In response to DNA damage, the cell-cycle checkpoint kinase CHEK2 can be activated by ATM kinase to phosphorylate p53 and BRCA1, which are involved in cell-cycle control, apoptosis, and DNA repair 12. A germline mutation 1100delC truncates the CHEK2 protein and was first observed in families exhibiting Li-Fraumeni syndrome 3. To address the possibility that this allele contributes to increased susceptibility to breast cancer, two recent studies have investigated the frequency of CHEK2*1100delC in cancer cases and controls. Vahteristo et al. 4 reported that the 1100delC protein-truncating mutation of CHEK2 was observed in 5.5% of 507 patients with a family history of breast cancer and no detectable BRCA mutation compared to 1.4% of 1,885 healthy Finnish controls. The CHEK2-Breast Cancer Consortium observed a 5.1% frequency of this allele in 1071 individuals with breast cancer derived from multiple-case families in which no mutations in BRCA1 or BRCA2 were detected, compared to a 1.1% of 1,620 healthy (mostly Northern European) controls 5. These two groups also reported an increased frequency of CHEK2*1100delC in kindreds with male breast cancer or bilateral breast cancer. We report here the frequency of CHEK2*1100delC in North American breast cancer patients and controls.

We have analyzed a North American sample of healthy controls participating in the New York Cancer Project, and two series of cases: those enrolled in a genetic counseling clinic as well as a separate series of cases unselected for family history ascertained at Memorial Sloan-Kettering Cancer Center. The cases and controls self-identified themselves as Jewish or non-Jewish and reported their ancestor's countries of origin. The population control samples were derived from the New York Cancer Project, an ongoing cohort study enrolling healthy volunteers. Of the 569 non-Ashkenazi controls participating in this study, 62 were African-American, 37 Hispanic, 20 Asian, 18 Other/Unknown, and 432 Caucasian. All individuals in this study provided informed consent for future DNA analysis. Genomic DNA was extracted and amplified utilizing CHEK2 external PCR primers:

9Fe 5' CTGTCATCTCAAGAAGAGGACT and

10R 5' ATTTGTGACTTCATCTAATCACCTCC

and internal PCR primers:

9F 5' TGGCAAGTTCAACATTATTCCC and

10Re 5' GAATAACTCCTAAACTCCAGC

Genotyping was performed by dHPLC (Transgenomics, USA) or Pyrosequencing (Pyrosequencing, Inc. Sweden); all mutations were confirmed by both of these methodologies. Non-Ashkenazi individuals were screened by DNA sequencing for mutations in the entire BRCA1 and BRCA2 coding sequence and Ashkenazi individuals were screened for the three founder alleles that account for >95% of mutations in this ethnic group 6.

We observed CHEK2*1100delC in 5/1665 (0.3%) of healthy New York City controls, significantly lower than the frequency in controls reported by Vahteristo et al. (p = 0.0004, Fisher's Exact test) or by the CHEK2 Consortium (p = 0.006). Within our sample, 3/1096 (0.3%) individuals of Jewish (predominantly Ashkenazi) descent carried this allele, a rate comparable to the 0.4% frequency (2/569) observed in individuals of non-Ashkenazi descent. We detected CHEK2*1100delC in 1/100 index cases from multiple-case breast cancer families in which mutations in BRCA1 or BRCA2 were not detected. As noted in Table 1, the one case heterozygous for the 1100delC was an individual of Ashkenazi descent seen at our Center who was previously included in the report by the CHEK2-Breast Cancer Consortium. Of these 100 families, 33 were of Ashkenazi descent and 67 were of non-Ashkenazi origin. CHEK2*1100delC was detected in 0/16 male individuals with breast cancer without mutations in BRCA2, and in 1/46 (2.2%) of clinic-ascertained bilateral breast cancer cases in individuals with a family history of breast cancer who were also of Ashkenazi Jewish descent compared to 0/46 clinic-ascertained unilateral breast cancer cases matched by age, ethnicity and family history. In a separate series of bilateral breast cancer cases unselected for family history, 0/46 carried CHEK2*1100delC, compared to 1/46 unilateral breast cancer cases matched to the bilateral cases by age and ethnicity. Taking all of the breast cancer cases combined, the frequency of the 1100delC was 3/300 (1.0%) compared to the population frequency of 5/1665 (0.3%) (p = 0.10).

Our current findings do not exclude the possibility that the CHEK2*1100delC variant increases the relative risk for breast cancer in North Americans. Because this variant occurs significantly less commonly in healthy New York controls compared to Northern European controls, the low frequency of this allele in multiple-case families, bilateral cases, and male breast cancer cases could still be compatible with a relative risk in the range of 3.5–12 as reported by Vahteristo et al. and the CHEK2-Breast Cancer Consortium.

A prior attempt to analyze the entire CHEK2 gene in a northern European cohort was unable to detect mutations in exons 10–14, including the 1100delC, in 79 BRCA1/2-negative individuals from Finnish families with three or more cases of breast cancer 7. The relative size of the population control samples in our studies was comparable to prior studies (1,620 population controls in the CHEK2-Consortium, 1,885 in Vahteristo et al, and 1,665 in the current study). We evaluated 100 BRCA mutation-negative breast cancer families, compared to 216 in the report by Vahteristo et al, and 718 in the Consortium study. However, our report also utilized the hypothesis proposed by Begg et al 8, of an "enrichment" of germline carriers of cancer predisposition alleles in individuals with bilateral breast cancer. Our series analyzed 92 cases of bilateral breast cancer and 92 controls, compared to 33 bilateral breast cancer cases included in the report by Vahteristo et al. While no enrichment was noted in our bilateral cases compared to unilateral cases, this may also reflect the lower population frequency of the allele in our population.

Further study of the frequency of the CHEK2*1100delC allele in North American breast cancer kindreds is warranted, however, the relative rarity of this allele in our population, plus the relatively low breast cancer penetrance reported for this allele, limit the clinical relevance in North American kindreds of CHEK2*1100delC as a cancer-predisposing allele.

The 0.3% frequency of the 1100delC CHEK2 allele in a New York population is significantly lower than the 1.1–1.4% rate observed in Northern European populations. This allele was infrequent in BRCA1/2 wild-type cases with family history of breast cancer or in cases with a personal history of unilateral or bilateral of breast cancer. The relatively low breast cancer penetrance, along with the low population frequency, limit the clinical relevance in North American kindreds of CHEK2*1100delC as a cancer predisposing allele.

None declared.

KO designed the study and wrote the final draft of the paper; HP helped in the design and coordinated specimen collection; TK, KN, and PK performed genotype analysis and wre supervised by NE who also participated in study design; BR wrote the first draft of the manuscript and coordinated data collection; PK and SG oversaw collection of control samples and participated in study design; OY ascertained the familial breast cases and helped with study design; HH helped with data collection and analysis, JS provided biostatistical analysis; MR and LS participated in study design and analysis.