8/10/2019 1-s2.0-S0959804913001482-main.pdf

1/6

Proven non-carriers in BRCA families have an earlier age of onsetof breast cancer

Janet R. Vos a,, Geertruida H. de Bock a, Natalia Teixeira b,Dorina M. van der Kolk c,Liesbeth Jansen d, Marian J.E. Mourits b, Jan C. Oosterwijk c

a Department of Epidemiology, University of Groningen, University Medical Center Groningen, Groningen, The Netherlandsb Department of Gynecologic Oncology, University of Groningen, University Medical Center Groningen, Groningen, The Netherlandsc Department of Genetics, University of Groningen, University Medical Center Groningen, Groningen, The Netherlandsd Department of Surgical Oncology, University of Groningen, University Medical Center Groningen, Groningen, The Netherlands

Available online 13 March 2013

KEYWORDS

BRCA

Non-carrier

Risk assessmentBreast cancerOvarian cancer

Abstract Background: Risk estimates for proven non-carriers in BRCAmutation families areinconsistent for breast cancer and lacking for ovarian cancer. We aimed to assess the age-related risks for breast and ovarian cancer for proven non-carriers in these families.

Methods:A consecutive cohort study ascertained 464 proven non-carriers who had a first-degree relative with a pathogenic BRCAmutation. KaplanMeier analyses were used to esti-mate the age-related cancer risks, and we calculated standardised incidence ratios.Results: In the 464 non-carriers, 17 breast cancers and two ovarian cancers were detected at amean age of 47 years (95% confidence interval (CI) 3261) and 49 years (95% CI 3267),respectively. Overall, by the age of 50, the breast and ovarian cancer risks among non-carrierswere 6.4% (95% CI 2.99.8%) and 0.4% (95% CI 01.3%), of which the breast cancer risk wasstatistically significantly higher than the risk in the general population. In particular, the num-ber of breast cancers among non-carriers in BRCA1families was higher than expected for thegeneral population (standardised incidence ratio (SIR) 2.0, 95% CI 1.13.3). In the BRCA1cohort, the mean number of breast cancer cases was higher in families in which non-carrierswere diagnosed before the age of 50 (p= 0.04).Conclusion:The age at diagnosis of breast cancer in non-carriers in BRCAmutation families is

younger than expected, yielding an increased risk in the fifth decade. This effect is most evi-dent in BRCA1 families. If our results are confirmed by others, this could affect the advicegiven on breast cancer screening to proven non-carriers between the age of 40 and 50 in suchfamilies.2013 Elsevier Ltd.

0959-8049 2013 Elsevier Ltd.

http://dx.doi.org/10.1016/j.ejca.2013.02.018

Corresponding author:Address: Department of Epidemiology, University Medical Center Groningen, P.O. Box 30.001, 9700 RB Groningen,The Netherlands. Tel.: +31 50 3615077.

E-mail address: j.r.vos01@umcg.nl(J.R. Vos).

European Journal of Cancer (2013) 49, 21012106

A v a i l a b l e a t w w w . s c i e n ce d i r e c t . c o m

j o u r n a l h o m e p a g e : w w w . e j c a n c e r . c o m

Open access under theElsevier OA license.

Open access under theElsevier OA license.

http://dx.doi.org/10.1016/j.ejca.2013.02.018mailto:j.r.vos01@umcg.nlhttp://dx.doi.org/10.1016/j.ejca.2013.02.018http://dx.doi.org/10.1016/j.ejca.2013.02.018http://dx.doi.org/10.1016/j.ejca.2013.02.018http://dx.doi.org/10.1016/j.ejca.2013.02.018http://www.sciencedirect.com/http://www.elsevier.com/open-access/userlicense/1.0/http://www.elsevier.com/open-access/userlicense/1.0/http://www.elsevier.com/open-access/userlicense/1.0/http://www.elsevier.com/open-access/userlicense/1.0/http://www.elsevier.com/open-access/userlicense/1.0/http://www.elsevier.com/open-access/userlicense/1.0/http://www.sciencedirect.com/http://dx.doi.org/10.1016/j.ejca.2013.02.018http://dx.doi.org/10.1016/j.ejca.2013.02.018mailto:j.r.vos01@umcg.nlhttp://dx.doi.org/10.1016/j.ejca.2013.02.018

8/10/2019 1-s2.0-S0959804913001482-main.pdf

2/6

1. Introduction

Female carriers of a pathogenic mutation in theBRCA1/2 genes have a high risk of developing breastand ovarian cancer: lifetime risks range from 45% to88% and 11% to 59%, respectively.16 Carriers are there-fore enrolled in an intensive screening program and mayopt to have preventive surgery. In the Netherlands, thisprogram runs from the age of 2560 years and consistsof an annual physical examination, annual breast MRIand, from the age of 30, annual mammography.7 Thereis no screening program for ovarian cancer, but BRCA1/2mutation carriers from the age of 3540 may opt forrisk-reducing salpingo-oophorectomy (RRSO).

It has been assumed that women who tested negativefor their family-specific BRCAmutation were not at anincreased risk.811 They are dismissed from intensivescreening and referred to our national breast cancerscreening program, which consists of biennial mammog-raphy from the age of 5075 years. However, the cancer

risk of these provennon-carriers inBRCA-positive fam-ilies is under debate12 and several studies have publishedcontradictory results on their residual breast cancerrisk.811,1316 The reported risk ratios ranges from 0.39to 5.3, while there are no figures available for residualovarian cancer risk.

It is thus uncertain whether non-carriers in BRCA-positive families are rightly being advised to stop screen-ing and to wait until they can enrol in the national pro-gram, and we do not know what advice to give themregarding their possible ovarian cancer risk. Our aimwas to evaluate the breast and ovarian cancer risks for

proven non-carriers who have a first-degree relative witha pathogenicBRCAmutation.

2. Methods

Our family cancer clinic in the University MedicalCenter Groningen provides genetic counselling andscreening to women who may carry a BRCA mutationbased on their personal and/or family history. If thepatient or family fulfils the Dutch criteria for genetictesting, comprehensive BRCA1 and BRCA2 mutationtesting is performed in one or more index cases.17 Once

a pathogenic BRCA mutation has been detected, tar-geted genetic testing for this mutation is offered to allrelatives, using a cascade protocol.18

Information was collected up to March 2008 and pre-viously used to calculate the breast and ovarian cancerpenetrance for BRCA mutation families3; this informa-tion was updated upto September 2011 for the currentstudy. Data were retrieved from patients medicalrecords and entered into a separate, anonymous, pass-word-protected database. Under the Dutch law, thismeans no further approval from our institutions EthicsReview Board was needed.

For all women, we collected information on the dateof birth and death or last contact, as well as data on thefamilial gene mutation, their breast- and ovarian cancerstatus, and if and when a risk-reducing mastectomy(RRM) and/or RRSO had been performed. Data aboutbreast- and ovarian cancer in the general population wasobtained from the Dutch Comprehensive CancerCentre.19,20

All analyses were performed using PASW Statistics18.0 software, and statistical significance was definedasp< 0.05. Descriptive statistics were applied to analysepatient characteristics, differences in continuous and cat-egorical variables were tested two-sided with the MannWhitney U-test and Fishers exact test, respectively.KaplanMeier survival analyses were used to calculatethe cumulative incidence rates. To calculate the breastcancer risk, right-censoring was applied at the age ofRRM (N = 1), the age at RRSO if performed beforethe age of 50 (N= 4), at the last contact, or age at death.In the ovarian cancer risk analyses, women were cen-

sored at the age at RRSO (N= 4), at the last contact,or age at death. Standardised incidence ratios (SIRs)were calculated for the age-specific breast cancer inci-dence, both with and without stratification by the BRCAgene. The numbers of observed cases were comparedwith the numbers of expected cases, which were calcu-lated usingdata from the Dutch Comprehensive CancerCentre.19,20 To account for possible ascertainment bias,SIRs were also calculated for the group of proven non-carriers expanded with the group of non-tested women.In unaffected women, the probability of testing negativeincreases with advancing age, so we divided women into

10-year age groups that had different probabilities ofbeing negative. The numbers in each age group weremultiplied by the probability of being negative and thenadded together to obtain the estimated number thatwould have tested negative.14

3. Results

In 365 BRCA-positive families (219BRCA1 and 146BRCA2families), 1524 women were tested for aBRCA1or BRCA2mutation. Of these, 464 women tested nega-tive for their family-specific BRCA mutation and were

included in this study. In total there were 700 untestedfirst-degree female relatives of 20 years or older, ofwhom 184 (26%) developed breast cancer.

With a mean age at last contact of 44.9 years for thegroup of non-carriers, 17 women had been diagnosedwith breast cancer. Thirteen of these cases were among283 non-carriers in BRCA1families (one case was bilat-eral) and four were among 181 non-carriers in BRCA2families. For the 17 cases the mean age at diagnosiswas 46.5 years (standard deviation (SD) 7.0): 44.8 years(SD 7.2) in BRCA1 families and 51.9 years (SD 2.9) inBRCA2families (p= 0.045). This mean age at diagnosis

2102 J.R. Vos et al. / European Journal of Cancer 49 (2013) 21012106

8/10/2019 1-s2.0-S0959804913001482-main.pdf

3/6

was lower than the mean age seen in the general popu-lation of 61.9 years (SD 14.3), but the difference wasnot significant (p= 0.17).3

Two cases of ovarian cancer had been detected, bothin non-carriers inBRCA1families. Their ages at diagno-sis were 43 and 55 years, which were much lower thanthe mean age in the general population (65.5, SD 13.7).

The overall breast cancer risk for non-carriers inBRCA2 families was lower than in BRCA1 families,but this was not significant (hazard ratio = 0.44 (95%confidence interval (CI) 0.141.35), p= 0.15). By theage of 60, the non-carriers had a breast cancer risk of9.5% (95% CI 5.014.1%), whereas womenin the generalpopulation had a risk of 5.2% (Table 1).19 At all ages,the breast cancer risk was higher in the proven non-carriers than in the age-matched cohort from the generalpopulation (Fig. 1). By the age of 50 the breast cancer

risk was significantly higher; the expected cumulativeincidence was outside the confidence interval of theobserved cumulative incidence. The ovarian cancer riskwas slightly higher in non-carriers than in the gen-eral population from age 50 upwards, but this was notsignificant (Table 1).

We compared the observed number of breast cancersto the expected number of cases in the general popula-tion (Table 2). There were significantly more casesobserved in the non-carriers in BRCA1 families in theage groups under 50 years: 3039 years SIR 11 (95%CI 3.029) and 4049 years SIR 4.5 (95% CI 1.89.2).

In non-carriers in BRCA2 families, the number ofobserved cases was not significantly higher thanexpected. Except for the significantly raised SIRs inthe BRCA2 group 4049 years and the total BRCAgroup, the SIRs did not change significantly when thecalculation was based on non-carriers and a proportionof the 700 women not tested (Table 2).

The 19 symptomatic non-carriers came from 14BRCA1 and 4 BRCA2 families. In 204 BRCA families,all the proven non-carriers were still so far free of breastand/or ovarian cancer, while another 143BRCAfamiliescontained no known non-carriers. To assess thedifference in cancer incidence due to genetic modifiersand/or environmental factors, we compared the relativesin BRCA families with symptomatic non-carriers tothose families without symptomatic non-carriers(Table 3). We observed no differences in the numbers

of cases or mean age at diagnosis, or in the cancer riskfor either all family members or for the carriers alone.However, among the BRCA1 families, the number ofbreast cancers was significantly higher in families withnon-carriers diagnosed with breast cancer before theage of 50 than in families with non-carriers affected afterthe age of 50, or in families with no affected non-carriers.

4. Discussion

We investigated the occurrence of breast and ovariancancer in 464 proven non-carriers in BRCAfamilies in a

Table 1Cumulative incidence of breast and ovarian cancer among proven BRCAnon-carriers and the general population.

Age (years) Breast cancer Ovarian cancer

Proven non-carriersa General population19 Proven non-carriersa General population19

40 1.3 (0.02.5) 0.6 0.0 (0.00.0) 0.150 6.4 (2.99.8) 2.5 0.4 (0.01.3) 0.160 9.5 (5.014) 5.2 1.4 (0.03.5) 0.4

a

95% confidence interval shown in brackets.

Fig. 1. Cumulative incidence of breast cancer (A) and ovarian cancer (B) for non-carriers (with 95% confidence interval) and age-matched womenfrom the general population.

J.R. Vos et al. / European Journal of Cancer 49 (2013) 21012106 2103

8/10/2019 1-s2.0-S0959804913001482-main.pdf

4/6

consecutive cohort study with a clinic-based ascertain-ment. We show that the cumulative incidence of breastcancer in proven non-carriers in their fifth decade is2.6 times higher than in the general population, with aSIR of 3.5 (95% CI 1.66.7) and a mean age at diagnosisthat was 15 years earlier. This study is the first to assessovarian cancer risks for proven non-carriers. Only twocases were diagnosed at ages much younger than the

mean age at diagnosis in the general population.Previously published estimates on breast cancer risk

in non-carriers in the BRCA families vary widely. Wefound an increased cancer risk for proven non-carriersat all ages compared with women from the general pop-ulation. This is in linewith threeother clinic-basedstud-ies from England,14 Canada15 and Poland16 thatreported a significantly increased risk of at least twofold.However, four other studies (two clinic-based studiesfrom the United States8,9 and one from Australia,11

and one population-based study from the UnitedStates10) report contradictory results that show that

the risk of proven non-carriers is not increased or atleast not twice as high as in the general population. Pos-sible explanations may be differences in: (1) ascertain-ment, (2) study design, or (3) national screeningprotocols.10 First, studies with ascertainment by a fam-ily cancer clinic will probably result in higher risk esti-mates, since families referred to these clinics have astronger history of breast cancer than the general popu-

lation. Second, although a prospective study design isfavored,27,28 it might result in lower risk estimates as asubstantial number of the family members have alreadybeen diagnosed before visiting the family cancer clinic,and non-symptomatic proven non-carriers are no longerfollowed-up in such clinics. As an illustration, when weapplied left-censoring at the moment of the individualsDNA test, the follow-up time was short and only a fewbreast cancer cases were observed in this period. Third,countries with more stringent inclusion criteria forgenetic counselling or those in which additional screen-ing for non-carriers is readily available will probably

Table 3Family history of cancer in BRCAcarriers and relatives at 50% risk but not tested, in families with and without affected proven non-carriers.

BRCA familiesa BRCA1 familiesb

Relatives of affectednon-carriersN= 115

Relatives of onlynon-affectednon-carriers

N= 1174

P Relatives of affectednon-carriers withBC < 50 years

N= 75

Relatives of(non-)affectednon-carriers

N= 1214

P

Mean age (standard deviation, SD)

Breast cancer 46 (12) 45 (12) 0.45 45 (12.1) 43 (12) 0.41Ovarian cancer 51 (11) 53 (12) 0.95 53 (12.0) 50(10) 0.29

Number (%)

Breast cancers 48 (42%) 417 (36%) 0.19 34 (47%) 264 (34%) 0.04Ovarian cancers 21 (18%) 169 (14%) 0.34 15 (21%) 131 (17%) 0.42

Abbreviations: BC, breast cancer.a Comparison ofBRCAfamilies with affected proven non-carriers with a diagnosis of breast or ovarian cancer and families with only proven non-

carriers with no diagnosis of breast or ovarian cancer.b Comparison ofBRCA1families with non-carriers with a breast cancer diagnosis before the age of 50, on the one hand, and BRCA1families

with affected non-carriers at over 50 and BRCA1families with no affected non-carriers, on the other hand.

Table 2Standardised incidence ratios of breast cancer by BRCAgene and age groupa in (A) non-carriers and (B) non-carriers and assumed non-carriers.

Age BRCA1 BRCA2 BRCAtotal

Obs. Exp. SIR 95% CI Obs. Exp. SIR 95% CI Obs. Exp. SIR 95% CI

(A) Non-carriers

3039 4 0.4 11 3.029 0 0.3 4 0.6 6.4 1.7164049 7 1.6 4.5 1.89.2 2 0.9 2.1 0.37.6 9 2.6 3.5 1.66.75059 2 2.6 0.8 0.12.8 2 1.8 1.2 0.14.4 4 4.5 0.9 0.22.3

Totalb 13 6.6 2.0 1.13.3 4 4.7 0.8 0.22.1 17 11.6 1.5 0.92.3

(B) Non-carriers and not-tested women

3039 6.5 0.6 11 5.819 0.5 0.3 1.6 0.111 7 0.9 7.5 3.7124049 8.6 1.5 3.4 1.95.7 4.5 1.6 2.8 1.14.7 13 4.2 3.1 2.05.05059 4 4.3 0.9 0.41.7 4 3.0 1.3 0.93.2 8 7.6 1.1 0.61.7Totalb 19 11 1.7 1.02.5 9.1 8.3 1.1 0.71.9 28 20 1.4 1.02.0

Abbreviations: obs. observed; exp. expected; SIR standarised incidence ratio; CI confidence interval.a No breast cancers were observed after the age of 59 years.b Ages:2069 years.

2104 J.R. Vos et al. / European Journal of Cancer 49 (2013) 21012106

8/10/2019 1-s2.0-S0959804913001482-main.pdf

5/6

8/10/2019 1-s2.0-S0959804913001482-main.pdf

6/6

References

1. Chen S, Parmigiani G. Meta-analysis of BRCA1 and BRCA2penetrance.J Clin Oncol2007;25:132933.

2. Antoniou A, Pharoah PD, Narod S, et al. Average risks of breastand ovarian cancer associated with BRCA1 or BRCA2 mutationsdetected in case series unselected for family history: a combinedanalysis of 22 studies. Am J Hum Genet 2003;72:111730.

3. van der Kolk DM, de Bock GH, Leegte BK, et al. Penetrance of

breast cancer, ovarian cancer and contralateral breast cancer inBRCA1 and BRCA2 families: high cancer incidence at older age.Breast Cancer Res Treat 2010;124:64351.

4. Evans DG, Shenton A, Woodward E, Lalloo F, Howell A, MaherER. Penetrance estimates for BRCA1 and BRCA2 based ongenetic testing in a clinical cancer genetics service setting: risks ofbreast/ovarian cancer quoted should reflect the cancer burden inthe family. BMC Cancer 2008;8:155.

5. Ford D, Easton DF, Stratton M, et al. Genetic heterogeneity andpenetrance analysis of the BRCA1 and BRCA2 genes in breastcancer families: the breast cancer linkage consortium. Am J HumGenet 1998;62:67689.

6. Easton DF, Ford D, Bishop DT. Breast and ovarian cancerincidence in BRCA1-mutation carriers: breast cancer linkageconsortium. Am J Hum Genet 1995;56:26571.

7. Comprehensive Cancer Centre the Netherlands. Mammacarcino-ma (national guideline, version 2.0). Oncolines, cancer clinicalpractice guidelines. Cited 05/01/2012. Available from: http://www.oncoline.nl/mammacarcinoom.

8. Domchek SM, Gaudet MM, Stopfer JE, et al. Breast cancerrisks in individuals testing negative for a known family mutationin BRCA1 or BRCA2. Breast Cancer Res Treat 2010;119:40914.

9. Korde LA, Mueller CM, Loud JT, et al. No evidence of excessbreast cancer risk among mutation-negative women from BRCAmutation-positive families. Breast Cancer Res Treat 2011;125:16973.

10. Kurian AW, Gong GD, John EM, et al. Breast cancer risk fornoncarriers of family-specific BRCA1 and BRCA2 mutations:findings from the breast cancer family registry. J Clin Oncol2011;29:45059.

11. Harvey SL, Milne RL, McLachlan SA, et al. Prospective study ofbreast cancer risk for mutation negative women from BRCA1 orBRCA2 mutation positive families. Breast Cancer Res Treat2011;130:105761.

12. Evans DG, Howell A. Breast cancer risk for noncarriers offamily-specific BRCA1 and BRCA2 mutations: more troublewith phenocopies. J Clin Oncol 2012;30:11423, author reply11434.

13. Katki HA, Gail MH, Greene MH. Breast-cancer risk in BRCA-mutation-negative women from BRCA-mutation-positive families.Lancet Oncol2007;8:10423.

14. Smith A, Moran A, Boyd MC, et al. Phenocopies in BRCA1 andBRCA2 families: evidence for modifier genes and implications for

screening. J Med Genet 2007;44:105.

15. Rowan E, Poll A, Narod SA. A prospective study of breast cancerrisk in relatives of BRCA1/BRCA2 mutation carriers. J MedGenet2007;44:e89, author reply e88.

16. Gronwald J, Cybulski C, Lubinski J, Narod SA. Phenocopies inbreast cancer 1 (BRCA1) families: implications for geneticcounselling. J Med Genet 2007;44:e76.

17. van der Hout AH, van den Ouweland AM, van der Luijt RB, et al.A DGGE system for comprehensive mutation screening ofBRCA1 and BRCA2: application in a Dutch cancer clinic setting.

Hum Mutat2006;27:65466.18. de Bock GH, Hesselink JW, Roorda C, et al. Model of care for

women at increased risk of breast and ovarian cancer. Maturitas2011;71:35.

19. Comprehensive Cancer Centre the Netherlands. Dutch cancerfigures. Cited 05/08/2012. Available from:http://www.cijfersover-kanker.nl/.

20. Curado MP, Edwards B, Shin HR, et al., editors.Cancer incidencein five continents, vol. IX. IARC Scientific Publications No. 160,Lyon, IACR.

21. Dorval M, Nogues C, Berthet P, et al. Breast and ovarian cancerscreening of non-carriers from BRCA1/2 mutation-positive fam-ilies: 2-year follow-up of cohorts from France and Quebec. Eur JHum Genet2011;19:4949.

22. Allweis TM, Sagi M, Peretz T. Sporadic breast cancer among

relatives of BRCA mutation carriers. Genet Med 2005;7:6556.23. Barnes DR, Antoniou AC. Unravelling modifiers of breast and

ovarian cancer risk for BRCA1 and BRCA2 mutation carriers:update on genetic modifiers.J Intern Med2012;271:33143.

24. Antoniou AC, Kuchenbaecker KB, Soucy P, et al. Commonvariants at 12p11, 12q24, 9p21, 9q31.2 and in ZNF365 areassociated with breast cancer risk for BRCA1 and/or BRCA2mutation carriers. Breast Cancer Res 2012;14:R33.

25. Murff HJ, Spigel DR, Syngal S. Does this patient have a familyhistory of cancer? An evidence-based analysis of the accuracy offamily cancer history. JAMA 2004;292:14809.

26. Sijmons RH, Boonstra AE, Reefhuis J, et al. Accuracy of familyhistory of cancer: clinical genetic implications. Eur J Hum Genet2000;8:1816.

27. Goldgar D, Venne V, Conner T, Buys S. BRCA phenocopies or

ascertainment bias? J Med Genet 2007;44:e86, author reply e88.28. Sasieni P. Phenocopies in families seen by cancer geneticists.J Med

Genet 2007;44:e82.29. van Ravensteyn NT, Miglioretti DL, Stout NK, et al. Tipping the

balance of benefits and harms favor screening mammographystarting at age 40 years. A comparative modeling study of risk.Ann Intern Med2012;156:60917.

30. Zonderland HM, Tuut MK, den Heeten GJ, et al. Revised practiceguideline screening and diagnosis of breast cancer. Ned TijdschrGeneeskd2008;152:23369.

31. Moser K, Sellars S, Wheaton M, et al. Extending the age range forbreast screening in England: pilot study to assess the feasibilityand acceptability of randomization. J Med Screen 2011;18:96102.

2106 J.R.Vos et al. / European Journal of Cancer 49 (2013) 21012106

http://www.oncoline.nl/mammacarcinoomhttp://www.oncoline.nl/mammacarcinoomhttp://www.cijfersoverkanker.nl/http://www.cijfersoverkanker.nl/http://www.cijfersoverkanker.nl/http://www.cijfersoverkanker.nl/http://www.oncoline.nl/mammacarcinoomhttp://www.oncoline.nl/mammacarcinoom