Li-Fraumeni Syndrome

A number sign (#) is used with this entry because Li-Fraumeni syndrome is caused by heterozygous mutation in the p53 gene (TP53; 191170) on chromosome 17p13.

Description

Li-Fraumeni syndrome (LFS) is a clinically and genetically heterogeneous inherited cancer syndrome. LFS is characterized by autosomal dominant inheritance and early onset of tumors, multiple tumors within an individual, and multiple affected family members. In contrast to other inherited cancer syndromes, which are predominantly characterized by site-specific cancers, LFS presents with a variety of tumor types. The most common types are soft tissue sarcomas and osteosarcomas, breast cancer, brain tumors, leukemia, and adrenocortical carcinoma. Classic LFS is defined as a proband with a sarcoma before the age of 45 years and a first-degree relative with any cancer before the age of 45 years and 1 additional first- or second-degree relative in the same lineage with any cancer before the age of 45 years or a sarcoma at any age (Li et al., 1988). Li-Fraumeni-like syndrome (LFL) is defined as a proband with any childhood cancer, or a sarcoma, brain tumor, or adrenocortical tumor before the age of 45 years, plus a first- or second-degree relative in the same lineage with a typical LFS tumor at any age, and an additional first- or second-degree relative in the same lineage with any cancer before the age of 60 years (Birch et al., 1994). A less restrictive definition of LFL is 2 different LFS-related tumors in first- or second-degree relatives at any age (Eeles, 1995). Approximately 70% of LFS cases and 40% of LFL cases contain germline mutations in the p53 gene on chromosome 17p13.1 (Bachinski et al., 2005).

Genetic Heterogeneity of Li-Fraumeni Syndrome

A second form of Li-Fraumeni syndrome (LFS2; 609265) is caused by mutation in the CHEK2 gene (604373).

Clinical Features

In reviewing medical records and death certificates of 648 childhood rhabdomyosarcoma patients, Li and Fraumeni (1969) identified 4 families in which sibs or cousins had a childhood sarcoma. These 4 families also had striking histories of breast cancer and other neoplasms, suggesting a new familial cancer syndrome of diverse tumors. Subsequent prospective studies confirmed the high risk in family members of the tumor types that comprise LFS (Li and Fraumeni, 1982). Studies in other geographic and ethnic groups by Birch et al. (1984, 1990) corroborated the syndrome. The spectrum of cancers in the syndrome was shown to include, in addition to breast cancer and soft tissue sarcomas, brain tumors, osteosarcoma, leukemia, and adrenocortical carcinoma.

Fraumeni et al. (1975) described a kindred in which in 1 sibship of 9 adults, 4 died of lymphocytic or histiocytic lymphomas and 1, a male, of Waldenstrom macroglobulinemia complicated by adenocarcinoma of the lung. In the next generation, 1 person died of Hodgkin disease; 4 of 9 healthy persons had impaired lymphocyte transformation with phytohemagglutinin, and 3 of these had polyclonal elevation of IgM. Subsequent to the studies, adenocarcinoma of the lung developed in 1 of those with an immune defect, a woman, and her 3-year-old grandson developed lymphocytic leukemia. This was the first suggestion of a genetic or immunologic basis of lung adenocarcinoma.

Pearson et al. (1982) reported 2 families resembling that reported by Li and Fraumeni (1969). In 1, the mother had breast cancer and 3 of her 4 children had adrenocortical carcinoma, medulloblastoma, and rhabdomyosarcoma; in the other, the mother had breast cancer and 2 of her 3 children had adrenocortical carcinoma and rhabdomyosarcoma.

Mulvihill (1982) used the designation sarcoma family syndrome of Li and Fraumeni for the familial association of breast cancer, soft tissue sarcoma, and other tumors. Hartley et al. (1987) referred to this as the SBLA syndrome, a designation derived from the tumors that occur in this cancer family syndrome: sarcoma, breast and brain tumors, leukemia, laryngeal and lung cancer, and adrenal cortical carcinoma (Lynch et al., 1978).

Possible component tumors of LFS are melanoma, gonadal germ cell tumors, and carcinomas of the lung, pancreas, and prostate (Strong et al., 1987; Li et al., 1988). The diverse tumor types in family members characteristically develop at unusually early ages, and multiple primary tumors are frequent. By segregation analysis, Strong et al. (1987) demonstrated that the observed cancer distribution in families best fit a rare autosomal dominant gene model. The model also predicted that in families at risk the probability of developing any invasive cancer (excluding carcinomas of the skin) reaches almost 50% by age 30, when only 1% of the general population has developed cancer. More than 90% of the gene carriers would develop cancer by age 70 (review by Malkin et al., 1990).

Hisada et al. (1998) quantified the incidence of second and third primary cancers in individuals from 24 LFS families originally diagnosed with cancer between 1968 and 1986. Among 200 LFS family members diagnosed with cancer, 30 (15%) developed a second cancer. Eight individuals (4%) had a third cancer, while 4 (2%) eventually developed a fourth cancer. Overall, the relative risk of occurrence of a second cancer was 5.3, with a cumulated probability of second cancer occurrence of 57% at 30 years after diagnosis of a first cancer. Relative risks of second cancers occurring in families with this syndrome were 83.0, 9.7, and 1.5 for individuals with a first cancer at ages 0 to 19 years, 20 to 44 years, and 45 years or more, respectively. Thirty (71%) of 42 subsequent cancers in this group were component cancers of LFS.

Masciari et al. (2011) reviewed 62 TP53 mutation-positive families for gastric cancer. There were 429 cancer-affected individuals. In the 62 families gastric cancer was the diagnosis in the lineages of 21 (4.9%) subjects from 14 families (22.6%). The mean and median ages at gastric cancer diagnosis were 43 and 36 years, respectively (range: 24-74 years), significantly younger compared with the median age at diagnosis in the general population based on Surveillance Epidemiology and End Results data (71 years). Five (8.1%) families reported 2 or more cases of gastric cancer, and 6 (9.7%) families had cases of both colorectal and gastric cancers. No association was seen between phenotype and type/location of the TP53 mutations. Pathology review of the available tumors revealed both intestinal and diffuse histologies. Masciari et al. (2011) concluded that early-onset gastric cancer seems to be a component of Li-Fraumeni syndrome, suggesting the need for early and regular endoscopic screening in individuals with germline TP53 mutations, particularly among those with a family history of gastric cancer.

Inheritance

Li-Fraumeni syndrome shows autosomal dominant inheritance. The lifetime penetrance is high: by age 50, women have an overall higher risk (93%) of developing cancer compared to men (68%), as well as an earlier age at onset (29 years in women vs 40 years in men). Gonzalez et al. (2009) identified TP53 mutations in 75 of 341 patients with early-onset cancer sent for TP53 testing. Family history was available for all 341 patients. Five (7%) of 75 patients with TP53 mutations were confirmed to have de novo mutations, and 4 (80%) of the 5 patients with de novo mutations had multiple primary cancers. Ten of 75 patients with TP53 mutations likely had de novo germline mutations by family history. Gonzalez et al. (2009) estimated that the frequency of de novo TP53 mutations resulting in Li-Fraumeni syndrome may be as high as 20% (15 of 75).

Clinical Management

Villani et al. (2016) introduced a clinical surveillance program (The Toronto Protocol) using physical examination and frequent biochemical and imaging studies (consisting of whole body MRI, brain MRI, breast MRI, mammography, abdominal and pelvic ultrasound, and colonoscopy) at 3 tertiary care centers in Canada and the USA on January 1, 2004, for carriers of TP53 pathogenic variants. After confirmation of TP53 mutation, participants either chose to undergo surveillance or chose not to undergo surveillance. Patients could cross over between groups at any time. The primary outcome measure was detection of symptomatic tumors by surveillance investigations. The secondary outcome measure was 5-year overall survival established from a tumor diagnosed symptomatically (in the nonsurveillance group) versus one diagnosed by surveillance. Villani et al. (2016) completed survival analyses using an as-treated approach. Between January 1, 2004 and July 1, 2015, Villani et al. (2016) identified 89 carriers of TP53 pathogenic mutations in 39 unrelated families, of whom 40 (45%) agreed to surveillance and 49 (55%) declined surveillance. Nineteen patients (21%) crossed over from nonsurveillance to the surveillance group, giving a total of 59 individuals (66%) undergoing surveillance for a median of 32 months (IQR 12-87). Forty asymptomatic tumors have been detected in 19 (32%) of 59 patients who underwent surveillance. Two additional cancers were diagnosed between surveillance assessments (false negatives) and 2 biopsied lesions were nonneoplastic entities on pathologic review (false positives). Among the 49 individuals who initially declined surveillance, 61 symptomatic tumors were diagnosed in 43 patients (88%). Twenty-one (49%) of the 43 individuals not on surveillance who developed cancer were alive compared with 16 (84%) of the 19 individuals undergoing surveillance who developed cancer (p = 0.012) after a median follow-up of 46 months for those not on surveillance and 38 months for those on surveillance. Five-year overall survival was 88.8% (95% CI 78.7-100) in the surveillance group and 59.6% (47.2-75.2) in the nonsurveillance group (p = 0.0132). Based on these findings, Villani et al. (2016) concluded that long-term compliance with a comprehensive surveillance protocol for early tumor detection in individuals with pathogenic TP53 variants is feasible and that early tumor detection through surveillance is associated with improved long-term survival. The authors suggested incorporation of this approach into clinical management of these patients.

Mapping

LFS1 results from mutation in the TP53 gene, which maps to chromosome 17p13.1.

To data on a set of kindreds with Li-Fraumeni syndrome, Shete et al. (2002) applied a method they developed to incorporate individual-specific liability classes into linkage analysis. The approach yielded higher lod scores and more accurate estimates of the recombination fraction in the families showing linkage.

Molecular Genetics

Because tumor suppressor genes had been found to be associated with familial neoplasms, Malkin et al. (1990) suspected mutation in this type of gene in LFS. The RB1 gene (614041) was an unlikely candidate for a germline mutation in LFS because retinoblastoma (180200) had not been observed in these families. On the other hand, the TP53 gene was a more likely candidate because inactivating mutations therein had been associated with sporadic osteosarcomas, soft tissue sarcomas, brain tumors, leukemias, and carcinomas of the lung and breast. Furthermore, transgenic mice carrying a mutant p53 gene have an increased incidence of osteosarcomas, soft tissue sarcomas, adenocarcinomas of the lung, and adrenal and lymphoid tumors--all tumors that occur as part of LFS. Such was the basis for the successful search for p53 mutations in this disorder (see 191170.0001).

To determine the frequency and distribution of germline p53 mutations in LFS families, Frebourg et al. (1995) sequenced the 10 coding exons of TP53 in lymphocytes and fibroblast cell lines derived from 15 families with the syndrome. Germline mutations were observed in 8 such families; of these, 6 were missense mutations located between exons 5 and 8. One mutation was a nonsense mutation in exon 6 and one was a splicing mutation in intron 4; each of these mutations generated aberrantly short p53 RNAs. The study indicated that most germline p53 mutations in LFS are located between exons 5 and 8 and that approximately 50% of patients with LFS have no germline mutations in the coding region of the p53 gene. Significantly, in 3 families, a mutation of the p53 gene was observed in a fibroblast cell line derived from the proband, but the mutation was not found in affected relatives in 2 families or in the blood from 1 of the probands. This indicated that the mutation probably occurred during cell culture; thus, it is necessary that analysis for germline p53 mutations be performed on cells that have not been grown in vitro.

Varley et al. (1997) stated that more than 50 families had been identified with LFS caused by germline TP53 mutations. LFS is defined by strict clinical criteria, described by Li et al. (1988). This definition has been relaxed to include Li-Fraumeni-like (LFL) cases (Birch et al., 1994). Varley et al. (1997) detected mutations in TP53 in approximately 70% of LFS and 20% of LFL families, when all exons, including noncoding regions, were sequenced.

Chompret et al. (2001) reported a family study of 2,691 children with a history of solid tumor before the age of 18. A subgroup of 239 children had in addition a family history of at least 1 cancer affecting a first- or second-degree relative before the age of 46 or multiple primary cancers in the proband. Among these 239, 211 had at least 1 first- or second-degree relative affected, 16 had at least 2 primary tumors, and 12 fulfilled both criteria. They performed genotyping of p53 and found mutations in 9 individuals from the first group, 1 in the second, and 5 in the third. Chompret et al. (2001) calculated sensitivity and predictive value of p53 mutation testing using various degrees of stringency of selection criteria and concluded that p53 mutation testing should be considered in families where there is (1) a proband affected by a narrow spectrum cancer (sarcoma, brain tumor, breast cancer, adrenocortical tumor) before 36 years, and at least 1 first- or second-degree relative affected by a narrow spectrum tumor (other than breast cancer if the proband is affected by breast cancer) before the age of 46, or by multiple primary tumors; (2) a proband with multiple primary tumors, 2 of which belong to the narrow spectrum and the first of which occurred before 36 years, whatever the family history; (3) a proband with adrenocortical carcinoma, whatever the age of onset or family history. Using such criteria, they expected to find a mutation in 20% of cases and to miss 20% of mutations that would be detected by the least stringent criteria.

Hwang et al. (2003) stated that germline mutations in the p53 gene had been identified in 50 to 70% of families with LFS. To characterize cancer risk in heterozygous p53 mutation carriers, they analyzed cancer incidence in 56 germline p53 mutation carriers and 3,201 noncarriers from 107 kindreds ascertained through patients with childhood soft tissue sarcoma. Members of these kindreds were systematically followed for more than 20 years for cancer incidence and their p53 gene status was evaluated. Hwang et al. (2003) identified 7 kindreds with germline p53 mutations that included missense and truncation mutation types. A significantly higher cancer risk was found in female carriers than in male carriers, a difference not explained by an excess of sex-specific cancer. The calculated standardized incidence ratio (SIR) showed that mutation carriers had a risk for all types of cancer that was much higher than that for the general population, whereas noncarriers had a risk for all types of cancer that was similar to that in the general population. The calculated SIRs showed a higher risk by more than 100-fold for sarcoma, female breast cancer, and hematologic malignancy for the p53 mutation carriers and agreed with the findings of an earlier segregation analysis based on the same cohort (Kleihues et al., 1997).

Varley (2003) reviewed the findings of Birch et al. (2001), who studied the distribution of cancers in carriers of germline TP53 mutations from 28 families in which all cancers were verified and the ages of all family members, affected and unaffected, were known. They found a highly significant difference from the expected cancer distribution in the general population. The tumors originally identified as being components of Li-Fraumeni syndrome were found to be strongly associated with a germline TP53 mutation, with the exception of leukemia, which was not found to be a major component. However, Wilms tumor and malignant phyllode tumors of the breast were found at significantly higher frequency. (Phyllode is a term applied to tumors that on section show a lobulated, leaf-like appearance.) The increased cancer risk was most marked at younger ages and decreased with age. The only common adult epithelial tumor apart from breast that could be found at an increased frequency was pancreas, with no increased risk of lung, ovary, bladder, bowel, or head and neck tumors. The latter observation was particularly interesting because somatic TP53 mutations are found in approximately 60% of sporadic tumors at these sites (Soussi et al., 2000). The tissue-cell specificity associated with inheritance of a germline TP53 mutation is very striking.

Bendig et al. (2004) sought to identify germline mutations in the TP53 gene in 5 index cases of German and Swiss origin with cancers typical of Li-Fraumeni syndrome. They identified 5 mutations, of which 3 were found in families with a strong history of LFS in several generations and 2 seemingly arose de novo. Bendig et al. (2004) concluded that the frequent identification of de novo germline mutations emphasizes the importance of mutation analyses of the TP53 gene in young patients with malignancies typical for LFS but without a positive family history of this tumor syndrome.

Wang et al. (2013) reported on members of families with Li-Fraumeni syndrome who carried germline mutations in the TP53 gene. As compared with family members who are not carriers and with healthy volunteers, family members with these mutations have increased oxidative phosphorylation of skeletal muscle. Basic experimental studies of tissue samples from patients with the Li-Fraumeni syndrome and a mouse model of the syndrome supported this in vivo finding of increased mitochondrial function. Wang et al. (2013) concluded that their results suggested that p53 regulates bioenergetic homeostasis in humans.

Other Genetic Abnormalities

Shlien et al. (2008) found that patients with TP53 mutations had significantly higher numbers of germline copy number variation (CNV) compared to controls. A microarray analysis counting CNVs on autosomal chromosomes comprising of 2 or more SNP probes identified 3,884 CNVs in genomic DNA from 770 healthy individuals. The median number of CNVs detected per control individual was 3, with 75% of the population having 4 or fewer CNVs. In individuals from 11 LFS families with TP53 mutations, the CNV mean was 12.19 per person, with 75% having 10 or fewer CNVs. The difference from controls was statistically significant (p = 0.01). The majority of specific CNVs observed in LFS families were acquired and not found in either parent. There was also a correlation between increased CNV number and cancer development among the LFS families, and analysis of tumor samples showed somatic increases in CNV. Shlien et al. (2008) suggested that CNVs represent regions of genomic instability and early neoplastic transformation.

Shlien et al. (2010) screened 4,524 patients with diverse clinical phenotypes for DNA dosage changes via array CGH or MLPA and identified 8 probands with a microdeletion on chromosome 17p13.1, at the TP53 (191170) locus. In 4 of the patients, who had childhood cancer and pedigrees consistent with Li-Fraumeni syndrome, deletions limited to the TP53 gene were found which deleted between 1 and 10 of the 11 exons. Another 4 patients, with a noncancer neurocognitive phenotype (613776), had larger deletions at 17p13.1, encompassing the entire TP53 gene and 26 to 85 other fully deleted genes. Shlien et al. (2010) demonstrated that mRNA expression levels of TP53 and TP53-dependent genes were altered in patients with partial, but not complete, deletions, which was consistent with mutant TP53-initiated tumorigenesis in the former group but not in the latter. The authors stated that their data supported a model in which partial deletions lead to the expression of a truncated protein, rather than the complete absence of it due to nonsense-mediated decay. Truncated and wildtype protein would oligomerize to form a defective TP53 tetramer, leading to a dominant-negative or gain-of-function effect similar to that observed with certain missense mutations, resulting in inhibition of wildtype TP53 function.

Genotype/Phenotype Correlations

Olivier et al. (2003) described a database for collecting information on families carrying a germline mutation in the TP53 gene and on families affected with Li-Fraumeni syndromes, both Li-Fraumeni and Li-Fraumeni-like syndromes. Data from the published literature was included. They described analysis of 265 families/individuals with LFS/LFL. In classic LFS families with a germline TP53 mutation (83 families), the mean age of onset of breast cancer was significantly lower than in LFS families (16 families) without a TP53 mutation (34.6 vs. 42.5 years; P = 0.0035). In individuals with a TP53 mutation, a correlation between the genotype and phenotype was found: brain tumors were associated with missense TP53 mutations located in the DNA-binding loop that contact the minor groove of DNA (P = 0.01), whereas adrenal gland carcinomas were associated with missense mutations located in the loops opposing the protein-DNA contact surface (P = 0.003). Finally, mutations likely to result in a null phenotype (absence of the protein or loss of function) were associated with earlier onset brain tumors (P = 0.004). These observations were considered to have clinical implications for genetic testing and tumor surveillance in LFS/LFL families.

Capponcelli et al. (2005) identified a mutation in the TP53 gene (Y220S; 191170.0039) in a mother and her 3 children with Li-Fraumeni syndrome. All affected family members had a very aggressive clinical phenotype associated with resistance to doxorubicin and early death from cancer. In vitro studies showed that the mutation conferred increased cellular resistance to doxorubicin treatment, perhaps by inducing expression of peroxiredoxin II (PRDX2; 600538) and thioredoxin (TXN; 187700), both of which reduce reactive oxygen species.

Other Features

Choriocarcinoma

Patrier-Sallebert et al. (2015) reported a gestational choriocarcinoma (CC) that developed in a female partner of a male patient with LFS; the CC carried a germline TP53 (191170) mutation initially detected in this LFS patient. The authors then identified 78 fathers who were carriers of a germline TP53 mutation. Among the 213 corresponding pregnancies, Patrier-Sallebert et al. (2015) found 2 other cases of gestational CC in the female partners, and estimated that gestational CC occurs in approximately 1% of the deliveries in female partners of TP53 mutation carriers.

Heterogeneity

Mutations in p53 had not been detected in approximately 30% of LFS families. To address the possibility either that TP53 mutations were missed or that another predisposing gene is altered in LFS, Evans et al. (1998) used a variety of methods to determine the TP53 status in a large LFS kindred. A transcriptional activation assay on exons 4 to 10 of TP53 excluded a mutation within the DNA-binding domain. SSCP analysis, using intronic primers and sequencing of all the coding exons and intron/exon junctions, also yielded no mutations. Finally, linkage analysis excluded potential mutations in the noncoding regions of TP53. The family had been ascertained through systematic surveys of cancer in relatives of 382 childhood osteosarcoma patients. Classic LFS was defined as the diagnosis of a sarcoma in an individual before 45 years of age, having 2 first-degree relatives with cancer before the age of 45 years. In the LFS kindred studied, in addition to osteosarcoma in the proband, a sister had osteosarcoma and the mother had breast cancer before the age of 45 years. There was also lung cancer in the family as well as cancers of the ovary, thyroid, tongue, and kidney. Thirteen members were affected in 7 sibships in 3 generations.

Lynch et al. (2000) reported extensive follow-up of the family described by Lynch et al. (1978); a remarkable excess of brain tumors became evident in the update. One patient in the direct genetic lineage had a rhabdomyosarcoma of the eyelid at age 29 months and, at age 14 years, was diagnosed with lymphoblastic lymphoma/acute lymphoblastic leukemia. This same patient also had Sturge-Weber syndrome (185300), which seemingly had not previously been identified in Li-Fraumeni syndrome. No p53 germline mutation was identified in any affected members of this family.

In affected members of a large pedigree with tumors suggestive of a Li-Fraumeni-like syndrome, in which mutation in the TP53 gene had been excluded, Evans et al. (2008) identified a deletion of exons 14 to 16 in the BRCA2 gene by multiplex ligation-dependent probe amplification (MLPA) assay. Noting that other BRCA2 families have been reported with LFS spectrum sarcomas, the authors concluded that BRCA2 clearly accounts for a proportion of LFS/LFL families negative for TP53 mutations, but that it is likely that TP53 is the only LF-specific gene and that TP53-negative families are due to mutations in a variety of other, mostly known, genes.

Exclusion Studies

In a Turkish family with Li-Fraumeni syndrome, Guran et al. (1999) demonstrated that the propositus with seminoma and his daughter with medulloblastoma had a hereditary TP53 mutation, lys292 to ile (K292I; 191170.0034), but also in analyses of tumor tissues had an ala94-to-glu missense mutation of the CDKN2A gene. Full blood analysis in the 2 cases revealed no CDKN2A mutation. This was the first time that a mutation in CDKN2A had been observed in Li-Fraumeni syndrome. Burt et al. (1999) excluded CDKN2A (600160) on chromosome 9 and PTEN (601728) on chromosome 10 as the cause of either LFS or LFL.

Animal Model

The heterogeneity in the tumor spectrum and latency in patients with LFS due to inherited mutations in p53 suggest risk modifiers at loci other than the major gene. Evans et al. (2004) developed a mouse model to investigate these risk modifiers. Inbred CE/J mice, which succumbed to multiple types of tumors similar to those found in LFS, were crossed with the p53 null mouse. In this cross, the authors found evidence for a genetic modifier of p53, Mop1, based on an unexpected mix of genotypes in the F2 progeny. A model in which a recessive CE/J allele in combination with p53 heterozygosity or homozygosity results in lethality most closely fitted the data. Using simple sequence length polymorphism analysis of the entire genome, Evans et al. (2004) identified a putative chromosomal region for this modifier of p53 on mouse chromosome 11 centromeric to p53.

History

Bachinski et al. (2005) studied a series of LFS kindreds with no p53 or CHEK2 (604373) mutations. Using a genomewide scan for linkage with complementing parametric and nonparametric analysis methods, they mapped a novel locus, previously designated Li-Fraumeni syndrome-3 (LFS3), to a 4-cM region on chromosome 1q23.