Li-Fraumeni Syndrome (TP53)

Literature and regulatory information updated: 4 October 2026.

Classic Li-Fraumeni syndrome, Li-Fraumeni-like syndrome, the Li-Fraumeni spectrum and heritable TP53-related cancer syndrome (hTP53rc), including the Brazilian founder variant p.Arg337His

0. Initial Reference Table

1. Introduction

Li-Fraumeni syndrome is an autosomal dominant hereditary cancer predisposition syndrome characterised by familial and individual clustering of a broad spectrum of malignancies arising in childhood and young adulthood, with frequent multiple primary tumours in the same individual. In most families, pathogenic germline variants in TP53, a tumour suppressor gene encoding a transcription factor central to cell-cycle and apoptosis regulation in response to DNA damage, cause the syndrome [5]. Breast cancer and sarcomas are the most frequent tumours in absolute terms, but the greatest relative risks compared with the general population concern adrenocortical carcinoma and breast phyllodes tumour, a fibroepithelial neoplasm ranging from benign (most cases) to malignant [5]. Common cancers of the lung, colon, bladder, prostate, cervix and ovary do not show a marked excess in classic families, although recent series have documented increased risk for several of these tumours [5, 10].

The syndrome's history began in 1969, when Frederick Li and Joseph Fraumeni, epidemiologists at the US National Cancer Institute (NCI), published “Soft-tissue sarcomas, breast cancer, and other neoplasms. A familial syndrome?” in the Annals of Internal Medicine [26]. Reviewing records of 648 children with rhabdomyosarcoma, they identified four families—two sibling pairs and two cousin pairs—in which childhood soft-tissue sarcoma coexisted with excess early breast cancer and other neoplasms in parents and other relatives, and proposed a familial syndrome. In 1988, Li, Fraumeni and colleagues described 24 families from the NCI Cancer Family Registry: cancer followed an autosomal dominant pattern in 151 blood relatives, 119 (79%) affected before 45 years, with 50 bone and soft-tissue sarcomas, 28 breast cancers, 14 brain tumours, nine leukaemias and four adrenocortical carcinomas before 45 years; six patients developed radiotherapy-associated second cancers, anticipating a treatment principle that remains central [13]. This work established the classic clinical criteria.

Two groups identified the molecular basis almost simultaneously in 1990. David Malkin and colleagues, in Stephen Friend's group at Massachusetts General Hospital, selected TP53 as a candidate—linkage analysis was impracticable because of the families' rarity and mortality—since the gene was inactivated in sporadic forms of most spectrum tumours, and detected germline variants in all five families studied (Science, 1990) [27]. Srivastava and colleagues described germline transmission of a mutated TP53 allele in an affected family that same year (Nature, 1990) [28]. Recognition that families with incomplete phenotypes could also carry variants subsequently led to the Birch (1994) and Eeles (1995) Li-Fraumeni-like definitions and, in 2001, the Chompret criteria for genetic test selection, revised in 2009 and 2015 [14–18].

In 2001, Ribeiro, Figueiredo, Zambetti and colleagues showed that 35 of 36 children from southern Brazil with adrenocortical carcinoma carried the same germline variant, p.Arg337His (arginine replaced by histidine at codon 337), in families lacking the cancer history that characterises the classic syndrome [29]. This observation opened the study of reduced-penetrance alleles and led, in the state of Paraná, to the first population newborn screening programme for a cancer predisposition variant [8]. Widespread multigene panels and tumour sequencing broadened the concept further: carriers without a suggestive family history, attenuated phenotypes, and mosaic variants or variants acquired through clonal haematopoiesis are now identified, which led ERN GENTURIS to propose, in 2020, the designation heritable TP53-related cancer syndrome (hTP53rc) and the IARC group to define the “Li-Fraumeni spectrum” [2, 3]. Intensive whole-body MRI surveillance, developed in Toronto from 2004, was the first intervention associated with improved survival in these patients [19, 30].

2. Molecular Genetics and Pathophysiology

2.1 The TP53 gene and p53 protein

TP53 lies at 17p13.1 and encodes p53 (393 amino acids), a transcription factor termed the “guardian of the genome”. Its successive functional domains comprise N-terminal transactivation domains, a proline-rich domain, the central DNA-binding domain (containing most pathogenic variants), the tetramerisation domain and a C-terminal regulatory domain [1]. Under basal conditions, the ubiquitin ligase MDM2 keeps p53 levels low by targeting it for proteasomal degradation. Cellular stress—DNA damage, oncogene activation, nutrient deprivation or oxidative stress—activates and stabilises p53, which accumulates, binds as a tetramer to response elements in target-gene promoters, and induces cell-cycle arrest, DNA repair, apoptosis, senescence or differentiation [1]. The “precancerous niche” hypothesis, still under investigation, proposes that p53 also participates in creating a tissue microenvironment conducive to tumour formation, potentially offering therapeutic implications [1].

p53 inactivation is among the most frequent events in sporadic cancer: approximately 50% of all tumours contain somatic TP53 variants, making it one of the most frequently mutated genes in human cancer [1]. In sporadic breast cancer, for example, somatic variants occur in approximately 29% of luminal B tumours, 72% of HER2-enriched tumours and 80% of triple-negative tumours [31]. This high somatic frequency has two practical implications: finding a TP53 variant in a tumour alone does not establish hereditary risk, and tumour sequencing is now a frequent route to suspecting the syndrome [1, 9].

2.2 Disease mechanism: two hits and dominant-negative effects

As with other tumour suppressors, p53 activity is lost through a two-hit process in sporadic tumours. In LFS, the first hit is constitutional; loss of heterozygosity (LOH), through deletion or mutation of the wild-type allele, initiates tumorigenesis [5]. Pathogenic germline variants impair p53 DNA binding constitutively, reducing the transcriptional response to DNA damage [1]. Some missense variants in the DNA-binding domain also have a dominant-negative effect: mutant protein enters tetramers and impairs the normal allele's product even before a second somatic hit, explaining severe phenotypes with a single germline variant [1]. The therapeutic consequence is increased sensitivity to genotoxic exposures: ionising radiation and, to a lesser extent, conventional cytotoxic chemotherapy increase second-cancer risk [1, 2].

Most germline variants are missense; sequencing the entire coding region (exons 2–11) detects approximately 95% of pathogenic TP53 variants [1]. In Birch's series, more than half were transitions at CpG dinucleotides, suggesting spontaneous mutation through deamination of methylated cytosines [17]. The six commonest complete loss-of-function or dominant-negative variants lie in the DNA-binding domain: p.Arg175His, p.Gly245Ser, p.Arg248Gln, p.Arg248Trp, p.Arg273His and p.Arg282Trp [1]. Intragenic deletions, including deletions of the promoter and non-coding exon 1, account for approximately 1% of variants [1].

2.3 Genotype–phenotype correlations

In the large French series of 415 carriers (214 families, 133 distinct alterations), mean age at first tumour was 21.3 years for dominant-negative missense variants, 28.5 years for other loss-of-function variants and 35.8 years for genomic rearrangements; affected children, apart from those with adrenocortical carcinoma, predominantly carried dominant-negative variants [16]. These data support management stratification by variant class. The most severe phenotype involves earlier first cancer, greater breast and other cancer incidence, and a higher probability of meeting classic or Chompret criteria; the German register found significantly lower childhood cancer incidence in partial loss-of-function carriers [1]. Other missense variants confer lower risk and later onset, including p.Pro47Ser, p.Gly334Arg, p.Asp49His, p.Arg181Cys, p.Arg181His and the 3' untranslated-region variant c.*1175A>C; p.Pro152Leu has been described as a reduced-penetrance variant [1]. Suspected low-penetrance variants often receive discordant laboratory classifications: among 140 families, 11% had a TP53 variant with clinically significant disagreement (for example, uncertain significance versus likely pathogenic) [1].

2.4 The p.Arg337His variant (NM_000546.6:c.1010G>A)

p.Arg337His lies in the tetramerisation rather than DNA-binding domain. The mutant tetramerisation domain adopts a near-native conformation but is less stable than the wild type; stability is highly pH sensitive within the physiological range, correlating with histidine 337 protonation. pH-dependent p53 dysfunction was proposed as the molecular basis of adrenocortical carcinoma in Brazilian children [32]. Carriers' tumours show deletion of the wild-type allele and nuclear accumulation of mutant protein, reproducing the pattern of LFS-associated tumours [29]. Early studies suggested independent origins, but detailed haplotype analysis of 12 unrelated carriers using 29 tagging SNPs (single-nucleotide polymorphisms) showed a shared haplotype, with an independent-occurrence probability of 3.1 × 10⁻⁹, demonstrating a founder effect [33]. Families are distributed along the road historically used by traders of Portuguese origin in the eighteenth and nineteenth centuries, and relatively low penetrance before 30 years may have helped maintain the variant in a large, open population [33].

2.5 Genetic modifiers

Modifiers contribute to variable LFS penetrance [1]. The TP53 p.Arg72Pro polymorphism (c.215G>C; rs1042522) increases MDM2 affinity, enhancing p53 degradation and advancing first-cancer onset. The MDM2 promoter variant rs2279744 (c.14+309T>G) increases MDM2 expression with the same effect. The MIR605 variant n.74T>C, affecting the p53–MDM2 circuit, was associated with a ten-year reduction in mean age at first tumour. The 16-base-pair duplication in TP53 intron 3 (PIN3; rs17878362) appears protective, with later first cancer. Telomere shortening across generations has been associated with earlier tumour development. WNT-pathway variants that may reduce risk, and inherited epimutations in ASXL1, ETV6 and LEF1 that may increase it, are also under investigation [1].

For p.Arg337His, the principal modifier is XAF1 p.Glu134Ter (rs146752602). Whole-genome sequencing identified it in a carrier subgroup; haplotypes were assessed in 203 cancer patients, 582 relatives and 42,438 newborns, and the compound mutant haplotype was enriched among patients, conferring sarcoma risk (p = 0.003) and subsequent-neoplasm risk (p = 0.006). Functionally, wild-type XAF1 enhances transactivation by wild-type p53 and hypomorphic variants, whereas XAF1 p.Glu134Ter has attenuated activity [34]. Double heterozygotes therefore have a more aggressive phenotype, with genetic counselling implications [1, 34].

2.6 Mosaicism, de novo variants and clonal haematopoiesis

The estimated frequency of de novo germline variants is 7–20% [1]. In a series of 328 patients with germline variants, 40 variants arose de novo, without a parental-age effect; Sanger sequencing revealed two mosaics (a child with adrenocortical carcinoma and an unaffected father of a child with medulloblastoma). Next-generation sequencing (NGS) at greater than 500× depth in 108 patients suggestive of LFS without a detectable variant identified six further mosaics: two of 49 children with adrenocortical carcinoma, two of 21 with choroid plexus tumours, one of 31 women with breast cancer before 31 years, and a patient with osteosarcoma at 12 years followed by breast carcinoma and sarcoma at 35 [6]. Parental gonadal mosaicism and sibling recurrence have also been reported [1].

The central interpretative problem is distinguishing constitutional variants from variants acquired in blood. Low-level somatic mosaicism for pathogenic TP53 variants in leucocytes is common and suggests an acquired postzygotic variant associated with clonal haematopoiesis of indeterminate potential (CHIP), a haematological neoplasm or preneoplasm, or circulating tumour cells [1, 35]. In a certified laboratory analysing 114,630 multigene panels and 1,454 targeted TP53 tests, 20% of 353 positive results had abnormal NGS metrics (allele fraction below 25% or more than two alleles); aberrant clonal expansion was confirmed in 91% of cases with additional material, mostly through clonal haematopoiesis. These individuals were older (50 versus 33.7 years), and only four met Chompret criteria [36]. At another large laboratory, approximately 40% of pathogenic TP53 variants detected in pan-cancer panels had allele fractions of 10–30%; even among apparently heterozygous carriers (30–70% allele fraction), the variant was absent from fibroblasts in 32.6% (15/46) of samples assessed [37]. Allele fraction alone cannot resolve the distinction.

Individuals whose pathogenic TP53 variant is absent from non-haematopoietic tissues and confirmed to arise from CHIP or another haematopoietic clonal expansion do not have LFS and should not be managed as such; conversely, constitutional or early postzygotic mosaicism increases spectrum-cancer risk [1, 35]. Recommended further assessment includes TP53 testing in non-haematopoietic tissue, such as cultured skin fibroblasts, and testing the proband's parents and offspring, following the NCCN approach table [1, 24].

3. Clinical Spectrum and Manifestations by Organ System

3.1 Overview and core tumours

Five tumour types account for most LFS cancers—adrenocortical carcinoma, breast cancer, central nervous system tumours, osteosarcoma and soft-tissue sarcomas—with additional leukaemia, colorectal, gastric, lung, pancreatic and prostate cancers, melanoma and paediatric head-and-neck cancers [1]. The Oxford Desk Reference grades associations as strong for breast cancer, soft-tissue sarcomas (e.g. leiomyosarcoma), osteosarcoma, brain tumours, adrenocortical carcinoma, Wilms tumour and phyllodes tumour; moderate for pancreatic and gastric cancer; and weak for leukaemias and neuroblastoma [5]. At least six families with Wilms tumour and a TP53 variant have been reported, but this is not a cardinal tumour and its absolute risk is low [38]. The ESMO handbook adds early-onset gastric cancer to the spectrum and includes LFS among hereditary colorectal cancer syndromes as a cause of early-onset colorectal cancer within pan-cancer predisposition [9, 39].

In the French series, 322 affected carriers developed 552 tumours; 43% had multiple neoplasms, mean age at first tumour was 24.9 years, and 41% had developed a tumour by 18 years [16]. Childhood tumours were predominantly osteosarcomas (30%), adrenocortical carcinomas (27%), CNS tumours (26%) and soft-tissue sarcomas (23%); in adulthood, breast cancer occurred in 79% of women and soft-tissue sarcomas in 27% of patients [16]. Mean age at first cancer was 17 years in men and, in women, 28 years including breast cancer and 13 years excluding it [1].

3.2 Age distribution of tumours

Analysis of the IARC (International Agency for Research on Cancer) database identified distinct age phases: the paediatric phase (0–15 years), accounting for 22% of all cancers, predominantly includes adrenocortical carcinoma, choroid plexus carcinoma, rhabdomyosarcoma and medulloblastoma; between 16 and 50 years (51% of cancers), breast cancer, osteosarcoma, soft-tissue sarcomas, leukaemia, astrocytoma, glioblastoma, colorectal and lung cancers predominate; between 51 and 80 years (27%), pancreatic and prostate cancers are most frequent [1, 11].

Source: Amadou and colleagues, 2018, and GeneReviews [1, 11].

3.3 Adrenal gland

Adrenocortical carcinoma occurs in 6–13% of carriers, mainly in childhood, although adults may be affected, usually before 40 years [1]. The French series found TP53 variants in 45% of children with adrenocortical carcinoma and no other suggestive features [16]. Typical paediatric presentation reflects androgen and/or glucocorticoid hypersecretion: virilisation (pubarche, clitoromegaly, penile enlargement, acne and accelerated growth), Cushingoid appearance, sudden weight gain or loss, and hypertension—supporting blood pressure measurement and pubertal assessment at every surveillance visit [1, 2]. Among p.Arg337His carriers, adrenocortical carcinoma accounted for 55% of childhood cancers (median age two years) and 23% of adult cancers in a Brazilian series [1, 40].

3.4 Breast

Breast cancer risk in women with LFS is 80–90%, exceeding that associated with BRCA1 and BRCA2 [1]; the ESMO handbook cites a cumulative risk of 40–79% and notes that pathogenic TP53 missense variants confer risks similar to truncating variants, whereas rare missense variants with limited functional effects have only a modest association [31]. A maximum-likelihood study of 146 families (4,028 individuals) estimated cumulative breast cancer risk by 50 years at 63.3% [10]; the NCI cohort found cumulative incidence by 70 years of 54% [41]. Almost all breast cancers occur before menopause; they are more commonly ductal, oestrogen- and progesterone-receptor positive, and HER2 amplified (human epidermal growth factor receptor 2) [1]. The TP53 variant detection rate among women with breast cancer before 31 years and no other suggestive features was 6% [16]. Breast-conserving surgery is associated with higher ipsilateral cancer rates, greater contralateral cancer incidence and shorter recurrence-free survival [1]. Malignant phyllodes tumour is also associated with the syndrome [1, 5]. Men do not appear to have increased breast cancer risk, although cases exist [1]; risk depends on sex assigned at birth and may be influenced by gender-affirming treatments [1].

3.5 Central nervous system

CNS tumours represent 9–14% of LFS cancers; cumulative incidence by 70 years was 6% in women and 19% in men [1, 41]. Onset is bimodal, with childhood and adult peaks, often before 40 years (median 16 years) [1]. Glioblastomas and astrocytomas are commonest, but ependymomas, choroid plexus carcinomas, supratentorial primitive neuroectodermal tumours and medulloblastomas have been reported; medulloblastomas more frequently belong to the sonic hedgehog (SHH)-activated subtype [1]. The TP53 variant detection rate in children with choroid plexus carcinoma was 42% [16].

3.6 Bone and soft tissues

Osteosarcomas account for up to 16% of cancers, peaking in childhood and adolescence, with cumulative incidence by 70 years of 5% in women and 11% in men [1, 41]. Soft-tissue sarcomas, particularly rhabdomyosarcoma (with a specific association with the anaplastic embryonal subtype), are the commonest childhood LFS cancers and account for 17–27% of all cancers, with cumulative incidence of 15% in women and 22% in men [1, 2, 41]. Childhood mandibular or maxillary osteosarcoma is a sentinel tumour [2].

3.7 Blood

Leukaemia occurs in approximately 4% of carriers, with peak risk between 0 and 15 years [1]. Low-hypodiploid acute lymphoblastic leukaemia (ALL; 32–39 chromosomes in blasts) is the most characteristic haematological malignancy; B-cell ALL is also frequent. Acute myeloid leukaemia and myelodysplastic syndrome may be primary or treatment related, and therapy-related chronic myeloid leukaemia has been reported in adults [1]. Current myeloid neoplasm classification includes a germline-predisposition category featuring TP53 alongside RUNX1, GATA2, CEBPA and DDX41; germline predisposition influences prognosis and treatment response [42]. LFS-associated leukaemias may have shorter responses and poorer outcomes [1]; lymphoma is also increased [1, 42].

3.8 Gastrointestinal tract, lung, skin and other organs

Gastrointestinal cancers account for approximately 5.1% of LFS neoplasms (2.8% colorectal and 1.3% oesophageal or gastric), arise earlier than in the general population, and include gastro-oesophageal junction cancers; gastric cancer risk is higher in individuals of Japanese and other Asian ancestry, probably owing to dietary factors and higher Helicobacter pylori infection prevalence [1]. Pancreatic cancer risk is also increased [1]. Early-onset lung cancer, with increased somatic EGFR variant frequency, melanoma, lymphoma and prostate cancer form part of the expanded spectrum, and risks of kidney, laryngeal, non-melanoma skin, ovarian, testicular and thyroid cancers may be increased [1]. The 2024 maximum-likelihood study identified significantly increased lifetime risks of colorectal, gastric, lung, pancreatic and ovarian cancers; any-cancer risk by 50 years was 92.4% (95% CI: 82.2–98.3) in women and 59.7% (95% CI: 39.9–81.3) in men [10].

3.9 Pregnancy and trophoblastic disease

Gestational choriocarcinoma and other forms of gestational trophoblastic disease have been reported in women with LFS and in women without LFS carrying a fetus with a paternally inherited TP53 variant [1].

3.10 Multiple primary tumours and treatment-induced neoplasms

Second-cancer risk reaches 50%, with a median interval of 10 years after the first diagnosis; approximately 49% of cancer patients in the NCI cohort developed at least one additional cancer, and age-specific second-cancer risk resembled first-cancer risk [1, 41]. Radiotherapy and chemotherapy for LFS-associated cancer may increase second-neoplasm risk, including sarcomas, angiossarcomas and basal cell carcinomas within irradiated fields, and therapy-related leukaemias or myelodysplasia [1, 2, 43].

3.11 Penetrance by sex and age

The sex difference principally reflects breast cancer; male risk concentrates in childhood and later adulthood [5, 41]. For classic LFS, cancer risk by 70 years is 80% [11].

4. Clinical Recognition and Red Flags

Recognition principally depends on personal and family history, as physical examination is usually uninformative [5]. History should cover three generations, specifically addressing all childhood and adult cancers in first-, second- and third-degree relatives, age at diagnosis and multiple primary tumours; wherever possible, full names, dates of birth and the treating hospital should be obtained for documentary confirmation [1, 5]. Family history may be incorrect or incomplete—cancer may be avoided as a subject, or a death may have caused estrangement—and collecting it is emotionally demanding given the numerous cancers and deaths among close relatives [1].

Red flags independent of family history include adrenocortical carcinoma at any age, choroid plexus carcinoma, anaplastic embryonal rhabdomyosarcoma, breast cancer before 31 years, and multiple spectrum primary tumours with the first before 46 years [1, 2]. Additional childhood and adolescent red flags include hypodiploid ALL, otherwise unexplained SHH-activated medulloblastoma, and mandibular or maxillary osteosarcoma, provided another predisposition syndrome, such as nevoid basal cell carcinoma syndrome (Gorlin syndrome), does not explain them [1, 2]. Other red flags include HER2-positive breast cancer before 40 years, a second primary in the radiotherapy field of a core tumour diagnosed before 46 years, and any childhood cancer in an individual of southern or southeastern Brazilian ancestry [1, 2]. A pathogenic tumour TP53 variant in an adult with a suggestive family history, or a child with cancer regardless of family history, warrants germline testing [1]; ESMO recommends considering germline testing after tumour-only detection, particularly when Chompret criteria are met or cancer occurs before 30 years [9]. European recommendations for germline follow-up of somatic panels restrict investigation of tumour-detected TP53 to tumours arising before 30 years, alongside RB1, APC and NF1 [44].

A child with adrenocortical carcinoma may initially present with endocrine features; virilisation, precocious puberty, Cushingoid appearance or hypertension in a young child warrants abdominal ultrasonography and hormonal assessment [1, 2]. In southern Brazil, median diagnosis age outside newborn screening was 5.2 years, with 15 weeks from symptoms to diagnosis, compared with 1.9 years and two weeks among children followed from newborn screening [45].

5. Diagnosis: Criteria, Tests and Algorithms

5.1 Classic clinical criteria

Clinical LFS diagnosis can be established in a proband meeting all three classic Li and colleagues criteria (1988): (1) sarcoma diagnosed before 45 years in the proband; (2) a first-degree relative with any cancer before 45 years; (3) another first- or second-degree relative in the same lineage with any cancer before 45 years or sarcoma at any age [1, 5, 13]. In families meeting these criteria, TP53 variant detection is approximately 75% in the Oxford Desk Reference and 90% in GeneReviews, reflecting newer techniques; detection is low in breast cancer/sarcoma families not meeting classic criteria [1, 5].

5.2 Li-Fraumeni-like criteria

Two Li-Fraumeni-like (LFL) definitions address incomplete phenotypes. Under Birch's definition (1994), the proband has childhood cancer or sarcoma, brain tumour or adrenocortical carcinoma before 45 years; a first- or second-degree relative has a typical LFS cancer (sarcoma, breast cancer, brain tumour, leukaemia or adrenocortical carcinoma) at any age; and another first- or second-degree relative in the same lineage has any cancer before 60 years [17]. Eeles' broader definition (1995) requires only two first- or second-degree relatives with spectrum neoplasms at any age [18]. In Birch's series, germline variants characterised families with children with rhabdomyosarcoma and/or adrenocortical carcinoma, being detected in six of nine such families [17].

5.3 Chompret criteria (2015 modified version)

The Chompret criteria were proposed in 2001 and revised in 2009 and 2015 to select patients for genetic testing [9, 14–16]. Under the modified version adopted by GeneReviews and ERN GENTURIS, any patient meeting one of the following criteria should be tested [1, 2]:

A pathogenic germline variant is detected in approximately 90% of individuals meeting classic criteria and 50% meeting modified Chompret criteria [1]. The original 2001 Chompret criteria demonstrated the sensitivity and predictive value of different selection rules in a French series [14].

5.4 European testing recommendations (ERN GENTURIS, 2020)

The ERN GENTURIS guideline extends Chompret criteria through numbered recommendations [2]. Testing is recommended for patients meeting modified Chompret criteria (recommendation 1); children and adolescents with hypodiploid ALL, otherwise unexplained SHH medulloblastoma or mandibular/maxillary osteosarcoma (recommendation 2); patients with a second primary in the radiotherapy field of a first core tumour arising before 46 years (recommendation 3); and children with any cancer from southern or southeastern Brazilian families, targeting p.Arg337His (recommendation 5). Women older than 46 years with breast cancer and no personal or family history meeting Chompret criteria should not undergo TP53 testing, while any isolated breast cancer patient lacking Chompret criteria in whom a variant is identified requires specialist multidisciplinary discussion (recommendation 4). Wherever possible, testing should precede treatment so that carriers can avoid radiotherapy and genotoxic chemotherapy and receive preferential surgical management [2].

The Japanese guideline reports pathogenic germline TP53 variants in 1.6% (median; range 0–6.7%) of children with cancer and 0.2% of adults with cancer [23].

5.5 Molecular diagnosis and techniques

Molecular diagnosis requires a heterozygous germline pathogenic or likely pathogenic TP53 variant; ACMG/AMP (American College of Medical Genetics and Genomics/Association for Molecular Pathology) guidelines regard both classes as diagnostic and suitable for guiding clinical decisions, whereas a variant of uncertain significance neither confirms nor excludes diagnosis [1]. Interpretation should follow ClinGen TP53 expert-panel specifications (version 2.3.0 cited in GeneReviews; the current version in ClinGen's CSpec registry is 2.4, updated 20 November 2025 [46]) and integrate epidemiological, phenotypic, bioinformatic and functional data [1, 2]. Testing may target TP53 or use a multigene panel; sequencing detects approximately 89% of variants in clinical series, and deletion/duplication analysis (MLPA, multiplex ligation-dependent probe amplification, quantitative PCR or targeted microarray) adds approximately 1% [1]. For southern or southeastern Brazilian ancestry, initial targeted p.Arg337His testing may be considered [1, 2].

DNA should come from tissue without tumour cells—leucocytes (except in leukaemia), saliva or skin—since TP53 missense variants are frequent in tumour tissue [1]. Tumour LOH analysis, useful in other syndromes, is uninformative in LFS and is not recommended [1].

5.6 Algorithm for a TP53 variant detected in blood

Sources: GeneReviews, Batalini and colleagues, Weitzel and colleagues, and Coffee and colleagues [1, 35–37].

5.7 Classification of the Li-Fraumeni spectrum

Analysis of the IARC germline dataset (3,034 individuals from 1,282 families) proposed four categories: (1) phenotypic LFS, where a family meets clinical criteria without a pathogenic TP53 variant; (2) LFS, where a pathogenic or likely pathogenic variant occurs in an individual or family meeting testing criteria; (3) attenuated LFS, where a variant occurs in an individual or family with cancer but without these criteria; (4) incidental LFS, where a variant is identified in an individual or family without a cancer history [3]. Carriers meeting criteria had more early adrenal (6.5% versus 0), brain, connective-tissue and bone tumours; those not meeting criteria had more breast cancers (38.2% versus 27.6%), 45% after 45 years, and 40.4% of their variants were classified as uncertain, conflicting, benign or unknown [3]. Attenuated classification may change as personal and family history evolves and currently does not alter recommended surveillance [1].

5.8 Initial assessment after diagnosis

Recommended assessment includes a comprehensive physical examination with heightened suspicion (blood pressure, neurological examination, growth, sudden weight change, Cushingoid appearance and childhood virilisation), non-contrast whole-body MRI, contrast-enhanced brain MRI, clinical breast examination and breast MRI from 20 years, upper gastrointestinal endoscopy and colonoscopy from 25 years, dermatological examination from 18 years, abdominal and pelvic ultrasonography, counselling by cancer genetics professionals, and assessment of family social and psychological support needs [1]. Where family history is present, each investigation should start at the specified age or 5–10 years before the family's earliest cancer, whichever is earlier [1].

5.9 Genetic test limitations and report content

Sequencing the full TP53 coding region (exons 2–11) and exon–intron junctions, combined with deletion/duplication analysis by MLPA or an equivalent method, is standard; promoter and non-coding exon 1 deletions are detected only if copy-number analysis covers these regions, while most panels omit deep intronic or regulatory variants [1]. Mosaicism sensitivity depends on read depth: only resequencing above 500× depth revealed mosaics missed by Sanger sequencing in the French series [6]. Allele fraction cannot reliably distinguish constitutional from acquired variants: in a Spanish series of 6,161 probands tested, among 39 carriers with established origin, 72% had constitutional variants, 10% mosaicism and 18% clonal haematopoiesis; two of seven clonal-haematopoiesis variants had allele fractions above 35%. Of constitutional carriers, 29% met no clinical criteria, and 2009 Chompret criteria, which include lung cancer, were more sensitive than the 2015 criteria (68% versus 54%) [47]. Negative blood testing does not exclude the syndrome in a classically affected family (phenotypic LFS, section 6), and tumour-only or circulating cell-free DNA sequencing does not replace germline testing [1, 9].

The report should state the gene; versioned reference transcript NM_000546.6 (protein NP_000537.3), confirmed as MANE Select in NCBI; normalised DNA and protein HGVS nomenclature; zygosity and allele fraction; tissue tested; ACMG/AMP classification under ClinGen TP53 expert-panel specifications; and technical limitations (coverage, copy number and mosaicism) [1, 46]. Historical codon notation (R337H, R175H, R248W) is common in the literature; text and reports should use normalised nomenclature, retaining historical names parenthetically where helpful, for example NM_000546.6:c.1010G>A, p.(Arg337His) (R337H), recorded in ClinVar (VCV000012379) as pathogenic, likely pathogenic and pathogenic with reduced penetrance, without conflicts [48]. This variant illustrates a limitation of current rules: specifications calibrated for high-penetrance variants inadequately classify atypical-penetrance variants that are causal but produce attenuated phenotypes, prompting a 2026 proposal for dedicated classification specifications and management recommendations [49]. A variant of uncertain significance does not become pathogenic because a carrier develops a spectrum cancer, responds to or experiences toxicity from treatment, or shows partial segregation in a small family; segregation and functional studies remain research, and management follows personal and family history until formal reclassification [1, 3].

6. Differential Diagnosis

Differential diagnosis includes other predisposition syndromes sharing spectrum tumours and laboratory findings mimicking LFS. The hereditary syndromes with phenotypic overlap are tabulated below [1].

AD: autosomal dominant; AR: autosomal recessive; ER/PR: oestrogen/progesterone receptors. Sources: GeneReviews (differential diagnosis table 2 and suggestive findings) [1].

In oncology practice, the commonest differential is now laboratory based: a TP53 variant detected on a germline panel may reflect clonal haematopoiesis, somatic mosaicism, an underlying haematological neoplasm or circulating tumour cells; only additional assessment distinguishes these from LFS [1, 35, 36]. Conversely, families meeting classic criteria without an identified variant (phenotypic LFS) remain at high risk: all at-risk relatives require counselling on potentially increased risk, surveillance and risk reduction options, and storage of proband DNA for future studies should be considered [1, 3]. For early breast cancer without other features, BRCA1, BRCA2 and PALB2 are more frequent causes than TP53; ESMO places TP53-associated breast cancer risk (40–79%) alongside BRCA1 (72%), BRCA2 (69%), PTEN (25–85%) and STK11 (45–50%) [31].

7. Treatment and Risk Management Strategies

7.1 General principles

No targeted treatment is approved for LFS. Management combines intensive lifelong early-detection surveillance, adaptation of cancer treatment to reduce genotoxic exposure, and surgical or behavioural risk reduction [1, 2, 23]. The Oxford Desk Reference emphasises the complexity of surveillance and the need for a unit experienced in the syndrome [5]. ClinGen rates clinical actionability as strong in both adults and children, based on cancer surveillance and radiotherapy avoidance [22].

7.2 The Toronto protocol and surveillance evidence

The Toronto protocol began on 1 January 2004 at three tertiary centres in Canada and the United States, combining physical examination, frequent biochemical and imaging investigations (whole-body, brain and breast MRI, mammography, abdominal and pelvic ultrasonography) and colonoscopy [19, 30]. In the 2016 update, 59 (66%) of 89 carriers from 39 families underwent surveillance for a median 32 months; 40 asymptomatic tumours were detected in 19 (32%) surveilled individuals, with two false negatives (interval cancers) and two false positives. Among 49 initially declining surveillance, 61 symptomatic tumours were diagnosed in 43 patients (88%). Of surveilled patients developing cancer, 16 of 19 (84%) were alive, compared with 21 of 43 (49%) without surveillance (p = 0.012); 5-year overall survival was 88.8% (95% CI: 78.7–100) with surveillance and 59.6% (95% CI: 47.2–75.2) without it (p = 0.0132) [19]. The authors concluded that sustained adherence was feasible and early detection associated with improved long-term survival, acknowledging the non-randomised observational design [19].

The LiFE consortium meta-analysis of 578 carriers from 13 cohorts in six countries (mean age 33.2 years) assessed baseline whole-body MRI: 225 lesions requiring follow-up were detected in 173 participants; 42 cancers were identified in 39 individuals, 35 localised and treated with curative intent, giving an estimated detection rate for new localised primary cancers of 7% (95% CI: 5–9%) [50]. Whole-body MRI's benefits (greater detection of diverse tumour types) must be balanced against cost, limited access, frequent false positives—particularly at baseline—and sedation requirements in young children [1]. The 2017 Oxford Desk Reference still regarded whole-body MRI as investigational, based on early British SIGNIFY findings [5]; subsequent guideline adoption reflects accumulating evidence [1, 2, 21]. Studies published in 2026 added data on successive rounds and surveillance burden. In the prospective observational British SIGNIFIED study, 54 carriers underwent baseline whole-body MRI and 50 repeated it 12 months later: baseline scanning led to biopsy or excision in 16.7% and four cancer diagnoses (7.4%), while the second scan identified six cancers (12%). Five interval cancers occurred outside the imaging field (upper-limb sarcoma, breast carcinoma) or became clinically apparent between scans; seven of ten detected cancers were early stage, and investigations for incidental findings decreased in the second round [51]. In the Swedish paediatric SWEP53 study (37 children with a confirmed variant or 50% risk, quarterly clinical examination, abdominal ultrasonography and urinary steroid profiling), adherence ranged from 77% to 97%, one of two cancers was surveillance detected, and false positives occurred at 12% per person-year, with 35% of children experiencing at least one [52].

7.3 Recommended surveillance schedules

GeneReviews adopts a modified Toronto protocol based on 2017 AACR (American Association for Cancer Research) recommendations and the Japanese guideline, acknowledging small differences across guidelines (AACR, NICE) because definitive effectiveness data remain unavailable [1, 20, 23].

Source: GeneReviews, table 4 [1]. NCCN suggests pancreatic surveillance only with at least one first- or second-degree relative with exocrine pancreatic cancer, whereas other authors consider it for all carriers [1, 24]; ESMO notes that the September 2024 NCCN update recommended considering pancreatic screening in high-risk individuals with specific germline variants [9].

The European ERN GENTURIS guideline is more selective and grades evidence [2]:

The European guideline defines a high-risk variant by childhood cancer in the index case, childhood cancers in the family, previous detection in other childhood-cancer families, or a dominant-negative missense variant [2]. ESMO summarises surveillance as annual diffusion-weighted whole-body MRI, upper and lower gastrointestinal endoscopy every 2–5 years from 25 years (or 5 years before the earliest familial colorectal or gastric cancer), annual breast MRI from 20–29 years with consideration of bilateral mastectomy, PSA from 40 years and annual skin surveillance [9].

The AACR working group's 2025 update, arising from the 2023 Childhood Cancer Predisposition Workshop, retains Toronto as the lifelong surveillance standard, proposes age-dependent targeting of selected modalities, and details adult surveillance, stressing preparation of patients and clinicians for transition from childhood to young-adult surveillance [21]. UK breast surveillance under NICE CG164, reproduced in the Oxford Desk Reference, offers annual MRI to TP53 carriers aged 20–49 years (and untested women with greater than 30% carrier probability) and advises against mammography at any age because of ionising radiation concerns; women with an initial probability of at least 30% who reach 50 years without cancer no longer qualify for MRI [53, 54]. This difference from the US schedule (mammography alternating with MRI from 30 years) requires discussion with each patient [1, 54].

7.4 Surveillance for p.Arg337His carriers

In Paraná, carriers identified through newborn screening received periodic clinical, laboratory and ultrasound assessments; imaging-detected tumours were resected [8]. All 11 adrenocortical carcinomas diagnosed in children followed from newborn screening were stage I and all children were alive, with median tumour weight 21 g versus 210 g in unscreened children, among whom five of 16 tumours were stage IV [45]. Estimated screening and surveillance cost per life-year saved was US$623 per patient [45]. A cohort of 708 adults (303 with p.Arg337His) showed lower cumulative cancer risk than for other variants (54% versus 78%), with a marked sex difference (65% in women and 30% in men) and greater female second-cancer risk; commonest cancers were breast cancer and soft-tissue sarcomas in women, and soft-tissue sarcomas and prostate cancer in men, supporting personalised surveillance [12]. Among 55 families of children with adrenocortical tumours, cumulative cancer risk by 45 years was 21% in carrier relatives and 2% in non-carriers, with breast and gastric cancers commonest [55].

7.5 Cancer treatment in LFS

Cancer treatment generally follows standard protocols with adaptations [1]. Bilateral mastectomy is frequently recommended instead of lumpectomy for breast cancer, reducing second primary breast cancer risk and avoiding adjuvant radiotherapy [1]. In an Institut Gustave Roussy series, six of eight patients with breast cancer as their first tumour received postoperative radiotherapy; these patients developed three ipsilateral recurrences, three contralateral cancers, two radiation-induced cancers and three new primaries (including thyroid carcinoma within the irradiated field), versus one event (contralateral cancer) among non-irradiated patients [43].

For other cancers, radiotherapy should be avoided wherever possible because radiation-induced tumours and leukaemias have been reported [1, 23]. Evidence is not uniform: in a multi-institutional register of 40 patients, 7 of 14 (50%) receiving curative radiotherapy developed a subsequent neoplasm (median 3.5 years), versus 46% of non-irradiated patients (median 5 years); four of seven in-field neoplasms shared the primary tumour's histology, suggesting local recurrence rather than new primaries [56]. Radiotherapy should therefore remain an option when clinically indicated, followed by close surveillance of the irradiated field [1]. Conventional cytotoxic chemotherapy, particularly alkylating agents, may increase second-cancer risk, but most experts recommend prioritising treatment efficacy over late-effect concerns after weighing risks and benefits; radiotherapy or alkylators may be unavoidable for tumours such as rhabdomyosarcoma [1, 23]. Hypodiploid ALL and other leukaemias follow risk-adapted protocols, acknowledging reported poorer outcomes [1].

7.6 Primary prevention and risk reduction

Risk-reducing bilateral mastectomy is an option for women with LFS [1, 9]. Colonoscopy provides both surveillance and primary colorectal cancer prevention through polypectomy, while dermatological surveillance enables removal of premalignant lesions [1]. Diagnostic and therapeutic radiation should be minimised, preferring ultrasonography or MRI to CT when clinically equivalent, and known carcinogens avoided, including unprotected sun exposure, tobacco, occupational exposures and excessive alcohol, since carcinogenic exposures and the germline variant may have cumulative effects [1].

7.7 Investigational treatments

No chemoprevention agent is approved for LFS [57]. Metformin is the most advanced candidate. In the NCI pilot study (NCT01981525), 26 non-diabetic adults (20 women, 6 men) received metformin for 14 weeks, starting at 500 mg once daily and increasing by 500 mg every 2 weeks to 2 g/day; common adverse effects were diarrhoea (50.0%), nausea (46.2%), dyspepsia (19%) and headache (30%), all grade 1 or 2, with no lactic acidosis or fasting glucose changes. Metformin suppressed hepatic mitochondrial function and reduced IGF-1 (insulin-like growth factor 1) and IGFBP3 [57, 58]. The British phase II, randomised, open-label MILI trial (ISRCTN16699730; EudraCT/CTIS 2022-000165-41) allocates 224 adults aged at least 16 years 1:1 to immediate-release metformin (500 mg once daily, increased by 500 mg every 14 days to 2 g/day, 1,000 mg twice daily, for up to 5 years) plus annual MRI or annual MRI alone; the primary endpoint is 5-year cancer-free survival, and the protocol mentions a parallel NCI study [57]. As of October 2026, MILI results were unpublished, and the only ClinicalTrials.gov-registered LFS metformin study was the completed NCT01981525 pilot [58, 59]. Metformin (Glucophage® and generics) is approved for type 2 diabetes mellitus; its use in LFS is experimental and should be confined to clinical trials. Protocol doses (up to 2 g per day) do not exceed the label maxima for type 2 diabetes in adults: the FDA label specifies starting at 500 mg twice daily or 850 mg once daily, increases of 500 mg per week or 850 mg every 2 weeks, and a maximum of 2,550 mg per day in divided doses (from 10 years of age, maximum 2,000 mg per day), and the Portuguese SmPC for Glucophage® specifies starting at 500 mg or 850 mg two or three times daily, adjustment after 10 to 15 days and a maximum of 3 g per day in three doses; the trials' escalation scheme and the indication are outside the label conditions [60, 61].

Liquid biopsy is another research approach: a multimodal cell-free DNA assay (targeted gene panel, low-coverage whole-genome sequencing and methylated DNA immunoprecipitation) applied to a longitudinal cohort of 89 patients detected cancer-associated signals before conventional surveillance diagnosis, with positive predictive value 67.6% and negative predictive value 96.5%; surveillance adoption requires further clinical validation [62].

7.8 Psychosocial support

Carriers recognise surveillance benefits, but intensive protocols and personal and familial cancer experiences warrant mental health support and dedicated resources [1]. Adolescents and young adults describe the burden of one or more cancer diagnoses, anticipatory fear and family communication difficulties [1]. Patient associations (LFSA, Living LFS) and research consortia (LiFE, LiFT Up study) provide useful resources [1]. ERN GENTURIS links European referral centres for genetic tumour risk syndromes, with patient participation through European associations, and children with cancer should receive care at ERN PaedCan-linked centres [2, 25].

7.9 Evidence hierarchy, regulatory status and access

Four levels of evidence are distinguished: peer-reviewed publications (trials, cohorts, meta-analyses and guidelines); preliminary results, protocols and trial registrations without published principal findings; regulatory opinions; and authorisation decisions. Marketing authorisation also does not ensure reimbursement or access in every country. No medicine has an FDA-, European Commission-, INFARMED- or ANVISA-approved LFS indication, and no pending authorisation application was identified; ClinicalTrials.gov, consulted in October 2026, principally listed imaging surveillance, liquid biopsy, registries and biobanks, a phase 1/2 nicotinamide riboside study completed in 2020, and a phase 2 arsenic trioxide trial for TP53-mutated childhood cancers that is not an LFS prevention trial [59].

Distinguishing evidence extrapolation from off-label use is particularly useful. Metformin for LFS is off-label at every age and in every jurisdiction because its label concerns diabetes; evidence also comes from non-diabetic adults (NCI pilot) and participants aged at least 16 years (MILI), so childhood, pregnancy or reduced-penetrance-carrier use would be both off-label and outside the studied population [57, 58]. Radiotherapy avoidance recommendations principally derive from high-penetrance-variant series; applying them equally rigorously to p.Arg337His carriers is an extrapolation requiring explicit acknowledgement and consideration of the tumour being treated [1, 12, 56]. Regarding access, Paraná p.Arg337His newborn screening operated as a research programme and was not part of mandatory state newborn screening in 2021 [45]; annual whole-body MRI depends on each health system's availability, and no published access data were identified for Portugal or Brazil.

7.10 Studies supporting surveillance and prevention

None is a randomised trial with published mortality results. The Toronto survival benefit derives from a non-randomised comparison between individuals accepting and declining surveillance, subject to selection, lead-time and overdiagnosis biases. Detection rates cannot be compared directly because participant ages, proportions previously treated, round type (baseline measuring prevalence versus subsequent rounds measuring incidence) and detected-cancer definitions differ; detection rates should accompany interval-cancer proportions (5 in 54 in SIGNIFIED) and false-positive rates (12% per person-year in Swedish children). Liquid biopsy predictive values depend on the studied cohort's cancer prevalence and cannot transfer to other populations. Paraná findings concern a reduced-penetrance variant with a childhood spectrum centred on adrenocortical carcinoma and cannot be extrapolated to high-penetrance variants or other countries. LIFSCREEN, the only randomised whole-body MRI trial, has no principal results published in the indexed literature consulted.

8. Reproductive Options and Family Counselling

LFS follows autosomal dominant inheritance, with a 50% chance of variant transmission to each offspring [1, 5]. Most probands inherited their variant; if neither parent's leucocytes carry it and biological parentage is confirmed, a de novo variant or parental gonadal mosaicism should be considered, and siblings tested because parental leucocyte testing misses some somatic mosaicism and gonad-restricted variants [1]. A parent with a relevant personal or family cancer history should be tested first; otherwise, both may be tested simultaneously [1]. If the variant arose postzygotically in the proband, other relatives have no increased risk [1]. When diagnosis is clinical without an identified variant, offspring are presumed at increased risk and a relative's “negative” result is uninformative [1].

Once the familial variant is identified, testing should be offered to all first-degree relatives, including children, and more distant relatives (aunts, uncles and cousins); for a child proband, cascade testing begins with siblings and parents [1]. Because risk increases from early childhood, testing is recommended at birth using cord blood or shortly thereafter; families declining testing in very young children should discuss the implications of not beginning surveillance [1]. Recommendations have evolved: the 2017 Oxford Desk Reference accepted diagnostic testing in an affected child but restricted predictive testing to adults [5], whereas ERN GENTURIS recommends routinely offering testing to adult first-degree relatives and, from birth, children with a high-paediatric-risk variant (childhood cancer in the index case or family, previous detection in other childhood-cancer families, or a dominant-negative variant); not routinely offering childhood testing for low-risk variants; and discussing testing with parents where familial cancers occurred before 31 years or evidence is insufficient, including surveillance burden and uncertain paediatric benefit [2]. Children and parents require appropriate pretest information, older children should be able to assent, and result communication should be agreed beforehand [1].

Genetic risk and reproductive options are best discussed before pregnancy, with counselling offered to affected or at-risk young adults [1]. Once the familial variant is known, prenatal diagnosis by chorionic villus sampling or amniocentesis and preimplantation genetic testing (PGT-M) are possible [1, 5]. Professional and family views differ, and decisions are personal [1]. During pregnancy, a woman with LFS should report cancer symptoms or signs and may continue clinical breast examination and indicated breast imaging; no specific surveillance recommendations exist for a carrier fetus, whose surveillance should begin after birth. Given the rare risk of gestational choriocarcinoma—including in women without LFS carrying a fetus with a paternal variant—beta-human chorionic gonadotrophin measurement may be considered when clinically appropriate [1]. DNA storage should be considered for probands without molecular diagnosis [1].

Maternal inheritance of p.Arg337His was observed in 72% of cases, suggesting preferential selection; one homozygous individual had no phenotype distinct from heterozygotes [1, 40]. Coexisting XAF1 p.Glu134Ter on the same haplotype should inform counselling because of its association with increased sarcoma and subsequent-neoplasm risk [34].

9. Special Considerations: Population Diversity and Epidemiology

9.1 Global epidemiology

Pathogenic germline TP53 variant prevalence in the general population is estimated at 1:3,000–1:10,000 [1]. A genome-first approach starting from large biobank genotypes rather than cancer-selected families identified 109 carriers across UK Biobank (n = 200,590), Geisinger (n = 170,503) and Penn Medicine Biobank (n = 43,731), with conservative estimates of 1:10,439, 1:3,790 and 1:2,983, respectively, after excluding relatives and heterozygotes with haematological neoplasms to limit clonal-haematopoiesis effects. p.Arg181His was commonest (8%), and 43% of 110 cancers in 71 carriers were haematological [7]. Cohort differences reflect database selection biases, and the authors emphasise prospective studies in young, diverse cohorts [7]. The Japanese guideline cites 1:500–1:20,000 [23]. Orphanet records point prevalence of 1–9/100,000, with estimates of 7/100,000 in the United Kingdom and 5/100,000 in the United States [4]. Carrier frequency greatly exceeds classic LFS family frequency, indicating that many potentially pathogenic-variant carriers never develop the full phenotype [11].

9.2 Asia

LFS-associated gastric cancer risk is higher in Japanese and other Asian carriers, probably because of diet and higher H. pylori prevalence, supporting upper gastrointestinal endoscopy in surveillance protocols [1]. The Japanese Society for Hereditary Tumors published its own guideline (version 1.1, 2019), recommending whole-body, brain and breast MRI, abdominal ultrasonography, and avoidance of radiation and alkylating agents wherever possible [23].

9.3 Brazil: the p.Arg337His founder variant

Southern and southeastern Brazil contain the world's largest concentration of germline TP53 variant carriers. Childhood adrenocortical carcinoma incidence is 10–15 times higher than elsewhere, and p.Arg337His was identified in 35 of 36 affected children [29]. The variant occurs in one in 375 people within a total population of approximately 100 million [64]; estimated frequency was 0.3% among 750 healthy women in a Porto Alegre breast screening programme, 10–20 times higher than other TP53 variants [65]; 0.27% in Paraná newborn screening [8]; 0.306% in Paraná and 0.245% in Santa Catarina [66]; and 0.21% in municipalities around Campinas, São Paulo [67]. A shared haplotype confirms the founder effect, distributed along a historical trade route of Portuguese origin [33]. Outside Brazil, the variant was detected in 0.05% of newborns in Paraguayan towns bordering Paraná [67].

The tumour spectrum is broader than initially assumed. Among 45 Brazilian LFS or LFL families, variants were found in 13 (28.9%), six carrying p.Arg337His, with breast (30.4%) and brain cancers (10.7%), soft-tissue sarcomas (10.7%) and adrenocortical carcinomas (8.9%) [68]. At a Campinas referral institution, carriers included 65 of 70 children (93%) with adrenocortical tumours, 9 of 13 (69%) with choroid plexus carcinoma and 3 of 41 (7.3%) with osteosarcoma—the latter having significantly poorer outcomes—but none with ALL, acute myeloid leukaemia, lymphoma, other CNS tumours, Ewing sarcoma or rhabdomyosarcoma [69]. Adult penetrance is incomplete and strongly sex dependent (65% in women, 30% in men), and cumulative risk for p.Arg337His is lower than for other variants (54% versus 78%) [12]. Adrenocortical carcinoma penetrance among carriers followed from birth to 12 years was 3.9% [66].

Environment modifies risk: annual adrenocortical carcinoma incidence was approximately 6.4 per million children younger than 10 years in Paraná and 4.15 per million in Santa Catarina, paralleling variant frequency; carriers in an agricultural subregion had lower risk than those in industrial and urban subregions (relative risk 2.4) [66]. XAF1 p.Glu134Ter also contributes to phenotypic heterogeneity [34].

9.4 The Paraná newborn screening programme

Free p.Arg337His newborn screening was offered at all Paraná hospitals. Of 180,000 newborns offered testing, 171,649 were screened and 461 (0.27%) carried the variant; parents of positive newborns were tested, and screening offered to relatives in the carrier lineage [8]. By April 2012, adrenocortical tumours had been diagnosed in 11 birth-screened carriers and only two non-carriers (p < 0.001), with six additional cases among 228 carrier relatives younger than 15 years (19 tumours overall); 347 (49.6%) of 699 carriers joined surveillance. Tumours were smaller among surveillance participants (p < 0.001) and more advanced among non-participants (four stage III tumours, two deaths). Birth-screened carriers also developed neuroblastoma, glioblastoma multiforme, choroid plexus carcinoma and Burkitt lymphoma; reconstruction of 353 families identified 1,704 carriers, in whom adrenocortical tumours were the commonest childhood cancer (n = 48) [8]. In a second 2016 cohort (42,438 newborns tested), educated families without periodic examinations detected one small tumour (21 g), without recurrence, among 112 carrier children [66].

Comparison of screened and unscreened children showed stage I in all 11 screened children (all alive), versus stages I, II, III and IV in five, five, one and five of 16 unscreened children, respectively; screening and surveillance cost per life-year saved was estimated at US$623, and the authors advocated including p.Arg337His in the state's mandatory universal newborn screening [45]. The programme remained outside mandatory state screening in 2021 [45]; no indexed publications documenting subsequent integration were identified.

9.5 Regional data gaps

No population studies of germline TP53 variant prevalence have been published for Africa, the Middle East or most Latin American countries, apart from Brazil and the Paraguayan border. p.Arg337His was identified in Portuguese and Spanish families sharing the Brazilian founder haplotype in an international study involving IPO Porto. Demographic modelling published in 2026 supports a single European introduction early in Portuguese colonisation (approximately 467 years), explaining rarity in Portugal and high Brazilian frequency through differing population growth [70, 71]; the reportedly low Portuguese population frequency has not been measured in population-based studies [33, 71]. Portuguese clinical series cannot establish prevalence: at IPO Porto, 3 of 149 women (2%) with HER2-positive breast cancer diagnosed by 40 years carried pathogenic TP53 variants, all with first breast cancer by 31 years and a first-degree relative with early cancer [72]. Available epidemiology for Portugal and most European countries derives from clinical series and registries, such as French and Spanish series, rather than population studies [4, 16, 47].

9.6 Special populations

Among adults identified through multigene panels without suggestive phenotypes, acquired variants (clonal haematopoiesis) are more likely, making non-haematopoietic tissue confirmation particularly important [36, 37]. In transgender people, breast cancer risk depends on sex assigned at birth but may be influenced by gender-affirming hormonal treatment [1].

10. Illustrative Clinical Case

10.1 Published case: radiation-induced angiosarcoma and a de novo variant

A 47-year-old Brazilian woman had bilateral breast cancer at 34 years without a family cancer history [73]. Initial treatment comprised bilateral mastectomy and left axillary dissection; right-sided invasive ductal carcinoma was hormone-receptor negative and HER2 positive (score 3+), pT1pN0; left-sided ductal carcinoma of no special type was hormone-receptor negative and HER2 positive, pT2pN1, with associated Paget disease. She received adjuvant docetaxel and cyclophosphamide with trastuzumab, followed by maintenance trastuzumab and bilateral chest-wall radiotherapy (50 Gy in 25 fractions), with subsequent implant reconstruction. Thirteen years later, progressive discomfort and lower-quadrant thickening developed; ultrasound and MRI showed nodules adherent to implant capsules, and bilateral biopsies confirmed bilateral breast angiosarcoma (FLI-1, CD31 and CD34 positive; Ki-67 40%). A germline panel found a heterozygous pathogenic TP53 variant previously unreported in the literature, NM_000546:c.788del, p.(Asn263Ilefs*82) (recorded in ClinVar on transcript NM_000546.6 as VCV000643332, pathogenic, by one submitter [48]); none of the tested relatives (parents, siblings and first cousins) carried it, indicating a de novo variant. Following multidisciplinary discussion, six neoadjuvant cycles of doxorubicin (75 mg/m²) and ifosfamide (9 g/m²) produced clinical and radiological responses, followed by removal of implants and capsules and excision of irradiated musculocutaneous tissues, achieving bilateral pathological complete response. The doxorubicin dose corresponds to the upper limit of the FDA label (60–75 mg/m² every 21 days as monotherapy; 40–75 mg/m² every 21–28 days in combination), and the ifosfamide dose lies within the range of the Portuguese SmPC for Holoxan® (8–12 g/m² per cycle, divided over 3 to 5 days, every 2 to 4 weeks, with mesna), which includes soft tissue sarcomas after failure of radiotherapy and surgery; the FDA label restricts ifosfamide to third-line treatment of testicular germ cell tumours, at 1.2 g/m² per day for 5 days (6 g/m² per cycle), so in the United States this regimen in angiosarcoma is off-label use [74–76]. Further radiotherapy was not indicated, and she remained disease free two years later, followed with whole-body MRI and low-dose chest CT [73]. In the open-access original article (https://pmc.ncbi.nlm.nih.gov/articles/PMC11671632/), figure 1 shows implant-capsule angiosarcoma histology, figure 2 positron emission tomography before and after chemotherapy, and figure 3 a p53 exon/domain diagram locating the variant.

HER2-positive breast cancer before 40 years and bilateral disease are independently suggestive of LFS [1]; absent family history does not exclude the syndrome given de novo variants [1, 6]; pre-treatment germline testing could have enabled radiotherapy avoidance, as ERN GENTURIS recommends [2]; and a second tumour in an irradiated field is itself a testing criterion [1, 2].

10.2 Composite illustrative case: adrenocortical carcinoma and p.Arg337His

This teaching case combines published Paraná data. A 22-month-old child from Curitiba presents with four weeks of pubarche, clitoral enlargement and accelerated growth, with mild hypertension. Abdominal ultrasound shows a 3 cm left adrenal mass and serum androgens are elevated. Complete adrenalectomy without capsule rupture confirms stage I adrenocortical carcinoma. Targeted germline testing identifies heterozygous p.Arg337His, inherited from her healthy 28-year-old mother; the maternal grandfather had soft-tissue sarcoma at 52 years and the maternal grandmother is healthy.

Management includes confirming that the family does not meet classic criteria, although adrenocortical carcinoma meets Chompret criteria regardless of family history [1, 2]; childhood surveillance with physical examination and abdominal ultrasonography every 3–4 months, annual whole-body MRI and brain MRI [1]; testing for XAF1 p.Glu134Ter, which affects sarcoma and second-tumour risk [34]; maternal surveillance with clinical examination, annual whole-body and breast MRI, and endoscopy from 25 years, considering the lower but substantial cumulative risk in adult female carriers (65% in women) [1, 12]; cascade testing of the 5-year-old sibling and maternal aunts and uncles [1]; and reproductive counselling on 50% transmission risk, prenatal diagnosis and PGT-M [1]. Paraná data indicate favourable outcomes for stage I detection [8, 45].

11. Prognosis and Long-Term Follow-Up

Lifelong cancer risk, multiple primary tumours and treatment-induced neoplasms determine prognosis. In the NCI cohort, half of female carriers had cancer by 31 years and half of male carriers by 46, with cumulative incidence approaching 100% by 70 years in both sexes; approximately 49% of cancer patients developed another after a median 10 years [41]. Multiple neoplasms occurred in 43% of affected carriers in the French series [16]. LFS-associated breast cancer had shorter recurrence-free and overall survival than breast cancer in other patients, and associated leukaemias may have poorer outcomes [1]. Osteosarcoma in p.Arg337His carriers had significantly poorer outcomes in a Brazilian series [69].

Surveillance changes prognosis. Toronto 5-year overall survival was 88.8% with surveillance versus 59.6% without [19], and all Paraná adrenocortical carcinomas detected through surveillance from newborn screening were stage I [45]. Follow-up should be lifelong, coordinated by a predisposition-experienced multidisciplinary team (clinical genetics, paediatric and adult oncology, radiology, breast surgery, gastroenterology, dermatology and psychology), with planned transition to adult surveillance [1, 21]. Each visit should review family history, potentially changing attenuated to classic classification; reassess variants against ClinGen specifications and databases (IARC, ClinVar); assess psychosocial and counselling needs; and reinforce avoidance of unnecessary radiation and carcinogens [1, 3]. Post-radiotherapy follow-up should include irradiated-field surveillance, including for basal cell carcinomas [1, 2].

After a first cancer, age-specific second-cancer risk resembles first-cancer risk, requiring surveillance to resume or continue after treatment [1, 41]. Ongoing research into metformin chemoprevention, liquid biopsy and variant-class-stratified surveillance may change follow-up in the medium term [16, 57, 62].

12. Summary and Concluding Remarks

Li-Fraumeni syndrome has the broadest tumour spectrum and earliest onset among cancer predisposition syndromes, caused by heterozygous germline TP53 variants with autosomal dominant inheritance. Its five core tumours—adrenocortical carcinoma, premenopausal breast cancer, CNS tumours, osteosarcoma and soft-tissue sarcomas—occur in distinct age phases; any-cancer risk reaches 92.4% in women and 59.7% in men by 50 years, with approximately half of survivors developing a second cancer.

Diagnosis has evolved from the 1988 classic criteria to modified Chompret criteria and 2020 European recommendations, including childhood sentinel tumours (adrenocortical carcinoma, choroid plexus carcinoma, anaplastic embryonal rhabdomyosarcoma, hypodiploid ALL, SHH medulloblastoma and mandibular osteosarcoma), breast cancer before 31 years and second tumours within irradiated fields. Multigene panels and tumour sequencing create two interpretative challenges: distinguishing constitutional variants, mosaicism and clonal haematopoiesis, requiring non-haematopoietic tissue confirmation without a suggestive phenotype; and recognising attenuated disease and reduced-penetrance variants.

Management centres on lifelong Toronto-based surveillance and adaptations (AACR 2017 and 2025, ERN GENTURIS 2020), including annual whole-body MRI from birth, associated with improved survival; cancer treatment adaptation favouring surgery, bilateral mastectomy and avoidance of radiotherapy and genotoxic chemotherapy wherever possible without compromising efficacy; and cascade testing from birth for high-paediatric-risk variants. No medicines are approved for the syndrome, and metformin remains under clinical investigation.

In Brazil, p.Arg337His, present in approximately 0.3% of the southern and southeastern population, confers high childhood adrenocortical carcinoma risk and lower, sex-dependent lifetime risks of other cancers; Paraná newborn screening and surveillance demonstrated low-cost stage I detection, providing a public-health genomics model for founder cancer predisposition variants.

13. References

1. Schneider K, Zelley K, Nichols KE, Schwartz Levine A, Garber J. Li-Fraumeni Syndrome. 1999 Jan 19 [updated 2024 Sep 5; revised 2025 May 1]. In: Adam MP, Bick S, Mirzaa GM, Pagon RA, Wallace SE, Amemiya A, editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993–2026. PMID: 20301488. Available at: https://www.ncbi.nlm.nih.gov/books/NBK1311/
2. Frebourg T, Bajalica Lagercrantz S, Oliveira C, Magenheim R, Evans DG; European Reference Network GENTURIS. Guidelines for the Li-Fraumeni and heritable TP53-related cancer syndromes. Eur J Hum Genet. 2020;28(10):1379–1386. doi:10.1038/s41431-020-0638-4
3. Kratz CP, Freycon C, Maxwell KN, Nichols KE, Schiffman JD, Evans DG, et al. Analysis of the Li-Fraumeni spectrum based on an international germline TP53 variant data set: an International Agency for Research on Cancer TP53 database analysis. JAMA Oncol. 2021;7(12):1800–1805. doi:10.1001/jamaoncol.2021.4398
4. Orphanet. Li-Fraumeni syndrome (ORPHA:524) [Internet]. Paris: INSERM; and Orphadata, prevalence dataset (en_product9_prev) [accessed 2026 Oct 4]. Available at: https://www.orpha.net/en/disease/detail/524
5. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. p. 584–585 (Li–Fraumeni syndrome).
6. Renaux-Petel M, Charbonnier F, Théry JC, Fermey P, Lienard G, Bou J, et al. Contribution of de novo and mosaic TP53 mutations to Li-Fraumeni syndrome. J Med Genet. 2018;55(3):173–180. doi:10.1136/jmedgenet-2017-104976
7. de Andrade KC, Strande NT, Kim J, Haley JS, Hatton JN, Frone MN, et al. Genome-first approach of the prevalence and cancer phenotypes of pathogenic or likely pathogenic germline TP53 variants. HGG Adv. 2024;5(1):100242. doi:10.1016/j.xhgg.2023.100242
8. Custódio G, Parise GA, Kiesel Filho N, Komechen H, Sabbaga CC, Rosati R, et al. Impact of neonatal screening and surveillance for the TP53 R337H mutation on early detection of childhood adrenocortical tumors. J Clin Oncol. 2013;31(20):2619–2626. doi:10.1200/JCO.2012.46.3711
9. Harrold E, Stadler Z. Cancer predisposition syndromes — cancer risks in BRCA1/2 carriers and Li–Fraumeni syndrome. In: ESMO Handbook — Cancer Genomics. Lugano: ESMO Press. p. 22–25.
10. Fortuno C, Feng BJ, Carroll C, Innella G, Kohlmann W, Lázaro C, et al. Cancer risks associated with TP53 pathogenic variants: maximum likelihood analysis of extended pedigrees for diagnosis of first cancers beyond the Li-Fraumeni syndrome spectrum. JCO Precis Oncol. 2024;8:e2300453. doi:10.1200/PO.23.00453
11. Amadou A, Achatz MIW, Hainaut P. Revisiting tumor patterns and penetrance in germline TP53 mutation carriers: temporal phases of Li-Fraumeni syndrome. Curr Opin Oncol. 2018;30(1):23–29. doi:10.1097/CCO.0000000000000423
12. Galante PAF, Guardia GDA, Pisani J, Sandoval RL, Barros-Filho MC, Gifoni ACLVC, et al. Personalized screening strategies for TP53 R337H carriers: a retrospective cohort study of tumor spectrum in Li-Fraumeni syndrome adult carriers. Lancet Reg Health Am. 2025;42:100982. doi:10.1016/j.lana.2024.100982
13. Li FP, Fraumeni JF Jr, Mulvihill JJ, Blattner WA, Dreyfus MG, Tucker MA, et al. A cancer family syndrome in twenty-four kindreds. Cancer Res. 1988;48(18):5358–5362. PMID: 3409256.
14. Chompret A, Abel A, Stoppa-Lyonnet D, Brugières L, Pagès S, Feunteun J, et al. Sensitivity and predictive value of criteria for p53 germline mutation screening. J Med Genet. 2001;38(1):43–47. doi:10.1136/jmg.38.1.43
15. Tinat J, Bougeard G, Baert-Desurmont S, Vasseur S, Martin C, Bouvignies E, et al. 2009 version of the Chompret criteria for Li Fraumeni syndrome. J Clin Oncol. 2009;27(26):e108–e109. doi:10.1200/JCO.2009.22.7967
16. Bougeard G, Renaux-Petel M, Flaman JM, Charbonnier C, Fermey P, Belotti M, et al. Revisiting Li-Fraumeni syndrome from TP53 mutation carriers. J Clin Oncol. 2015;33(21):2345–2352. doi:10.1200/JCO.2014.59.5728
17. Birch JM, Hartley AL, Tricker KJ, Prosser J, Condie A, Kelsey AM, et al. Prevalence and diversity of constitutional mutations in the p53 gene among 21 Li-Fraumeni families. Cancer Res. 1994;54(5):1298–1304. PMID: 8118819.
18. Eeles RA. Germline mutations in the TP53 gene. Cancer Surv. 1995;25:101–124. PMID: 8718514.
19. Villani A, Shore A, Wasserman JD, Stephens D, Kim RH, Druker H, et al. Biochemical and imaging surveillance in germline TP53 mutation carriers with Li-Fraumeni syndrome: 11 year follow-up of a prospective observational study. Lancet Oncol. 2016;17(9):1295–1305. doi:10.1016/S1470-2045(16)30249-2
20. Kratz CP, Achatz MI, Brugières L, Frebourg T, Garber JE, Greer MC, et al. Cancer screening recommendations for individuals with Li-Fraumeni syndrome. Clin Cancer Res. 2017;23(11):e38–e45. doi:10.1158/1078-0432.CCR-17-0408
21. Achatz MI, Villani A, Bertuch AA, Bougeard G, Chang VY, Doria AS, et al. Update on cancer screening recommendations for individuals with Li-Fraumeni syndrome. Clin Cancer Res. 2025;31(10):1831–1840. doi:10.1158/1078-0432.CCR-24-3301
22. ClinGen Actionability Working Group. Li-Fraumeni syndrome — TP53 (report AC068), adult and paediatric contexts (released 2024 Mar 24). Clinical Genome Resource [Internet]; [accessed 2026 Oct 4]. Available at: https://actionability.clinicalgenome.org/ac/Adult/api/sepio/doc/AC068 and https://actionability.clinicalgenome.org/ac/Pediatric/api/sepio/doc/AC068
23. Kumamoto T, Yamazaki F, Nakano Y, Tamura C, Tashiro S, Hattori H, et al. Medical guidelines for Li-Fraumeni syndrome 2019, version 1.1. Int J Clin Oncol. 2021;26(12):2161–2178. doi:10.1007/s10147-021-02011-w
24. National Comprehensive Cancer Network. NCCN Clinical Practice Guidelines in Oncology: Genetic/Familial High-Risk Assessment: Breast, Ovarian, Pancreatic, and Prostate [Internet]. Plymouth Meeting (PA): NCCN; [accessed 2026 Oct 4]. Available at: https://www.nccn.org/guidelines/guidelines-detail?category=2&id=1503
25. Engels M, Urbanczyk K, Hölzenspies J, Röhl C, Geverink N, Hoogerbrugge N, et al. The European Reference Network on Genetic Tumour Risk Syndromes (ERN GENTURIS): benefits for patients, families, and health care providers. Fam Cancer. 2025;24(2):33. doi:10.1007/s10689-025-00457-9
26. Li FP, Fraumeni JF Jr. Soft-tissue sarcomas, breast cancer, and other neoplasms. A familial syndrome? Ann Intern Med. 1969;71(4):747–752. doi:10.7326/0003-4819-71-4-747
27. Malkin D, Li FP, Strong LC, Fraumeni JF Jr, Nelson CE, Kim DH, et al. Germ line p53 mutations in a familial syndrome of breast cancer, sarcomas, and other neoplasms. Science. 1990;250(4985):1233–1238. doi:10.1126/science.1978757
28. Srivastava S, Zou ZQ, Pirollo K, Blattner W, Chang EH. Germ-line transmission of a mutated p53 gene in a cancer-prone family with Li-Fraumeni syndrome. Nature. 1990;348(6303):747–749. doi:10.1038/348747a0
29. Ribeiro RC, Sandrini F, Figueiredo B, Zambetti GP, Michalkiewicz E, Lafferty AR, et al. An inherited p53 mutation that contributes in a tissue-specific manner to pediatric adrenal cortical carcinoma. Proc Natl Acad Sci U S A. 2001;98(16):9330–9335. doi:10.1073/pnas.161479898
30. Villani A, Tabori U, Schiffman J, Shlien A, Beyene J, Druker H, et al. Biochemical and imaging surveillance in germline TP53 mutation carriers with Li-Fraumeni syndrome: a prospective observational study. Lancet Oncol. 2011;12(6):559–567. doi:10.1016/S1470-2045(11)70119-X
31. Collet L, et al. Breast cancer — genomics. In: ESMO Handbook — Cancer Genomics. Lugano: ESMO Press. p. 82–87.
32. DiGiammarino EL, Lee AS, Cadwell C, Zhang W, Bothner B, Ribeiro RC, et al. A novel mechanism of tumorigenesis involving pH-dependent destabilization of a mutant p53 tetramer. Nat Struct Biol. 2002;9(1):12–16. doi:10.1038/nsb730
33. Garritano S, Gemignani F, Palmero EI, Olivier M, Martel-Planche G, Le Calvez-Kelm F, et al. Detailed haplotype analysis at the TP53 locus in p.R337H mutation carriers in the population of Southern Brazil: evidence for a founder effect. Hum Mutat. 2010;31(2):143–150. doi:10.1002/humu.21151
34. Pinto EM, Figueiredo BC, Chen W, Galvao HCR, Formiga MN, Fragoso MCBV, et al. XAF1 as a modifier of p53 function and cancer susceptibility. Sci Adv. 2020;6(26):eaba3231. doi:10.1126/sciadv.aba3231
35. Batalini F, Peacock EG, Stobie L, Robertson A, Garber J, Weitzel JN, et al. Li-Fraumeni syndrome: not a straightforward diagnosis anymore — the interpretation of pathogenic variants of low allele frequency and the differences between germline PVs, mosaicism, and clonal hematopoiesis. Breast Cancer Res. 2019;21(1):107. doi:10.1186/s13058-019-1193-1
36. Weitzel JN, Chao EC, Nehoray B, Van Tongeren LR, LaDuca H, Blazer KR, et al. Somatic TP53 variants frequently confound germ-line testing results. Genet Med. 2018;20(8):809–816. doi:10.1038/gim.2017.196
37. Coffee B, Cox HC, Bernhisel R, Manley S, Bowles K, Roa BB, et al. A substantial proportion of apparently heterozygous TP53 pathogenic variants detected with a next-generation sequencing hereditary pan-cancer panel are acquired somatically. Hum Mutat. 2020;41(1):203–211. doi:10.1002/humu.23910
38. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. p. 620–621 (Wilms tumour).
39. Harrold E, Stadler Z. Cancer predisposition syndromes — polyposis syndromes and colorectal cancer. In: ESMO Handbook — Cancer Genomics. Lugano: ESMO Press. p. 27.
40. Ferreira AM, Brondani VB, Helena VP, Charchar HLS, Zerbini MCN, Leite LAS, et al. Clinical spectrum of Li-Fraumeni syndrome/Li-Fraumeni-like syndrome in Brazilian individuals with the TP53 p.R337H mutation. J Steroid Biochem Mol Biol. 2019;190:250–255. doi:10.1016/j.jsbmb.2019.04.011
41. Mai PL, Best AF, Peters JA, DeCastro RM, Khincha PP, Loud JT, et al. Risks of first and subsequent cancers among TP53 mutation carriers in the National Cancer Institute Li-Fraumeni syndrome cohort. Cancer. 2016;122(23):3673–3681. doi:10.1002/cncr.30248
42. Genomics in haematological malignancies. In: ESMO Handbook — Cancer Genomics. Lugano: ESMO Press. p. 109–113.
43. Heymann S, Delaloge S, Rahal A, Caron O, Frebourg T, Barreau L, et al. Radio-induced malignancies after breast cancer postoperative radiotherapy in patients with Li-Fraumeni syndrome. Radiat Oncol. 2010;5:104. doi:10.1186/1748-717X-5-104
44. Harrold E, Stadler Z. Cancer predisposition syndromes — identification from somatic panels and conclusion. In: ESMO Handbook — Cancer Genomics. Lugano: ESMO Press. p. 29–30.
45. Tosin KCF, Legal EF, Pianovski MAD, Ibañez HC, Custódio G, Carvalho DS, et al. Newborn screening for the detection of the TP53 R337H variant and surveillance for early diagnosis of pediatric adrenocortical tumors: lessons learned and way forward. Cancers (Basel). 2021;13(23):6111. doi:10.3390/cancers13236111
46. ClinGen TP53 Variant Curation Expert Panel. ClinGen TP53 Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines, version 2.4 (GN009) [Internet]. Clinical Genome Resource, Criteria Specification Registry; updated 2025 Nov 20 [accessed 2026 Oct 4]. Available at: https://cspec.genome.network/cspec/ui/svi/doc/GN009
47. Rofes P, Castillo-Manzano C, Menéndez M, Teulé Á, Iglesias S, Munté E, et al. TP53 germline testing and hereditary cancer: how somatic events and clinical criteria affect variant detection rate. Genome Med. 2025;17(1):3. doi:10.1186/s13073-025-01429-5
48. National Center for Biotechnology Information. ClinVar: VCV000012379 (NM_000546.6(TP53):c.1010G>A, p.Arg337His) and VCV000643332 (NM_000546.6(TP53):c.788del) [Internet]. Bethesda (MD): NCBI [accessed 2026 Oct 4]. Available at: https://www.ncbi.nlm.nih.gov/clinvar/variation/12379/ and https://www.ncbi.nlm.nih.gov/clinvar/variation/643332/
49. Kratz CP, Frone MN, Khincha PP, Hatton JN, Penkert J, James PA, et al. Urgent need to recognize that disease-causing TP53 variants with atypical penetrance require distinct clinical recommendations. J Natl Cancer Inst. 2026 Jan 24:djag015. doi:10.1093/jnci/djag015
50. Ballinger ML, Best A, Mai PL, Khincha PP, Loud JT, Peters JA, et al. Baseline surveillance in Li-Fraumeni syndrome using whole-body magnetic resonance imaging: a meta-analysis. JAMA Oncol. 2017;3(12):1634–1639. doi:10.1001/jamaoncol.2017.1968
51. Finn E, Sohaib A, Withey SJ, Sardo Infirri S, Bunce C, Clarke CS, et al. SIGNIFIED: whole-body MRI screening in Li-Fraumeni syndrome in the UK. ESMO Open. 2026;11(3):106045. doi:10.1016/j.esmoop.2025.106045
52. Sun Zhang A, Omran M, Wille J, Ek T, Sabel M, Pal N, et al. Surveillance adherence and clinical findings in children with confirmed or familial TP53 variants: the Swedish multicenter constitutional TP53 study. Genet Med. 2026 Jul 7:102643. doi:10.1016/j.gim.2026.102643
53. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. p. 552–554 (Cancer surveillance methods; box 4.4 and table 4.6, NICE CG164).
54. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. p. 546–550 (Breast cancer).
55. Mastellaro MJ, Seidinger AL, Kang G, Abrahão R, Miranda ECM, Pounds SB, et al. Contribution of the TP53 R337H mutation to the cancer burden in southern Brazil: insights from the study of 55 families of children with adrenocortical tumors. Cancer. 2017;123(16):3150–3158. doi:10.1002/cncr.30703
56. Hendrickson PG, Luo Y, Kohlmann W, Schiffman J, Maese L, Bishop AJ, et al. Radiation therapy and secondary malignancy in Li-Fraumeni syndrome: a hereditary cancer registry study. Cancer Med. 2020;9(21):7954–7963. doi:10.1002/cam4.3427
57. Dixon-Zegeye M, Shaw R, Collins L, Perez-Smith K, Ooms A, Qiao M, et al. Cancer Precision-Prevention trial of Metformin in adults with Li Fraumeni syndrome (MILI) undergoing yearly MRI surveillance: a randomised controlled trial protocol. Trials. 2024;25(1):103. doi:10.1186/s13063-024-07929-w
58. Walcott FL, Wang PY, Bryla CM, Huffstutler RD, Singh N, Pollak MN, et al. Pilot study assessing tolerability and metabolic effects of metformin in patients with Li-Fraumeni syndrome. JNCI Cancer Spectr. 2020;4(6):pkaa063. doi:10.1093/jncics/pkaa063
59. ClinicalTrials.gov. Condition search for “Li-Fraumeni” (API v2), including records NCT01464086 (LIFSCREEN), NCT01981525, NCT03789175 and NCT06088030 [Internet]. Bethesda (MD): National Library of Medicine [accessed 2026 Oct 4]. Available at: https://clinicaltrials.gov/search?cond=Li-Fraumeni
60. U.S. Food and Drug Administration. Metformin hydrochloride: official prescribing information (section 1, Indications and Usage), openFDA [Internet; accessed 2026 Oct 4]. Available at: https://api.fda.gov/drug/label.json?search=openfda.generic_name:"METFORMIN+HYDROCHLORIDE"
61. INFARMED. Glucophage 500 mg comprimidos revestidos por película (cloridrato de metformina) — summary of product characteristics (text revised 03/2025) [Internet]. Lisbon: INFARMED, Infomed; [accessed 2026 Oct 26]. Available at: https://extranet.infarmed.pt/INFOMED-fo/
62. Wong D, Luo P, Oldfield LE, Gong H, Brunga L, Rabinowicz R, et al. Early cancer detection in Li-Fraumeni syndrome with cell-free DNA. Cancer Discov. 2024;14(1):104–119. doi:10.1158/2159-8290.CD-23-0456
63. Caron O, Frebourg T, Benusiglio PR, Foulon S, Brugières L. Lung adenocarcinoma as part of the Li-Fraumeni syndrome spectrum: preliminary data of the LIFSCREEN randomized clinical trial. JAMA Oncol. 2017;3(12):1736–1737. doi:10.1001/jamaoncol.2017.1358
64. Achatz MI, Zambetti GP. The inherited p53 mutation in the Brazilian population. Cold Spring Harb Perspect Med. 2016;6(12):a026195. doi:10.1101/cshperspect.a026195
65. Palmero EI, Schüler-Faccini L, Caleffi M, Achatz MI, Olivier M, Martel-Planche G, et al. Detection of R337H, a germline TP53 mutation predisposing to multiple cancers, in asymptomatic women participating in a breast cancer screening program in Southern Brazil. Cancer Lett. 2008;261(1):21–25. doi:10.1016/j.canlet.2007.10.044
66. Costa TEJ, Gerber VKQ, Ibañez HC, Melanda VS, Parise IZS, Watanabe FM, et al. Penetrance of the TP53 R337H mutation and pediatric adrenocortical carcinoma incidence associated with environmental influences in a 12-year observational cohort in Southern Brazil. Cancers (Basel). 2019;11(11):1804. doi:10.3390/cancers11111804
67. Falcon-de Legal E, Ascurra M, Vega-Paredes R, Sade E, Monteiro M, Paraízo M, et al. Germline TP53 p.R337H and XAF1 p.E134* variants: prevalence in Paraguay and comparison with rates in Brazilian State of Paraná and previous findings at the Paraguayan-Brazilian border. Curr Oncol. 2025;32(6):333. doi:10.3390/curroncol32060333
68. Achatz MI, Olivier M, Le Calvez F, Martel-Planche G, Lopes A, Rossi BM, et al. The TP53 mutation, R337H, is associated with Li-Fraumeni and Li-Fraumeni-like syndromes in Brazilian families. Cancer Lett. 2007;245(1–2):96–102. doi:10.1016/j.canlet.2005.12.039
69. Seidinger AL, Mastellaro MJ, Paschoal Fortes F, Godoy Assumpção J, Aparecida Cardinalli I, Aparecida Ganazza M, et al. Association of the highly prevalent TP53 R337H mutation with pediatric choroid plexus carcinoma and osteosarcoma in southeast Brazil. Cancer. 2011;117(10):2228–2235. doi:10.1002/cncr.25826
70. Pinto EM, Fridman C, Figueiredo BC, Salvador H, Teixeira MR, Pinto C, et al. Multiple TP53 p.R337H haplotypes and implications for tumor susceptibility. HGG Adv. 2024;5(1):100244. doi:10.1016/j.xhgg.2023.100244
71. Pinto EM, Muzzi JCD, Yunes JA, de Figueiredo BC, Ribeiro RC, Zambetti GP. Reconstructing the origin and demographic expansion of the TP53 p.R337H founder variant in Brazil. Cancer Epidemiol Biomarkers Prev. 2026;35(2):210–216. doi:10.1158/1055-9965.EPI-25-1148
72. Escudeiro C, Pinto C, Vieira J, Peixoto A, Pinto P, Pinheiro M, et al. The role of TP53 pathogenic variants in early-onset HER2-positive breast cancer. Fam Cancer. 2021;20(3):173–180. doi:10.1007/s10689-020-00212-2
73. David BBL, Silva SSA, Dinoa V, Diniz T, Pereira E, Bustamante C, et al. New germline TP53 variant detected after radiotherapy-induced angiosarcoma of the chest wall in a previously treated breast cancer patient: a case report and review of Li-Fraumeni syndrome and radiotherapy-induced sarcoma. Case Rep Oncol Med. 2024;2024:6640468. doi:10.1155/crom/6640468
74. U.S. Food and Drug Administration. Doxorubicin hydrochloride injection, for intravenous use — prescribing information (revision of 2026-04-30) [Internet]. openFDA Drug Label API; [accessed 2026 Oct 26]. Available at: https://api.fda.gov/drug/label.json?search=openfda.generic_name:"doxorubicin+hydrochloride"
75. INFARMED. Holoxan 1000 mg pó para solução injetável (ifosfamida) — summary of product characteristics (approved 03/04/2023) [Internet]. Lisbon: INFARMED, Infomed; [accessed 2026 Oct 26]. Available at: https://extranet.infarmed.pt/INFOMED-fo/download-ficheiro.xhtml?med_id=4309&tipo_doc=rcm
76. U.S. Food and Drug Administration. Ifosfamide for injection, for intravenous use — prescribing information (ANDA 076078; revision of 2026-04-09) [Internet]. openFDA Drug Label API; [accessed 2026 Oct 26]. Available at: https://api.fda.gov/drug/label.json?search=openfda.generic_name:"ifosfamide"