Differences in patient ascertainment affect the use of gene‐specified ACMG/AMP phenotype‐related variant classification criteria: Evidence for TP53

The American College of Medical Genetics and Genomics/Association for Molecular Pathology (ACMG/AMP) guidelines for variant classification are widely used for clinical interpretation of gene test results. These guidelines may be specified to genes/syndromes of interest to improve their utility in the clinical setting. As part of these specifications, phenotype‐related criteria can be detailed and weighted depending on the personal history of disease for a given variant carrier. We investigated how ascertainment can affect the significance and/or weight of patient phenotype as a predictor of germline‐variant pathogenicity, using the Li–Fraumeni Syndrome gene TP53 as an example. Likelihood ratios in favor of variant pathogenicity were determined for a report of the personal history of several TP53‐related cancers, using data from 2,656 probands undergoing single‐gene testing (SGT) and 15,483 undergoing multi‐gene panel testing (MGPT). Overall, TP53‐associated cancers were more predictive of pathogenicity, and demonstrated greater evidence weight, in the MGPT versus SGT dataset. This observation is almost certainly explained by differences in proband ascertainment for the two streams of testing, and these findings have implications for germline‐variant classification using ACMG/AMP guidelines.     

A quantitative model to predict pathogenicity of missense variants in the TP53 gene

Germline pathogenic variants in the TP53 gene cause Li‐Fraumeni syndrome, a condition that predisposes individuals to a wide range of cancer types. Identification of individuals carrying a TP53 pathogenic variant is linked to clinical management decisions, such as the avoidance of radiotherapy and use of high‐intensity screening programs. The aim of this study was to develop an evidence‐based quantitative model that integrates independent in silico data (Align‐GVGD and BayesDel) and somatic to germline ratio (SGR), to assign pathogenicity to every possible missense variant in the TP53 gene. To do this, a likelihood ratio for pathogenicity (LR) was derived from each component calibrated using reference sets of assumed pathogenic and benign missense variants. A posterior probability of pathogenicity was generated by combining LRs, and algorithm outputs were validated using different approaches. A total of 730 TP53 missense variants could be assigned to a clinically interpretable class. The outputs of the model correlated well with existing clinical information, functional data, and ClinVar classifications. In conclusion, these quantitative outputs provide the basis for individualized assessment of cancer risk useful for clinical interpretation. In addition, we propose the value of the novel SGR approach for use within the ACMG/AMP guidelines for variant classification.     

Improved,ACMG‐compliant,in silico prediction of pathogenicity for missense substitutions encoded by TP53 variants

Clinical interpretation of germline missense variants represents a major challenge, including those in the TP53 Li–Fraumeni syndrome gene. Bioinformatic prediction is a key part of variant classification strategies. We aimed to optimize the performance of the Align‐GVGD tool used for p53 missense variant prediction, and compare its performance to other bioinformatic tools (SIFT, PolyPhen‐2) and ensemble methods (REVEL, BayesDel). Reference sets of assumed pathogenic and assumed benign variants were defined using functional and/or clinical data. Area under the curve and Matthews correlation coefficient (MCC) values were used as objective functions to select an optimized protein multisequence alignment with best performance for Align‐GVGD. MCC comparison of tools using binary categories showed optimized Align‐GVGD (C15 cut‐off) combined with BayesDel (0.16 cut‐off), or with REVEL (0.5 cut‐off), to have the best overall performance. Further, a semi‐quantitative approach using multiple tiers of bioinformatic prediction, validated using an independent set of nonfunctional and functional variants, supported use of Align‐GVGD and BayesDel prediction for different strength of evidence levels in ACMG/AMP rules. We provide rationale for bioinformatic tool selection for TP53 variant classification, and have also computed relevant bioinformatic predictions for every possible p53 missense variant to facilitate their use by the scientific and medical community.     

Suspected clonal hematopoiesis as a natural functional assay of TP53 germline variant pathogenicity

PurposeSome variants identified by multigene panel testing of DNA from blood present with low variant allele fraction (VAF), often a manifestation of clonal hematopoiesis. Research has shown that the proportion of variants with low VAF is especially high in TP53, the Li-Fraumeni syndrome gene. Based on the hypothesis that variants with low VAF are positively selected as drivers of clonal hematopoiesis, we investigated the use of VAF as a predictor of TP53 germline variant pathogenicity.MethodsWe used data from 260,681 TP53 variants identified at 2 laboratories to compare the distribution of pathogenic and benign variants at different VAF intervals.ResultsLikelihood ratios toward pathogenicity associated with a VAF < 26% equated to the American College of Medical Genetics/Association of Molecular Pathology strong strength level and were applicable for 1 in 5 variants of unknown significance.ConclusionIn conclusion, detection of variants with low VAF in blood can be considered an in vivo functional assay to aid assessment of TP53 variant pathogenicity.     

Modeling the ACMG/AMP variant classification guidelines as a Bayesian classification framework

PurposeWe evaluated the American College of Medical Genetics and Genomics/Association for Molecular Pathology (ACMG/AMP) variant pathogenicity guidelines for internal consistency and compatibility with Bayesian statistical reasoning.MethodsThe ACMG/AMP criteria were translated into a naive Bayesian classifier, assuming four levels of evidence and exponentially scaled odds of pathogenicity. We tested this framework with a range of prior probabilities and odds of pathogenicity.ResultsWe modeled the ACMG/AMP guidelines using biologically plausible assumptions. Most ACMG/AMP combining criteria were compatible. One ACMG/AMP likely pathogenic combination was mathematically equivalent to pathogenic and one ACMG/AMP pathogenic combination was actually likely pathogenic. We modeled combinations that include evidence for and against pathogenicity, showing that our approach scored some combinations as pathogenic or likely pathogenic that ACMG/AMP would designate as variant of uncertain significance (VUS).ConclusionBy transforming the ACMG/AMP guidelines into a Bayesian framework, we provide a mathematical foundation for what was a qualitative heuristic. Only 2 of the 18 existing ACMG/AMP evidence combinations were mathematically inconsistent with the overall framework. Mixed combinations of pathogenic and benign evidence could yield a likely pathogenic, likely benign, or VUS result. This quantitative framework validates the approach adopted by the ACMG/AMP, provides opportunities to further refine evidence categories and combining rules, and supports efforts to automate components of variant pathogenicity assessments.     

Integration of functional assay data results provides strong evidence for classification of hundreds of BRCA1 variants of uncertain significance

PurposeBRCA1 pathogenic variant heterozygotes are at a substantially increased risk for breast and ovarian cancer. The widespread uptake of testing has led to a significant increase in the detection of missense variants in BRCA1, the vast majority of which are variants of uncertain clinical significance (VUS), posing a challenge to genetic counseling. Here, we harness a wealth of functional data for thousands of variants to aid in variant classification.MethodsWe have collected, curated, and harmonized functional data for 2701 missense variants representing 24.5% of possible missense variants in BRCA1. Results were harmonized across studies by converting data into binary categorical variables (functional impact versus no functional impact). Using a panel of reference variants we identified a subset of assays with high sensitivity and specificity (≥80%) and apply the American College of Medical Genetics and Genomics/Association for Molecular Pathology (ACMG/AMP) variant interpretation guidelines to assign evidence criteria for classification.ResultsIntegration of data from validated assays provided ACMG/AMP evidence criteria in favor of pathogenicity for 297 variants or against pathogenicity for 2058 representing 96.2% of current VUS functionally assessed. We also explore discordant results and identify limitations in the approach.ConclusionHigh quality functional data are available for BRCA1 missense variants and provide evidence for classification of 2355 VUS according to their pathogenicity.     

Proposition of adjustments to the ACMG‐AMP framework for the interpretation of MEN1 missense variants

In 2015, the ACMG‐AMP guidelines provided a general procedure for the objective and reproducible classification of genomic variants. While the benefits of this framework are of major importance, its adaptation for locus‐specific use is needed. Multiple Endocrine Neoplasia type 1 (MEN1) occurs due to inactivating mutations in the tumour suppressor gene MEN1, including 20% of missense variants. The classification of these variants may be extremely challenging. Here, we compared the interpretation of the 122 MEN1 missense variants, identified in the French population over the past 15 years by the TENGEN network (French oncogenetics network of neuroendocrine tumors) versus by using the ACMG‐AMP guidelines, and analyzed the causes of discordance. A total of 59.8% of missense variants were termed as (likely)‐pathogenic variants by TENGEN versus only 28.7% using ACMG‐AMP guidelines. Actually, 53.4% (39/73) of TENGEN (likely)‐pathogenic variants were declassified in variant of uncertain significance (VUS) by using ACMG‐AMP guidelines, thereby affecting the clinical management of patients and their families. Twenty of these ACMG‐AMP VUS were found in patients with a clinically authentic MEN1 disease. Here, TENGEN proposes adjustments to the ACMG‐AMP framework for the interpretation of MEN1 missense variants. These propositions merge both the classification systems, and are particularly interesting, as MEN1 is included in the ACMG secondary findings list for reporting in clinical genomic sequencing.     

Clinical impact of re-evaluating genes and variants implicated in dilated cardiomyopathy

PurposeAccurate interpretation of variants detected in dilated cardiomyopathy (DCM) is crucial for patient care but has proven challenging. We applied a set of proposed refined American College of Medical Genetics and Genomics/Association for Molecular Pathology (ACMG/AMP) criteria for DCM, reclassified all detected variants in robust genes, and associated these results to patients' phenotype.MethodsThe study included 902 DCM probands from the Maastricht Cardiomyopathy Registry who underwent genetic testing. Two gene panel sizes (extended n = 48; and robust panel n = 14) and two standards of variant classification (standard versus the proposed refined ACMG/AMP criteria) were applied to compare genetic yield.ResultsA pathogenic or likely pathogenic (P/LP) variant was found in 17.8% of patients, and a variant of uncertain significance (VUS) was found in 32.8% of patients when using method 1 (extended panel (n = 48) + standard ACMG/AMP), compared to respectively 16.9% and 12.9% when using method 2 (robust panel (n = 14) + standard ACMG/AMP), and respectively 14% and 14.5% using method 3 (robust panel (n = 14) + refined ACMG/AMP). Patients with P/LP variants had significantly lower event-free survival compared to genotype-negative DCM patients.ConclusionStringent gene selection for DCM genetic testing reduced the number of VUS while retaining ability to detect similar P/LP variants. The number of genes on diagnostic panels should be limited to genes that have the highest signal to noise ratio.     

The ClinGen Brain Malformation Variant Curation Expert Panel: Rules for somatic variants in AKT3, MTOR,PIK3CA,and PIK3R2

PurposePostzygotic (somatic) variants in the mTOR pathway genes cause a spectrum of distinct developmental abnormalities. Accurate classification of somatic variants in this group of disorders is crucial for affected individuals and their families.MethodsThe ClinGen Brain Malformation Variant Curation Expert Panel was formed to curate somatic variants associated with developmental brain malformations. We selected the genes AKT3, MTOR, PIK3CA, and PIK3R2 as the first set of genes to provide additional specifications to the 2015 American College of Medical Genetics and Genomics/Association for Molecular Pathology (ACMG/AMP) sequence variant interpretation guidelines, which currently focus solely on germline variants.ResultsA total of 24 of the original 28 ACMG/AMP criteria required modification. Several modifications used could be applied to other genes and disorders in which somatic variants play a role: 1) using variant allele fraction differences as evidence that somatic mutagenesis occurred as a proxy for de novo variation, 2) incorporating both somatic and germline evidence, and 3) delineating phenotype on the basis of variable tissue expression.ConclusionWe have established a framework for rigorous interpretation of somatic mosaic variants, addressing issues unique to somatic variants that will be applicable to many genes and conditions.     

Higher‐than‐expected population prevalence of potentially pathogenic germline TP53 variants in individuals unselected for cancer history

Li–Fraumeni syndrome (LFS) is an autosomal‐dominant cancer predisposition disorder associated with pathogenic germline variants in TP53, with a high penetrance over an individual's lifetime. The actual population prevalence of pathogenic germline TP53 mutations is still unclear, most likely due to biased selection of cancer affected families. The aim of this study was to estimate the population prevalence of potentially pathogenic TP53 exonic variants in three sequencing databases, totaling 63,983 unrelated individuals. Potential pathogenicity was defined using an original algorithm combining bioinformatic prediction tools, suggested clinical significance, and functional data. We identified 34 different potentially pathogenic TP53 variants in 131 out of 63,983 individuals (0.2%). Twenty‐eight (82%) of these variants fell within the DNA‐binding domain of TP53, with an enrichment for specific variants that were not previously identified as LFS mutation hotspots, such as the p.R290H and p.N235S variants. Our findings reveal that the population prevalence of potentially pathogenic TP53 variants may be up to 10 times higher than previously estimated from family‐based studies. These results point to the need for further studies aimed at evaluating cancer penetrance modifiers as well as the risk associated between cancer and rare TP53 variants.     

Three germline mutations in the TP53 gene

Three germline mutations in the TP53 tumor-suppressor gene are reported, two of which are not reported previously. A missense mutation at codon 265 of TP53 was found in three patients of a family that complied with the definition of the Li-Fraumeni syndrome. A nonsense mutation in codon 306 was found in a woman who had had a rhabdomyosarcoma at age 4 and a subsequent breast cancer at age 22. She was part of a Li-Fraumeni-like family, but the parental origin of the mutation could not be traced. Finally, while screening for somatic alterations in TP53 in a series of 141 sporadic breast tumors, we detected a constitutional missense mutation in codon 235 in a woman diagnosed with breast cancer at age 26 and a recurrence 4 years later. The recurrence, but not the primary tumor, showed an additional missense mutation at codon 245 as well as loss of the wild-type allele. This suggests that the 245 mutation was particularly important for tumor progression and that there might exist heterogeneity in terms of cancer predisposition potential among the various germline TP53 mutations. Hum Mutat 9:157–163, 1997. © 1997 Wiley-Liss, Inc.     

Adapting ACMG/AMP sequence variant classification guidelines for single-gene copy number variants

PurposeThe ability of a single technology, next-generation sequencing, to provide both sequence and copy number variant (CNV) results has driven the merger of clinical cytogenetics and molecular genetics. Consequently, the distinction between the definition of a sequence variant and a CNV is blurry. As the 2015 American College of Medical Genetics and Genomics/Association for Molecular Pathology (ACMG/AMP) standards and guidelines for interpretation of sequence variants address CNV classification only sparingly, this study focused on adapting ACMG/AMP criteria for single-gene CNV interpretation.MethodsCNV-specific modifications of the 2015 ACMG/AMP criteria were developed and their utility was independently tested by three diagnostic laboratories. Each laboratory team interpreted the same 12 single-gene CNVs using three systems: (1) without ACMG/AMP guidance, (2) with ACMG/AMP criteria,and (3) with new modifications. A replication study of 12 different CNVs validated the modified criteria.ResultsThe adapted criteria system presented here showed improved concordance and usability for single-gene CNVs compared with using the ACMG/AMP interpretation guidelines focused on sequence variants.ConclusionThese single-gene CNV criteria modifications could be used as a supplement to the ACMG/AMP guidelines for sequence variants, allowing for a streamlined workflow and a step toward a uniform classification system for both sequence and copy number alterations.     

Seshat: A Web service for accurate annotation,validation, and analysis of TP53 variants generated by conventional and next‐generation sequencing

Accurate annotation of genomic variants in human diseases is essential to allow personalized medicine. Assessment of somatic and germline TP53 alterations has now reached the clinic and is required in several circumstances such as the identification of the most effective cancer therapy for patients with chronic lymphocytic leukemia (CLL). Here, we present Seshat, a Web service for annotating TP53 information derived from sequencing data. A flexible framework allows the use of standard file formats such as Mutation Annotation Format (MAF) or Variant Call Format (VCF), as well as common TXT files. Seshat performs accurate variant annotations using the Human Genome Variation Society (HGVS) nomenclature and the stable TP53 genomic reference provided by the Locus Reference Genomic (LRG). In addition, using the 2017 release of the UMD_TP53 database, Seshat provides multiple statistical information for each TP53 variant including database frequency, functional activity, or pathogenicity. The information is delivered in standardized output tables that minimize errors and facilitate comparison of mutational data across studies. Seshat is a beneficial tool to interpret the ever‐growing TP53 sequencing data generated by multiple sequencing platforms and it is freely available via the TP53 Website, http://p53.fr or directly at http://vps338341.ovh.net/ .     

Variant curation expert panel recommendations for RYR1 pathogenicity classifications in malignant hyperthermia susceptibility

PurposeAs a ClinGen Expert Panel (EP) we set out to adapt the American College of Medical Genetics and Genomics (ACMG)/Association for Molecular Pathology (AMP) pathogenicity criteria for classification of RYR1 variants as related to autosomal dominantly inherited malignant hyperthermia (MH).MethodsWe specified ACMG/AMP criteria for variant classification for RYR1 and MH. Proposed rules were piloted on 84 variants. We applied quantitative evidence calibration for several criteria using likelihood ratios based on the Bayesian framework.ResultsSeven ACMG/AMP criteria were adopted without changes, nine were adopted with RYR1-specific modifications, and ten were dropped. The in silico (PP3 and BP4) and hotspot criteria (PM1) were evaluated quantitatively. REVEL gave an odds ratio (OR) of 23:1 for PP3 and 14:1 for BP4 using trichotomized cutoffs of ≥0.85 (pathogenic) and ≤0.5 (benign). The PM1 hotspot criterion had an OR of 24:1. PP3 and PM1 were implemented at moderate strength. Applying the revised ACMG/AMP criteria to 44 recognized MH variants, 29 were classified as pathogenic, 13 as likely pathogenic, and 2 as variants of uncertain significance.ConclusionCuration of these variants will facilitate classification of RYR1/MH genomic testing results, which is especially important for secondary findings analyses. Our approach to quantitatively calibrating criteria is generalizable to other variant curation expert panels.     

Comprehensive profiling of BRCA1 and BRCA2 variants in breast and ovarian cancer in Chinese patients

The identification and interpretation of germline BRCA1/2 variants become increasingly important in breast and ovarian cancer (OC) treatment. However, there is no comprehensive analysis of the germline BRCA1/2 variants in a Chinese population. Here we performed a systematic review and meta‐analysis on such variants from 94 publications. A total of 2,128 BRCA1/2 variant records were extracted, including 601 from BRCA1 and 632 from BRCA2. In addition, 414, 734, 449, and 307 variants were also recorded in the BIC, ClinVar, ENIGMA, and UMD databases, respectively, and 579 variants were newly reported. Subsequent analysis showed that the overall germline BRCA1/2 pathogenic variant frequency was 5.7% and 21.8% in Chinese breast and OC, respectively. Populations with high‐risk factors exhibited a higher pathogenic variant percentage. Furthermore, the variant profile in Chinese is distinct from that in other ethnic groups with no distinct founder pathogenic variants. We also tested our in‐house American College of Medical Genetics‐guided pathogenicity interpretation procedure for Chinese BRCA1/2 variants. Our results achieved a consistency of 91.2–97.6% (5‐grade classification) or 98.4–100% (2‐grade classification) with public databases. In conclusion, this study represents the first comprehensive meta‐analysis of Chinese BRCA1/2 variants and validates our in‐house pathogenicity interpretation procedure, thereby providing guidance for further PARP inhibitor development and companion diagnostics in the Chinese population.     

Cancer predisposition and germline CTNNA1 variants

Hereditary Diffuse Gastric Cancer (HDGC) is a cancer predisposing syndrome mainly caused by germline inactivating variants in CDH1, encoding E-cadherin. Early-onset diffuse gastric cancer (DGC) and/or invasive lobular breast cancer (LBC) are the main phenotypes in CDH1-associated HDGC. CTNNA1, encoding for α-E-catenin, and E-cadherin-partner in the adherens junction complex, has been recently classified as a HDGC predisposing gene. Nevertheless, little is known about CTNNA1 tumor spectrum in variant carriers and variant-type associated causality. Herein, we systematically reviewed the literature searching for CTNNA1 germline variants carriers, further categorized them according to HDGC clinical criteria (HDGC vs non-HDGC), collected phenotypes, classified variants molecularly and according to CDH1 ACMG/AMP guidelines and performed genotype-phenotype analysis. We found 41 families carrying CTNNA1 germline variants encompassing in total 105 probands and relatives. All probands from 13 HDGC families presented DGC and their average age of onset was 40 ± 17 years; 10/13 (77%) HDGC families carried a pathogenic (P) variant. Most probands from 28 non-HDGC families developed unspecified-BC, as well as most of their relatives; 4/28 (14%) carried a P variant, 16/28 (57%) carried a likely pathogenic (LP) variant, 7/28 (25%) carried variants of unknown significance (VUS) and 1/28 (4%) carried a likely benign variant. Regardless of clinical criteria, 97% (32/33) of probands and relatives from P variant-carrier families had DGC/unspecified-GC. In LP variant-carrier families, 82% (28/34) of probands and relatives had unspecified-BC. Only 2/105 individuals had LBC. A cluster of frameshift and nonsense variants was found in CTNNA1 last exon of non-HDGC families and classified as VUS. In conclusion, current available data confirms an association of CTNNA1 P variants with early-onset DGC, but not with LBC. We demonstrate that in ascertained cohorts, CTNNA1 P variants explain <2% of HDGC families and support the use of ACMG/AMP CDH1 specific variant curation guidelines, while no specific guidelines are developed for CTNNA1 variant classification. Moreover, we demonstrated that truncating variants at the CTNNA1 NMD-incompetent last exon have limited deleteriousness, and that CTNNA1 LP variants have lower actionability than CDH1 LP variants. Current knowledge supports considering only CTNNA1 P variants as clinically actionable in HDGC carrying families.     

Variable population prevalence estimates of germline TP53 variants: A gnomAD‐based analysis

Reports of variable cancer penetrance in Li–Fraumeni syndrome (LFS) have raised questions regarding the prevalence of pathogenic germline TP53 variants. We previously reported higher‐than‐expected population prevalence estimates in sequencing databases composed of individuals unselected for cancer history. This study aimed to expand and further evaluate the prevalence of pathogenic and likely pathogenic germline TP53 variants in the gnomAD dataset (version r2.0.2, n = 138,632). Variants were selected and classified based on our previously published algorithm and compared with alternative estimates based on three different classification databases: ClinVar, HGMD, and the UMD_TP53 database. Conservative prevalence estimates of pathogenic and likely pathogenic TP53 variants were within the range of one carrier in 3,555–5,476 individuals. Less stringent classification increased the approximate prevalence to one carrier in every 400–865 individuals, mainly due to the inclusion of the controvertible p.N235S, p.V31I, and p.R290H variants. This study shows a higher‐than‐expected population prevalence of pathogenic and likely pathogenic germline TP53 variants even with the most conservative estimates. However, these estimates may not necessarily reflect the prevalence of the classical LFS phenotype, which is based upon family history of cancer. Comprehensive approaches are needed to better understand the interplay of germline TP53 variant classification, prevalence estimates, cancer penetrance, and LFS‐associated phenotype.     

In cis TP53 and RAD51C pathogenic variants may predispose to sebaceous gland carcinomas

Pathogenic variants in TP53 have been classically thought to cause Li-Fraumeni syndrome (LFS), a cancer predisposition with high risks for various childhood- and adult-onset malignancies. However, increased genetic testing has lately revealed, that pathogenic variant carriers exhibit a broader range of phenotypes and that penetrance may be dependent both on variant type and modifiers. Using next generation sequencing and short tandem repeat analysis, we identified germline pathogenic variants in TP53 and RAD51C located in cis on chromosome 17 in a 43-year-old male, who has developed a rare sebaceous gland carcinoma (SGC) but so far no tumors of the LFS spectrum. This course mirrors a Trp53-Rad51c-double-mutant cis mouse-model, which similarly develops SGC, while the characteristic Trp53-associated tumor spectrum occurs with significantly lower frequency. Therefore, we propose that co-occurent pathogenic variants in RAD51C and TP53 may predispose to SGC, reminiscent of Muir-Torre syndrome. Further, this report supports the diversity of clinical presentations associated with germline TP53 alterations, and thus, the proposed expansion of LFS to heritable TP53-related cancer syndrome.Subject terms: Genetics research, Cancer genetics     

Assessment of TP53 Polymorphisms and MDM2 SNP309 in Premenopausal Breast Cancer Risk

Germline polymorphic variants in cancer predisposition genes such as TP53 have been shown to impact the risk of premenopausal cancer. Accordingly, the aim of this study was to assess the spectrum of polymorphisms in TP53 and its negative regulatory gene, MDM2 (SNP309:T>G) in patients with premenopausal breast cancer. Our findings in a cohort of 40 female patients demonstrate no significant correlation between the studied polymorphisms and risk of premenopausal breast cancer. Although one polymorphism is found in high frequency in this cohort (rs1800372:A>G, 9.0%), it was not associated with the risk of developing cancer before the age of 35 years in an extended cohort of 1,420 breast cancer cases. Functional studies of the rs1800372:A>G polymorphic allele reveal that it does not affect p53 transactivation function. Further study of variants or mutations in other cancer susceptibility genes is warranted to refine our understanding of the germline contribution to premenopausal breast cancer susceptibility.