The relevance of CHEK2 mutations as a screening target for an elevated risk of breast cancer is of interest. Given the variability among different populations the frequency within a given population needs to be ascertained to qualify relevance. Our contribution to this area of investigation was to establish relevance in the BRCA1/2-negative high-risk breast/ovarian cancer Pakistani population by assessing the prevalence of CHEK2 mutations. While only one Asian study conducted in China has investigated the frequency of CHEK2 germ line mutations using the whole gene screen , our study will provide additional information on genetic variability of the CHEK2 gene in an Asian population from Pakistan.
The risk conferred by CHEK2 germ line mutations to breast and ovarian cancer in Asian populations is not well defined since most studies have determined the frequency of the c.1100delC mutation but did not evaluate other mutations. This holds true for studies conducted in Korea , Malaysia , Singapore , Japan , China [22, 32], India [29, 30], Pakistan and the Philippines . In the present study we investigated the entire CHEK2 coding sequence including the exon-intron junctions using DHPLC analysis followed by direct DNA sequencing of variant fragments. Two potentially deleterious missense mutations, novel p.P92R and p.R406C, were identified in two unrelated patients. Both mutations are rare as they were not detected in a second group of 229 BRCA1/2-negative early-onset and familial breast/ovarian cancer patients and 150 healthy controls suggesting that they may be associated with the disease. Our findings suggest that CHEK2 missense mutations may not contribute significantly to breast/ovarian cancer susceptibility in Pakistan. In a study conducted in China that looked at CHEK2 mutations based on whole gene analysis, a novel recurrent missense mutation, p.H371Y, was identified in 5/118 (4.2%) familial and 16/909 (1.8%) unselected breast cancer patients and in 9/1,228 (0.7%) controls . The results of this study and the present one suggest that CHEK2 germ line mutations may contribute to breast cancer susceptibility in populations from East Asia, but may play a minor role in a South Asian population from Pakistan. Our data on the absence of the c.1100delC mutation in Pakistan are in line with those of previous studies [20–23, 28–32] supporting the notion that c.1100delC does not contribute to breast cancer susceptibility in Asian populations.
In contrast to the single study of a whole CHEK2 gene screen in an Asian population, many studies have been performed in Caucasian populations reporting different mutations and mutation frequencies in the range from 0% to 12%. In the French-Canadian population no mutations were detected in 25 (0%) BRCA1/2-negative familial breast cancer patients . In France mutations were found in 15/507 (2.9%) BRCA1/2-negative familial breast cancer cases . In the United States mutations were observed in 5/169 (3%) early-onset breast cancer patients (≤40 years)  and in another study mutations were identified in 51/1,303 (3.9%) early-onset patients (≤45 years) from the United States, Canada and Australia . The highest mutation frequencies were reported in 30/516 (5.8%) BRCA1/2-negative familial breast cancer patients from Germany , 10/172 (5.8%) Ashkenazi Jewish familial BRCA1/2-founder mutation-negative breast and/or ovarian cancer patients or unaffected women , 8/89 (8.9%) familial breast cancer patients from the UK, North America and the Netherlands  and in 10/82 (12.2%) familial BRCA1/2-founder mutation-negative breast and/or ovarian cancer patients from Finland .
Mutation screening was performed using the combined approach of DHPLC analysis followed by DNA sequencing of variant fragments. DHPLC has been commonly used for mutation screening of various genes as it is rapid, cost-effective and highly sensitive detecting 93 to 100% of mutations that were observed by DNA sequencing . However, since the DHPLC mutation detection rate is not 100% for each gene, it cannot be ruled out that some mutations were missed. CHEK2 mutation analysis is hampered by the presence of multiple homologous copies for exons 10–14 and requires an adapted primer design and amplification conditions. In the present study we employed a nested PCR strategy that specifically amplified the functional copy of the CHEK2 gene .
The CHEK2 mutation frequency of 2/145 (1.4%) observed in this study may be an underestimate as the sensitivity of DHPLC can be below 100% and screening for regulatory mutations residing outside the coding region and for large genomic rearrangements was not performed. Previously a large genomic deletion of 5,395 bp, which results in loss of exons 9 and 10, has been identified in two families of Czechoslovakian ancestry, in a series of Czech and Slovak breast cancer patients (8/631, 1.3%) enriched for familial cases and was absent in 367 controls . In Poland the same deletion was detected in 19/1,978 (1.0%) unselected breast cancer cases, 28/3,228 (0.9%) early-onset cases and in 24/5,496 (0.4%) controls . Thus further analyses for genomic rearrangements in Asian populations may reveal additional CHEK2 mutations as well.
In the present study a novel missense mutation, p.P92R, was identified in a young Pakistani breast cancer patient of Punjabi origin from a family with multiple breast cancers, intestinal cancer, lung cancer, cancer of unknown type and Hodgkin lymphoma. The mutation was also found in another paternal female relative diagnosed with breast cancer at 53 years of age. Since CHEK2 is known to be a multi-organ cancer susceptibility gene , the mutation may have co-segregated with the disease in this family. However, this could not be tested since DNA samples of relevant family members were not available for analysis. The other mutation, p.R406C, was found in a young Punjabi patient diagnosed with ovarian papillary serous carcinoma, who reported a family history of breast cancer. The same mutation has previously been observed in a Spanish familial breast cancer patient and in an early-onset breast cancer patient (≤45 years) of unknown ancestry, but was absent in 400 healthy controls from Spain and 1,109 controls from Canada, the USA and Australia [43, 44]. Due to the lack of DNA samples of relevant family members, co-segregation of this mutation with cancer could not be studied.
The breast tumor of the patient harboring the p.P92R missense mutation was positive for ER, which is consistent with previous findings that breast tumors linked with CHEK2 frame shift and missense mutations (c.1100delC, c.IVS2 +1G>A, del5395, p.I157T) are predominantly ER-positive [48–50]. Additionally, an association of the CHEK2 p.I157T missense mutation has been reported with lobular carcinoma [50, 51]. This link was not observed with the novel p.P92R missense mutation in the present study given the only p.P92R-associated tumor was invasive ductal carcinoma. Given the solitary finding, no interpretation could be rendered.
Our study included cases and controls from only one ethnic group providing uniformity of subjects, adding to the robustness of data. All patients had been selected for BRCA1/2 testing by virtue of having an early-onset of disease or a family history suggestive of hereditary breast/ovarian cancer . Genetic variability in CHEK2 was analyzed in a whole gene screen and was not restricted to the few regions with known CHEK2 mutations. Limitations of the present study are that the functional effects p.P92R and p.R406C and the frequency of large genomic rearrangements were not investigated.