Epidemiology

The incidence of FA is difficult to ascertain. The Fanconi Anemia Research Fund, Inc., estimates that about 30 persons with FA are born in the United States each year; this would indicate that the prevalence of FA is about 2.5 per million and the carrier frequency is approximately 1 in 180. The male-to-female ratio is 1.2 : 1. FA has been reported in all races and ethnic groups. The world’s highest prevalence of FA, caused by a founder effect, is seen in Spanish gypsies, who have a carrier frequency of 1 in 70. A high prevalence of FA is also found in individuals of Ashkenazi Jewish decent, who have an approximate heterozygous frequency of 1 in 77, and in South African Afrikaans, in whom carrier frequency is estimated to be 1 in 83.

Clinical Manifestations

The clinical manifestations of FA are heterogeneous (variable penetrance and expressivity; for definitions, see Box 7-2 ). Affected family members can have a wide variety of abnormalities (variable expressivity). Even monozygotic twins have been described to differ in their physical and hematologic manifestations. However, the occurrence of malformations is nonrandom. Siblings are usually concordant for the presence or absence of multiple congenital abnormalities; likewise, the age at onset of hematologic manifestations is similar within an individual family. Thus the variability in clinical findings appears to be to some extent linked to the different genotypes. The emergence of revertant cells leading to hematopoietic mosaicism further contributes to disease variability (see the section on Somatic Mosaicism, Functional Reversion, and Attenuation of Fanconi Anemia Cells). Originally, the diagnosis of FA was based on the presence of both aplastic anemia and a specific, though broad, range of physical abnormalities. With the availability of more sensitive and specific tests for the diagnosis of FA, including the availability of genetic testing, it has become evident that a significant proportion of patients lack congenital abnormalities (25% to 40%) or fail to develop aplastic anemia. Nevertheless, these individuals are still at risk for late complications such as leukemia and cancer.

Congenital Abnormalities

The frequency and nature of congenital abnormalities within the FA population is summarized in Box 7-3 . Figure 7-2, A shows a 1-year-old girl with FA, and Figure 7-2, C shows a 12-year-old boy with several classic phenotypic features, including short stature, microencephaly, broad nasal bridge, epicanthal folds, micrognathia, café-au-lait spots and hypopigmentation, and one absent thumb and one triphalangeal thumb. FA patients may have a combination of the congenital anomalies listed or none of them. The phenotypic variability makes a diagnosis of FA difficult when it is based only on clinical features.

A , Infant female with Fanconi anemia, who exhibited several classic phenotypic features, including short stature, microcephaly, broad nasal base, epicanthal folds, micrognathia, and café-au-lait spot (left forehead). B , Cytogenetic findings in peripheral blood lymphocytes in Fanconi anemia. C and D , A 12-year-old boy with Fanconi anemia. Classic phenotypic features are microcephaly, broad nasal base, epicanthal folds, micrognathia, café-au-lait spot (chin and left temple), hyperpigmentation, and an absent thumb on one hand and a triphalangeal thumb on the other hand.

( B , Courtesy of Shashikanat Kulkarni, Washington University.)

Endocrinopathies

Data from the International Fanconi Anemia Registry and from the Cincinnati FA cohort indicate that endocrine abnormalities are present in about 80% of children and adults with FA, primarily glucose/insulin abnormalities with a blunted first-phase insulin release, mild hypothyroidism, and male hypogonadism.

Hematologic Abnormalities

Progressive BMF is one of the hallmarks of FA, although the hematologic findings vary. Thrombocytopenia and macrocytosis often precede anemia and neutropenia. Elevated hemoglobin F levels and elevated expression of “i” antigen on red cells usually coincide with macrocytosis and are suggestive of stress hematopoiesis. Serum alpha-fetoprotein levels are consistently elevated in FA patients irrespective of the presence of liver abnormalities. Erythropoietin levels are usually elevated in anemic FA patients. The first hematologic abnormalities in individuals with FA are detected at a median age of 7 years. The majority of FA patients already have pancytopenia at the time of diagnosis (53%). By the age of 40, the cumulative incidence of hematologic abnormalities is 90% to 98%.

In rare cases, thrombocytopenia may be present at birth and progress to pancytopenia in the neonatal period or infancy. Thus FA also has to be considered in cases of aplastic anemia in neonates and infants. In other patients, hematologic abnormalities never become clinically evident, and the diagnosis of FA is made later in life as a result of late FA-associated complications such as cancer or leukemia. The emergence of revertant cells leading to hematopoietic mosaicism is often associated with a mild or absent hematologic phenotype (see later). Finally, in a small proportion (2% to 5%) of FA patients, MDS or leukemia is diagnosed but the underlying BMF never becomes clinically apparent. Bone marrow examination generally shows reduced cellularity; however, it may also be normocellular or hypercellular, particularly in disease that evolves into MDS or MDS/acute myeloid leukemia (AML). The cumulative incidence of BMF is approximately 50% by 40 to 50 years of age. BMF is usually progressive. Fifty percent of individuals who are initially found to have thrombocytopenia progress to pancytopenia within 3 to 4 years.

Myelodysplastic Syndrome and Acute Leukemia in Patients with Fanconi Anemia

The cumulative incidence of MDS in FA patients is approximately 30%, and of acute leukemia about 10% by the age of 40 to 50 years. In the literature the definition of MDS and acute leukemia in FA patients varies, which makes a strict comparison difficult. Clonal hematopoiesis, as suggested by nonrandom X-inactivation, may be present even before the onset of MDS. Clonal cytogenetic abnormalities are found in 34% to 48% of FA patients. Fluctuations are frequent, including the disappearance of clones, the appearance of new clones, and clonal evolution. Chromosomal abnormalities in MDS most frequently involve chromosomes 1q+, 3q+, −7/7q, 11q− 20q−, 6, 13, and 21q− ( Fig. 7-3 ). Individuals in whom clonal cytogenetic abnormalities develop have a reduced survival rate of 35% to 40% as opposed to 92% in FA patients without such abnormities. An exception is patients with an isolated 1q+ that may be present without morphologic MDS and is not associated with progression to MDS/AML. MDS morphology, as defined by dysplasia in multiple blood cell lineages, portends a poor outcome, with a 5-year survival rate of 9%. Gain of the distal portion of chromosome 3q(26-29) is frequently present in FA patients and is associated with a poor prognosis, decreased survival times, and an increased frequency of MDS/AML.

A recurrent pattern of somatic copy number abnormalities and uniparental disomy (UPD) in Fanconi anemia bone marrow cells. All acquired copy number and UPD regions found in the bone marrow cells of 57 individuals with Fanconi anemia are shown. Green lines represent deletions; red lines, gains; and blue lines, regions of UPD (somatic copy-neutral loss of heterozygosity). Constitutional copy-neutral homozygosity regions are not shown. The RUNX1 gene location at 21q22 is shown.

(From Quentin S, Cuccuini W, Ceccaldi R, et al: Myelodysplasia and leukemia of Fanconi anemia are associated with a specific pattern of genomic abnormalities that includes cryptic RUNX1/AML1 lesions. Blood 117:e161–170, 2011.)

Leukemia is the most common malignant disease in patients with FA. AML is the most frequent leukemic phenotype, although patients with acute lymphocytic leukemia and chronic myelomonocytic leukemia have been described. Leukemia in patients with FA shows many molecular and clinical similarities to leukemia occurring in other IBMFSs; cases of FA-associated leukemia are notoriously difficult to treat, and most patients die within 6 months of diagnosis. Leukemia in patients with FA may arise de novo or develop from MDS. In some FA patients, AML may be the first hematologic manifestation of disease. Leukemic cells often show complex cytogenetic abnormalities, including abnormalities similar to those seen in MDS. Chromosomal translocations and mutations in oncogenes and tumor-suppressor genes frequently found in de novo AML are rare in FA MDS/AML with the exception of RUNX1 mutations, which are seen in approximately 20% of FA patients with MDS/AML. Abnormalities of chromosome 7 (monosomy, isochromosome, or other structural rearrangement), followed by abnormalities of 1q (usually duplications), are frequent in FA-associated leukemia.

Predisposition to Malignancy

The risk of cancer developing is about 500 to 700 times higher than in the normal population ( Table 7-2 ). By the age of 40 years, the cumulative incidence of solid tumors is estimated to be 26% to 30%. The median age at development of cancer is 16 years, versus 68 years for the same types of cancer in the general population. The risk of solid tumors and leukemia developing increases with age. Thus patients with mild hematologic and phenotypic abnormalities and the longest survival times are at the highest risk for the development of tumors or leukemia. However, with increased survival after hematopoietic stem cell transplantation (HSCT), the frequency of cancer might also increase in patients with more severe disease. The most frequent cancers occur in the neck, head, and upper esophagus, followed by the vulva, anus, and lower esophagus. Human papillomavirus (HPV), particularly HPV16, is frequently isolated from patients with FA. Although the FA DNA repair pathway is involved in limiting HPV replication and patients with FA are frequent carriers of HPV, the role of HPV in the pathogenesis of squamous cell cancer in patients with FA remains controversial. HPV vaccination at an early age should be considered for all patients with FA. The risk for development of squamous cell cancer of the head and neck and particularly the tongue is higher in FA patients after HSCT than in those without HSCT and correlates with the presence and severity of graft-versus-host disease. Adenoma of the liver and hepatocarcinoma were found mainly in patients undergoing androgen therapy. Hepatoadenomas usually regress after cessation of androgen therapy.

The risk and type of cancer vary for different FA subtypes (also see the section on Risk of Malignant Disease). FA patients belonging to complementation group D1 (FA-D1) with biallelic mutations in FANCD1/BRCA2 have a high risk for the development of malignant disease at a very young age, frequently before manifestations of BMF appear, and the cumulative risk for AML or other cancers is 97% by the age of 5.2 years. In addition, the spectrum of cancer is distinct, with medulloblastoma being the predominant malignant disease, possibly because BRCA2 participates in DNA repair outside the FA pathway. An increased risk for childhood cancer with a very similar spectrum has also been found in families belonging to the FA-N subtype. FANCN/PALB2 binds to BRCA2 and facilitates its binding to chromatin. The interaction between these two proteins explains the similar cancer predisposition in FA individuals belonging to the FA-D1 and FA-N subgroups.

All FA cancer patients commonly have a low tolerance for DNA-damaging chemotherapeutic agents. Accordingly, chemotherapeutic regimens often need to be modified, given at low dosage, or avoided in favor of alternative or surgical approaches.

Diagnosis

The most widely used diagnostic test for FA is hypersensitivity to the clastogenic (chromosome-breaking) effect of diepoxybutane (DEB) or mitomycin C (MMC). The increased spontaneous chromosomal instability of FA cells leads to distinctive chromosomal breaks and gaps and various chromatid interchanges, which were previously used as a cellular marker for FA. The frequency of spontaneous chromosomal instability in FA cells is highly variable, and it is thought to be responsible for the increased susceptibility of FA patients to cancer. The sensitivity of FA cells to the clastogenic effect has been studied for numerous cross-linking agents, but DEB and MMC are the agents most widely used for the diagnosis of FA. Figure 7-2, B shows the increased spontaneous chromosome fragility in peripheral lymphocytes in a patient with FA. The diagnosis of FA is made when DEB-treated cultured lymphocytes demonstrate a 3- to 10-fold increase in chromosomal breakage over normal controls. The tests are highly sensitive and fairly specific for FA and are used for the diagnosis of FA before the development of clinical disease and for prenatal diagnosis. Possible rare exceptions that can result in a pattern of DEB- or MMC-induced chromosomal breakage similar to that found in FA patients are individuals with specific mutations in the NBS1 gene in Nijmegen breakage syndrome and patients with cohesinopathies, that is, those with Roberts syndrome and Warsaw breakage syndrome. However, more frequently, the presence of revertant cells in the specimen may complicate the interpretation of DEB or MMC test results (see the section on Somatic Mosaicism, Functional Reversion, and Attenuation of Fanconi Anemia Cells). It is estimated that 10% to 25% of FA patients exhibit two populations of phytohemagglutinin-stimulated lymphocytes, one that is sensitive to the clastogenic effects of DNA cross-linking agents and one that is resistant to them. Thus the presence of only a few metaphases with multiple structural chromosomal changes might indicate FA with somatic mosaicism. In these cases, demonstration of increased chromosomal breakage in DEB- or MMC-treated fibroblasts might help confirm the diagnosis of FA.

Immunoblotting for FANCD2 Monoubiquitination

Biallelic mutations in 10 of the 15 FA genes ( FANC genes) lead to failure of FANCD2 monoubiquitination, and failure of FANCD2 monoubiquitination is specific for a defect in the FA pathway (see the section on Function of FANC Proteins). Testing for FANCD2 monoubiquitination by immunoblotting of peripheral blood lymphocytes and fibroblasts is therefore used in some laboratories as an alternative first-line tool for the diagnosis of FA. Biallelic mutations in FANCD1/BRCA2 , FANCJ , FANCN/PALB2 , FANCO/RAD51C, and FANCP/SLX4 that have normal FANCD2 monoubiquitination but an abnormal DEB or MMC test (or both) are missed with this assay and, therefore, additional testing is required.

Genetic Testing for FANC Gene Mutations

Only the identification of biallelic mutations in a FANC gene definitively confirms the diagnosis of FA. Currently genetic testing for FANC gene mutations is not usually recommended as a first-line tool for the diagnosis of FA. Possible exceptions are families with a known FANC gene mutation and FA in a population with a high frequency of a founder mutation, for example, a case of FA in Spanish gypsies. The indications, advantages, and difficulties of genetic testing for biallelic mutations in FANC genes are discussed later. As costs decrease, targeted or whole-genome next-generation sequencing is likely to become the diagnostic test of choice for patients clinically suspected to have FA (see the section on Molecular Diagnosis of Fanconi Anemia).

Cellular Phenotype of Fanconi Anemia Cells

In addition to increased spontaneous and induced chromosomal instability, FA cells have several other phenotypic characteristics ( Table 7-3 ). FA cells have increased sensitivity to other DNA-damaging agents, such as ionizing radiation and oxygen radicals. FA cells have an abnormality in cell cycle distribution, with an increased number of cells with 4N DNA content, suggesting a delay in the G ₂ /M or late S phase (or both) of the cell cycle. FA cells have a failure in cytokinesis forming an increased number of ultrafine DNA bridges (UFB) during the M phase of the cell cycle, leading to increased chromosome instability and binucleated cells that undergo apoptosis. The increase in cells with 4N DNA content is accentuated after treatment with interstrand DNA cross-linking agents. Flow cytometric measurements of the increase in the proportion of cells with 4N DNA content after such treatment may be used as an additional criterion for the diagnosis of FA. However, similar results have been obtained in patients with ataxia-telangiectasia. FA cells show accelerated telomere shortening in vitro and an increase in telomere free ends and chromosome end fusions, suggesting a defect in maintenance of telomere integrity and increased breakage at telomeric sequences. Cells deficient in the FA repair pathway are hypersensitive to both formaldehyde and acetaldehyde.

Differential Diagnosis

Clinically, FA has to be differentiated from other IBMFSs, particularly dyskeratosis congenita (DC), Diamond-Blackfan anemia (DBA), Shwachman-Diamond syndrome (SDS), Wiskott-Aldrich syndrome (WAS), thrombocytopenia with absent radii (TAR), and other nonclassified forms of BMF. Clinical manifestations, associated physical abnormalities, family history (pattern of inheritance), and increased chromosomal fragility in the DEB or MMC test performed on cultured peripheral lymphocytes or fibroblasts, or both, are helpful in strengthening the diagnosis of FA. However, only identification of pathogenic mutations on both copies of an FA gene (biallelic mutation) may definitely consolidate the diagnosis of FA. Congenital abnormities in FA may have an overlap with malformations seen in other disorders with genetic and nongenetic causes. Thrombocytopenia and radial ray abnormalities are also characteristic of TAR. However, in FA, if the radii are affected, the thumbs are always abnormal; in TAR, in which the radii are absent, thumbs are always present. Other syndromes that have phenotypic overlap with FA include Holt-Oram syndrome (MIM 142900), which is characterized by thumb anomalies and atrial septal defects and caused by mutations in the TBX5 gene, and Baller-Gerold syndrome (MIM 218600) and Rothmund-Thomson syndrome (MIM 278400), which are caused by mutations in the RECQL4 gene and characterized by radial defects, craniosynostosis, and cancer predisposition but not BMF. FA patients may present with VATER (vertebral defects, imperforate anus, tracheoesophageal fistula, and radial and renal dysplasia) or VACTERL (vertebral abnormalities, anal atresia, cardiac abnormalities, tracheoesophageal fistula and/or esophageal atresia, renal agenesis and dysplasia, and limb defects) (MIM 192350) or the IVIC syndrome (MIM 147750), which is an acronym for Instituto Venezolano de Investigaciones Cientificas, the institution at which a family with autosomal dominant radial ray defects, hearing impairment, internal ophthalmoplegia, and thrombocytopenia was first described. In patients presenting with these associated anomalies, FA should be excluded.

The increased chromosomal breakage in FA has to be distinguished from that seen in Bloom syndrome (MIM 210900), another single-gene disorder with a predisposition for leukemia as a result of mutation in the helicase RecQ protein-like-3 gene ( RECQL3 ). Chromosomal aberrations are in general more numerous in FA cells than in Bloom cells. Furthermore, in Bloom syndrome, most interchanges occur between homologous chromosomes, whereas in FA they are more frequently seen between nonhomologous chromosomes.

Nijmegen breakage syndrome (NBS; MIM 251260), caused by mutations in the NBS1 gene and characterized by immunodeficiency and a predisposition to lymphoma, is another autosomal recessive chromosomal breakage disease. A small number of patients with NBS may have FA-like features, including BMF and the development of myeloid leukemia, and specific mutations in the NSB1 gene cause an increased clastogenic response to MMC and DEB testing.

A third disease associated with spontaneous chromosomal breakage and hematopoietic malignant disease is ataxia-telangiectasia (MIM 208900), which is inherited in an autosomal recessive manner. Cells from patients with ataxia-telangiectasia are hypersensitive to ionizing radiation but not to cross-linking agents. Furthermore, the disease is associated with immunodeficiency and progressive cerebellar neuronal degeneration, which clearly distinguish this condition from FA.

Seckel syndrome ( SCKL1 , MIM 210600), caused by a defect in the signaling pathway dependent on ataxia-telangiectasia and Rad3-related (ATR), may have some clinical similarities to FA, including anemia and an increased risk for AML.

Patients deficient in DNA ligase IV ( LIG4, MIM 606593) are characterized by microcephaly, growth retardation starting in utero, distinctive facial appearance (birdlike face), developmental delay, immunodeficiency, pancytopenia, and pronounced clinical and cellular radiosensitivity. LIG4 deficiency is another rare IBMFS caused by defective DNA repair.

Two of the cohesinopathies, Roberts syndrome (MIM 268300) (caused by mutations in ESCO2, encoding an acetyltransferase involved in sister chromatid cohesion) and Warsaw breakage syndrome (MIM 613398) (caused by mutations in DDX11/ChlR1, encoding an XPD-like DNA helicase), show excessive MMC-induced chromosomal breakage in primary lymphocyte cultures and an increased arrest of primary skin fibroblasts in the G ₂ /M phase of the cell cycle that need to be distinguished from FA.

Genetic Characteristics

Fanconi Genes ( FANC Genes).

FANC genes have been identified by expression cloning, positional cloning, protein complex purification, or a combination thereof and are summarized in Table 7-4 . Figure 7-4 schematically represents the 15 FANC genes and their functional domains.

Schematic representation of the Fanconi anemia genes and their functional domains. BTB/POZ , Broad-complex, tramtrack, and bric-a-brac/poxvirus and zinc finger; CCD , conserved C-terminal domain; dsDNA , double-stranded DNA; MLR , MEI9 ^XPF -interaction–like region; NES , nuclear export sequence; NLS , nuclear localization signal; OB , oligonucleotide/oliosaccharide-binding fold; ssDNA , single-stranded DNA; TD , tower domain; TRP , tetratricopeptide repeat; UBZ , ubiquitin-binding zinc-finger domain.

(Modified from Taniguchi T, D’Andrea AD: Molecular pathogenesis of Fanconi anemia: recent progress. Blood 107:4223–4233, 2006.)

Function of FANC Proteins.

The FANC proteins have been shown to act mainly during the S phase of the cell cycle. Their main role is in the repair of DNA interstrand cross-links that impede replication and transcription by inhibiting DNA strand separation. It is generally thought that FANC proteins sense, promote, and coordinate three DNA repair pathways: Repair is initiated by recognition of a lesion by nucleotide excision repair components, followed by incisions made flanking the interstrand cross-links; the gapped intermediate that is produced is filled by translesion synthesis or by repair of broken replication forks by homologous recombination. Thus it is thought, and schematically summarized in Figure 7-5 , that in normal cells the FA pathway serves to minimize the severity of mutational outcomes by favoring error-free DNA repair and error-prone DNA repair that results in base pair substitutions or small deletions over DNA repair that results in larger deletions and chromosomal rearrangements ( Box 7-4 summarizes the main DNA repair pathways).

Presumed function of the Fanconi anemia DNA repair pathway. In normal cells the Fanconi anemia pathway ensures correct DNA repair by homologous recombination, whereas error-prone alternative pathways using nonhomologous end joining are rarely used. In Fanconi anemia cells the error-free DNA repair by homologous recombination is impaired, so alternative DNA repair pathways are used more frequently, leading to genomic instability and cell death. Suppression of nonhomologous end joining by deleting a component of this pathway or by using DNA-dependent protein kinase catalytic subunit inhibitors promotes the repair of the double-strand breaks by FA-independent pathways of homologous recombination and improves in some experimental systems the sensitivity to DNA cross-linking agents (dashed arrow) . HR, Homologous recombination; ICLs, intrastrand cross-links; NHEJ, non-homologous end joining.

(Modified from Deans AJ, West SC: DNA interstrand crosslink repair and cancer. Nat Rev Cancer 11:467–480, 2011.)

Box 7-4

DNA Repair Pathways

DNA repair pathways are highly regulated pathways that involve detection of DNA lesions, signaling to recruit DNA repair factors, activation of DNA repair enzymes, and disassembly or degradation of DNA repair factors after repair.

• •
  

  EXCISION OF DAMAGED AND MISMATCHED BASES

  
  
  
  
  • •
    

    Nucleotide excision repair, including global genome repair and transcription-coupled repair

    
    
  • •
    

    Base pair excision repair

    
    
  • •
    

    Mismatch repair

  
  
  
  
• •
  

  REPAIR OF DOUBLE-STRANDED BREAKS

  
  
  
  
  • •
    

    Homologous recombination (generally error free)

    
    
  • •
    

    Nonhomologous end joining (generally error prone)

  
  
  
  
• •
  

  DNA STRESS REPLICATION PATHWAY, REPLICATION PATHWAY OF DAMAGED TEMPLATES DURING DNA SYNTHESIS

  
  
  
  
  • •
    

    Translesion DNA synthesis (error prone)

  
  

The current model suggests that FANCM recognizes the DNA interstrand cross-links and initiates DNA repair by recruiting eight FA core complex proteins (FANCA, FANCB, FANCC, FANCE, FANCF, FANCG, FANCL, and FANCM) to the DNA damage site, stabilizing the stalled replication fork, and initiating ATR mediated checkpoint signaling. The FA core complex functions as an ubiquitin ligase, of which FANCL is the catalytic subunit. FANCE recruits FANCD2 to the FA core complex, whereas disease-associated FANCE mutations disrupt the FANCE-FANCD2 interaction. Figure 7-6 schematically summarizes the current view of the FANC protein interconnections in the FA DNA repair pathway. During activation, multiple FA proteins undergo phosphorylation by ATR-CHK1 checkpoint kinases, highlighting the close interconnection of the FA pathway with DNA damage response signaling. The FA core complex catalyzes monoubiquitination of the FANCI-FANCD2 complex, which is a key regulator of the FA pathway. A defect in any of these core proteins results in failure to monoubiquitinate the FANCI-FANCD2 complex, underscoring the importance of this modification in the FA pathway. FANCI and FANCD2 are dependent on each other for their respective monoubiquitination. The FANCL subunit is a ubiquitin E3 ligase, and in concert with the UBE2T E2 enzyme, it conjugates a single ubiquitin to FANCD2 and to FANCI. The monoubiquitinated FANCI-FANCD2 complex is relocalized to the DNA lesion, where it coordinates cross-link repair together with downstream FA proteins. Monoubiquitinated FANCD2 acts as a platform to recruit multiple nucleases, including FANCP, to coordinate nucleolytic incisions flanking the interstrand cross-links. The incision creates a DSB, which is repaired by homologous recombination. Downstream FA proteins, including FANCD1, FANCJ, FANCN, and FANCO, promote RAD51-dependent strand invasion and the resolution of homologous recombination intermediates.

Schematic overview of protein interconnections in the Fanconi anemia DNA repair pathway (see also text). The Fanconi anemia pathway maintains genomic stability in response to stalled replication forks, particularly in the context of interstrand DNA cross-links (DNA ICL). Fork stalling is detected by the FANCM sensor complex and triggers ATR (ataxia-telangiectasia and rad3-related)–mediated checkpoint signaling, resulting in the phosphorylation of several Fanconi anemia proteins. FANCM becomes a platform for the other Fanconi anemia core complex proteins, including the FANCL E3 ubiquitin ligase. The core complex monoubiquinates FANCD2 and FANCI (Ub) to promote interstrand DNA cross-link repair. The incision complex, including FANCP/SLX, is essential for resolving the cross-link. Repair and replication fork restart is further accomplished using translesion synthesis and homology-directed repair, which involves BRCA2/FANCD1, PALB2/FANCN, and RAD51C/FANCO. The USP1-UAF1 DUB complex removes monoubiquitin from FANCD2-I and completes the repair.

(Modified from Kee Y, D’Andrea AD: Molecular pathogenesis and clinical management of Fanconi anemia. J Clin Invest 122:3799–3806, 2012; and Hays LE, Meyer S, van de Vrugt HJ: A molecular, genetic, and diagnostic spotlight on Fanconi anemia. Anemia 2012:650730, 2012. For further review see also Crossan GP, Patel KJ: The Fanconi anaemia pathway orchestrates incisions at sites of crosslinked DNA. J Pathol 226:326–337, 2012; Deans AJ, West SC: DNA interstrand crosslink repair and cancer. Nat Rev Cancer 11:467–480, 2011; Kim H, D’Andrea AD: Regulation of DNA cross-link repair by the Fanconi anemia/BRCA pathway. Genes Dev 26:1393–1408, 2012; Kottemann MC, Smogorzewska A: Fanconi anaemia and the repair of Watson and Crick DNA crosslinks. Nature 493:356–363, 2013.)

The FA core complex is assembled sequentially from subcomplexes, which may have additional functions aside from DNA repair. Several FA proteins, including FANCA, FANCE, FANCG, FANCD2, FANCI, and FANCM, are posttranslationally modified by phosphorylation, and two FA proteins, FANCC and FANCD2, are regulated through a caspase-mediated proteolytic process. FANC proteins have been reported to interact with a number of other proteins, including proteins involved in transcription, the cell cycle, oxidative metabolism, cell signaling, and cytokine response, suggesting that FANC proteins/subcomplexes might have a separate role in cell cycle control, apoptosis, hematopoiesis, and tumorigenesis that is independent of DNA damage recognition/repair.

Bone Marrow Failure in Fanconi Anemia

Although the impairment of DNA repair seems to explain the genomic instability and possibly the increased risk for malignant disease in patients with FA, its link to the pathogenesis of BMF is less obvious, because mutations associated with genomic instability outside the FA DNA repair pathway are not usually associated with BMF. There is an ongoing debate on whether BMF is a direct consequence of the FA DNA repair defect or is caused by impairment of functions apart from DNA repair. Naturally, FA DNA repair and non–DNA repair pathways in the pathogenesis of BMF may not be mutually exclusive but, rather, may overlap. However, two recent findings point strongly towards defective DNA repair being at least the initiating cause of BMF in FA. In a recent study Ceccaldi and colleagues demonstrated that the basal and MMC-induced levels of p53 and its downstream target p21 are elevated in the blood, bone marrow, and liver of fetuses with FA, with a simultaneous accumulation of cells in phase G ₀ /G ₁ . These findings suggest that in FA patients the pool of available hematopoietic stem cells and progenitor cells is compromised, with fewer cells and less proliferation of cells than in normal individuals. The impairment is intrinsic, occurs even in the fetus, and leads to overt BMF later in life. Although cross-linking agents have been used experimentally to demonstrate the FA defect in DNA repair, experiments by Ridpath and colleagues and Langevin and colleagues suggest that naturally occurring aldehydes might be a major source of endogenous cross-linking and FA pathway activation under steady-state conditions in vivo. Interestingly, HSCs express aldehyde hydrogenase type 2 ( Aldh2 ), a condition that is essential for the protection of HSCs against acetaldehyde toxicity. The experiments were performed in mice and need to be confirmed in patients with FA, but the findings suggest that the FA pathway is needed to protect HSCs from aldehyde-mediated genotoxicity (see the section on Mouse Models for Fanconi Anemia). Moreover, these observations explain the tissue specificity of HSC and solve the conundrum of why HSC is affected in FA but is not affected, or not to the same extent, in other non-FA DNA repair disorders.

FANC Gene Mutations in Fanconi Anemia

Detailed information on mutations identified in the FANC genes in patients with FA is maintained in the FA Mutation Database at Rockefeller University through the International Fanconi Anemia Registry (available at rockefeller.edu/fanconi/ ) and in the Leiden Open Variation Database; chromium.liacs.nl/lovd/ ). Frequent FA mutations and mutations with founder effects are presented in Table 7-5 .

FANCA is the gene most frequently mutated in FA patients. The mutations are spread throughout the gene and include missense and nonsense mutations, small deletions, insertions, and duplications, as well as splicing mutations. Of note is the substantial number of large genomic, usually intragenic DNA deletions, which account for about 40% of all pathogenic mutations within the FANCA gene and are thought to be due to the frequent occurrence of Alu repeats in this genomic region. Large gene deletions are not generally detected by routine nucleotide sequence–based mutation detection methods but require more specialized analysis (haplotype analysis) or molecular methods specifically designed for the detection of large gene deletions. Two small deletions (deletion of three nucleotides between positions 3788 and 3790, with c.3788_3790delTCT leading to deletion of phenylalanine at position 1263, 1263delF, and a 4–base pair (bp) deletion at position 1115, c.1115-1118delTTGG, leading to a frameshift and premature stop codon) are the most common pathogenic mutations in the FANCA gene, with a frequency of 8.8% and 5.5% in the non-Brazilian and 51.1% and 2.2% in the Brazilian mutated FANCA alleles (see Table 7-5 ).

Mutations in FANCC genes account for 10% to 15% of all FA cases. The splicing mutation c.711+4A→T (old nomenclature, IVS4+4A→T) in the FANCC gene accounts for the majority (80%) of FA cases in Ashkenazi Jewish families. Mutations in FANCG are responsible for about 8% to 9% of patients with FA. Several mutations with founder effects have been identified in specific populations (see Table 7-5 ).

Biallelic mutations in BRCA2 are found in individuals with FA-D1. Biallelic FANCD1 / BRCA2 mutations account for about 3% to 4% of all FA patients. The majority of FANCD1 / BRCA2 mutations in FA lead to frameshifts or truncations (see Table 7-5 ). Interestingly, however, FA-D1 cells are not biallelic for FANCD1 / BRCA2 null alleles but have at least one hypomorphic FANCD1 / BRCA2 allele expressing a FANCD1/BRCA2 protein with some residual activity. Biallelic null FANCD1 / BRCA2 mutations have never been identified and in mice are embryonically lethal. Biallelic mutations in FANCD1 / BRCA2 have a very high risk of early malignant disease. The splice-site mutations c.859+1G→A (IVS7+1G→A) and c.859+1G→A (IVS7+1G→A) are associated with a high frequency of leukemia at a very young age, whereas the frameshift mutations c.886delGT and c.6174delT are associated with brain tumors.

Mutation in FANCD2 accounts for 3% to 6% of FA-affected patients. Malformations are frequent in these patients, and hematologic manifestations appear early and progress rapidly. Although hypomorphic mutations exist, patients with FANCD2 mutations have a relatively severe form of FA. Mutation analysis revealed 66 mutated alleles, 34 causing aberrant splicing. Many mutations are recurrent and have ethnic associations and shared allelic haplotypes. There were no biallelic null mutations, suggesting that in contrast to the Fancd2 knockout mouse, complete absence of FANCD2 does not exist in patients with FANCD2 mutations.

Mutations in FANCN/PALB2 were identified in families with the new subtype FA-N and cancer in early childhood. Similar to biallelic BRCA2 mutations, biallelic FANCN/PALB2 mutations confer a high risk for childhood cancer.

Molecular Diagnosis of Fanconi Anemia

Molecular diagnosis is not routinely performed for FA, because it is expensive, labor intensive, and, in some cases, possible only in specialized laboratories using research tests that are currently not approved by Clinical Laboratory Improvement Amendments. However, only identification of biallelic mutations in a FANC gene (or hemizygosity for a mutation in the FANCB gene) definitively establishes the diagnosis of FA. Furthermore, molecular diagnostic testing enables the identification of mutation carriers, facilitates prenatal diagnosis for an existing pregnancy, and allows preimplantation genetic diagnosis for partners identified as mutation carriers. For a definition of preimplantation genetic diagnosis see Box 7-5 . In addition, because the clinical manifestations and severity of disease may vary depending on the affected FANC gene and the nature of individual mutations, molecular diagnostic testing may influence medical management not only for FA individuals carrying biallelic FANC gene mutations but possibly also for heterozygous FANC gene mutation carriers (see the section on Heterzygous Carriers of FANC Gene Mutations). Currently, an educated candidate gene approach is used to identify the affected FANC gene by taking into account the origin and ethnicity of the affected FA individual. After exclusion of the most frequent mutations, it might be necessary to determine the complementation group and perform mutational screening of the entire coding sequence, including methods that will detect large deletions in compound heterozygotes. Protein expression analysis and functional assays are used to test for the pathogenicity of unclassified variants. Multiplexed next-generation sequencing with massively parallel sequencing is becoming increasingly available for clinical genetic testing and has been shown to be an effective molecular diagnostics approach for FA. With the decreasing costs of next-generation sequencing, customized targeted, whole-exome, or whole-genome sequencing is likely to become the future diagnostic standard for patients with FA, enabling genetic molecular diagnosis to be performed at a reasonable price in a short turnaround time ( Box 7-6 ). Although genetic mutation analysis is likely to be successful in the majority of FA patients, in some, the responsible gene mutations might remain unclear. In addition to the large number of mutations in FANC genes associated with FA, a number of nonpathogenic mutations and mutations of unknown pathogenic significance are identified and require further functional investigations. The multitude of FA complementation groups, the presence of pseudogenes, the heterogeneity of the mutation spectrum, and the frequency of large intragenic deletions present a considerable challenge for the current and future molecular diagnosis of FA.

Heterozygous Carriers of FANC Gene Mutations

Heterozygous carriers of FANC gene mutations are not at risk for the development of BMF. Similarly, cell lines heterozygous for FANC gene mutations do not show the chromosomal instability (either spontaneously or in response to cross-linking agents) characteristic of FA cells with biallelic FANC gene mutations. However, an increased frequency of skeletal and genitourinary abnormalities and an increased risk of malignant disease have been reported in heterozygous FANC D1/J/N/O and possibly FANCP and female FANCB gene mutation carriers.

Somatic FANC Gene Mutations in Cancer

Cancer cells often accumulate a large number of mutations within an individual cancer cell, many of which promote cell growth and survival. It is therefore thought that a cancer cell with an increased mutation rate, a mutator phenotype, will have a selective advantage in tumor growth and development. Inactivation or mutation of genes involved in maintaining genetic stability is thought to be an early event in cancer development. The FA pathway has been observed to be one of the pathways that maintains genomic stability and is targeted in several forms of sporadic cancer.

Identification of acquired abnormalities in the FA pathway might become clinically important because these tumors are expected to have increased sensitivity to cross-linking agents such as cisplatin, cyclophosphamide, MMC, and other novel agents. Thus identification of abnormalities in the FA pathway may serve as a biomarker that identifies a subgroup of patients with cancers selectively sensitive to specific chemo­therapeutic agents. Similarly, inhibition of the FA pathway in cancer may render cancer cells sensitive to chemotherapy.

Genotype-Phenotype Correlation

Clinical findings in FA are highly variable, although within an individual family the presence or absence of multiple congenital abnormalities or the age at onset of other disease manifestations is usually concordant. More recent studies have demonstrated that disease severity and manifestations are determined, at least in part, by the FANC gene affected (complementation group) and by the nature of the mutations. For example, patients with mutations in the FANCA gene usually have a milder form of disease and BMF develops at a later age than it does in FA individuals with mutations in the FANCC or FANCG gene, whereas in FA patients with biallelic FANCD1 or FANCN mutations, cancer develops even before the manifestation of BMF. Furthermore, splice-site mutations in FANCD1 are more frequently associated with the development of leukemia, whereas frameshift mutations within the same gene more frequently lead to brain tumors (see the section on FANC Gene Mutations in Fanconi Anemia). Malformations are frequent in FANCD2 patients, and hematologic manifestations appear earlier and progress more rapidly than they do in non- FANCD2 patients.

FA-B is X-linked, in contrast to all other forms of FA, which are autosomal. Thus a single mutation is sufficient to cause a phenotype and disease in males. In females, because of inactivation of one of the two X chromosomes, a phenotype is expressed only in cells with the FANCB mutation on the active X chromosome. Because X-chromosome inactivation occurs at random in early embryogenesis and, once determined, persists clonally, one might expect female FANCB mutation carriers to have a mild form of FA-B. Interestingly, however, to date no female FA-B patient with clinical manifestations of FA has been reported, and careful clinical examination of females heterozygous for FANCB gene mutations showed no manifestations of disease. Molecular studies investigating X-inactivation revealed that only the X chromosome with the normal FANCB gene is expressed in blood and bone marrow and also in fibroblasts. This finding indicates a proliferative disadvantage of cells expressing the mutant FANCB allele and complete outgrowth by phenotypically normal cells, not only in bone marrow but possibly also in all other tissues. This situation differs from other X-linked IBMFSs such as X-linked DC (see the section on Dyskeratosis Congenita). For individuals carrying a FANCB mutation, however, longer follow-up and examination of additional mutation carriers will be necessary to determine whether some FA-B cells persist in female carriers and whether these persisting FA-B cells are at risk for malignant transformation. Whether somatic FANCB inactivation can be found in malignant disease (only a single mutation on the active X chromosome is required) remains to be determined.

Somatic Mosaicism, Functional Reversion, and Attenuation of Fanconi Anemia Cells

About 10% to 25% of FA patients have been found to have mosaicism in the peripheral blood inasmuch as two populations of lymphocytes can be detected, one with the classic increased sensitivity to cross-linking agents characteristic of FA cells and the other usually a slowly increasing second population that is resistant to treatment with these agents. For a definition of genetic mosaicism see Box 7-7 . The presence of mosaicism is usually considered if the clinical course of the disease is milder than expected in comparison with individuals with the same genetic abnormality, if the course of the disease improves spontaneously rather than worsens, or in the presence of classic phenotypic abnormalities but rather mild or absent hematologic disease. The molecular mechanisms of genetic reversion include intragenic recombination (in compound heterozygotes), mitotic gene conversion (nonreciprocal exchange of genetic information as a result of heteroduplex formation between nonsister chromatids), DNA slippage, second-site compensating mutations, and site-specific correction by an as yet unknown mechanism. Gene reversion may occur in vivo or in vitro. In FA patients the reverting mutation is thought to occur in progenitor cells with self-renewing capability, possibly stem cells, thereby conferring a proliferative advantage to the revertant cells; this may lead to expansion of the revertant clone and progressive replacement of the FA cells in bone marrow. Genetic reversion in FA has been documented to date only in hematopoietic cells, not in fibroblasts. However, the fact that females heterozygous for FANCB mutations have biased X-inactivation in fibroblasts suggests that an intact FA pathway confers a proliferative advantage to nonhematopoietic cells. Thus the frequency of genetic reversion is largely unknown and so are its clinical consequences. In many patients genetic reversion is associated with hematologic improvement. However, genetic reversion has also been found in leukemic cells from FA patients, indicating that genetic reversion may confer a selective growth advantage and resistance of leukemic or preleukemic cells to specific chemotherapy drugs. Furthermore, genetic reversion in hematopoietic cells has no effect on cancer frequency in nonhematopoietic tissues. The strong selection for a functional FANC pathway may be used to advantage in gene therapy approaches. Genetic mosaicism is responsible at least in part for phenotypic variability, particularly within an individual family, and may create difficulty in interpretation of the chromosomal breakage testing with DEB and MMC (see the section on Diagnosis) or even lead to false-negative testing results. DEB or MMC testing of fibroblasts should therefore be considered in patients with clinical or genetic suspicion of FA but a negative DEB or MMC test result for peripheral blood lymphocytes.

The term attenuation has been introduced by Ceccaldi and associates for FA patients whose peripheral blood cells lack monoubiquitination of FANCD2 and have the characteristic increased chromosomal breaks after MMC treatment but whose cells, unlike those of most patients, do not show the characteristic G ₂ arrest after MMC treatment. Most of these patients developed MDS or leukemia later in life. Attenuation was accompanied by clonal hematopoiesis, implying that all of the peripheral blood cells were derived from a single progenitor cell in which a molecular event took place that gave rise to attenuated G ₂ arrest. The molecular mechanism underlying attenuation remains to be determined. Attenuation in patients with FA contributes to the variability in disease expression and identifies patients with a milder presentation of BMF but who might need continuous, intensified tumor surveillance.

Mouse Models for Fanconi Anemia

Orthologues of FANCD2, FANCM, and FANCL are found in nonvertebrates, whereas the other members of the FA core complex have evolved in vertebrates. To study clinical phenotypes and develop possible new therapeutic strategies, several mouse models for FA have been generated by introducing mutations into murine Fanc genes, including Fanca , Fancc , Fancg , Fancd2 , Fancd1, Fancf, Fancl, Fancm, Fancn, Fanco, and Fancp. Homozygous null mutations were lethal for Fancd1, Fancn, Fanco, and some Fancl embryonic mice, depending on the genetic background; this finding confirms what had already been suspected from the mutation analysis in patients with FA, namely that some residual function in these FANC genes is required for a living organism. Some models recapitulate the clinical features seen in patients with FA, specifically microphthalmos, germ cell defects, chromosomal instability, and increased sensitivity to cross-linking agents. In some models tumors develop with increased frequency, and tumor formation is more pronounced in the absence of p53. These models underline the importance of the FA pathway in tumorigenesis. Hematopoietic progenitor cells from mice homozygous for a null mutation in the Fancc gene demonstrate a distinct hypersensitivity to IFN-gamma that leads to cell death independent of the DNA repair pathway mediated by FAS-induced apoptosis, suggesting that IFN-gamma hypersensitivity may play a major role in the pathogenesis of BMF in patients with FA. Interestingly, however, although Fancd1- and Fancp- deficient mice have a reduced number of HSCs and decreased white blood cell and platelet counts, spontaneous development of aplastic anemia was not observed in these mice. Mice deficient in Fancd1 are exquisitely sensitive to excess acetaldehydes from endogenous and exogenous sources. Aldehydes are produced endogenously during normal metabolism or as a by-product of ethanol metabolism and can induce DNA-DNA and DNA-protein cross-links. Aldehydes are detoxified by aldehyde dehydrogenases. Interestingly, hematopoietic stem cells in mice express the aldehyde dehydrogenase Aldh2 , which is essential for protection against acetaldehyde toxicity, suggesting that the FA pathway is needed to protect from aldehyde-mediated genotoxicity. Although these findings need to be confirmed in humans, they would explain why hematopoiesis is preferentially affected in FA but is not affected, or not to the same extent, in other non-FA DNA repair disorders.

Induced Pluripotent Stem Cells Derived from Patients with Fanconi Anemia

Induced pluripotent stem cells (iPSCs) are derived from human somatic cells through ectopic expression of transcription factors. Since their discovery these cells have become a widely used tool for modeling and studying IBMFSs. Their unlimited self-renewal capacity, pluripotency, and ease of accessibility are very promising for disease modeling, cell-based therapy, and drug discovery. Interestingly, the generation of iPSCs from FA patients is challenging due to a low reprogramming efficiency in the presence of a defective FA pathway, suggesting that cellular reprogramming or the methodology (integrating viruses) of reprogramming requires an intact FA pathway. FA iPSC lines are capable of undergoing hematopoietic differentiation, but the hematopoietic progenitors undergo increased apoptosis and have a reduced clonogenic potential. Further investigation using the iPSC FA disease model is likely to provide exciting insights into the disease pathology and future treatment options.

Clinical Management

Once the diagnosis is confirmed, the family should be referred to a clinical geneticist for counseling and careful examination of family members. Genetic testing or chromosome breakage analysis should be offered to all family members. Early diagnosis is important for correct management of hematologic complications, diagnosis and appropriate treatment of coexisting congenital abnormalities and associated endocrinopathies, and identification of affected but asymptomatic family members and unaffected family members or pregnancies that may serve as potential hematopoietic stem cell donors. In addition, early diagnosis allows targeted cancer surveillance, not only in patients with FA but also in some cases in heterozygous mutation carriers (also see earlier). Management of patients in whom FA is diagnosed depends on the age at diagnosis and the presence or absence of congenital or hematologic abnormalities (see also Fanconi Anemia: Guidelines for Diagnosis and Management, Third Edition, 2008; available from the Fanconi Anemia Research Fund at fanconi.org/). All diagnosed FA patients should have a full hematologic assessment, including examination of bone marrow, and HLA typing should be performed in anticipation of possible HSCT. Other investigations at diagnosis should include audiometry, ophthalmic examination, renal ultrasound studies, and endocrine assessment. Referral to specialized surgeons might be necessary for correction of radial ray defects.

Management of Hematologic Disease

Initially, in the absence of hematologic abnormalities or clinical and laboratory signs of mild BMF, monitoring of peripheral blood values and yearly examination of bone marrow might be sufficient. Bone marrow examination should include a trephine biopsy and marrow aspirate for examination of marrow cellularity and morphologic changes characteristic of MDS, as well as for cytogenetic examination by G banding and fluorescent in situ hybridization (FISH) analysis, including probes that identify the most frequent cytogenetic abnormalities seen in patients with FA and progression to MDS or AML (see the section on Myelodysplastic Syndrome and Acute Leukemia in Patients with Fanconi Anemia). Plans for possible HSCT should be in place because hematologic complications may develop rapidly. With progression of BMF it will become necessary to increase the frequency of hematologic monitoring. HSCT is usually considered in patients with moderate BMF and an available HLA-matched unaffected sibling donor or when BMF becomes severe and an unrelated donor is available. Preventive HSCT in the absence of hematologic disease is highly controversial and currently considered investigational because the risk and time course of disease progression are highly variable and have to be weighed against the morbidity and mortality associated with HSCT. However preventive HSCT might be considered in patients with biallelic BRCA2 mutations associated with a higher risk of leukemic transformation.

Management of FA patients with MDS or AML, or both, is less standardized. The significance of transient clonal cytogenetic abnormalities limited to a single chromosome is unclear, but if they are persistent or complex, such abnormalities may predict a poor prognosis; the exception is an isolated trisomy 1q that might be present years before the development of MDS (see the section on Myelodysplastic Syndrome and Acute Leukemia in Patients with Fanconi Anemia). HSCT with or without induction chemotherapy or enrollment in clinical trials for MDS or MDS/AML is usually considered. Androgens such as oxymetholone, methyl testosterone, or danazol and hema­topoietic growth factors transiently improve BMF in 50% to 60% of FA patients. Side effects of androgen therapy include masculinization, acne, hyperactivity, growth spurt followed by premature closure of the epiphyseal growth plates, high serum lipid levels, liver enzyme abnormalities, hepatic adenomas (with danazol), and a risk for hepatic adenocarcinomas (with androgens) that requires frequent monitoring of liver function and liver ultrasound scanning. The effect of pretransplantation androgen use is controversial. Hematopoietic growth factors [granulocyte colony-stimulating factor (G-CSF) or granulocyte-macrophage colony-stimulating factor (GM-CSF)] may transiently improve neutrophil counts and, in rare FA patients, hemoglobin levels and platelet numbers. The use of hematopoietic growth factors in patients with clonal cytogenetic abnormalities is controversial because of the potential risk of inducing or promoting leukemia. Antiinflammatory agents such as anti-TNF-alpha and antioxidants have been shown to improve hematopoiesis in vitro and in mouse models and are currently being tested in clinical trials for efficacy and safety in patients with FA. Red cell transfusion therapy and iron chelation (after multiple red cell transfusions) are frequently used in symptomatic FA patients. Platelet transfusions may be indicated in thrombocytopenic patients with bleeding or before surgical procedures.

HSCT from an HLA-matched sibling donor is accepted as the best treatment to cure BMF in FA patients and to prevent progression to MDS and AML. Early experiences with HSCT for the treatment of BMF were disastrous because of excessive regimen-related toxicity. Early conditioning regimens consisted of cyclophosphamide alone or in combination with total-body irradiation (TBI). Unfortunately, severe toxicities (e.g., gastrointestinal hemorrhage, hemorrhagic cystitis, cardiomyopathy, and skin burns) were common. Gluckman and colleagues proposed the use of a less intensive conditioning regimen consisting of low-dose cyclophosphamide and thoracoabdominal irradiation for patients with FA. This approach proved to be a major advance in the treatment of patients who had HLA-matched sibling donors. However, it did not afford consistent engraftment in recipients of unrelated donor transplants, with graft rejection rates exceeding 20%, and head and neck cancer was frequent after HSCT. Consequently, HSCT without irradiation is increasingly used for patients with FA and a matched sibling donor using fludarabine and cyclophosphamide and with T cell depletion either of the graft or in vivo with rabbit ATG or alemtuzumab to prevent graft-versus-host disease (GVHD). Recent studies suggest that fludarabine-based conditioning regimens are associated with improved survival rates in patients undergoing related or unrelated donor HSCT and have significantly improved survival rates for alternative donor transplant recipients. The use of irradiation for alternative donor transplant varies among countries and transplant centers. Transplant-related mortality and GVHD remain a problem. Furthermore chronic GVHD significantly increases the risk of squamous cell carcinoma in FA patients who undergo HSCT.

PGD may be used in parents with a previously affected child to select embryos that are both unaffected by FA (if a FANC gene mutation is known) and have HLA matched to the affected sibling. HSCT is performed with the HLA-matched umbilical cord blood of the unaffected newborn.

Unrelated cord blood as an alternative HSC source showed an overall survival rate of 53% with fludarabine in the conditioning regimen (vs. a rate of 19% without fludarabine). There was a relatively high incidence of GVHD and transplant-related mortality rates of 32% and 59%, respectively, suggesting that unrelated cord blood transplant should be considered only when other alternative donor stem cell sources cannot be found. Alternatively a mismatched or haploidentical family donor may be a valid option if no matched unrelated donor can be found.

Cancer Surveillance and Treatment

For FA patients who survive the hematologic complications, follow-up surveillance for solid malignant tumors becomes increasingly important. There are no clear guidelines on the treatment of cancer in FA patients, which is complicated because the increased sensitivity to irradiation and certain chemotherapeutic agents may result in enhanced toxicity and an increased risk of therapy-induced secondary malignant diseases.

Gene Therapy

HSCT may cure hematologic disease in FA patients; however, HLA-matched sibling donor cells are available only for a minority of patients. Gene therapy to correct HSCs, followed by autologous transplantation, offers an attractive alternative for the treatment of hematologic disease. Indeed, gene transfer corrects the hypersensitivity of most FA cell lines to DNA cross-linking agents in vitro (with the exception of FANCM, most likely because of the insolubility of the protein) and the hypersensitivity of FA knockout mice in vivo. The high incidence of somatic mosaicism and the often ameliorated course of hematologic disease in these patients (frequently referred to as the gene therapy of nature ) suggest that FA cells corrected by gene therapy are likely to have a selective growth and survival advantage. However, gene therapy attempts in FA patients have thus far been rather disappointing, mainly because of the lack of permanent gene correction and the difficulty in culturing FA bone marrow cells in vitro. Multiple investigations are currently addressing the major difficulties of gene therapy for FA, including the scarcity of hematopoietic stem cells in patients with BMF, the extraordinary fragility of FA hematopoietic progenitor cells, and the increased sensitivity of FA hematopoietic progenitor cells to oxidative stress, which makes it difficult to culture and transduce FA hematopoietic progenitor cells in vitro. There is also a risk that the transduced hematopoietic cells might already have acquired secondary somatic, and possibly oncogenic, mutations and that correction by gene therapy would promote rather than prevent leukemogenesis. The International FA Gene Therapy Working Group was formed to accelerate the transition from research to clinical gene therapy trials and coordinating activities on an international scale.

Clinical Manifestations

Originally, DC was clinically recognized by the classic diagnostic triad of reticular pigmentation of the skin, nail dystrophy, and mucosal leukoplakia ( Figure 7-7 ). Other signs and symptoms found frequently in classic DC are summarized in Table 7-6 . The classic clinical course is mucocutaneous features first appearing in childhood, BMF developing in adolescence, and death occurring in young adulthood. Cytopenias due to BMF, immune deficiency, pulmonary fibrosis, and malignant disease are the major causes of death.

Physical findings in a 31-year-old patient with classic dyskeratosis congenita and homozygosity for a TERT gene mutation. A , Reticular skin pigmentation. B , Dystrophic fingernails, rather mild in this case. C , Leukoplakia of the tongue. D , Hypocellular myelodysplastic syndrome in a 41-year-old patient with atypical dyskeratosis congenita due to a TERC gene mutation. E , Hypocellular bone marrow.

(Photographs courtesy of Susan Bayliss, Washington University.)

The identification of nine of the genes responsible for DC and the realization that these genes converge into a common pathway have changed the perception of this disease. It is now thought that individuals in whom DC is diagnosed on the basis of BMF and the classic cutaneous manifestations represent only a fraction of those who suffer from this condition. Whether affected individuals in the absence of cutaneous disease should be labeled with the diagnosis of “dyskeratosis congenita” or whether the disease should be renamed with a term better reflecting the pathogenetic pathway is still under debate. Here we will use the conventional nomenclature and clinically classify patients as having classic, atypical, or silent DC. Patients who have two or more of the three diagnostic mucocutaneous features or who have one of the mucocutaneous features and hypocellular bone marrow with all three blood cell lineages affected are those with classic DC. Patients with BMF but lacking mucocutaneous features or patients with clinical mani­festations of DC other than hematologic disease or mucocutaneous features, such as pulmonary fibrosis, are those with atypical DC. Individuals with no clinical manifestations but carrying a pathogenic mutation are classified as having silent DC. Our knowledge about the natural history of DC is biased because early studies focused on patients with the classic clinical manifestations. Our experience is still limited about the course of disease, response to treatment, and late complications in individuals with atypical or silent DC.

Classic Dyskeratosis Congenita

Other Clinical Features.

Pulmonary complications are common (>20%) and include pulmonary fibrosis and vascular abnormalities. Pulmonary fibrosis is usually a more prominent finding in older patients or after BMT. In some patients with atypical DC, pulmonary fibrosis may be the only clinical manifestation. Figure 7-8 shows a family with autosomal dominant DC and pulmonary fibrosis. Families with hereditary pulmonary fibrosis of adult onset should be investigated for DC, particularly when pulmonary fibrosis is associated with macrocytic anemia or other cytopenias. In families with pulmonary fibrosis and BMF, a pathogenic DC gene mutation is almost always responsible for both conditions; in such kindreds pulmonary disease typically manifests in the sixth decade of life, whereas BMF manifests in the first or second decade (see Fig. 7-8 ).

A , Pedigree of a family with dyskeratosis congenita due to a TERC gene mutation. Note, in addition to myelodysplastic syndrome, pulmonary fibrosis is a prominent feature of dyskeratosis in this family (to protect privacy, gender identification has been removed). B , Chromatogram demonstrating that the parent and the sibling are heterozygous for the TERC gene mutation, whereas the child has not inherited the mutation. C , Telomere length of the affected parent (below and outside of telomere length distribution of normal controls) and the nonaffected child (normal telomere length). Lines indicate the 10th, 25th, 50th 75th, and 90th percentile of telomere length determined in 105 healthy controls between the ages of 1 and 95 years.

Liver fibrosis may be a late manifestation of DC. As with pulmonary fibrosis, liver fibrosis tends to occur in older patients or after BMT. In some patients liver fibrosis might be the only or first manifestation of disease. Excessive alcohol consumption or acute hepatitis is thought to accelerate liver fibrosis and severity in patients with a pathogenic DC mutation. The presence of a DC mutation does not preclude the development of hepatocellular carcinoma.

Ocular findings include epiphora (excessive tearing caused by blocked lacrimal ducts), blepharitis, cataracts, loss of eyelashes, conjunctivitis, ectropion, glaucoma, strabismus, ulcers, ocular albinism, and exudative retinopathy ( Coats retinopathy ). Interestingly, Coats retinopathy is specific for Revesz syndrome (caused by mutations in TINF2 ) and Coats- plus syndrome (caused by mutations in CTC1 ); exudative retinopathy is not typical of other forms of DC. Corneal limbal insufficiency owing to impaired corneal stem cell function has also been described in patients with DC.

Sparse hair and early graying are common. Dilated cardiomyopathy due to cardiac fibrosis has been associated with DC.

Multiple dental caries early loss of teeth, and a bifid uvula are common oral manifestations of DC. Gastrointestinal abnormalities include web formation in the upper esophagus, esophageal stricture, severe necrotizing ulcerative pancolitis in infants, and chronic diarrhea with malabsorption due to mucosal villous blunting in young adults. Patients have an increased risk of esophageal, rectal, and possibly gastric cancers.

Genitourinary abnormalities include hypospadias, phimosis, and urethral strictures. Underdeveloped testes and azoospermia have been reported in more severe and advanced forms of DC.

Cognitive deficiencies such as learning disorders are common in classic DC. In rare cases an association with schizophrenia has been described. Other neurologic manifestations may include cerebellar ataxia due to cerebellar hypoplasia, intracranial calcification, and hypoplasia of the corpus callosum, pituitary gland, or brainstem.

Osteoporosis and avascular necrosis of the bones, in particular the hips, is frequent and probably underappreciated and may complicate the course of the disease. In some forms of DC avascular necrosis of the bones occurs at a very young age and may be the dominating clinical feature. Short stature and an elf-like appearance have been associated with classic DC.

Predisposition to Malignant Disease

DC is a syndrome that predisposes to cancer. After BMF and pulmonary complications, malignant disease is the third leading cause of death. The cumulative incidence has been estimated to be 20% to 30% by the age of 50 years. The most frequent malignant tumors in patients with DC are squamous cell carcinoma of the head and neck (especially the tongue but also the lip, mouth, palate, cheek, nasopharynx, and larynx), the upper and lower gastrointestinal tract (esophagus and anus), and the female genital tract. Other reported malignant tumors include skin cancer and adenocarcinoma of the stomach, colon, lung, and pancreas. Whether all forms of DC have equal risks of predisposition to cancer remains to be determined. Malignant tumors usually develop in the third decade and are therefore more frequently found in individuals with milder and late-onset forms of the disease (see Table 7-1 ).

MDS and MDS/AML are common in patients with atypical DC and in patients with late-onset of the disease (see Fig. 7-7, E ). The cumulative incidence of MDS has been estimated to be about 30% and that of MDS/AML approximately 10% by the age of 50 years. MDS or MDS/AML may be the first hematologic manifestation of disease. Although DC is rarely a cause of MDS or MDS/AML in children, it should be considered in the differential diagnosis in young adults with MDS or MDS/AML. Abnormalities of chromosome 7 (monosomy, isochromosome, or other structural rearrangements) are the most common cytogenetic abnormalities in patients with DC-associated MDS or MDS/AML.

Little is known about the biology of leukemia or other cancers in patients with DC and to what extent these malignant diseases differ from those seen in patients without DC. The cancer recurrence rate is high in patients with DC.

Laboratory Findings

Common laboratory findings are summarized in Box 7-9 . In most patients, the initial hematologic abnormalities are thrombocytopenia or macrocytic anemia followed by pancytopenia. Bone marrow examination often reveals increased cellularity at the outset, suggesting an element of hypersplenism. However, hypocellularity then ensues, consistent with the diagnosis of aplastic anemia. Macrocytosis and elevated hemoglobin F levels (stress erythropoiesis) are common. The number of circulating hematopoietic progenitor cells is decreased, indicative of the impairment in hematopoiesis at the stem cell level.

Bone marrow examination usually shows hypocellular marrow affecting all three blood cell lineages. The hypocellularity of the bone marrow is often striking in comparison with rather moderate cytopenias in the peripheral blood. Hypocellular MDS or MDS/AML is increasingly being found in patients with atypical DC or a late onset of disease. Monosomy 7 is the most frequent clonal cytogenetic abnormality in patients with MDS or MDS/AML.

Long-term culture assays of bone marrow cells show a decreased number of progenitor cells and an impaired capability for colony formation, indicative of a quantitative and qualitative defect.

Immunologic abnormalities may include humoral and cellular defects (summarized in Box 7-9 ). Severe combined immunodeficiency is characteristic of severe forms of DC.

DC cells have increased spontaneous chromosomal instability that leads to distinctive chromosomal breaks, rearrangements, and sister chromatid exchanges. Chromosomal instability increases as DC cells are passaged in vitro. Chromosome breakage does not increase after treatment with alkylating agents such as DEB, MMC, or nitrogen mustard, a finding that distinguishes DC from FA. Chromosomal abnormalities characteristic of DC include end-to-end fusions, dicentric or tricentric chromosomes, anaphase or telophase bridges, short telomeres, and telomere free ends.

Paternally or maternally biased X-inactivation in blood cells, as well as in fibroblasts from buccal mucosa and skin biopsy samples in female mutation carriers, is characteristic of X-linked DC.

The presence of short telomeres in circulating peripheral blood cells is a hallmark of patients with BMF and DC. Telomere length in individuals with BMF secondary to DC is far below the normal distribution of telomere length in age-matched healthy control individuals ( Box 7-10 and Fig. 7-9 ). In contrast to other IBMFSs, which may have short telomeres in circulating myeloid cells, telomeres are exquisitely short in the myeloid and lymphoid cells of patients with DC-associated BMF. Telomere length measurements of circulating peripheral blood lymphocytes (or total blood cells, although less sensitive) are therefore used to screen patients with BMF for DC; normal telomere length makes DC an unlikely cause of BMF. Similarly, very short telomeres in circulating lymphocytes are highly suggestive of DC, but the diagnosis should be reserved for patients with classic clinical DC features or an identified pathogenic DC mutation.

Telomere length in peripheral blood mononuclear cells in individuals with dyskeratosis congenita and bone marrow failure. In all individuals with dyskeratosis congenita and bone marrow failure, telomere lengths are below and outside of telomere length distribution of normal controls. The black lines indicate the 1st, 10th, 25th, 50th 75th, 90th, and 99th% percentile of telomere length determined in 148 healthy controls between the ages of 1 and 95 years. The gray dots indicate telomere length in 148 patients with aplastic anemia due to other causes than dyskeratosis congenita.

(Modified from Du HY, Pumbo E, Ivanovich J, et al: TERC and TERT gene mutations in patients with bone marrow failure and the significance of telomere length measurements. Blood 113:309–316, 2009.)

Cellular Phenotype of Dyskeratosis Congenita Cells

In addition to increased spontaneous chromosomal instability, DC cells exhibit several other phenotypic characteristics, as summarized in Box 7-9 . DC cells have been reported to have increased sensitivity to ionizing radiation and certain chemotherapeutic agents (e.g., bleomycin), although these findings were not consistent among all investigators. Decreased proliferation capacity in vitro and premature replicative senescence are also characteristic of DC cells (see Box 7-9 ). Cells from classic and more severe forms of DC (X-linked DC) fail to immortalize after Epstein-Barr virus–induced transformation or after transfection of a plasmid expressing the telomerase catalytic subunit TERT. Similarly the reprogramming of DC skin cells into iPS cells is difficult, and reprogrammed iPS cells are not immortal.

Differential Diagnosis

Clinically, BMF in patients with DC has to be distinguished from other forms of BMF, including acquired aplastic anemia, familial aplastic anemia, and IBMFSs such as FA (MIM 227650), Seckel syndrome (MIM 606593), DNA ligase IV deficiency (MIM 606593), and Shwachman-Diamond syndrome (SDS, MIM 260400). Reticulate hyperpigmented disorders including Naegeli-Franceschetti-Jadassohn syndrome (MIM 161000), dermatopathia pigmentosa reticularis (MIM 125595), X-linked reticulate pigmentary disorder, incontinentia pigmenti (MIM 308300), and poikiloderma with neutropenia (MIM 604173) need to be distinguished from DC. Clinical manifestations, associated physical abnormalities, the family history (pattern of inheritance), and the lack of increased chromosomal fragility in the DEB or MMC test performed on cultured peripheral lymphocytes or fibroblasts (or both) are helpful in strengthening the diagnosis of DC. As noted earlier, telomere length below the normal distribution strongly suggests DC, whereas normal telomere length makes DC an unlikely cause for BMF in these patients. X-inactivation studies may be helpful in identifying female carriers of the X-linked form of DC.

Genetic Characteristics

Dyskeratosis Congenita Genes and Their Functions

DKC1 and Dyskerin.

X-linked DC is caused by mutations in the DKC1 gene at Xq28. DKC1 was identified by positional cloning in families with affected males and biased X-inactivation in female carriers. The DKC1 -encoded protein dyskerin is a 58-kD nucleolar protein orthologous to proteins in yeast, flies, and rats (Cbf5p, Nop60B/minifly, and NAP57, respectively).

Dyskerin is associated with box H/ACA RNA molecules in ribonucleoprotein (RNP) complexes. Box H/ACA RNP complexes consist of four different proteins (dyskerin, GAR1, NOP10, and NHP2) and an integral RNA species. The box H/ACA RNA determines the function of the RNP complex. The majority of box H/ACA RNAs are small nucleolar RNAs (snoRNAs) found in the nucleolus. These snoRNAs guide the protein complex containing dyskerin by complementary base pairing to specific uridine residues of newly synthesized ribosomal RNA (rRNA) molecules, which are then modified by pseudouridylation. In snoRNP dyskerin is the enzyme, a pseudouridine synthase, that actually catalyzes the pseudouridylation reaction. The role of pseudouridylation is not known, but pseudouridines tend to be found in the conserved sequences of rRNA and transfer RNA (tRNA) in regions of high secondary structure, and it is likely that these modifications stabilize the secondary and tertiary structures of RNA molecules. Other box H/ACA RNAs that associate with dyskerin and the three other proteins of the complex participate in ribosome biogenesis downstream of rRNA modification or messenger RNA (mRNA) splicing or are orphan RNAs, the function of which remains to be determined.

Telomerase RNA ( TERC , telomerase RNA component), which along with TERT (telomerase reverse transcriptase) forms the core of the telomerase enzyme, also associates with dyskerin and the other three proteins present in H/ACA-class snoRNP ( Fig. 7-10, C ). The conserved secondary structure of the 3′ end of telomerase RNA is very similar to that of H/ACA RNAs. This part of the telomerase RNA and its association with H/ACA RNP components are found only in vertebrates, whereas the H/ACA RNP components themselves are highly conserved among all eukaryotes. The four proteins may be involved in transport, stability, and assembly of the telomerase complex in the nucleolus. There is no evidence that the H/ACA part of telomerase RNA functions as a guide RNA or that dyskerin exhibits any pseudouridylation activity as part of the telomerase complex.

A , Telomeres and telomere structure. Telomeres are complex DNA protein structures that protect the ends of chromosomes in every cell. The organization of DNA ends at telomeres is highly conserved. Telomeric DNA is composed of short, direct repeats. B , In quiescent cells the telomeres adopt a looplike structure, the T loop, which hides the end of chromosomes from degradation and end-to-end fusion (see also Box 7-10 ). C , A number of additional telomeric proteins and protein complexes are responsible for maintaining this unique telomeric DNA structure. Highlighted here are the shelterin complex, telomerase, and CST1 and RTEL1 enzymes involved in telomere elongation and replication. Patients with dyskeratosis congenita have mutations in one of eight gene products participating in telomere maintenance (highlighted in color; mutations in the telomerase complex are TERT, TERC, DKC1, NHP2, and NOP10; mutations in the shelterin complex are TIN2; mutations in enzymes involved in the replication of telomeres are CTC1 and RTEL1. D , Telomerase is a ribonucleoprotein complex. It uses its RNA component, TERC, as a template to synthesize telomeric repeats on the 3′ end of telomeres, thereby elongating the ends of chromosomes (see also Box 7-10 ). E , Graphic illustration of the telomeric regions detected using the most frequently used telomere measurement methods. Telomere length measured by terminal restriction fragment includes subtelomeric DNA and telomeric repeats of the telomere. Both the quantitative fluorescent in situ hybridization and flow fluorescent in situ hybridization techniques use probes that are specific for pure telomeric repeats. Quantitative polymerase chain reaction compares amplification of pure telomere reads using telomeric primers (green arrows). The table below summarizes the different methods and their advantages and disadvantages.

( B, Tungsten shadow-cast visualization of a mammalian telomeric T loop using transmission electron microscopy courtesy of Jack Griffith, Department of Biochemistry, University of North Carolina, Chapel Hill, N.C.; E, Figure modified from Aubert G, Hills M, Lansdorp PM: Telomere length measurement: caveats and a critical assessment of the available technologies and tools. Mutat Res 730, 59–67, 2012; Table modified from Vera E, Blasco MA: Beyond average: potential for measurement of short telomeres. Aging [Albany, N.Y.] 4:379–392, 2012.)

The diverse function of dyskerin in humans raises the interesting question of whether and to what extent the pathogenesis of X-linked DC is caused by defects in telomerase activity or in other functions of RNP, such as rRNA metabolism.

Other Genes.

For about 50% of patients in whom DC and BMF are diagnosed, the mutations or the genes mutated have not been identified. Like patients with DC and BMF caused by mutations in known genes, most of these patients, but not all, also have very short telomeres, suggesting that the affected gene or genes participate in the same pathway.

Unraveling of the molecular pathology of DC has greatly advanced our understanding of the disease. The finding that all forms of DC in which the pathogenesis is understood are caused by disruption of telomere maintenance highlights the importance of telomeres in the disease (see Fig. 7-9 ). Figure 7-11 summarizes our current model of the pathogenesis of DC.

Model of the pathogenesis of disease in patients with dyskeratosis congenita. Our hypothesis is that excessive telomere shortening and dysfunctional telomeres play a central role in the pathogenesis of dyskeratosis congenita. Dysfunctional telomeres lead to cell cycle arrest, cell senescence, and cell death. In patients with DKC1, NHP2, or NOP10 mutations, TERC RNA stability and accumulation are decreased, which leads to decreased telomerase activity and telomere shortening. Impaired ribosome biogenesis may also limit cell survival and proliferation. TERC and TERT mutations cause telomere shortening through haploinsufficiency. The inheritance of preshortened telomeres is the molecular basis of anticipation in autosomal dominant dyskeratosis congenita. TCAB1 mutations prevent the recruitment of telomerase to the telomeres. RTEL1 and CTC1 are important for the proper replication of telomeres. Dysfunctional telomeres lead to genomic instability that may contribute to the cancer predisposition in patients with dyskeratosis congenita.

Dyskeratosis Congenita Gene Mutations

Mutations in telomere maintenance genes that have been found in patients with DC are listed on the comprehensive Telomerase Database, which is updated regularly.

DKC1 Mutations.

The mutations in the DKC1 gene that cause DC are mainly single–amino acid missense mutations ( Fig. 7-12, A ) and can be inherited or occur de novo. The A353V mutation is the most frequent de novo DKC1 mutation and is found in approximately 40% of patients with DC and DKC1 mutations. DKC1 mutations are not randomly distributed but are concentrated in the N-terminal region of dyskerin. The crystal structure of the archaeal homologue of the H/ACA RNP complex, the Cbf5-Nop10 complex, has enabled modeling of human dyskerin. In this model most of the residues mutated in cases of DC are tightly clustered in or near the PUA domain, which is predicted to bind H/ACA RNA, including TERC RNA (see Fig. 7-12, B ). The mutations would thus be expected to decrease binding of the cognate RNA and lead to instability. Only two mutations have been found in the TruB catalytic domain, one of which, S121G, is located three residues from the essential aspartate required for pseudouridylation and, interestingly, is associated with a severe form of DC known as Hoyeraal-Hreidarsson syndrome (HHS, MIM 300240). The absence of mutations that would predict a “null” allele, such as frameshifts and nonsense mutations, suggests that a dyskerin null mutation is incompatible with life. In support of this notion, a dyskerin null mutation is embryonically lethal in mice.

A , Mutations in the DKC1 gene. Schematic representation of the DKC1 gene, below the schematic representation of dyskerin protein and its functional domains. A series of patient-derived mutations are shown in bold. An asterisk indicates recurrent mutations. The A353V mutation is shown in red; this mutation is a frequent de novo mutation and accounts for about one third of DKC1 mutations. Mutations found in patients with the severe form of dyskeratosis congenita known as Hoyeraal-Hreidarsson syndrome are shown in purple. B , Clustering of pathogenic mutations in dyskerin. Crystal structure of the archaeal homologue of dyskerin. In this model the majority of residues mutated in cases of dyskeratosis congenita are tightly clustered in or near the pseudouridine synthase and archaeosine transglycosylase (PUA) domain, which is predicted to bind box H/ACA RNAs, including the TERC RNA. C , Schematic representation of the tertiary structure of the telomerease RNA (TERC). The pseudoknot and the CR4-CR5 domain are important for the association of TERC with the catalytic subunit TERT and for telomerase activity. Mutations found in families with autosomal dominant dyskeratosis congenita are shown, including two large deletions, one affecting the 5′ end and the other the 3′ end of TERC. D , Mutations in the TERT gene. Schematic representation of the 16 exons of the TERT gene, below the schematic representation of the TERT protein with its functional domains. Patient-derived mutations are shown in bold. Sequence variants of uncertain pathogenicity are shown in gray. Small deletions are indicated below the diagram. L1 and L2 are linker regions; Sf indicates a frameshift, which results in a premature stop codon X. The asterisk indicates recurrent mutations. E , TINF2 mutations are tightly clustered in exon 6. Mutations at 5′ of the TRF1 binding site have been found in families with familial aplastic anemia, whereas mutations at 3′ of the TRF1 binding site are sporadic and cause a severe phenotype of disease with early onset, intrauterine growth retardation, cerebellar hypoplasia, immune deficiency, and retinal bleeding; this was previously described as Revesz syndrome but is in fact a severe form of dyskeratosis congenita. For a comprehensive listing of mutations see telomerase.asu.edu/diseases.html . L1 and L2 , Linker regions; NLS , nuclear localization singnal; Sf , frameshift; TruB and PUA , pseudouridine synthase domains.

( B , Modified from Rashid R, Liang B, Baker DL, et al: Crystal structure of a Cbf5-Nop10-Gar1 complex and implications in RNA-guided pseudouridylation and dyskeratosis congenita. Mol Cell 21:249–260, 2006.)

The functional consequences of a subset of pathogenic DKC1 mutations have been investigated in fibroblasts or mononuclear cells from patients with X-linked DC and in XY mouse embryonic stem cells harboring a mutated Dkc1 allele. Analysis of fibroblasts or mononuclear cells from patients with X-linked DC showed a reduced level of TERC RNA and decreased telomerase activity, whereas analysis in mouse cells revealed that not only the level of Terc RNA but also the levels of other specific snoRNAs were affected and that overall pseudouridylation activity—in particular, site-specific pseudouridylation involving the affected guide snoRNA—was markedly decreased. Controversy continues as to whether X-linked DC is due solely to defective telomere maintenance or whether defective ribosome biogenesis contributes to the pathogenesis. If the latter is the case, other relevant genes may not be effectively translated and thereby contribute to the pathophysiology (see later discussion of Diamond-Blackfan Anemia [DBA]).

Biochemical and structural studies of the assembly and structure of H/ACA RNPs, including telomerase, have so far failed to explain the structural and functional relationships in the effects of the DKC1 mutations. Early in the biogenesis of H/ACA RNPs, prior to introduction of NAF1 and before assembly with the RNA component, dyskerin forms a tight connection with a protein SHQ1 and many DKC1 mutations are in the region of dyskerin that interacts with SHQ1, suggesting that the mutations may affect biogenesis of the complex. It will be interesting to see if telomerase is preferentially affected by these mutations.

Genotype-Phenotype Correlation

The disease phenotype of patients with DC is highly variable and is influenced by the gene mutated, the pattern of inheritance (X-linked, autosomal dominant, or autosomal recessive), the nature of the mutation, and the history of inheritance (number of generations that the mutations have been passed on to). A distinctive feature of DC is that clinical manifestations differ with the age of disease onset (see Fig. 7-13 ).

Clinical presentation of dyskeratosis congenita differs with age. The most severe forms of the disease, Hoyeraal-Hreidarsson syndrome and Revesz syndrome, manifest in early childhood and are mainly caused by TINF2 , RTEL1 , and DKC1 mutations. Dyskeratosis congenita with the classic cutaneous features occurs in the teens and 20s and is mainly caused by DKC1 mutations or homozygous TERT , TERC , or CTC1 mutations. Aplastic anemia in the adolescent and young adult as well as pulmonary fibrosis and liver fibrosis and osteoporosis later in life are mainly caused by heterozygous TERC and TERT gene mutations. The risk of cancer and leukemia is highest in the young adult with moderate to severe disease.

Anticipation.

AD DC is one of a small number of diseases that show genetic anticipation (i.e., the disease becomes more severe and has an earlier age at onset in later generations). Anticipation became apparent when asymptomatic parents of affected individuals were found to also have TERC gene mutations (see Box 7-11 ). In several cases, examination of these asymptomatic parents revealed that they had signs of mild anemia, such as raised mean corpuscular volume. In other pedigrees, anemia develops in the parents at an older age than in the children. The mechanism of anticipation appears to be that telomere shortening increases in later generations carrying a defective TERC allele. Figure 7-14 shows the inheritance of preshortened telomeres in a nuclear family with AD DC caused by a TERC gene deletion. Thus AD DC develops if an individual inherits a defective TERC or TERT allele and shortened telomeres. It is not known how many generations are needed to produce telomeres short enough to cause the disease. Studies in individuals with a de novo deletion of 5p that includes the TERT gene (5p− syndrome) indicate that in the first generation with TERT haploinsufficiency, telomere shortening is minimal. Interestingly, in individuals with DC and BMF, the telomeres are all equally short, suggesting that there is a minimal length of telomeres required for a progenitor cell to contribute to the peripheral blood cell pool.

Inheritance of preshortened telomeres in a family with autosomal dominant dyskeratosis congenita due to a TERC gene deletion. Metaphase spreads of peripheral blood lymphocytes. The red signals at the end of chromosomes show the telomeric repeats hybridized to a fluorescent probe. Below, the graphic representation of the telomere length distribution of each family member. Note that both the affected and the nonaffected child have inherited preshortened telomeres; according to their young age their telomeres should be longer than the telomeres in their parents.

(From Goldman F, Bouarich R, Kulkarni S, et al: The effect of TERC haploinsufficiency on the inheritance of telomere length. Proc Natl Acad Sci U S A 102:17119–17124, 2005.)

Molecular Diagnosis of Dyskeratosis Congenita

Currently, molecular diagnostic tests are not routinely performed as the first line of investigation for DC because they are expensive, labor intensive, and informative in only half of patients. Nevertheless, identification of pathogenic mutations in one of the DC genes is the only way to definitively establish the diagnosis of DC. Furthermore, molecular diagnostic tests enable the identification of mutation carriers, facilitate prenatal diagnosis for an existing pregnancy, and allow preimplantation genetic diagnosis for partners identified as mutation carriers. In addition, because the clinical manifestations, severity of disease, prognosis, and response to treatment may vary, depending on the affected gene and the nature of the individual mutation, molecular diagnostic tests may also influence medical management not only for the individual carrying the mutation but also for unaffected family members. Currently, an educated candidate gene approach, in which the gender ages at disease onset and family history of the affected individual are taken into account, is used to identify the affected gene (see also Fig. 7-13 ). In an affected male, after exclusion of the most frequent DKC1 mutation (A353V), it might be necessary to perform a mutational screen of the entire coding sequence of all three DC genes (usually only TER C and TERT in an affected female). For TERC and TERT the screening methods should include techniques that will detect large deletions. In addition to the large number of mutations in DC genes associated with DC, a number of nonpathogenic mutations and mutations of unknown pathogenic significance have been identified. Determining the pathogenicity of a particular mutation may be difficult.

Telomere measurement in peripheral blood cells and peripheral blood lymphocytes has emerged as a sensitive screening tool for DC in patients with BMF. Whether telomere measurements performed on peripheral blood cells may also identify DC mutation carriers or individuals with DC but nonhematopoietic disease manifestations is controversial.

Diseases Related to Dyskeratosis Congenita

Mechanism of Disease Pathogenesis in Dyskeratosis Congenita

Dysfunctional telomeres are the likely central theme in the pathogenesis of DC. Figure 7-11 summarizes our current model of the pathogenesis of disease in DC.

Clinical Management

Most of what we know regarding therapy and prognosis comes from our experience with patients with classic DC. Care must be taken when extrapolating such experience to all DC patients. The prognosis for patients with classic or severe DC is usually poor. In males with X-linked and sporadic DC, the median age at death was 20 years. Patients with AD DC have milder disease with a better survival rate. Deaths were usually due to complications from aplastic anemia, BMT, pulmonary fibrosis, or cancer.

Only identification of a pathogenetic mutation can confirm the diagnosis. Once confirmed, the family should be referred to a clinical geneticist familiar with new developments in DC for counseling and careful examination of family members. Genetic testing and possibly telomere measurement should be offered to all family members. Early diagnosis is important for correct management of hematologic complications; appropriate treatment of associated pulmonary, gastrointestinal, ocular, dental, and bone disease; identification of affected presymptomatic family members and unaffected family members or pregnancies that may serve as potential hematopoietic stem cell donors; and targeted cancer surveillance, not only in the patient with DC but also in mutation carriers.

Unfortunately, there is currently no consensus on how to manage patients with DC. Management depends on the age at diagnosis, the presence or absence of clinical signs, the gene affected, the molecular characteristic of the mutation identified, and possibly the telomere length. All patients with DC and hematologic manifestations should have a full hematologic assessment, including examination of bone marrow, and in anticipation of possible HSCT, HLA typing should be performed. Other investigations at diagnosis should include pulmonary evaluation and examination of the oropharynx and anal region; in females, cytologic examination of the vagina and cervical epithelium is needed. The prognosis and life expectancy of family members who have inherited only the short telomeres but not the gene mutation are unknown. Whether these family members should be excluded as potential sibling donors of an affected family member is also unknown.

Management of Hematologic Disease

Treatment of BMF in patients with classic DC is similar to that in patients with FA. Initially, in the absence of hematologic abnormalities or clinical and laboratory signs of mild BMF, monitoring of peripheral blood values might be sufficient. In the presence of hematologic abnormalities, yearly bone marrow examination should include a trephine biopsy and marrow aspiration for the examination of marrow cellularity and morphologic changes characteristic of MDS, as well as for cytogenetic examination by G banding and FISH analysis; testing should include probes that identify the most frequent cytogenetic abnormalities seen in patients with DC and progression to MDS or AML, or both (see earlier also). Plans for possible HSCT should be in place because hematologic complications may develop rapidly. With progression of BMF it will become necessary to increase the frequency of hematologic monitoring, and it may have to be individualized. The anabolic steroid oxymetholone can produce improvement in hematopoietic function. Approximately two thirds of patients with DC will respond to oxymetholone; in some cases this response can last several years and involve all lineages. Patients with DC may respond to a dose as low as 0.25 mg of oxymetholone/kg/day, which can be increased, if necessary, to 2 to 5 mg/kg/day. Hematopoietic growth factors usually, if at all, have only a transient effect and should not be used in combination with androgens because of the risk of peliosis and splenic rupture. Unlike idiopathic aplastic anemia, immunosuppressive agents such as antithymocyte globulin and cyclosporine are ineffective or only have a transient response in cases of DC and BMF.

The ultimate treatment of BMF is allogeneic HSCT. However, HSCT for DC has been associated with high morbidity and mortality rates, more so than in other BMF syndromes. Pulmonary and vascular complications are the major causes of death associated with HSCT in patients with DC. Reduced-intensity nonmyeloablative fludarabine-based conditioning regimens with no or minimal irradiation are associated with lower toxicity and fewer complications. Long-term survival and late complication rates after HSCT in patients with DC remain to be determined.

Management of DC patients with MDS or AML (or both) is even less standardized. The significance of transient clonal cytogenetic abnormalities is unclear, but if persistent or complex, such abnormalities usually predict a poor prognosis. Success is unlikely with early hematopoietic stem cell harvest for autologous transplantation at a later time because of the limited repopulating ability of the affected stem cells.

Management of Pulmonary Disease

There is currently no treatment of pulmonary disease in DC. Cigarette smoke is known to accelerate disease onset and severity. Thus patients with DC should refrain from smoking, avoid drugs with pulmonary toxicity (e.g., bleomycin), and have their lungs shielded from irradiation during BMT or cancer treatment. Pulmonary fibrosis in patients with DC does not respond to immunosuppressive therapy. Lung transplantation might be considered in individuals with otherwise mild manifestations of disease.

Cancer Surveillance and Treatment

Patients with confirmed DC should avoid excessive exposure to ultraviolet light and use barrier sun creams. Avoidance of smoking and alcohol should be emphasized, particularly in patients with oral leukoplakia. DC patients should also undergo regular dental and hygienist review to prevent early tooth loss and reduce the occurrence of leukoplakia. It is important to screen for malignant disease, especially of the gastrointestinal system. Management of malignant disease may be complicated by a number of factors. The diagnosis may be delayed in some cases because of malignant change occurring within an area of persistent leukoplakia, resulting in a delay in treatment. There is also a high potential for local recurrence and multiple cancers in these patients. In addition, the presence of BMF makes chemotherapy very difficult and can limit the indications for surgery. The mucosa also has high sensitivity to irradiation, which can lead to severe mucosal damage and may result in early termination of radiotherapy.

Management of oral malignant disease usually includes a regular review and biopsy of the premalignant hyperkeratotic lesion and surgical laser excision of any suspicious areas. Neoadjuvant radiotherapy or surgical resection of confirmed malignant tumors, along with ipsilateral node clearance with or without adjuvant radiotherapy in accordance with staging of the disease, is the mainstay of treatment.

Gene Therapy

HSCT may improve or cure hematologic disease in DC patients, but HLA-matched sibling donor cells are available for only a minority of patients; furthermore, immune reconstitution is often delayed. Gene therapy to correct hematopoietic stem cells followed by autologous transplantation might offer an attractive alternative for the treatment of hematologic disease in DC patients. Biased X-inactivation in female DKC1 mutation carriers suggests that DC cells corrected by gene therapy are likely to have a selective growth and survival advantage. Indeed, gene transfer of TERT or TERC (or both) corrects the telomere defect in most DC cell lines. However, as in many other IBMFSs, the few surviving hematopoietic stem cells might have already acquired secondary somatic and possibly oncogenic mutations, so correction by gene therapy could promote rather than prevent leukemogenesis in these patients. Interestingly, in rare patients with DC due to TERC gene mutations, gene reversion by mitotic recombination in hematopoietic stem cells leads to two normal copies of the TERC genes in the majority of circulating blood cells. Revertant cells contribute to the myeloid and B lymphocyte lineages but were not detectable in T lymphocytes, suggesting that mitotic recombination occurred in a stem cell committed to myeloid and B-cell differentiation or that the selective advantage for gene revertant cells is less in T cells than in myeloid and B cells. Indeed, T cells are less severely affected by DC mutations (see also above) and have a longer survival time in circulation. Revertant mosaicism in DC seems to be less frequent than reported in FA and currently has only been described in DC patients with TERC gene mutation. However, as in FA, DC revertant mosaicism might mask an inherited TERC gene mutation when genetic testing is performed in peripheral blood cells. DNA from skin fibroblasts or buccal swabs might have to be sent for genetic testing in patients with classic DC features but no BMF. Revertant mosaicism in patients with DC but no BMF powerfully illustrates the feasibility of gene therapy and shows that the restoration of a normal TERC gene in a single hematopoietic stem cell in these patients can restore blood cell production.