Recent advances in Wilms’ tumor predisposition

Abstract

Wilms’ tumor (WT), the most common childhood kidney cancer, develops in association with an underlying germline predisposition in up to 15% of cases. Germline alterations affecting the WT1 gene and epigenetic alterations affecting the 11p15 locus are associated with a selective increase in WT risk. Nevertheless, WT also occurs in the context of more pleiotropic cancer predispositions, such as DICER1, Li-Fraumeni and Bloom syndrome, as well as Fanconi anemia. Recent germline genomic investigations have increased our understanding of the host genetic factors that influence WT risk, with sequencing of rare familial cases and large WT cohorts revealing an expanding array of predisposition genes and associated genetic conditions. Here, we describe evidence implicating WT1, the 11p15 locus, and the recently identified genes CTR9, REST and TRIM28 in WT predisposition. We discuss the clinical features, mode of inheritance and biological aspects of tumorigenesis, when known. Despite these described associations, many cases of familial WT remain unexplained. Continued investigations are needed to fully elucidate the landscape of germline genetic alterations in children with WT. Establishing a genetic diagnosis is imperative for WT families so that individuals harboring a predisposing germline variant can undergo surveillance, which should enable the early detection of tumors and use of less intensive treatments, thereby leading to improved overall outcomes.

Introduction

Wilms’ tumor (WT), the most common childhood kidney cancer, has an estimated incidence of one in 10 000 children less than 15 years of age and accounts for 6% of all childhood cancers (1–3). The disease typically occurs before 7 years of age, with the mean age at diagnosis being 3.7 years in unilateral cases and 2.6 years in bilateral cases (4). The earlier onset of bilateral tumors and occurrence of familial cases (5) has long supported the notion that hereditary predisposition plays a role in WT development. Indeed, statistical analyses of familial/bilateral versus sporadic/unilateral WT cases by Alfred Knudson and colleagues in 1972 reinforced the seminal ‘two-hit’ model of tumorigenesis that was originally proposed 1 year earlier to explain the hereditary basis of bilateral versus unilateral retinoblastoma (6,7). Knudson’s theory regarding WT was proven true in 1990 with the discovery of WT1, the first WT predisposition gene (8). Since this discovery, several additional genes associated with WT development have been identified and it is now estimated that up to 15% of cases are due to an underlying predisposition (9). The most common genetic conditions known to confer an increased risk for WT are described in Table 1. Although a full discussion of each of these syndromes is beyond the scope of this report, the reader is referred to the references contained in Table 1 and to an excellent recent review (10).

Table 1

WT predisposing conditions

Syndrome	OMIM reference	Genetic lesion	Estimated WT risk	Phenotype	References
WT1-associated Wilms’ tumor predisposition syndromes
Denys–Drash syndrome (DDS)	194 080	WT1 pathogenic variant affecting exon 8 or 9	~75%	Ambiguous genitalia, diffuse mesangial sclerosis	(104)
Frasier syndrome	136 680	WT1 pathogenic variant affecting intron 9 donor splice site	Low; 1 case reported	Ambiguous genitalia, streak gonads, focal segmental glomerulosclerosis	(105–107)
WAGR syndrome (Wilms’ tumor, aniridia, genital anomalies, retardation)	194 072	Deletion of 11p13 containing WT1	~50%	Aniridia, genitourinary anomalies, delayed-onset renal failure	(108,109)
WT1-associated WT predisposition	194 070	WT1 pathogenic variant	~30%	Median age at WT diagnosis is 1.3 years (range: 0.6 to 4.5).	(21,110–112)
Other syndromes that predispose to Wilms’ tumor
Beckwith–Wiedemann syndrome (BWS)	130 650	LOM at maternal IC2 at 11p15.5GOM at maternal IC1 at 11p15.5Paternal UPD at 11p15.5Pathogenic variant in CDKN1CNegative molecular test	~0.2%~24%~8%~1.4%~4%	Organomegaly, large birth weight, macroglossia, omphalocele, hemihypertrophy, ear pits and creases, neonatal hypoglycemia	(51,113)
Bloom syndrome	210 900	Biallelic BLM pathogenic variants	~3%	Short stature, microcephaly, growth deficiency, immune abnormalities, sensitivity to sunlight	(114,115)
Fanconi anemia	605 724, 610 832	Biallelic BRCA2 or PALB2 pathogenic variants	20–40%	Short stature, abnormal skin pigmentation, skeletal malformations, microcephaly, bone marrow failure	(116,117)
Hyperparathyroid-jaw tumor syndrome	145 001	CDC73 pathogenic variant	~3%	Primary hyperparathyroidism, ossifying fibroma of the maxilla and/or mandible	(68,69)
Mosaic variegated aneuploidy	257 300, 617 598	Biallelic BUB1B or TRIP13 pathogenic variants	>85%	Microcephaly, growth deficiency, developmental delay, eye anomalies, mild dysmorphism	(118–121)
Perlman syndrome	267 000	Biallelic DIS3L2 pathogenic variants	~65%	Organomegaly, large birth weight, developmental delay	(122)
Simpson–Golabi–Behmel syndrome	312 870	GPC3 pathogenic variant	4–9%	Overgrowth, congenital heart defects	(123,124)
Genetic variants that predispose to non-syndromic Wilms’ tumor
CTR9-associated WT predisposition		CTR9 pathogenic variant	4 families reported; 9/14 individuals with a pathogenic CTR9 variant developed WT.	Median age at WT diagnosis is 1.3 years (range: 0.6–3.3). All reported variants are paternally inherited.	(65,66)
DICER1 syndrome		DICER1 pathogenic variant	Low; 5 families in which 6/22 individuals with a pathogenic DICER1 variant developed WT.	Lung cysts, cystic nephroma, nodular hyperplasia of the thyroid, nasal chondromesenchymal hamartoma, ciliary body medulloepithelioma and increased cancer risks including pleuropulmonary blastoma (PPB), ovarian sex cord-stromal tumors, pineoblastoma and pituitary blastoma.	(125–129)
Li-Fraumeni syndrome (LFS)	151 623	TP53 pathogenic variant	Low; 12 cases reported	Around 50% of individuals with LFS develop cancer by the age of 30 years, with a lifetime risk of > 70%. The tumors most closely associated with LFS include soft tissue sarcomas, osteosarcoma, premenopausal breast cancer, brain tumors, and adrenocortical carcinoma.	(130–133)
REST-associated WT predisposition	616 806	REST pathogenic variant	4 families reported; 7/14 individuals with a pathogenic REST variant developed WT. Ten presumed sporadic cases reported.	Median age at WT diagnosis is 3 years (range: 0.5–6). Other clinical features reported include primary ovarian failure, café-au-lait macules, and mild developmental delay.	(78,79)
TRIM28-associated WT predisposition		TRIM28 pathogenic variant	9 families reported; 18/24 individuals with a pathogenic TRIM28 variant developed WT; however, few unaffected family members were tested. 17 presumed sporadic cases reported.	Median age at WT diagnosis is 1.1 years (range: 0.4–9.8). All reported variants are maternally inherited. Histology is frequently epithelial type.	(92–95)

GOM, gain of methylation; IC, imprinting center; LOM, loss of methylation; UPD, uniparental disomy

WT1-associated predisposition to WT

Normal kidney development results from the interaction of mesenchymal cells that originate from the intermediate mesoderm with epithelial cells from the invading ureteric bud. The mesenchymal cells then condense around the bud tips to form an aggregate that undergoes a mesenchyme-to-epithelial transition (MET), resulting in the formation of nephrons, the filtering units of the kidneys (Fig. 1A) (11). WT1 encodes a zinc-finger transcription factor that is required for normal development of the genitourinary system and mesothelial tissues (12). WT1 is vital for the survival and proliferation of mesenchymal cells through direct transcriptional regulation of the fibroblast growth factor (FGF) and bone morphogenetic protein (BMP)-small mothers against decapentaplegic (SMAD) signaling pathways (13). During normal kidney development, WT1 expression levels are low in undifferentiated mesenchymal cells and increase as these cells condense around the ureteric bud prior to MET (14). In mice, complete disruption of Wt1 expression results in embryonic lethality, with Wt1 mutant embryos deficient in kidney and gonadal development (15). In contrast, deletion of Wt1 in mid-gestation impairs metanephric mesenchyme differentiation and reduces the number of mature nephrons (16,17). Accordingly, WTs harboring pathogenic germline or somatic WT1 alterations are thought to arise from impaired WT1-dependent transcriptional regulation (Fig. 1B) that disrupts the MET, resulting in disorganization of nephrogenic structures (18,19) and setting the stage for tumor formation.

Figure 1

Normal kidney development and processes affected by WT predisposing lesions. (A) During normal kidney development, mesenchymal cells condense around the ureteric bud tips (Stage 1) to form an aggregate that undergoes a MET (Stage 2), resulting in the formation of nephrons, and after branching and proliferation (Stage 3), a healthy kidney. (B) WT predisposing lesions alter genomic imprinting (11p15, TRIM28), chromatin modification (REST) and transcriptional regulation (CTR9, WT1), leading to perturbations in normal kidney differentiation and generation of a cellular state that is poised for malignant transformation upon acquisition of cooperating somatic genetic lesions.

Germline WT1 alterations have been identified in an estimated 2–4% of presumed sporadic WT cases (20,21) and rarely in non-syndromic familial cases (22–24). These alterations confer variable degrees of WT risk and a broad spectrum of non-oncologic manifestations (Table 1) (10). The median age of tumor-onset for patients with germline WT1 alterations is earlier than that of patients without such alterations (1.0 years versus 3.0 years, respectively) (25), and approximately 40% of cases are bilateral (25). While most families harboring germline WT1 variants exhibit an autosomal dominant mode of inheritance with incomplete penetrance, interestingly, one recent study reported a penetrance of 100% in cases of paternal inheritance (9 of 9 cases) and 67% in cases of maternal inheritance (4 of 6 cases) (26).

Beckwith–Wiedemann syndrome

Contemporaneous with the discovery of WT1, studies of Beckwith–Wiedemann syndrome (BWS), an overgrowth condition also associated with WT development, revealed causal germline methylation defects at 11p15 with recent studies demonstrating that up to 80% of BWS patients harbor such defects (27). The 11p15 locus is normally imprinted, and it is proposed that epigenetic alterations at this locus lead to the dysregulated expression of specific genes that then spur WT formation. Among these genes is IGF2, which encodes the growth promoting protein insulin-like growth factor 2. Notably, IGF2 is one of the most commonly altered genes in WT (28–30), with somatic changes affecting IGF2 methylation identified in over 50% of cases (31,32). The abnormal methylation seen in BWS and sporadic WT cases is associated with activation of a normally silent, maternally inherited, IGF2 allele (29,33), resulting in increased IGF2 expression. This increase in IGF2 is believed to contribute to tumor formation by activating the insulin, phosphoinositide 3-kinase (PI3K) and mitogen activated protein kinase (MAPK) signaling pathways, culminating in cell cycle progression, reduced apoptosis and increased cell survival (34,35). Further supporting a role for IGF2 in WT formation are mouse studies showing that genetic ablation of Dis3l2, the gene associated with Perlman syndrome (another WT predisposing condition, Table 1), leads to upregulated Igf2 expression in nephron progenitor cells (36). Similarly, forced expression of Plag1, a microRNA target gene that is overexpressed in WT, results in upregulation of Igf2 in developing kidney mesenchyme (37). Finally, ablation of mouse Wt1 along with overexpression of Igf2 in nephron progenitor cells results in WT formation (38). Despite these studies supporting a role of IGF2 in Wilms’ tumorigenesis, the observation of 11p15 loss of heterozygosity (LOH) in normal tissues from some WT patients (39) and the lack of tumors in mice with biallelic Igf2 expression (17) suggest that upregulation of IGF2 alone is inadequate for tumorigenesis (40).

Other genes impacted by 11p15 methylation defects include CDKN1C, which encodes the critical cell cycle regulator Cyclin Dependent Kinase Inhibitor 1C, and H19, which encodes a long non-coding RNA that functions as a regulator of growth and body weight (41–43). Notably, a recent study examining the origin of WT from embryonic precursor cells reported that H19 hypermethylation was prevalent in clonal WT precursor lesions, but not in normal kidney cells (44). From this work, the authors suggest that H19 hypermethylation represents an epigenetic mark that is essential for WT formation. These results are consistent with the finding that 11p15 LOH is one of the commonest somatic findings in WT (45,46). Altogether, these data suggest that epigenetic or copy number alterations at 11p15 lead to the dysregulated expression of one or more genes that create a cellular state primed for subsequent malignant transformation.

Clinically, BWS is characterized by varying degrees of overgrowth, macroglossia, hemihypertrophy, hyperinsulinemia, abdominal wall defects and development of embryonal tumors (most often WT, but hepatoblastoma, neuroblastoma, rhabdomyosarcoma, adrenocortical and other tumors have also been reported) (47–49). BWS affects around 1 in 10 000 live births (50), with approximately 7% of patients developing WT (1,51,52). While the majority of BWS cases are associated with altered methylation and typically not inherited, about 10–15% are caused by germline loss-of-function (LOF) variants in CDKN1C, pathogenic single nucleotide or copy number variants at 11p15 or deletions/duplications of 11p15 itself (27,53). These latter cases can exhibit autosomal dominant inheritance with penetrance dependent on whether the allele was transmitted maternally or paternally (as summarized in (51)).

Recent discoveries in WT predisposition

While the studies above have provided important insights into normal kidney development and tumor formation, much remains to be learned about the inherited and somatic genetic events associated with WT development. Toward this end, additional familial WT loci have been mapped by genome-wide linkage analysis to chromosomes 17q12-q21 (FWT1) (54,55), 19q13 (FWT2) (56) and to other loci (22,23,57,58); however, the causative genes at these loci have not yet been identified. Recently, several reports have described novel genes in which pathogenic germline variants confer an increased risk for WT, including CTR9, REST and TRIM28.

CTR9-associated predisposition to WT

CTR9 encodes a component of the Polymerase-Associated Factor 1 (PAF1) complex, which associates with RNA polymerase II (59), a large protein complex that transcribes DNA into messenger RNA and several small nuclear RNAs. The human PAF1 complex is comprised of several subunits including CTR9, CDC73, LEO1, PAF1, RTF1 and SKI8 (60). Studies of the Paf1 complex in yeast have revealed multiple roles including gene regulation, transcriptional elongation and chromatin modifications (61,62). The Paf1 complex is recruited to the transcriptional machinery by CDC73 (63) or CTR9 (64) binding to RNA Pol II.

CTR9 was first implicated in familial WT in 2014 (65) and to date, four unrelated families harboring germline variants in CTR9 have been reported (65,66). These pedigrees all exhibited an autosomal dominant mode of inheritance with nine of 14 CTR9 variant carriers developing WT (median age at WT diagnosis is 1.3 years (range: 0.6–3.3 years; Table 1, Fig. 2). Interestingly, the pathogenic CTR9 allele was inherited from the father in all nine of these cases. Three of the identified CTR9 variants (Fig. 3A) are in consensus splice site sequences and predicted to result in aberrant splicing and in-frame deletion of exon 9, suggesting that exon 9 deletion may be a common mechanism of tumorigenesis. The fourth CTR9 variant truncates the protein in exon 2 (c.106C > T, p.Gln36^*).

Figure 2

Genetic, clinical and demographic features of WT cases with germline lesions in CTR9, REST and TRIM28. Values for age and gender distribution, histology, familial or presumed sporadic occurrence, inheritance and secondary somatic findings were derived from the following reports: CTR9-associated WT predisposition (n = 9) (65, 66); REST-associated WT predisposition (n = 17) (78, 79); TRIM28-associated WT predisposition (n = 35) (92–95). The n-value indicates the number of cases for which data were available for each feature. Age is described in years. For histology counts, bilateral cases were counted as two tumors (e.g. left tumor of epithelial subtype and right tumor blastemal subtype) if laterality data were available. Cohort type corresponds to either familial cases, where multiple individuals with WT are reported in a family, or presumed sporadic, where the individual represents the only reported case. Inheritance was derived for cases in which parental genetic testing was performed. Somatic second hits in individuals with germline variants are described when available. The asterisks (^*) indicate that bar height corresponds to the percentage of cases in each bin, such that the bins total 100% and the denominator is the number of cases with reported information regarding that feature. Other; other mixed or undefined tumor types. LOH; loss of heterozygosity.

Figure 3

Protein structures and distribution of reported pathogenic germline variants in individuals with WT. (A) CTR9 variants all result in deletion of exon 9. (B) REST variants are primarily clustered within the DNA-binding domain. (C) TRIM28 variants are mostly nonsense of LOF frameshift variants expected to result in nonsense-mediated decay and distributed throughout the protein. Variants are labelled with counts of individuals with WT harboring each variant and colored per their classification, splice (purple), frameshift (red), nonsense (orange), missense (blue) and stop loss (black). Known domains of the proteins are as shown. Variant panels were generated using ProteinPaint (103).

Descriptions of the WTs developing in CTR9 families are somewhat limited. Tumor features were only available for one family comprising three cases, one of which was bilateral. All three tumors showed alteration of the second CTR9 allele (Fig. 2), two demonstrating tumor LOH and the third with a second somatic mutation, supporting the notion that CTR9 functions as a tumor suppressor. Although the exact mechanism by which CTR9 LOF contributes to kidney tumor formation is yet to be determined, as the PAF1 complex is involved in regulation of gene expression, DNA repair and cell cycle, it is possible that a disturbance in one or more of these functions leads to Wilms’ tumorigenesis (Fig. 1B) (62,67). Of note, the PAF1 complex has also been implicated in the hyperparathyroidism-jaw tumor syndrome (HPT-JT), which is caused by germline pathogenic variants in the CDC73 gene and is rarely associated with WT development (68,69). Three individuals harboring a CDC73 variant have been reported to develop WT, one of whom presented at 53 years of age (68,69). The increased risk of WT in individuals harboring pathogenic CTR9 variants versus those harboring CDC73 variants may be due to the fact that unlike CDC73, CTR9 is involved in organ development during embryogenesis (70) and the maintenance of embryonic stem cell identity (71). Further supporting potential differences in protein function and resulting clinical features, none of the phenotypes related to HPT-JT have been reported in any of the 14 known CTR9 variant carriers.

REST-associated predisposition to WT

RE1-silencing transcription factor (REST; also known as Neuron-Restrictive Silencer Factor) encodes a Krüppel-associated box (KRAB) zinc-finger transcription factor made up of two repressor domains (RD1, RD2) and a DNA-binding domain (DBD). REST serves as a focal point for the recruitment of chromatin-modifying enzymes that silence the expression of target genes (72,73) and play a critical role during embryonic development and neurogenesis (74,75). In mice, ubiquitous deletion of Rest causes upregulation of neuronal transcripts in non-neural tissues and embryonic lethality (76), while conditional deletion in the common progenitors of glia and neurons causes genomic instability and premature expression of neuronal transcripts (77).

REST was first implicated in WT predisposition in 2015 when four unrelated families with pathogenic variants in REST were reported (78). These pedigrees exhibited an autosomal dominant mode of inheritance with seven of 14 carriers developing WT. An additional 10 non-familial, presumably sporadic, cases have also been reported (78,79). The median age at WT diagnosis in carriers of pathogenic REST variants is 3 years (range: 0.5–6 years; Table 1, Fig. 2). Interestingly, 12 of 13 specified variants were clustered within the DNA-binding domain (5 frameshift, 5 missense, 1 nonsense and 1 splice), and one nonsense variant was located C-terminal to the DBD and predicted to remove the RD2 zinc finger (78) (Fig. 3B). These findings suggest that pathogenic WT predisposing variants likely result in altered gene regulation (Fig. 1B).

REST-associated WT predisposition does not seem to be associated with a specific histologic subtype; of the nine cases where histologic data were provided, four were triphasic, three blastemal predominant, one of mixed blastemal/epithelial histology and one of non-classical histology with very little stroma and minimal differentiation (78) (Fig. 2). Information on laterality was available for only one case, which was reported as unilateral (79). While the mechanism(s) by which germline variants in REST contribute to WT remain to be determined, some KRAB zinc-finger proteins can mediate transcriptional repression by recruiting the co-regulator TRIM28 (80), a gene in which pathogenic germline variants also predispose to WT.

TRIM28-associated predisposition to WT

Tripartite motif containing 28 (TRIM28; also known as KAP1) serves as a co-regulator for the KRAB proteins (81). TRIM28-associated complexes contribute to many aspects of cellular biology, including proliferation, genome stability, immune response, early embryonic development and embryonic stem cell pluripotency (82–85). Importantly, TRIM28 controls genomic imprinting through distinct mechanisms at different developmental stages (86). In mice, maternal deletion of Trim28, achieved by deletion from oocytes and subsequent fertilization, causes a severe phenotype associated with aberrant H19 methylation and embryonic lethality (87,88). Interestingly, Igf2 was downregulated in 4 of 6 maternally deleted Trim28 mutant embryos, whereas H19 was upregulated, indicating that maternal Trim28 protects paternal H19 from aberrant DNA demethylation (87). Zygotic Trim28 deletion also results in downregulation of Igf2, upregulation of H19 (86) and embryonic lethality (89). While the downregulation of Igf2 is the opposite of what is observed in many blastemal or triphasic WTs, it is important to note that TRIM28-associated WT often contains a predominance of epithelial cells, which generally express lower levels of IGF2 (90). Thus, IGF2 upregulation may not be as crucial for the formation of epithelial-predominant TRIM28-associated tumors. This is further supported by reports of lower IGF2 expression and normal imprinting of 11p15 in a subset of tumors exhibiting epithelial-predominant histology (91).

TRIM28 was first implicated in WT predisposition in 2018 (92) and to date, nine unrelated families harboring TRIM28 variants have been reported (92–94). Of note, TRIM28 is in proximity to the FWT locus, 19q13.4, discovered through linkage analysis of five familial WT pedigrees; however, it has not been confirmed to be the causative gene (56). Affected families exhibit an autosomal dominant mode of inheritance with 18 of 24 TRIM28 variant carriers developing WT. In 18 patients whose parents were also tested, all reported alleles were either maternally inherited (13 of 18) or de novo (5 of 18) (Fig. 2). An additional 17 non-familial, presumably sporadic, WT cases harboring germline TRIM28 variants have also been reported (92,95). The median age at WT diagnosis is 1.1 years (range: 0.4–9.8 years; Table 1, Fig. 2). Except for one missense variant segregating with WT in a kindred containing six affected individuals, a de novo stop loss variant (94), and two canonical splice variants, all reported TRIM28 variants are nonsense or LOF frameshifts expected to result in nonsense-mediated decay (Fig. 3C).

Twenty-three of 30 WTs with pathology data exhibit a predominantly epithelial histology (Fig. 2). Among 17 tumors with genetic data, 15 show LOH at the TRIM28 locus (93), while one tumor harbored a 2nd somatic hit (93), and another exhibited methylation of exon 1 (92). Collectively, these findings suggest that TRIM28 functions as a classical tumor suppressor. While the mechanisms by which TRIM28 inactivation induces Wilms’ tumorigenesis remain to be elucidated (Fig. 1C), TRIM28 plays an important role in the developing kidney. Toward this end, the silencing of Trim28 in rat kidney rudiments results in branching arrest of ureteric bud structures (96). Consistent with models of kidney development in which bud loss is linked to nephrogenic failure (97,98), it is possible that the branching arrest of bud structures results in failed growth inhibition of early epithelial structures derived from the undifferentiated metanephric mesenchyme. It is also possible that the interaction of TRIM28 with WTX/AMER1 promotes tumorigenesis (99).

Evaluation and management of children with hereditary predisposition to WT

A genetics referral is recommended for all children with WT who have a positive family history of cancer, bilateral kidney involvement or presence of syndrome-specific features. Investigators in Canada have developed the McGill Interactive Pediatric OncoGenetic Guidelines (MiPOGG) to assist in determining whether a child with WT should undergo a genetics referral (79). The MiPOGG algorithm consists of five criteria including 1) diagnosis before 2 years of age, 2) bilateral or multifocal tumors, 3) stromal-predominant histology, 4) presence of nephrogenic rests and 5) overgrowth features. These criteria should be taken into consideration along with the family history and presence of congenital anomalies and/or dysmorphic features.

In the absence of syndromic features or a family history suggestive of any specific cancer predisposition syndromes, clinical germline testing should include sequencing and deletion/duplication analysis of CTR9, DICER1, REST, TP53, TRIM28 and WT1; other genes can be added at the discretion of the genetics provider, based on personal medical or family history features. If the child is found to harbor a pathogenic or likely pathogenic variant in a WT predisposition gene, his or her parents and close relatives can also be offered testing. Affected individuals should be counseled about the risks for additional neoplasms and non-oncologic manifestations, as appropriate, as well as the risks for recurrence in future offspring.

Children testing positive should be offered surveillance throughout the period of increased WT risk, which is typically up to 8 years of age but may vary depending upon the condition (10). The goal of surveillance is to detect WTs while they are low-stage and more likely to be cured using fewer intensive therapies. Luckily, it is easy to visualize the kidneys by radiologic methods such as ultrasound, which is a readily available, safe and easy and relatively inexpensive procedure. As WTs can double in size every week (1), it is recommended that an abdominal ultrasound be completed once every 3 months. Data describing the impacts of renal surveillance on survival and preservation of renal function are areas of active investigation (100). Additional modes of surveillance can be considered for individuals with predisposition to a wider spectrum of cancers, such as those with Li-Fraumeni (101) or DICER1 syndrome (102).

Summary and future opportunities

Via a variety of mechanisms, WT predisposing variants perturb the expression of genes required for normal kidney development and create a state that is primed for malignant transformation upon the acquisition of additional somatic second hits. The existence of unsolved familial and bilateral cases indicates that additional predisposing lesions remain to be identified. In these remaining cases, genetic and/or epigenetic discoveries may be challenging as causal lesions could reside in non-coding regions of the genome or involve more complex mechanisms, such as structural variants (including intronic deletions or inversions), digenic or polygenic inheritance and events that occur post-zygotically. All the same, the continued study and collation of data from additional families and sporadic cases is essential to identify the responsible germline predisposing events, elucidate the underlying mechanisms of tumorigenesis, estimate the age-specific cancer risks and use this information to optimize WT surveillance and treatment.

Acknowledgements

The authors wish to thank Joshua Stokes in Biomedical Communications at St. Jude for his assistance in preparing Figure 1.

Conflict of Interest

The authors declare no conflict of interest.