However, the generalizability of these findings also may be influenced by the extent to which women whose samples were available for testing may not be representative of all eligible women in the study.11 If there was an association between mutation status and survival, with mutations in BRCA1 or BRCA2 leading to either increased or reduced survival, our results could overestimate or underestimate the true frequency of mutations. No mutations were found among the controls with a family history of a first-degree relative with breast carcinoma. This is not unexpected, given the estimates of mutation frequency in the general population, the minimal family history that many controls had, and the relatively small number of controls with a family history of a first-degree relative with breast carcinoma that were available for analysis. In addition there are technical considerations. When performed correctly, SSCP and techniques like it can detect ⬎ 80% of mutations.34,35 Indeed, in an analysis of 237 families, each with at least 4 cases of breast carcinoma, Ford et al.33 assessed and compared the sensitivity of completely sequencing the coding region of the BRCA1 gene with the sensitivity of other gel-based genomic screening techniques and found 1401 them to be the same. Nevertheless, some sequence changes will be missed by gel-based techniques such as SSCP, the majority of which are single base pair changes that lead to missense and nonsense changes. Considerable care was taken in the current study to optimize primers and gel conditions so that even subtle variants were detected routinely, but in the absence of complete gene sequencing for every woman it must be assumed that some mutations are missed. An additional factor that limits the generalizability of these findings is the fact that mutations affecting the expression, splicing, or stability of the transcript will not be detected by the majority of screening methods currently in use. Finally, in the absence of a definitive functional assay, the significance of most missense changes remains unclear. Although we estimate that as many as four BRCA1 or BRCA2 changes observed in the current study might have been disease-associated based solely on their position in conserved domains of potential functional significance, this cannot be stated with certainty in the absence of a clear functional assay, and these changes remain as unclassified variants. Therefore, little information can be provided to patients regarding the significance of single amino acid changes in disease predisposition. Overall the current study data suggest that, even in two groups of women selected for their perceived increased risk of carrying germline mutations, the majority do not carry such mutations. It appears likely, given current constraints on health care resources, that BRCA1 and BRCA2 screening will have its strongest impact when targeted to specific, at-risk populations. REFERENCES 1. 2. 3. 4. 5. 6. Hall JM, Lee MK, Newman B, Morrow JE, Anderson LA, Huey B, et al. Linkage of early-onset familial breast cancer to chromosome 17q21. Science 1990;250:1684 –9. Narod SA, Feunteun J, Lynch HT, Watson P, Conway T, Lynch J, et al. Familial breast-ovarian cancer locus on chromosome 17q12-q23. Lancet 1991;338:82–3. Wooster R, Neuhausen SL, Mangion J, Quirk Y, Ford D, Collins N, et al. Localization of a breast cancer susceptibility gene, BRCA2, to chromosome 13q12-13. Science 1994;265: 2088 –90. Castilla LH, Couch FJ, Erdos MR, Hoskins KF, Calzone K, Garber JE, et al. Mutations in the BRCA1 gene in families with early-onset breast and ovarian cancer. Nat Genet 1994; 8:387–91. Friedman LS, Ostermeyer EA, Szabo CI, Dowd P, Lynch ED, Rowell SE, et al. Confirmation of BRCA1 by analysis of germline mutations linked to breast and ovarian cancer in ten families. Nat Genet 1994;8:399 – 404. Friedman LS, Szabo CI, Ostermeyer EA, Dowd P, Butler L, Park T, et al. Novel inherited mutations and variable expressivity of BRCA1 alleles, including the founder mutation 185delAG in Ashkenazi Jewish families. Am J Hum Genet 1995;57:1284 –97. 1402 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. CANCER March 15, 2000 / Volume 88 / Number 6 Neuhausen SL, Mazoyer S, Friedman L, Stratton M, Offit K, Caligo A, et al. Haplotype and phenotype analysis of six recurrent BRCA1 mutations in 61 families: results of an international study. Am J Hum Genet 1996;58:271– 80. Serova OM, Mazoyer S, Puget N, Dubois V, Tonin P, Shugart YY, et al. Mutations in BRCA1 and BRCA2 in breast cancer families: are there more breast cancer-susceptibility genes? Am J Hum Genet 1997;60:486 –95. Struewing JP, Abeliovich D, Peretz T, Avishai N, Kaback MM, Collins FS, et al. The carrier frequency of the BRCA1 185delAG mutation is approximately 1 percent in Ashkenazi Jewish individuals. Nat Genet 1995;11:198 –200. Szabo CI, King MC. Population genetics of BRCA1 and BRCA2 [editorial]. Am J Hum Genet 1997;60:1013–20. Malone KE, Daling JR, Thompson JD, O’Brien CA, Francisco LV, Ostrander EA. BRCA1 mutations and breast cancer in the general population: analyses in women before age 35 years and in women before age 45 years with first-degree family history. JAMA 1998;279:922–9. Newman B, Mu H, Butler LM, Millikan RC, Moorman PG, King MC. Frequency of breast cancer attributable to BRCA1 in a population-based series of American women. JAMA 1998;279:915–21. Oddoux C, Struewing JP, Clayton CM, Neuhausen S, Brody LC, Kaback M, et al. The carrier frequency of the BRCA2 6174delT mutation among Ashkenazi Jewish individuals is approximately 1%. Nat Genet 1996;14:188 –90. Struewing JP, Hartge P, Wacholder S, Baker SM, Berlin M, McAdams M, et al. The risk of cancer associated with specific mutations of BRCA1 and BRCA2 among Ashkenazi Jews. N Engl J Med 1997;336:1401– 8. Thorlacius S, Struewing JP, Hartge P, Olafsdottir GH, Sigvaldason H, Tryggvadottir L, et al. Population-based study of risk of breast cancer in carriers of BRCA2 mutation. Lancet 1998;352:1337–9. Brinton LA, Daling JR, Liff JM, Schoenberg JB, Malone KE, Stanford J, et al. Oral contraceptives and breast cancer risk among younger women. J Natl Cancer Inst 1995;87:827–35. Daling JR, Malone KE, Voigt LF, White E, Weiss NS. Risk of breast cancer among young women: relationship to induced abortion. J Natl Cancer Inst 1994;86:1584 –92. Friedman LS, Gayther SA, Kurosaki T, Gordon D, Noble B, Casey G, et al. Mutation analysis of BRCA1 and BRCA2 in a male breast cancer population. Am J Hum Genet 1997;60: 313–9. Kirkpatrick H, Waber P, Hoa-Thai T, Barnes R, OsborneLawrence S, Truelson J, et al. Infrequency of BRCA2 alterations in head and neck squamous cell carcinoma. Oncogene 1997;14:2189 –93. Langston AA, Malone KE, Thompson JD, Daling JR, Ostrander EA. BRCA1 mutations in a population-based sample of young women with breast cancer. N Engl J Med 1996;334: 137– 42. Zhang H, Somasundaram K, Peng Y, Tian H, Zhang H, Bi D, et al. BRCA1 physically associates with p53 and stimulates its transcriptional activity. Oncogene 1998;16:1713–21. 22. Mazoyer S, Dunning AM, Serova O, Dearden J, Puget N, Healey CS, et al. A polymorphic stop codon in BRCA2 [letter]. Nat Genet 1996;14:253– 4. 23. Couch FJ, DeShano ML, Blackwood MA, Calzone K, Stopfer J, Campeau L, et al. BRCA1 mutations in women attending clinics that evaluate the risk of breast cancer. N Engl J Med 1996;336:1409 –15. 24. Easton DF, Ford D, Bishop DT. Breast and ovarian cancer incidence in BRCA1-mutation carriers. Breast Cancer Linkage Consortium. Am J Hum Genet 1995;56:265–71. 25. Narod SA, Ford D, Devilee P, Barkardottir RB, Lynch HT, Smith SA, et al. An evaluation of genetic heterogeneity in 145 breast-ovarian cancer families. Am J Hum Genet 1995;56: 254 – 64. 26. Schubert EL, Lee MK, Mefford HC, Argonza RH, Morrow JE, Hull J, et al. BRCA2 in American families with four or more cases of breast or ovarian cancer: recurrent and novel mutations, variable expression, penetrance, and the possibility of families whose cancer is not attributable to BRCA1 or BRCA2. Am J Hum Genet 1997;60:1031– 40. 27. Tavtigian SV, Simard J, Rommens J, Couch F, ShattuckEidens D, Neuhausen S, et al. The complete BRCA2 gene and mutations in chromosome 13q-linked kindreds. Nat Genet 1996;12:333–7. 28. Shattuck-Eidens D, Oliphant A, McClure M, McBride C, Gupte J, Rubano T, et al. BRCA1 sequence analysis in women at high risk for susceptibility mutations. Risk factor analysis and implications for genetic testing. JAMA 1997;278: 1242–50. 29. Krainer M, Silva-Arrieta S, FitzGerald MG, Shimada A, Ishioka C, Kanamaru R, et al. Differential contributions of BRCA1 and BRCA2 to early-onset breast cancer. N Engl J Med 1997;336:1416 –21. 30. Claus EB, Risch N, Thompson WD. Genetic analysis of breast cancer in the cancer and steroid hormone study. Am J Hum Genet 1991;48:232– 42. 31. Ford D, Easton DF, Peto J. Estimates of the gene frequency of BRCA1 and its contribution to breast and ovarian cancer incidence. Am J Hum Genet 1995;57:1457– 62. 32. Whittemore AS, Gong G, Itnyre J. Prevalence and contribution of BRCA1 mutations in breast cancer and ovarian cancer: results from three U.S. population-based case-control studies of ovarian cancer. Am J Hum Genet 1997;60:496 –504. 33. Ford D, Easton DF, Stratton M, Narod S, Goldgar D, Devilee P, et al. Genetic heterogeneity and penetrance analysis of the BRCA1 and BRCA2 genes in breast cancer families. Am J Hum Genet 1998;62:676 – 89. 34. Jordanova A, Kalaydjieva L, Savov A, Claustres M, Schwarz M, Estivill X, et al. SSCP analysis: a blind sensitivity trial. Human Mutat 1997;10:65–70. 35. Sheffield VC, Beck JS, Kwitek AE, Sandstrom DW, Stone EM. The sensitivity of single-strand conformation polymorphism analysis for the detection of single base substitutions. Genomics 1993;16:325–32.