TP53 germline testing and hereditary cancer: how somatic events and clinical criteria affect variant detection rate

Abstract

Background: Germline heterozygous pathogenic variants (PVs) in TP53 cause Li-Fraumeni syndrome (LFS), a condition associated with increased risk of multiple tumor types. As the associated cancer risks were refined over time, clinical criteria also evolved to optimize diagnostic yield. The implementation of multi-gene panel germline testing in different clinical settings has led to the identification of TP53 PV carriers outside the classic LFS-associated cancer phenotypes, leading to a broader cancer phenotypic redefinition and to the renaming of the condition as "heritable TP53-related cancer syndrome" (hTP53rc). Germline TP53 variant interpretation is challenging due to the diverse nature of TP53 PVs, variable penetrance of the syndrome, possible occurrence of TP53 somatic mosaicism, and TP53 involvement in clonal hematopoiesis of indeterminate potential (CHIP). Here we aim to assess the relevance and impact of these issues on the diagnostic routine, and to evaluate the sensitivity of the different LFS clinical criteria to identify hTP53rc.

Methods: TP53 was analyzed in 6161 suspected hereditary cancer non-related patients categorized into three subgroups: (1) 495 patients fulfilling any LFS/Chompret clinical criteria; (2) 2481 patients diagnosed with early-onset breast/colorectal cancer; (3) 3185 patients without clinical criteria suggestive of hTP53rc. Ancillary tests were performed when TP53 PVs were identified in individuals not meeting LFS/Chompret criteria and/or when the variant was identified at low variant allele frequency (VAF).

Results: TP53 PVs were identified in blood DNA of 45 probands. Variant origin was elucidated in 39 of these: 72% patients had a constitutional PV, 10% were mosaics, and 18% had CHIP-associated PVs. Notably, two of the seven CHIP-TP53 PVs identified were detected at high allelic frequencies (VAF > 35%). Twenty-nine percent of germline TP53 PV did not meet any of the LFS clinical criteria. Among the clinical criteria, Chompret 2009 showed the highest sensitivity in our cohort (68% vs. 54% for Chompret 2015), highlighting the relevance of considering lung cancer in the criteria.

Conclusions: Our data supports performing TP53 ancillary testing for the identification of potential mosaicisms and CHIP-associated PVs, particularly in patients not meeting clinical criterial for LFS, irrespective of the VAF, and the application of clinical criteria that include lung cancer diagnosis.

Keywords: TP53; Clonal hematopoiesis; Hereditary cancer; Heritable TP53-related cancer syndrome; Li-Fraumeni syndrome; Mosaicism.

Fig. 1

Patient cohorts and workflow for the evaluation of TP53 variant origin. A Pie chart displaying the distribution of 6161 patients in three different cohorts, based on the fulfillment (or not) of TP53 clinical testing criteria. TP53 mutational rate is indicated by cohort. B Workflow summarizing the assessment of the origin of 45 TP53 pathogenic variants identified in our cohort. Considerations: ¹One biallelic carrier was identified in our cohort (VAF = 100%); ²The identification of an additional heterozygous carrier within the family by cascade testing also confirms germline origin. ³Variant origin could not be assessed for 8 TP53 pathogenic variants. For patients 28 and 43, origin was presumed from genotype/clinical data. Abbreviations: BC: breast cancer; CHIP: clonal hematopoiesis of indeterminate potential; CRC: colorectal cancer; LFS: Li-Fraumeni syndrome; NA: Not assessed; PV: pathogenic variant; VAF: variant allele frequency

Fig. 2

Upset plot illustrating the overlap of TP53 testing criteria met by germline pathogenic variant carriers. Four classification criteria are defined: classic 1988 Li-Fraumeni syndrome, Chompret 2001, Chompret 2009, and Chompret 2015. Horizontal bars represent the total number of patients for each category (set sizes). Vertical bars display the intersection sizes, with dots and connecting lines below indicating the criteria involved in each intersection. Patient IDs for each intersection are labeled above the bars. Abbreviations: LFS: Li-Fraumeni syndrome

Fig. 3

Spectrum of germline and somatic TP53 pathogenic variants identified in our cohort. Variant location is displayed by lollipop structures. Color and shape coding correspond to variant consequence and origin as indicated in the figure key. A copy-number variant encompassing exon 1 (non-coding), found in one index case, is not represented in the scheme. TP53 protein domains are shown in colored boxes with an amino acid numbered scale on top. TP53 transcript: NM_000546.5. Abbreviations: aa: amino acid; CHIP: clonal hematopoiesis of indeterminate potential; DBD: DNA binding domain; TAD: transactivation domain; TD: tetramerization domain

Fig. 4

Age at first tumor diagnosis (panel A) and overall survival (panel B) categorized by TP53 variant effect. Variants are categorized into five groups (see Methods section for further details). Color coding corresponds to the variant effect categories as shown in the figure key. A The x-axis represents the categories of TP53 pathogenic variants, while the y-axis shows the age at first tumor diagnosis. The central line in each box represents the median, and each patient is depicted as an individual point. Statistical significance was determined using the Kruskal–Wallis or the Mann–Whitney test. *p-value < 0.05; **p-value < 0.01; ***p-value < 0.001; ****p-value < 0.0001. B The x-axis represents the overall survival in years, while the y-axis shows the survival rate. Statistical significance was determined using the Mantel-Cox test. Abbreviations: DNE: dominant negative; LOF: loss-of-function