Our study suggests that the prevalence of the TP53 R337H mutation in breast cancer patients diagnosed in clinics of Rio de Janeiro, Brazil is low. Only two of the 390 cases were found to carry this mutation (~0.5%) with an age of diagnosis of 35 and 39 (bilateral breast cancer). This estimate is substantially lower than that obtained in previous studies conducted in southern Brazil (range 2.4-11%) which included study participants from Rio de Janeiro, Sao Paolo and Porto Alegre [5, 8]. Sao Paolo and Porto Alegre are 442 km and 1546 km from Rio de Janeiro, respectively. They did not evaluate frequency based on city; however, all three cities are located on the eastern side of Brazil and close to the Atlantic ocean. Despite similarities in geographic regions and the mutation of interest, these studies were substantially different with respect to the selection of their study populations and may explain why our mutation frequency was substantially lower. For example, the current study included unselected cases of breast cancer, while the two other studies included subjects with families suggestive of Li-Fraumeni syndrome [5, 8].
Achatz et al. screened 45 Brazilian subjects from unrelated families with cancer histories suggestive of Li-Fraumeni syndrome and 53 healthy subjects for the R337H mutation . They found a TP53 mutation in 13 cases, six of which were the R337H mutation. None of the 53 healthy subjects carried the mutation. In addition, the most common type of tumor in these families was that of the breast (~30.4%) . In additional studies, the same researchers identified the R337H mutation in two out of 750 healthy women (0.3%) enrolled in a breast screening program in Porto Alegre, southern Brazil . Both cases had a family history of cancer but did not match the Li-Fraumeni syndrome clinical profile.
Assumpcao et al. estimated the prevalence of this mutation in 123 breast cancer cases and 223 age-matched controls . Three of the cases (2.4%) and none of the controls were carriers of the R337H mutation (P = 0.04). All three cases had an early age at diagnosis (range 19-44 years old) and two of the three cases had a family history suggestive of Li-Fraumeni syndrome. Interestingly, all three tumors showed loss of heterozygosity of the mutant allele versus the wild-type allele in contrast to pediatric adrenal cortical carcinomas that normally retain the mutant allele [3, 10]. Given that not all families with the R337H mutation present with other cancers, the authors tested for the presence of other genetic variants within TP53 (R72P and ins16) and MDM2 (SNP309) that have been reported to affect tumor susceptibility and age at onset in R337H mutation carriers, respectively [11–14]. Two of the cases were heterozygous both the TP53 SNPs, while the third patient was homozygous for both R72P. Two of the cases also carried the MDM2 polymorphism. Based on these findings, it has been suggested that the R337H mutation in combination with other SNPs may affect p53 function and ultimately, breast cancer predisposition.
Functional analysis has shown the mutant p53 protein to be pH-sensitive and may adopt a mutant phenotype only under particular physiological conditions [3, 10]. Further, the authors have shown that the mutated tumor cells demonstrate a loss of heterozygosity of the wild-type allele with retention of the mutant allele and accumulation of mutant p53 protein in the nucleus. The penetrance of adrenal cortical carcinomas among carriers of this mutation is 9.9% (95%CI 8.7-11.1%) by the age of 12 years . The frequency of this mutation in southern Brazil has been estimated to be approximately 0.2%-0.3% . Due to this high population frequency the R337H mutation, along with the high incidence of adrenal cortical carcinomas, has led to the implementation of screening all newborns in the state of Parana in southern Brazil .
Strengths of the current study include the large sample size and extensive information on family history. All the participants in this study were recruited from public and private clinics in Rio de Janeiro, Brazil and thus our results are generalizeable to population of Brazil.