Correspondence
Breast Cancer and BRCA1 Mutations
N Engl J Med 1996; 334:1197-1200May 2, 1996
- Article
To the Editor:
The paper by FitzGerald et al. (Jan. 18 issue)1 provides important data on the frequency of common mutations of the BRCA1 breast-cancer–susceptibility gene in the Ashkenazi Jewish population. However, in their discussion of the use of these data in counseling, the authors' calculation of the “predictive value” of the 185delAG is not as useful to clinicians as the “age-specific penetrance.”
The predictive value of a test is the probability of a disease given a positive test result.2 According to the application of Bayes' theorem, it is the product of the sensitivity of the test and the prevalence of the disease, divided by (sensitivity) (prevalence) + [(1 - specificity)(1 - prevalence)]. As calculated by FitzGerald et al., the prevalence of breast cancer among Jewish women by the age of 40 was estimated as 1 in 217, the estimate of the probability of the mutation in this population was 8 in 39, and the estimate of the false positive rate (1 - specificity) was taken as 8 in 858 from a separate population studied by Struewing et al.3 A difficulty with the last assumption is that for the population studied by Struewing et al., information was not included on age, sex, or medical history,3 since the study was not designed to define the sensitivity or specificity of the 185delAG-mutation test. If one assumes, for example, that there was an excess of male subjects in that series or that some of the female carriers of the mutation had a history of breast cancer, the calculated predictive value of a positive 185delAG test in the series by FitzGerald et al. would be considerably greater than the 9 percent calculated.1
Even if derived precisely, the predictive value for “early onset” breast cancer is of limited clinical value. It does not address, for example, the risk of disease in women between the ages of 40 and 60, which is the period of greatest increase in breast-cancer risk for those heterozygous for BRCA1. 4 These observations underscore the difference between the epidemiologic concept of predictive value and the genetic concept of penetrance. Penetrance, defined as the age-specific risk of disease in a person with a specific mutation, is of paramount interest to family members and clinicians. As also pointed out by Collins (Jan. 18 issue),5 it will be defined by further prospective studies of families with germ-line mutations of BRCA1.
5 ReferencesKenneth Offit, M.D., Ph.D.
Memorial Sloan-Kettering Cancer Center, New York, NY 100211
FitzGerald MG ,MacDonald DJ ,Krainer M , et al. Germ-line BRCA1 mutations in Jewish and non-Jewish women with early-onset breast cancer. N Engl J Med 1996;334:143-149
Full Text | Web of Science | Medline2
Knapp RG, Miller MC. Clinical epidemiology and biostatistics. Baltimore: Williams & Wilkins, 1992.
3
Struewing JP ,Abeliovich D ,Peretz T , et al. The carrier frequency of the BRCA1 185delAG mutation is approximately 1 percent in Ashkenazi Jewish individuals. Nat Genet 1995;11:198-200
CrossRef | Web of Science | Medline4
Ford D, Easton DF, Bishop DT, Narod SA, Goldgar DE, Breast Cancer Linkage Consortium
CrossRef | Web of Science | Medline5
Collins FS . BRCA1 -- lots of mutations, lots of dilemmas. N Engl J Med 1996;334:186-188
Full Text | Web of Science | Medline
To the Editor:
Langston et al. (Jan. 18 issue)1 report germ-line mutations in the BRCA1 gene that confer susceptibility to breast and ovarian cancer in young women with sporadic cases of breast cancer. We do not dispute the existence of BRCA1 mutations in these women, but we consider the authors' classification of the sequence change in one of these patients as a “definite mutation” to be at best premature. The mutation is a duplication of 12 nucleotides in intron 20, a region of the gene not likely to be involved in RNA processing. RNA was not available to test the effect of the variant; its credentials as a pathogenic mutation are based on its absence in 73 controls and its presence in another patient with breast and ovarian cancer and a family history of breast cancer. The extent to which the control population of subjects provided by the Centre d'Etude du Polymorphisme Humain (CEPH) matches the ethnic origins of the patients in whom the duplication was detected is also not clear.
The interpretation of rare sequence variants (which may be nonpathogenic) as “definite mutations” could have serious consequences if the information is used for predictive testing in unaffected relatives of the index patient. This is particularly true with respect to genetic predisposition to breast cancer, since at least one other major gene, BRCA2, predisposes people to breast cancer.2 In the absence of a functional assay for the BRCA1 protein, rigorous criteria should be applied to the assessment of the pathogenic credentials of a mutation. These should include a strong presumption of an adverse effect on the protein product (such as truncation, a nonconservative amino acid substitution, or mutation of a highly conserved sequence involved in RNA processing) and, in the case of missense mutations, absence of the mutation in a substantial number of ethnically matched controls.
2 ReferencesChristopher G. Mathew, Ph.D.
Ellen Solomon, Ph.D.
Shirley V. Hodgson, F.R.C.P.
Guy's Hospital, London SE1 9RT, United Kingdom1
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;334:137-142
Full Text | Web of Science | Medline2
Wooster R ,Bignell G ,Lancaster J , et al. Identification of the breast cancer susceptibility gene BRCA2. Nature 1995;378:789-792
CrossRef | Web of Science | Medline
To the Editor:
Large studies from independent centers are required to evaluate the contribution of germ-line mutations of BRCA1 to the incidence of breast cancer in the general population. Langston et al.1 and FitzGerald et al.2 reported on the frequency of BRCA1 mutations in U.S. women with early-onset breast cancer and no ascertained family history of the disease. We report the results of a study of 83 unrelated French women with invasive breast carcinoma before the age of 36 years (mean age at diagnosis, 31) who were treated at the Institut Curie between 1990 and 1994.
Forty-nine probands had no family history of breast or ovarian cancer, 27 had a family history of breast cancer alone, and 7 had a family history of breast and ovarian cancer. For each patient, the probability of being a carrier of a deleterious BRCA1 allele was evaluated with the genetic model proposed by Easton et al.3 According to this calculation, about 14 BRCA1 mutations would be expected in this group.
After giving informed consent, all patients provided a blood sample. A search for DNA variants in the entire coding sequence and in intron–exon junctions of the BRCA1 gene was performed by using denaturing-gradient-gel electrophoresis (unpublished data).
Sequencing of both strands of variant polymerase-chain-reaction (PCR) fragments revealed six frame-shift and two nonsense mutations predicted to lead to truncated proteins. In addition, one missense mutation, Cys47Tyr, was identified that was likely to result in a defective protein, since this cysteine is a major constituent of the BRCA1 ring finger. Finally, eight rare DNA variants whose effect on the phenotype was unknown were identified (Table 1Table 1
BRCA1 Variants Identified among 83 Women with Breast Cancer Diagnosed before the Age of 36 Years.). The observed prevalence of deleterious BRCA1 mutations was a minimum of 11 percent, a result in accord with the predicted value and similar to those of Langston et al.1 and FitzGerald and colleagues.2 The high predictive value of a family history of breast or ovarian cancer is underscored by the finding of 4 mutations among the 7 patients with such a family history (57 percent) — an observation that contrasts with the detection of 2 mutations among the 27 patients with a family history of breast cancer only (7.4 percent). Interestingly, three BRCA1 mutations (i.e., a third of the detected mutations) were identified in young patients with breast cancer who had unaffected first- and second-degree relatives. The relative contributions of new mutations, chance segregation, incomplete penetrance, and male transmission of this finding of the absence of a family history must now be evaluated.
3 ReferencesGhislaine Ithier, M.D.
Marc Girard, M.D.
Dominique Stoppa-Lyonnet, M.D., Ph.D.
Institut Curie, 75231 Paris, France1
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;334:137-142
Full Text | Web of Science | Medline2
FitzGerald MG ,MacDonald DJ ,Krainer M , et al. Germ-line BRCA1 mutations in Jewish and non-Jewish women with early-onset breast cancer. N Engl J Med 1996;334:143-149
Full Text | Web of Science | Medline3
Easton DF, Bishop DT, Ford D, Crockford GP, Breast Cancer Linkage Consortium
Web of Science | Medline
To the Editor:
It has been estimated that 1 in 100 women of Ashkenazi origin carries the 185delAG frame-shift mutation in the BRCA1 gene. In contrast, none of the 815 non-Ashkenazi controls tested were found to carry this mutation.1
We recently saw a non-Ashkenazi Jewish woman who carries the 185delAG mutation. The patient, a Jewish woman of Iraqi origin, received a diagnosis of infiltrative ductal breast carcinoma at 29 years of age. Her mother had received a diagnosis of breast cancer at the age of 38 years. There are no other known cases of breast or ovarian cancer in the family.
DNA analysis was performed with the use of allele-specific oligonucleotide hybridization,2-4 with the wild-type labeled oligonucleotide sequence 5'AATCTTAGAGTGTCCCA3' and the mutant labeled oligonucleotide sequence 5'AAAATCTTAGTGTCCCAT3'. The test was repeated on a second blood specimen, and the finding of the 185delAG mutation was confirmed.
Until now, the 185delAG mutation has been considered specific to the Ashkenazi population, and women of non-Ashkenazi origin who are at risk for breast cancer have not usually been offered DNA testing, because data on the families are generally uninformative for linkage analysis. Our finding suggests that all women of Jewish ancestry who have an increased risk of breast cancer should be tested for this mutation.
Population-based studies are needed to determine the frequency rate of carriers among Iraqi Jews, as well as in other non-Ashkenazi Jewish populations. These frequency rates are essential for planning screening programs.
4 ReferencesCarron Sher, M.B., Ch.B.
Limor Sharabani-Gargir
Mordechai Shohat, M.D.
Rabin Medical Center, 49100 Petah Tikva, Israel1
Struewing JP ,Abeliovich D ,Peretz T , et al. The carrier frequency of the BRCA1 185delAG mutation is approximately 1 percent in Ashkenazi Jewish individuals. Nat Genet 1995;11:198-200
CrossRef | Web of Science | Medline2
FitzGerald MG ,MacDonald DJ ,Krainer M , et al. Germ-line BRCA1 mutations in Jewish and non-Jewish women with early-onset breast cancer. N Engl J Med 1996;334:143-149
Full Text | Web of Science | Medline3
Struewing JP ,Brody LC ,Erdos MR , et al. Detection of eight BRCA1 mutations in 10 breast/ovarian cancer families, including 1 family with male breast cancer. Am J Hum Genet 1995;57:1-7
Web of Science | Medline4
Friedman LS ,Szabo CI ,Ostermeyer EA , 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;57:1284-1297
Web of Science | Medline
Author/Editor Response
The authors reply:
To the Editor: We agree with Dr. Offit that many unanswered questions remain regarding the 185delAG mutation in BRCA1 in the Ashkenazi Jewish population, and we welcome his call for prospective, population-based studies. The need for such studies was noted both in our paper and in the accompanying editorial. However, prospective studies of a mutation that confers a lifelong risk of cancer will take time, and important insights can be gained by retrospective analyses. In estimating a predictive value for the 185delAG mutation, we used the only available figures for the prevalence of this mutation in the Ashkenazi Jewish population. The prevalence of 8 in 858 for 185delAG reported by Struewing et al.1 may not be derived from a completely unbiased population sample, although it was based on presumably healthy young Jewish adults who sought screening to determine their Tay–Sachs carrier status. This 1 percent prevalence of 185delAG is, of course, unaffected by the age or sex of the sample population. According to our calculations, the specificity of the mutation test would be minimally altered, even in the unlikely event that breast cancer had developed in a substantial fraction of 185delAG carriers by the time they underwent Tay–Sachs screening. Nonetheless, in our paper we stated that the predictive value of 185delAG for the development of breast cancer before the age of 40 is 9 percent and may be higher, depending on the accuracy of available epidemiologic measurements.
In most women with a BRCA1 mutation, breast cancer is diagnosed between the ages of 40 and 60. However, the risk of cancer at an earlier age is significant, and it should be considered in the testing of clinical programs for surveillance and prevention. From an epidemiologic standpoint, using early onset of breast cancer as a criterion for identifying women at increased risk because of genetic predisposition can complement information obtained from the family history, which is often difficult to interpret, given the small size of most families and the high frequency of sporadic breast cancer. The observation that Jewish women in whom breast cancer is diagnosed before the age of 40 have a 20 percent chance of having the 185delAG mutation, irrespective of family history, therefore contributes to the definition of a high-risk subgroup in the general population. Finally, the proportion of BRCA1-associated cases of breast cancer is greater among young women, making it possible to compare the observed frequency of 185delAG mutations with that predicted from age-specific penetrance rates derived from pedigrees with a history of breast cancer.1,2 This comparison suggests that 185delAG is a fully penetrant allele. Our study of early-onset breast cancer therefore predicts that the lifetime risk of breast cancer in 185delAG carriers may be as high as that observed in pedigrees with a history of breast cancer.2 As noted by Dr. Offit, long-term prospective studies will be required to confirm this prediction.
2 ReferencesDaniel A. Haber, M.D., Ph.D.
Massachusetts General Hospital Cancer Center, Boston, MA 02129Judy E. Garber, M.D., M.P.H.
Dana–Farber Cancer InstituteDianne M. Finkelstein, Ph.D.
Harvard School of Public Health, Boston, MA 021151
Struewing JP ,Abeliovich D ,Peretz T , et al. The carrier frequency of the BRCA1 185delAG mutation is approximately 1 percent in Ashkenazi Jewish individuals. Nat Genet 1995;11:198-200
CrossRef | Web of Science | Medline2
Easton DF ,Bishop DT ,Ford D ,Crockford GP , Breast Cancer Linkage consortium. Genetic linkage analysis in familial breast and ovarian cancer: results from 214 families. Am J Hum Genet 1993;52:678-701
Web of Science | Medline
Author/Editor Response
We agree with Mathew et al. that in the absence of complementary DNA (cDNA) analysis it is difficult to say with absolute certainty that the insertion of 12 base pairs that we noted in intron 20 is in fact a mutation. We discussed this issue in our paper, stating that in the absence of cDNA it was not possible to determine the effects of this sequence change on splicing or the stability of messenger RNA. We opted to list this variant as a mutation, since it had also been found in a patient with both breast and ovarian cancer who had a strong family history (five affected relatives) of breast cancer.1 In addition, we did not observe it in the 73 CEPH controls we examined. This result has now been substantiated by our analysis of 164 additional controls from the same geographic region who are unrelated and ethnically matched, none of whom have this variation (unpublished data).
Despite our differing viewpoint with regard to this single variant or mutation, we agree wholeheartedly with Mathew et al. that the issue of classification as rare sequence variants or mutations is extremely important. Like Mathew et al., we favor a very conservative approach in defining missense changes as rare variants until a great deal more is known about the BRCA1 protein itself. For this reason it is our view that the publication of polymorphisms and variants found in the BRCA1 gene is as important as the publication of mutations.
1 ReferencesAmelia A. Langston, M.D.
Kathleen E. Malone, Ph.D.
Elaine A. Ostrander, Ph.D.
Fred Hutchinson Cancer Research Center, Seattle, WA 981041
Takahashi H ,Behbakht K ,McGovern PE , et al. Mutation analysis of the BRCA1 gene in ovarian cancers. Cancer Res 1995;55:2998-3002
Web of Science | Medline
- Citing Articles (1)
Citing Articles
1
H Dolk. (1999) Dietary vitamin A and teratogenic risk: European Teratology Society discussion paper. European Journal of Obstetrics & Gynecology and Reproductive Biology 83:1, 31-36
CrossRef
