Clinical and genetic spectrum of Fanconi anemia in Australia and New Zealand

Abstract

Purpose

Fanconi anemia (FA) is a rare genetic condition that predisposes to progressive bone marrow failure, a specific spectrum of malignancies, including head and neck squamous cell carcinoma, and an array of other clinical manifestations. The clinical and genetic spectrum of FA in Australia and New Zealand remains relatively undescribed in the literature and is limited to case reports.

Methods

In this study, we conducted a comprehensive investigation of FA within this region combining cohort data and case reports, aiming to elucidate its diagnostic patterns, clinical manifestations, and genetic characteristics.

Results

Our findings reveal a positive correlation between national testing rates and case detection, across states and territories of Australia, suggesting that targeted testing strategies may enhance the identification of FA cases. Furthermore, our analysis demonstrates that the physical and genetic profiles of people with FA in Australia and New Zealand resemble those observed in other international cohort studies but also exhibit nuanced differences.

Conclusion

This study emphasizes the continued significance of cytogenetic testing for FA while stressing the need for heightened awareness among medical professionals. FA, once perceived primarily as a pediatric condition, now demands vigilance across all age groups because of advances in medical care and complementary detection methods, such as high-throughput sequencing. This study also underscores the heightened susceptibility of individuals with FA in Australia and New Zealand to head and neck squamous cell carcinoma, consistent with observations in other regions worldwide.

Introduction

Fanconi anemia (FA) is a rare genetic condition caused by pathogenic variants in genes involved in a common DNA repair pathway. FA predisposes affected individuals to a spectrum of disorders, such as progressive bone marrow failure, early onset cancer, endocrine issues, physical abnormalities, and reduced fertility.^1,2 The traditional congenital anomalies that can be seen in FA include those of the vertebral anomalies, anal atresia, cardiac anomalies, tracheoesophageal fistula, esophageal atresia, renal structural anomalies, limb anomalies, and hydrocephalus (VACTERL-H) association. The acronym PHENOS (skin pigmentation including café au lait spots, small head: microcephaly, small eyes: microphthalmia, nervous system, otology, and short stature) can also be used in consideration of the diagnosis of FA because these are phenotypic features that can be considered a “signal” of FA to highlight the features that are not part of the VACTERL-H association.^3,4 The clinical presentation of these associations is heterogeneous between individuals, FA siblings with the same pathogenic variants, and even identical twins, which can often make diagnosis difficult. Three or more of the 8 congenital abnormalities described in the VACTERL-H association have been reported in 5% to 30% of individuals with FA.3, 4, 5, 6 Four or more of the 6 PHENOS features were reported in 9% of cases of FA in a comprehensive literature review.³ In the same study, 21% of people with FA did not have an abnormal phenotype,³ which can lead to cases going undetected. Therefore, cases of FA that are not diagnosed in early childhood because of an absence of overt malformations would have insufficient surveillance for life-threatening diseases to which they are predisposed, such as bone marrow failure and specific malignancies, such as acute myeloid leukemia or head and neck squamous cell carcinoma.^1,7, 8, 9, 10, 11, 12

The median age of diagnosis for FA is 6.7 years, based on data from 1497 case reports,¹¹ mainly due to progressive bone marrow failure. However, the diagnosis of FA may be earlier for more severe cases. Diagnostic testing for FA uses a chromosome breakage assay, which remains the gold standard functional test.^1,13, 14, 15, 16, 17, 18 The chromosome breakage assay performed with the patient’s cells is a useful diagnostic tool because it provides a functional assessment of genomic instability—the underlying molecular defect in FA—in the presence of specific DNA crosslinking agents, diepoxybutane or mitomycin C. Another functional test is based on the finding that FA cells arrest in the G2/M phase of the cell cycle because of interstrand crosslinks.^19,20 The assay for G2/M phase arrest may be easier to implement than the chromosome breakage test; however, it is not unique for loss of function of the FA pathway.21, 22, 23 Therefore, the combination of the clinical suspicion and a positive chromosome breakage test remains mandatory to make the diagnosis FA. Genetic testing can be used to identify the gene containing the FA-causing variants. However, a challenge with diagnostics assays that use peripheral blood as the source of cells or DNA is the issue of somatic mosaicism that occurs in 15% to 25% of people with FA^15,24, 25, 26, 27, 28, 29, 30, 31, 32, 33 because any spontaneous mutation or recombination events that restore the function of the mutated gene in a hematopoietic stem cell results in that cell having a proliferative advantage and becoming the dominant genotype in the peripheral blood. The diagnostic challenge associated with somatic mosaicism may be overcome by skin biopsy and fibroblasts assay for diagnostic tests. A skin fibroblast biopsy is routinely taken for the ZERO2, the clinical trial component of the Australian Zero Childhood Cancer Program,³⁴ and for testing for genome sequencing as part of the inherited bone marrow failure and related disorders study³⁵; therefore, all people with BMF can have germline genomic testing via this method once they have consented.

When a positive chromosome breakage test result is obtained, it is prudent to identify the genetic cause of FA because it is complementary evidence supporting the diagnosis. Furthermore, pathogenic variant identification may assist with (1) the testing of potential related stem cell donors and family planning^1,36,37 and (2) predicting the course of the disease in an individual and surveillance for specific malignancies.³⁸ The identification of pathogenic variants that cause FA may also be useful to determine eligibility for gene therapy trials and new therapeutic trials.^39,40

FA is caused by the loss of function of any one of the known 22 FANC genes, including FANCS (BRCA1, HGNC:1100) and FANCD1 (BRCA2, HGNC1101).^1,8 FA is an autosomal recessive disease except for FANCB (HGNC:3583), which is X-linked,⁴¹ and RAD51 (FANCR, HGNC:9817), which is a rare dominant subtype.⁴² The FANC genes act in a common genetic pathway, sometimes referred to as the FA/BRCA pathway, which are known traditionally for their role in interstrand crosslink repair⁴³ but also have established roles in other types of DNA repair, including fork protection, R-loop metabolism, and meiosis.44, 45, 46 This pathophysiology makes people with FA exquisitely sensitive to agents that cause interstrand crosslinks in DNA, such as platinum-based chemotherapies. The downstream genes of the FA/BRCA DNA repair pathway are also involved in the repair of double strand breaks via homologous recombination.^47,48

Carrier rates for FA vary between populations but are estimated to range from 1 in 67 to 1 in 181.^1,49,50 No publications are yet to provide the prevalence of FA in the Australian and New Zealand populations. These populations are composed of a unique mixture of people from Australian Aboriginal and Torres Strait Islander groups, Maori and Pacific Islands peoples, European ancestry, and countries throughout Asia, among other regions of the world. This article presents the clinical spectrum of available cases of FA from Australia and New Zealand and provides an estimate of FA cases in Australia from national testing data.

Materials and Methods

Chromosome breakage testing

We obtained chromosome breakage test data for the periods 2010 to 2023 from Queensland Health. We aimed to estimate how many cases of FA might be in Australia. We reviewed chromosome breakage test results from 2010 to 2023 from a laboratory that receives samples from the majority of Australia’s 6 states and 2 territories, with an estimated coverage of 82.2% of Australia’s population. Tests were for the populations of Queensland (QLD), New South Wales (NSW), Victoria (VIC), Tasmania (Tas), the Australian Capital Territory, and the Northern Territory, which accounts for 21,908,200 residents of Australian’s population of 26,638,500 residents (source Australian Bureau of Statistics, June 30, 2023). The populations of South Australia (SA) and Western Australia (WA) were 1782 and 2660 thousand residents, respectively.

Ethics, participant recruitment, and clinical records

This study was approved by the Royal Children’s Hospital Human Research Ethics Committee (2019-282). Study participants provided written informed consent. We recruited 12 participants with an FA diagnosis from Australia (n = 8) and New Zealand (n = 4), and we refer to this group of individuals as the Fanconi Anemia Australia and New Zealand (FAANZ) cohort. The study has been recruiting participants since 2020. Individuals with FA and their relatives are eligible to participate. All published FA cases from Australia and New Zealand were also compiled.^35,51, 52, 53, 54, 55 Consenting study participants completed questionnaires and provided access to medical records. Information presented in this manuscript corresponds to the initial data collection for the FAANZ cohort. Individual chromosome breakage test results and pathogenic FANC gene variant information was obtained where possible. The self-reported overt phenotypic anomalies and those described in clinical records representing the lower bound for the prevalence of congenital malformations in this cohort. This approach may underestimate the true frequency of malformations because subtle or undocumented anomalies could be present. Details of any statistical tests that were used are described in the text in context.

Screening for oral potentially malignant lesions

Noninvasive oral inspections and cytopathology for oral potentially malignant lesions were performed as described previously.⁵⁶ Oral screening was performed opportunistically with participants of the FAANZ cohort at regional support group meetings by members of the study team. All ages were seen to establish a baseline in this new and unscreened population. Members of the medical team of participants also brushed visible oral lesions for cytological investigations.

Gene symbols

Gene symbols are used according to HGNC-approved nomenclature. A full list of genes mentioned in this manuscript, along with their HGNC IDs, is provided in Supplemental Table 1.

Results

Testing results in Australia and prevalence

National testing data from 927 chromosome breakage tests were aggregated from 3 laboratories. There were 843 tests performed over the period from 2010 to 2023 in a laboratory in Queensland. From South Australia 61 test results were aggregated from 2018 to 2024, 1 being positive. From WA, 23 tests were performed in 2023 with 1 being positive. Tissue type information was available for 770 tests. Most tests were performed on peripheral blood (95.7%, n = 737/770), followed by fibroblasts (3.8%, n = 29/770), amniotic fluid (0.3%, n = 2/770), or cerebrospinal fluid (0.3%, n = 2/770). In the available data, at least 568 individuals were tested most of whom were tested once (74%, n = 423/568), with 18% tested twice (n = 107/568), and 6.7% tested 3 or more times (n = 38/568). There were 59 unique individuals tested with peripheral blood samples that returned an inconclusive result. Positive tests occurred at a rate of 6.1% (n = 57/927), whereas 7.9% (n = 73/927) were inconclusive (Figure 1). Of these inconclusive tests from peripheral blood, 7 individuals were also tested for FA with fibroblast samples. Four of these people also had inconclusive test results from their fibroblast samples. None of these individuals tested positive as a result of the fibroblast test, and no individuals who tested normal with peripheral blood samples tested inconclusive with fibroblast-derived tests. In the 2021 Australian Census, 3.2% of the population identified as Aboriginal (2.92%), Torres Strait Islander (0.13%), or both (0.14%). In the testing data, in which ethnicity information was available, we found that Aboriginal people represented 6.2% (21/341), Torres Strait Islanders represented 0.59% (2/341), and people who identified as both Aboriginal and Torres Strait Islander represented 1.2% (4/341). Nonindigenous people represented 92.1% (314/341).

Figure 1.

Chromosome breakage testing results. Chromosome breakage test results are summarized. A. Most test results are normal or “negative” for FA. B. Test reports sometimes contain information about the ethnicity of the people tested. Testing of first nations people appears to be proportional to their fraction of the Australian population based on census data. C. Most abnormal or “positive” cases for FA have no information about their ancestry, followed by those that are nonindigenous. FA, Fanconi anemia.

The median age at which individuals tested positive was 7 years (Figure 2). Using these data, we attempted to estimate the prevalence of FA in Australia. With several conservative assumptions, we extrapolated the number of cases of FA in Australia to be approximately 189 or a frequency of 1 in 141,169 people (see Supplemental Text for calculations and assumptions). Among the people for whom biological sex information was available, 377 were female, 446 were male, and 104 tests did not have this information. The rate of testing of females and males appeared to be different (X-squared = 5.8, df = 1, P value = .016), showing that females are tested less frequently than males (see Discussion). However, no significant difference was detected in the number of positive cases (males n = 27, females n = 21, NA = 9, X-squared = 0.75, df = 1, P value = .39).

Figure 2.

Testing ages for Fanconi anemia in Australia. A. All tests and the ages of the individuals, where available (n = 831), are presented as a histogram. B. The same tests are separated as a function of biological sex. C. The ages of all individuals who returns a positive test presented as a histogram with the median represented by a vertical red line. D. The same positive tests are presented separating results into the biological sexes of the individuals as either male or female. There is a male who tests positive at age 55. Details of this individual with a homozygous pathogenic variant in FANCJ is discussed below and previously published.⁵²

The highest rate of testing per million people came from the state of Queensland (Figure 3). Notably, this is the state where the laboratory is located (see Discussion). The highest rate of positive tests per million people in the population also came from Queensland. We noticed that increased testing in a state or territory tends to positively correlate with more cases of FA being detected in Australia, with no particular indication that the highest test rates in this data set have reached saturation. Testing rates were noticed to increase over time in the testing data from the Queensland laboratory. Specifically in the year 2017, there appeared to be an increase in testing numbers, which was sustained in the subsequent years. From 2010 to 2016 the average number of tests was 28 ± 17, and from 2017 to 2023 the average number of tests was 82 ± 23, which represents a 2.9-fold increase in tests per year. In the corresponding periods, the number of positive tests was 2.7 ± 1.4 (2010-2016) and 4.1 ± 1.7 (2017-2023) (see Discussion).

Figure 3.

Increased testing rates positively correlate with detection of FA cases. More testing for FA in the states and territories of Australia positively correlates with more cases being detected without signs of saturation, based on weighted linear regression (red line) using total test numbers per state. Note that tests for WA were only from 1 year (2023) from the WA laboratory, tests from SA were from 6 years (2018-2024), and tests from other states and territories are from the Qld laboratory during 2010 to 2023. ACT, Australian Capital Territory; FA, Fanconi anemia; NSW, New South Wales; NT, Northern Territory; QLD, Queensland; SA, South Australia; Tas, Tasmania; VIC, Victoria; WA, Western Australia.

Demographics of individuals with FA in Australia and New Zealand

A total of 12 individuals with FA from Australia and New Zealand provided informed consent to share information. The median age of diagnosis in this special cohort was only 1 year of age (n = 12). Enriching these data with the identified published Australian and New Zealand cases (n = 8) the median shifts to 4.5 years of age (n = 20; 8 females, 12 males) (Figure 4A). These median ages of diagnosis are unexpectedly younger than what is observed in the national testing data, which highlights a potential sampling bias with respect to who participates in research (see Discussion). Congenital abnormalities were skewed toward radial defects and the kidneys (Figure 4B). We therefore considered if there might be a correlation between the number of congenital abnormalities and the age of diagnosis. This analysis suggested a modest negative association between diagnosis age and abnormality count (Poisson regression analysis, coefficient = −0.01392, P value = .25), with each additional year of age at diagnosis corresponding to a slight decrease in the expected number of abnormalities. However, the strength of this relationship was weak, and further investigation would be needed to confirm its clinical significance as more participants are recruited. We investigated if those with an earlier diagnosis (under 5 years of age, n = 10, 7 males, 3 females) had a stronger association between age of diagnosis and number of congenital malformations. A Poisson regression was conducted to examine the relationship between diagnosis age and abnormality count, including age group (before vs after 5 years). The results indicated that the age of diagnosis was not significantly associated with the number of congenital malformations (z = 0.541, P = .588), whereas being diagnosed after 5 years was marginally associated with fewer abnormalities (z = −1.756, P = .079). We explored, qualitatively, the radial and renal abnormalities as a function of age of diagnosis (Figure 4C). We considered the age of diagnosis and whether an individual had complete loss-of-function variants or had a hypomorphic variant(s), which suggested that individuals with 2 null variants have an earlier diagnosis on average (Figure 4D) (Welch 2 sample t test, t = −2.8, df = 9.2, P = .02).

Figure 4.

Physical manifestations of FA are dominated by upper limb and renal issues, as well as short stature. A. Histogram of the age of diagnosis of the combined cohort and cases from the literature. The median is represented by a dotted vertical line. B. Radial anomalies are the most common physical feature of FA seen in most people in this cohort, followed by short stature in half of the people. Renal issues are also common. C. Radial and renal abnormalities are plotted as a function of age of diagnosis. D. Patient variants and age of diagnosis are plotted, with the relevant gene indicated for individuals. Classification of variants as null or hypomorphic used the same criteria as Altintas et al.¹⁰ Similarly, patients with biallelic or homozygous null alleles were groups as “Null,” and patients with 2 hypomorphic alleles or a hypomorphic and a null allele were grouped as “Hypomorphic.”¹⁰ FA, Fanconi anemia.

Time to onset of bone marrow failure, hematopoietic stem cell transplant, and cancer

A total of 9 in 12 people had experienced decreased blood counts, which most frequently presented between 4 to 10 years of age (Table 1). A total of 5 in 11 people had a bone marrow transplant. Regarding cancer, a total of 6 in 20 people had a malignancy. The cumulative incidence of solid tumors by the age of 40 was 20% (4/20), including individuals under 40, some of whom may not have reached the age at which the event could manifest, thus reflecting both observed cases and censored data. The most common solid malignancy was HNSCC, which occurred in 3 of 20 cases, all of which were in young adults. The increased risk of HNSCC is consistent with the risk seen in international cohorts with adults with FA.^1,9,12,56, 57, 58, 59 In total, 4 in 12 people self-reported having persistent oral lesions. In total, 8 individuals from the FAANZ cohort were screened for oral potentially malignant lesions. Three people had 10 lesions brushed. Three lesions, all from the same individual, showed neoplastic features, including squamous cell carcinoma, representing a case of early detection. There were 2 cases of bowel cancer, including gastrointestinal adenocarcinoma.⁵¹ There was a separate case of gastric tubular adenoma.⁵¹ There were 2 cases of hepatocellular carcinoma.^35,52 One person presented with a melanoma. One person had T-acute lymphoblastic leukemia.⁵³ One person had metastatic head and neck squamous cell carcinoma, which was excised with surgery and radiation in addition to vulvar SCC and cervical precancer. Both SCC were HPV negative.⁶⁰

Table 1.

Characteristics of the people

Characteristic	N	Value	Percentage	Median Age	Cohort
Positive age, median age at FA testing from pathology laboratory	770	-	-	7	National testing data
Positive cases, median age at FA testing of cohort and literature	20	-	-	4.5	FAANZ cohort and literature
Positive cases, median age at FA testing of cohort	12	-	-	1.5	FAANZ cohort
Positive cases, median age at FA testing of literature	8	-	-	34	Literature
Biological sex - male	20	8	40%		FAANZ cohort and literature
Biological sex - female	20	12	60%		FAANZ cohort and literature
Age at onset of hematological manifestations (0-3 yo)	12	3	25%	-	FAANZ cohort
Age at onset of hematological manifestations (4-10 yo)	12	5	42%	-	FAANZ cohort
Age at onset of hematological manifestations (>10yo)	12	1	8%	-	FAANZ cohort
Percentage transplanted	11	5	45%	-	FAANZ cohort
Median age at bone marrow transplant	5	-	-	7	FAANZ cohort
Neoplastic features	9	-	-	32	FAANZ cohort and literature
Physical abnormalitiesa	19	14	74%	-	FAANZ cohort and literature
No congenital malformations	19	3	16%	-	FAANZ cohort and literature
1-2 congenital malformations	19	7	37%	-	FAANZ cohort and literature
3+ congenital malformations	19	9	47%	-	FAANZ cohort and literature

Aggregated information is presented about the characteristics of the participants in this study. “N” refers to the number of participants for whom information was available for a particular characteristic; in some instances, data could not be obtained for all fields for all participants. “Value” refers to how many people present a given characteristic from those for whom information was available. Percentage is (N/Value) × 100.

FA, Fanconi anemia.

^aPhysical abnormalities include skin abnormalities such as cafe au lait spots and short stature.

Population diversity in Australia and New Zealand

Australia and New Zealand’s populations are composed of substantial diversity with a blend that is unique based on its geography and pre- and postcolonial timelines. According to the Australian Bureau of statistics 2021 Census data, ancestries in Australia are Oceanian (which includes Australian Aboriginal, Torres Strait Islander, Australian, New Zealand, and Pacific Islands), North-West Europe, Southern and Eastern European, North African and Middle Eastern, South-East Asian, North-East Asian, Southern and Central Asian, Peoples of the Americas, and Sub-Saharan African.⁶¹ Participants ancestries were skewed toward broadly European backgrounds (Figure 5), which is consistent with the nonindigenous status of most individuals in the diagnostic testing data (Figure 1). Two individuals in our study had Aboriginal ancestry. In New Zealand, according to the official data agency, Stats NZ Tatauranga Aotearoa, the major ethnic groups are European, Maori, Pacific Peoples, Asian, MELAA (Middle Eastern/Latin American/African), and other ethnicity. No individuals recruited into this study reported Maori heritage. However, a person with FA with Maori ethnicity was identified but had died of acute myeloid leukemia before the study (personal communication, Siobhan Cross).

Figure 5.

Ancestral backgrounds of study participants are predominantly European. Self-reported ancestries of the cases and literature (n = 15). Individuals can report more than 1 ancestry.

Genetic subtypes of individuals with FA in Australia and New Zealand

The genes with pathogenic variants were skewed toward FANCA (HGNC:3582) (7/14 resolved cases). The second most common gene was FANCD2 (HGNC:3585) (3/14), and other rare subtypes made up the rest (Figure 6, Table 2). The people with pathogenic variants in FANCA and FANCD2 do not appear to be related based on the FANC gene variants and should not account for the overrepresentation of either of these genes. The only exception was a single case of FA attributable to homozygous variants in FANCA. The published twins with a homozygous pathogenic variant in FANCJ (HGNC20473) were reported to be from nonconsanguineous parents. We identified 1 novel FANCD2 variant, (NM_001018115.3:c.1691_1692del p.(Ser564Ter)), which is predicted to introduce a premature stop codon in the protein (Figure 7, Table 2).

Figure 6.

The most frequently affected gene in FA cases is FANCA consistent with other cohorts. Genetic subtypes from cases of FA are plotted in descending frequencies. Observed frequencies from the 20 Australian and New Zealand cases are similar to the global data from Frohnmayer et al. FA, Fanconi anemia.

Table 2.

Genetic variants in FANC genes in Australian and New Zealand people with Fanconi anemia

ID	Gene	HGNC ID	Variant 1	Genomic Coordinates (hg38)	Protein	Pubmed ID	Variant 2	Genomic	Protein	Pubmed ID	Reference Sequence	Study
FAANZ002	FANCA	HGNC:3582	c.(1900+20_1973)_(2810_2952)del	NC_000016.10:g.(89758606_89761991)_(89773312_89775722)del	p.?	-	c.(?_-15)_(570_642)del	NC_000016.10:g.(?_89816651)_(89805347_89808320)del	p.?	-	NM_000135.4	This study
FAANZ003	FANCA	HGNC:3582	c.2602-2A>T	NC_000016.10:g.89765068T>A	p.?	10090479	c.(3765+1_3766-1)_(3828+1_3829-1)del	NC_000016.10:g.(89740100_89740803)_(89740867_89742799)del	p.?	-	NM_000135.4	This study
FAANZ004	FANCA	HGNC:3582	c.(1006+1_1007-1)_(3066+1_3067-1)del	NC_000016.10:g.(89749903_89752137)_(89792548_89795905)del	p.?	-	c.1226-2A>G	NC_000016.10:g.89791538T>C	p.?	19367192	NM_000135.4	This study
FAANZ006	FANCD2	HGNC:3585	c.3707G>A	NC_000003.12:g.10090315G>A	p.(Arg1236His)	11239453	c.1691_1692del	NC_000003.12:g.10060328_10060329del	p.(Ser564Ter)	Novel	NM_001018115.3	This study
FAANZ007	FANCD2	HGNC:3585	c.2444G>A	NC_000003.12:g.10067267G>A	p.(Arg815Gln)	17436244	c.3453_3456del	NC_000003.12:g.10087251_10087254del	p.(Asn1151LysfsTer46)	17436244	NM_001018115.3	This tudy
FAANZ008	FANCG	HGNC:3588	c.652C>T	NC_000009.12:g.35077096G>A	p.(Gln218Ter)	11093276	c.1642C>T	NC_000009.12:g.35074489G>A	p.(Arg548Ter)	11093276	NM_004629.2	This study
FAANZ010	FANCA	HGNC:3582	c.331_334dup	NC_000016.10:g.89811021_89811024dup	p.(Leu112HisfsTer10)	22778927	c.(1900+20_1973)_(2810_2952)del	NC_000016.10:g.(89758606_89761991)_(89773312_89775722)del	p.?	22778927	NM_000135.4	This study
FAANZ011	FANCD1/BRCA2	HGNC:1101	c.4415_4418del	NC_000013.11:g.32338770_32338773del	p.(Lys1472ThrfsTer6)	15918047	c.9082G>C	NC_000013.11:g.32379878G>C	p.(Ala3028Pro)	34687993	NM_000059.4	Ip et al,51 2022
FAANZ012	FANCD2	HGNC:3585	c.904C>T	NC_000003.12:g.10043065C>T	p.(Arg302Trp)	11239453	c.2715+1G>A	NC_000003.12:g.10073363G>A	p.?	17436244	NM_001018115.3	This study
FAANZ013	FANCJ	HGNC:20473	c.751C>T	NC_000017.11:g.61808634G>A	p.(Arg251Cys)	23613520	c.751C>T	NC_000017.11:g.61808634G>A	p.(Arg251Cys)	23613520	NM_032043.3	Stevens et al,52 2016
FAANZ015	FANCM	HGNC:23168	c.1972C>T	NC_000014.9:g.45167133C>T	p.(Arg658Ter)	21681190	c.5101C>T	NC_000014.9:g.45189123C>T	p.(Gln1701Ter)	25288723	NM_020937.4	Ryland et al,53 2020
FAANZ017	FANCA	HGNC:3582	c.2980A>G	NC_000016.10:g.89758578T>C	p.(Ser994Gly)	32054657	c.2980A>G	NC_000016.10:g.89758578T>C	p.(Ser994Gly)	32054657	NM_000135.4	Blombery et al,35 2021patient ID 39
FAANZ018	FANCO/RAD51C	HGNC:9820	c.773G>A	NC_000017.11:g.58709926G>A	p.(Arg258His)	20400963	c.773G>A	NC_000017.11:g.58709926G>A	p.(Arg258His)	20400963	NM_058216.3	Blombery et al,35 2021 patient ID 44
FAANZ019	FANCI	HGNC:25568	c.3184C>T	NC_000015.10:g.89305240C>T	p.(Gln1062Ter)	32054657	c.3041G>A	NC_000015.10:g.89303898G>A	p.(Cys1014Tyr)	22720145	NM_001113378.2	Blombery, et al,35 2021patient ID 85
FAANZ020	FANCA	HGNC:3582	c.2852G>A	NC_000016.10:g.89761949C>T	p.(Arg951Gln)	12697994	c.3971C>T	NC_000016.10:g.89739517G>A	p.(Pro1324Leu)	10521298	NM_000135.4	Blombery, et al,35 2021patient ID 96
FAANZ039	FANCA	HGNC:3582	c.2303T>C	NC_000016.10:g.89770179A>G	p.(Leu768Pro)	15643609	c.2303T>C	NC_000016.10:g.89770179A>G	p.(Leu768Pro)	15643609	NM_000135.4	Gille et al,55 2012 patient ID 8
FAANZ040	FANCA	HGNC:3582	c.427-8_427-5del	NC_000016.10:g.89810808_89810811del	p.?	22778927	c.1771C>T	NC_000016.10:g.89778948G>A	p.(Arg591Ter)	22778927	NM_000135.4	Gille, et al,55 2012 patient ID 21
FAANZ041	FANCA	HGNC:3582	c.3491C>T	NC_000016.10:g.89746606G>A	p.(Pro1164Leu)	22778927	c.3491C>T	NC_000016.10:g.89746606G>A	p.(Pro1164Leu)	22778927	NM_000135.4	Gille, et al,55 2012 patient ID 22

For variant 2 from FAANZ002, the maximum size of the deletion cannot be determined because the deletion includes exon 1 of FANCA, and there is no probe covering the upstream 5¢ UTR.

Figure 7.

Mapping of genetic variants onto the proteins. Genetic variants are shown on each of the respective proteins.

Discussion

Age of diagnosis

This study describes a cohort of people with FA from Australia and New Zealand and substantial testing data from Australia. The main findings are that diagnostic rates of FA in most of Australia appear to be similar to that of the United States, when considering the median age of diagnosis, which is 7.

How does the diagnosis of FA in Australia and/or New Zealand compare with other countries?

The testing data from the cytogenetic lab suggest that with a median age of diagnosis of 7, Australia is similar to the United States.¹¹ Nevertheless, higher testing rates in across states and territories of Australia positively correlate with more cases being detected per million people. This is consistent with the hypothesis that more cases of FA will be detected if there is more testing for FA when there is a lower index of suspicion, and guidelines from the Fanconi Cancer Foundation provide a useful resource to consider who could be tested for FA (Supplemental Figure 1).¹ The highest testing rates in Australia and the highest number of cases detected per capita was in Queensland. It is unclear what drives this increased testing. One speculative possibility is that it could be due to referral systems, which is in some way tied to the physical proximity to the testing laboratory. Another non-exclusive speculative possibility is that maybe there is a group of doctors who elected to test more frequently for FA because of the particular awareness of this condition.

We observed that the number of tests for FA increased from 2017 onward. Interestingly, this approximately 3-fold increase in tests performed correlated with a 53% increase in cases detected per year. One possible reason for this large increase in testing numbers is because these tests are requested by bone marrow transplant teams who are systematically testing all people diagnosed with aplastic anemia for FA before a bone marrow transplant as part of clinical protocols and followed updated guidelines⁶² suggesting testing for FA before transplant in adults with inherited aplastic anemia. Therefore, this highly prudent and sensible test would only identify a small number of previously undetected cases of FA. The value, however, of those who it does identify as having FA is extremely high because it means that these individuals will receive the appropriate conditioning regimen for their transplant, as well as appropriate follow-up and care after the transplant.

We have provided an estimate of Australia’s rate of FA cases to be 1 in 141,169. This is relatively similar to some historical estimates from other countries. However, our estimation is conservative for the following reasons: (1) inconclusive testing: there are more test results that are inconclusive than abnormal/positive in the data. A portion of these tests may reasonably be due to nonexclusive reasons, such as technical limitations, borderline chromosome breakage test results, or genetic factors, such as hypomorphic variants in FANC genes. Hypomorphic variants can lead to partial FA pathway dysfunction. Documented examples of such a hypomorphic variant includes FANCA c.3624C>T p.(Ser1208∗), which was associated with delayed onset of FA symptoms (median diagnosis at 28 years of age) and reduced chromosome fragility⁶³ compared with loss-of-function variants from large deletions in FANCA.⁶⁴ Other examples of hypomorphic variants leading to improved outcomes compared with complete loss-of-function variants for the same complementation group are documented.^59,65,66 Nevertheless these individuals with partial loss-of-function variants still have an underlying genomic instability phenotype and predisposition to the manifestation of FA. (2) We do not know if all tests from the states and territories discussed in this study go to these Australian testing laboratories or to international laboratories and perhaps for challenging cases in which second opinions are sought, potentially leading to another mechanism of underestimating case numbers. (3) Cases of FA may be underreported because of the variable clinical presentation of the condition. Approximately 25% of people with FA do not exhibit classical phenotypes and reverse mosaicism can contribute to FA going undetected. These occult cases of FA may only receive a diagnosis because of health issues in adulthood, such as head and neck cancer,⁶⁶ hematological manifestations,^35,67 or infertility.⁶⁸

Possible sex bias in testing

We have found in the national FA testing data that males have higher rates than females, particularly in their first year of life. Although we cannot be certain what causes the male bias, we offer the hypothesis that dysmorphologists may be ordering these tests for any of the following reasons: males from large studies of the general population in US-based studies, were reported to have a higher congenital malformation prevalence than females,^69,70 which is not strictly linked to FA but therefore may require FA to be considered. More specifically, congenital abnormalities of the genitourinary system are more common in males as observed during pregnancies and assessment during the perinatal period in a US-based study.⁷¹ Furthermore, in a detailed study of individuals with FA, it has been reported that there was a high prevalence of genitalia abnormalities in males (50%-72.7%) compared with females (10.8%-14.3%).¹⁰ There, the detection of congenital malformations of the genitourinary system, particularly if radial abnormalities are also present, might lead to more tests for FA in newborn males in Australia than females. Another possibility is that male children with growth faltering can be more likely to be referred for evaluation,⁷² and this may be in addition to, or in parallel to, the higher rate of congenital malformations in males. We consider it unlikely that the bias is coming from pediatric hematologists-oncologists because it is rare to have cytopenias in infancy.¹

Genetics of FA in Australia and New Zealand

The distribution of genetic subtypes, (ie, which genes are mutated) is similar to what has been seen in other studies. FANCA is the most common subtype. FANCD2 was the second most frequently occurring subtype, which has also been observed as the second most common gene affected in a Spanish cohort of 227 people with FA.^16,59 However, with the small number of samples, it is unclear if this would be maintained as the cohort increases in size prospectively because FANCC (HGNC:3584) is the second most common subtype in a US-based cohort.¹³ However, it is possible that the US-based cohort has a founder effect, which accounts for the higher proportion of individuals with FA with pathogenic variants in FANCC.⁷³ Notably, the 3 cases from our cohort with pathogenic FANCD2 variants are not related.

Cancer in the FAANZ cohort

Head and neck cancer is of particular concern to people with FA. Oral lesions have been reported by 4 individuals, and 3 young adults have been diagnosed with HNSCC. This is exceptionally higher than the lifetime risk of the general population. Most of the cancers described here are from the cases added from literature searches. The cancer cases from the literature skews toward less common genetic subtypes of (BRCA2 [FANCD1], FANCJ, FANCM [HGNC:23168]) and are published cases in which the genetic diagnosis was not made until the person had experienced chemotherapy-induced toxicity. It might have been reasonable to expect myeloid malignancies in this cohort, based on international literature; however, this was not the case.

Early diagnosis and barriers to access health care system

Individuals with FA have 500- to 800-fold higher than general population in developing HNSCC.⁹ Therefore, the current guidelines suggest that routine surveillance should begin at the age of 10, which is the earliest age reported for HNSCC diagnosis in the literature, and should be coordinated with specialized oral pathologists.¹ People with FA often have numerous oral lesions, and it requires a health care provider with extensive experience in evaluating and managing HNSCC to distinguish between suspicious and noncancerous oral lesions in these individuals. Oral clinicians, such as those with backgrounds in oral medicine, oral surgery, and head and neck surgery, may serve as frontline health care professionals in the detection and monitoring of oral lesions. This has been recognized because oral manifestations can be early indicators of the condition in the subset of individuals who pass through childhood without detection of their condition and who have hypomorphic variants.^1,66 In addition, oral-brush-biopsy-based cytology has been found to be a reliable adjunct in the surveillance of individuals diagnosed with oral potentially malignant disorders,⁷⁴ including those with FA.⁵⁶

Limitations of the study

• •Inconclusive chromosome breakage test results likely result in an underestimation of the prevalence of FA in the Australian population.
• •It is not possible to establish how many tests might have been sent to laboratories overseas, which may also lead to an underestimation of FA prevalence.
• •Questionnaires were completed by parents of children with FA or adults with FA and where possible, complemented with clinical notes. This can lead to incomplete descriptions of physical features associated with the individual cases of FA. It would be ideal for future studies to perform a standardized dysmorphology assessment for all participants.

Conclusion

We have described here a cohort of individuals with FA in Australia and New Zealand, including nationwide FA testing data from Australia. We also described FA testing data. Our findings suggest that the detection rate of FA increases with more testing, which is of huge clinical importance for individuals who would otherwise be missed and therefore not receive appropriate care. We also found that the presentation of FA in Australia and New Zealand is particularly similar to other countries when considering features such as physical features, genetic subgroups, and an increased rate of bone marrow failure and HNSCC compared with the general population. Therefore, because Australia and New Zealand do not have their own clinical guidelines for the diagnosis and management of FA but clinical features are similar to what is already described, it would be reasonable to use the international Fanconi Anemia Clinical Care Guidelines.

Data Availability

The data supporting this study are available in aggregate form within the manuscript and supplemental materials. Because of the rarity of the condition studied and concerns regarding potential reidentification, individual-level data cannot be publicly shared. Researchers with a legitimate interest may request access to deidentified data, subject to appropriate ethical and institutional approvals.

Declaration of AI and AI-Assisted Technologies in the Writing Process

During the preparation of this work, the author(s) used ChatGPT to efficiently produce or edit code for plotting using R in Rstudio and to produce code for running statistical tests. After using this tool/service, the author(s) reviewed and edited the content as needed and take(s) full responsibility for the content of the publication.

Conflict of Interest

The authors declare no conflicts of interest.