Wilms tumour

Wilms tumour (WT) is a childhood embryonal tumour that is paradigmatic of the intersection between disrupted organogenesis and tumorigenesis. Many WT genes play a critical (non-redundant) role in early nephrogenesis. Improving patient outcomes requires advances in understanding and targeting of the multiple genes and cellular control pathways now identified as active in WT development. Decades of clinical and basic research have helped to gradually optimize clinical care. Curative therapy is achievable in 90% of affected children, even those with disseminated disease, yet survival disparities within and between countries exist and deserve commitment to change. Updated epidemiological studies have also provided novel insights into global incidence variations. Introduction of biology-driven approaches to risk stratification and new drug development has been slower in WT than in other childhood tumours. Current prognostic classification for children with WT is grounded in clinical and pathological findings and in dedicated protocols on molecular alterations. Treatment includes conventional cytotoxic chemotherapy and surgery, and radiation therapy in some cases. Advanced imaging to capture tumour composition, optimizing irradiation techniques to reduce target volumes, and evaluation of newer surgical procedures are key areas for future research. Wilms tumour (WT) is the most common renal tumour in infants and young children. This Primer reviews the epidemiology, mechanisms, diagnosis and management of WT. In addition, the authors outline potential opportunities to translate novel biological targets to improve clinical outcomes.

Wilms tumour (WT) is the most common renal tumour of infants and young children 1,2 . WT is inti mately linked to early nephrogenesis, which it resem bles morphologically 3 and transcriptionally 4,5 . WT may occur sporadically or in the context of bilateral tumours, multifocal disease and specified genetic predisposition syndromes that frequently include either genitourinary malformation or overgrowth 3 . Beyond genetic pre disposition, external causative factors for WT are not yet defined. The molecular drivers frequently involve blockade of genetic pathways that guide normal embryo genesis of the genitourinary tract but are not restricted to these. Indeed, the cancer genes that underpin WT are diverse and surprisingly involve ~40 genes.
The implementation of international co operative group trials and studies across North America, Australia, New Zealand, Europe and Brazil has contributed sig nificantly to improving outcomes [6][7][8] . Two interna tional multidisciplinary cooperative consortia -the Children's Oncology Group (COG) Renal Tumour Committee, previously known as the National Wilms Tumour Study Group (NWTSG), and the International Society of Paediatric Oncology (SIOP) Renal Tumour Study Group (RTSG) -have conducted large multi centre studies since 1969 and 1971, respectively, which have defined the current diagnostic and therapeutic approach to patients with WT (Fig. 1). These groups continue research to optimize disease and patient risk classification and treatment strategies 9-11 . In the COG, WTs are treated with primary resection (if possible), followed by risk adapted adjuvant therapy, whereas in the context of SIOP cooperation, neoadju vant chemotherapy followed by resection and adjuvant therapy is the preferred treatment approach. Regardless of the initial approach, the overall survival of children with WT is remarkable with rates of >90%. Such satisfy ing survival rates have been achieved at the same time as fine tuning treatment by adopting well studied prognos tic factors, leading to a two drug regimen (vincristine and actinomycin D) prescribed in nearly two thirds of affected children 7,10 . Notably, striking survival disparities still exist within countries 12 and between different parts of the world, which remain to be addressed 13,14 . However, 20% of patients relapse after first line therapy and up to 25% of survivors report severe late morbidity of treatment 15,16 . Addressing the long term effect of radical nephrectomy on renal function and cardiovascular func tion will probably drive more attention on expanding the role of nephron-sparing surgery (NSS) 17 .
Molecular studies are expanding the landscape of cancer genes implicated in WT beyond exclusive roles in nephrogenesis 3 . The use of next generation integrative genomic and epigenetic tumour analy sis has provided important insights into WT biology. Comparisons of the regulation of progenitor cells in the fetal kidney with the disrupted regulation of their counterparts in WT should provide further insights into tumour formation 18 . Targeting WT tumour genes with a non redundant role in nephrogenesis and targeting the fetal renal transcriptome warrant further therapeu tic exploration. Interventions that could prevent the evolution of nephrogenic rests to malignant WT could transform therapy in this setting and could even lead to preventive strategies in children known to be at high risk of developing WT.
This Primer describes our current understanding of WT epidemiology, disease susceptibility and mecha nisms, as well as elements of clinical care, including diag nostics and risk stratified treatment of newly diagnosed disease. In addition, we also outline potential opportu nities to further translate new biological insights into improved clinical outcomes. We discuss how the wide spread implementation of standardized diagnostics and treatments for as many children as possible, regardless of socioeconomic status or geographical region of origin, may propel further clinical advances.

Global disease burden
Malignant renal tumours comprise 5% of all cancers occurring before the age of 15 years 19 . Every year ~14,000 children (0-14 years of age) are diagnosed worldwide, and 5,000 children die from these diseases, with regional variation in mortality 20 (Fig. 2). The incidence of child hood renal tumours is not associated with economic status, but mortality is higher in low income areas than high income areas (0.5 per million in high income areas versus 7.5 per million in low income areas).
WT is the most common renal tumour in children 1 and studies have found variation in incidence between regions or ethnicities 2,21 (Fig. 3). The annual incidence of WT in East Asia is lower than in North America or Europe (4.3 per million versus 8-9 per million) 2 . In the USA, children with African American ances try have the highest incidence (9.7 per million) whilst those with Asian Pacific Islander ancestry have the lowest (3.7 per million) 2 . However, owing to the lack of population based childhood cancer registries in resource constrained regions, or because of the low quality of the data (that is, not all cancers are reported or not all children are reported), the estimation of global incidence has been difficult 14,22,23 . In addition, 50% of patients from areas with less resources have metastases at diagnosis 24 .
Up to 17% of WT occur as part of a recognizable malformation syndrome 25 , 10% of which are asso ciated with known WT predisposition 26 (TAble 1). Overgrowth syndromes, in particular Beckwith-Wiedemann syndrome, carry ~5% risk of developing WT, ranging from 0.2% to 24% according to the under lying genetic cause [27][28][29] . Syndromes involving genito urinary anomalies combined with aniridia and variable intellectual disability, or with nephrotic syndrome, are associated with mutations of the gene WT1 on chromo some 11p13 and these patients have a greatly increased risk of developing WT 3,30,31 .
No temporal trends in the incidence of WT were observed within the period 1996-2010 (reF. 2 ), suggest ing that environmental factors play a marginal role in WT aetiology. Nevertheless, modifiable risk factors for WT are not well understood. the lower total incidence in East Asian populations is consistent with genetic factors primarily driving WT. Studies with large samples from many countries and different ethnic groups will be needed to validate the likelihood that the genetic heterogeneity of WT explains this variation in clinical features by ethnicity.
Mechanisms/pathophysiology WT is an embryonal malignancy thought to arise through abortive or disrupted development 36 . During kidney embryogenesis, intermediate mesoderm dif ferentiates into metanephric mesenchyme, which con denses around the branching ureteric bud structures. This metanephric mesenchyme undergoes a mesen chymal to epithelial transformation to form renal ves icles, which expand and give rise to the majority of cell types of the functional kidney 37 . In WT, this process can be disrupted at different levels, leading to variable mixtures of blastemal, epithelial and stromal cells that may even exhibit myogenic differentiation. Histology is partly shaped by the underlying genetic defects but may also reflect the timing of divergence from normal nephrogenesis (Fig. 5).
Our understanding of the genetic causes of WT has long been limited to mutations of WT1, CTNNB1 and WTX as well as loss of H19-IGF2 imprinting, but these alterations only explain a subset of cases 38 . Additional features such as allele loss on chromosomes 1p and 16q or gain of 1q may underpin aggressive clinical behaviour in some cases, but do not provide mechanistic insights into tumour development or therapeutic targets 39-41 .
Next generation sequencing analyses have unveiled many additional drivers, mostly chromatin modifying and transcription factors as well as microRNA (miRNA) processing genes, many of which are involved in nor mal renal development [42][43][44] (TAble 2; bOx 1). A surpris ingly large fraction of WT (up to 17%) occur in the context of genetic malformation syndromes associated with tumour predisposition 25 (TAble 1). The paradigms are WAgr syndrome and Beckwith-Wiedemann syn drome, which led to the understanding that defects in the tumour suppressor gene WT1 and loss of H19-IGF2 imprinting predisposes to WT.

WT1, CTNNB1 and stromal WT
WT1 was originally identified through homozygous deletions in WT 45,46 . Nevertheless, the functions of this zinc finger protein are more complex -germline inac tivation leads to male genitourinary anomalies, such as hypospadias, cryptorchidism, through haploinsufficiency and is associated with an increased risk of developing WT (>50%) 47 . Additionally, dominant negative muta tions, especially of the zinc finger domains that abro gate DNA binding, lead to Denys-Drash syndrome with intersex and renal failure due to diffuse mesangial scle rosis and the risk of developing WT increased >90% 48 . Of note, Frasier syndrome, in which intronic mutations alter the balance of WT1 splice isoforms rather than altering the WT1 protein amino acid sequence, includes different forms of intersex and renal failure, and car ries a risk of gonadoblastoma in streak gonads rather than WT 49,50 .
6. Omission of lung radiotherapy proved safe in patients with complete metastatic response to induction chemotherapy

Hypospadias
An anatomical congenital malformation of the male external genitalia, characterized by abnormal development of the urethral fold and the ventral foreskin of the penis that causes abnormal positioning of the urethral opening.

Cryptorchidism
The absence of at least one testicle from the scrotum. Mutations in WT1 are often paired with frequent alterations of CTNNB1, which lead to constitutive Wnt signalling 51 . In most cases, point mutations or deletions are observed in the phosphodegron motif in exon 3, leading to β catenin stabilization and nuclear accumu lation, where it acts as co activator with the TCF-LEF transcription factors. These tumours usually exhibit stromal predominant histology, decreased response to preoperative chemotherapy and represent up to 15% of cases in Caucasian populations 52 . Notably, although the incidence of WT in Japanese children is only 50% of that found in Caucasian children, an increased rate of WT1 mutations (81%) is observed in bilateral WTs, which points to differences in genetic constitutions 53 . WT1 driven stromal tumours occur at a median age of 22 months and are characterized by the presence of intralobar nephrogenic rests as presumed precursor lesions. WTX may likewise facilitate Wnt signalling as it is part of the β catenin degradation complex. Mutations or loss of expression of WTX are observed in up to 30% across histological subtypes, but with intratumoural heterogeneity, suggestive of a late event 54 .

H19-IGF2 imprinting
Chromosome 11p15.5 is frequently altered in WT through copy neutral loss of heterozygosity with invar iant loss of the maternal allele or loss of imprinting with epigenetic changes on the maternal allele. The net out come being hypermethylation of the imprinting centre IC1 with elevated expression of the neighbouring growth factor gene IGF2, among others. With ~70% incidence of such alterations, overexpression of IGF2 is the most frequent change in WT 43,44,52 . However, tumour initiation may require additional events as loss of imprinting on its own in Beckwith-Wiedemann syndrome only modestly increases WT risk 27 .

microRNA biogenesis mutations
An unexpected addition to WT driver genes is miRNA processing genes. miRNA biogenesis covers a stepwise maturation process from pri miRNA via pre miRNA to mature miRNA. Mutations in WT primarily affect the so called microprocessor genes DROSHA and DGCR8, which are involved in pri miRNA processing 43,44,55,56 . Heterozygous DROSHA mutations tend to inactivate the catalytic core of one of the two RNAse III domains that processes pri miRNA molecules. DGCR8 mutations affect a single amino acid (E518K) in one of the double stranded RNA binding domains and the mutations occur homo zygously or with monoallelic expression of the mutant. The net result is a reduced and unbalanced miRNA processing. DGCR8 mutations have been observed predominantly in girls, which remains to be explained. DICER1, which encodes the second RNAse III type enzyme, is rarely mutated in WT, but predisposes to pleuro pulmonary blastoma and is implicated in the very rare entity, anaplastic sarcoma of the kidney 57,58 . The cata lytic centre is often mutated on the single functional allele, leading to a partial processing defect with a deficiency in miRNA5p and largely unaffected miRNA3p levels.

Frasier syndrome
A rare autosomal recessive disorder that presents with male pseudohermaphroditism with gonadal dysgenesis, renal failure in early adulthood and increased risk of developing gonadoblastoma.  234 . Reprinted from GLOBOCAN 2020, International Agency for Research on Cancer, Estimated age-standardized mortality rates (world) in 2020, kidney, both sexes, ages 0-14, copyright 2020 (reF. 234 ). and downstream let-7 miRNA modulators such as DIS3L2 or LIN28B at a lower frequency in the 1% range but the mechanistic details are as yet unclear. Nevertheless, almost all steps of miRNA biogenesis can be critically altered to drive WT formation and several of these mutations are rather specific to WT. The fact that most mutations do not fully abrogate miRNA pro duction implies that specific miRNA subsets are impor tant to control deviation from regular developmental progression or cell proliferation and survival.

MYCN and transcriptional control
Alterations in MYCN may contribute to WT biol ogy in several ways. Elevated expression levels have been described, especially in relapsing and fatal cases. Furthermore, studies have identified specific P44L point mutations or copy number gains with one or a few addi tional copies 44,55,59 . Proline 44 is located immediately upstream of the conserved MYC box I that interacts with AURKB, FBXW7 (reF. 60 ) and GSK3 to control N Myc sta bility. Stabilization occurs through dephosphorylation at threonine58 by the phosphatase EYA1, which is recruited to the nucleus by the transcription factor SIX2 (reF. 61 ). This process provides a direct link to the paralogous genes SIX1 and SIX2 that control early kidney development. SIX1 and SIX2 can be found as drivers of blastemal predominant WT if their DNA binding domain becomes subtly altered by stereotypic Q177R mutations 44 .
Intriguingly, several MYC interacting protein com plexes can be targets of mutations in WT. The obligate heterodimerization partner MAX can exhibit R60Q mutations in the helix-loop-helix domain to alter its transcriptional potency. N Myc exerts its effects on transcriptional control through a multitude of interac tions with the core transcriptional machinery to regulate polymerase pausing. The PAF1 transcription complex is one of the critical interactors in this respect and sev eral of its components such as CDC73, MLLT1-ENL and CTR9 have been shown to be mutated in familial and sporadic cases of WT 62 . Collectively, these data indicate that MYCN hyperactivity through various means can contribute to WT through a multitude of mechanisms.

Epigenetic modifiers
A striking genotype-phenotype correlation is observed in epithelial predominant WT, which is often driven by inactivating TRIM28 mutations (reFs 63-66 ). Gene expression analyses have identified these tumours as more mature, post induction tumours with excellent prognosis. TRIM28 is part of a chromatin silencing complex that has an important function in the repres sion of endogenous retroviral transcript in embryonic cells. Indeed, these tumours show strong induction of transcripts from repetitive elements, but the mechanistic links to oncogenic transformation in these tumours with otherwise few mutations remains to be established.
Besides TRIM28, studies have shown that a whole array of epigenetic regulators are potential drivers in WT. These regulators include REST, RERE  truncating or missense mutations, some being heritable. Intriguingly, BCOR is also the main culprit for driving clear cell sarcoma of the kidney, another childhood renal tumour. In this case the gene is not inactivated as in WT, but harbours small carboxy terminal tandem duplica tions corresponding to ~30 amino acids that encompass the binding domains for PCGF1 and KDM2B as part of the polycomb repressive complex (PRC1) that controls, for example, mesodermal differentiation 67,68 .

TP53 and anaplasia
WTs generally have a low mutation load that increases with patient age, and karyotypes tend to be stable 43,44 . However, the mutation load is different in diffuse ana plastic WT, which frequently harbours oncogenic TP53 mutations. In most cases, the wild type allele is lost and the cells are characterized by chromosomal instability including chromothripsis and deregulation of cell cycle and DNA repair genes 59,69-71 . TP53 mutated tumours exhibit strong positive p53 staining owing to the accu mulation of the mutant protein, although a smaller fraction demonstrate negative staining due to null muta tions. TP53 alterations are secondary events in WT pro gression, in line with WT being reported as a rare feature in li-Fraumeni syndrome 72 . Whether TP53 mutation confers an additional risk independent of the high risk status of morphological anaplasia is still unknown. Several other genes that fall into the category of genome maintenance and repair, such as BRCA2, PALB2 and TRIP13, have been found to be mutated in WT 73 . Whether such mutations likewise increase mutation load or chromosomal aberrations remains to be determined, and no reports are available on aneuploidy yet.

Nephrogenic rests
The underlying genetic defects also have an impact on the presence of nephrogenic rests in the kidney that occur in 30-40% of patients 74,75 . These precursor lesions are foci of embryonic renal cells that abnormally persist beyond 36 weeks of gestation. Nephrogenic rests are his tologically and anatomically classified as either perilobar Chromothripsis A catastrophic chromosomal shattering event associated with random rejoining.

Li-Fraumeni syndrome
An inherited autosomal dominant cancer predisposition disorder that is usually associated with abnormalities in TP53 located on chromosome 17p13.

Anaplasia
Cells with hyperchromatic, pleomorphic nuclei that are three times larger than adjacent cells and have abnormal mitotic figures. Anaplasia is associated with a poor response to chemotherapy.  74 . WT1 related WTs frequently carry few intralobar nephrogenic rests, centrally located within or adjacent to the renal medulla, suggestive of an early developmental lesion. TRIM28 associated or Beckwith-Wiedemann syndrome associated WTs tend to harbour perilobar nephrogenic rests in the adjacent kidney tissue rather than intralobar nephrogenic rests. These perilobar nephrogenic rests may even encompass the entire renal cortex in extreme cases. Although few samples have been assayed thus far, nephrogenic rests seem to carry even fewer mutations than their adjacent WT 5,76,77 .

Bilateral Wilms tumour
Almost one in ten children present with bilateral WT or bilateral disease (WT with nephrogenic rests or nephro blastomatosis visible on imaging in the contralateral kid ney), especially in syndromic cases 32 . WT1 is the most prominent driver in these cases 32,52 , together with spe cific imprinting abnormalities at 11p15 affecting IGF2, although neither explain the majority of cases. TRIM28 inactivation is also frequent in bilateral and familial tumours 64-66, 73 . Importantly, bilateral tumours can be due to early postzygotic founder mutations in somatic cells that emerge before the divergence of left and right kidney primordia 5 . Individual clones may expand to yield mosaic kidneys with molecular evidence of clonal (mosaic) nephrogenesis. Thus, it may be justified to compare bilateral or multifocal tumours with blood and surrounding normal kidney tissue as controls to differentiate putative germline mutations, postzygotic mosaic events or single tumours with metastatic disease.

Heterogeneity and subclassification
Molecular analysis has unveiled intratumoural hetero geneity of WTs, with either chromosomal copy number alterations or mutations, for example, in WTX or TP53, being present in only a fraction of cells as evidence of tumour evolution 78 . These differences may become clearer with single cell or single nucleus analyses, which already highlighted a great cellular diversity 4 . Even the main driver genes stratify WTs according to the age of the child (for example, TRIM28, WT1 in WT occurring at younger ages (generally <2 years of age), TP53 (generally >4 years of age) and Beckwith-Wiedemann syndrome occurring at older ages (3-4 years of age)), location of nephrogenic rests (intralobar versus perilobar) or his tology (WT1 in stromal WTs, TRIM28 in epithelial WTs and TP53 in anaplastic WTs) (Fig. 5). Nevertheless, the majority of triphasic or blastemal predominant WTs do not carry defining genetic alterations.

Liquid biopsies
Although WT represents >80% of paediatric renal tumours, other intrarenal tumours exist that are important to differ entiate as therapeutic approaches are markedly diverse 31 . These non Wilms renal tumours are often characterized by rather specific molecular alterations (Fig. 5). These tumours may become amenable to liquid biopsy diagnostics looking for diagnostic changes or entity specific patterns of methylation 79 . In particular, if a neoadjuvant chemother apy approach is planned and the clinical pattern is unusual, such tests will become helpful to rule out non WT from the start, or to follow response to treatment during follow up. The fact that patients with paediatric kidney tumours often have large amounts of circulating tumour DNA makes this approach rather promising 80 .

Tumour models
Modelling WT in the mouse has been difficult with only a few successful scenarios, the first using Wt1 ablation together with Igf2 upregulation 81 . Other researchers have successfully employed Lin28 overexpression or Dis3l2 mutation 82,83 . On the other hand, prototypic Drosha mutations or Wtx deletions did not yield evidence of WT formation but led to either kidney agenesis or aberrant kidney development and functional impairment 84,85 .
Modelling efforts, including patient derived xeno grafts (PDX) 86 , can now be complemented with spheroid and organoid techniques to grow tumour cells in vitro for genetic and histological characterization and for therapeutic compound testing [87][88][89] . These models will become an invaluable resource to test novel agents in patients with relapse who respond poorly to conven tional regimens, provided a time frame suitable for clinical feedback can be accomplished 90 .

Diagnosis, screening and prevention Clinical presentation
Most children with WT are asymptomatic at presenta tion and predominantly have a distended abdomen with a palpable mass 91 . Frequently, the parents notice such a mass during dressing or cuddling. Alternatively, WT is identified by the general practitioner or the paediatrician during a regular clinical assessment of a well child or a child with non specific symptoms. WT usually reveals a non tender, large flank mass, which does not move with respiration in contrast to splenomegaly. Approximately only one in five children have distinct symptoms; pain, haematuria, fever, hypertension, urinary tract infec tions, constipation and weight loss are among the most common complaints at presentation 31, 91 . Although rare, symptoms related to metastases, such as dyspnoea (lung), abdominal pain (liver) or tumour thrombus in the renal vein or vena cava or varicocele, may occur 92 . Ultimately, a few children with severe subcapsular haem orrhage may present with rapid abdominal enlargement, anaemia and severe pain. Age at presentation is typi cally in the range 2-5 years and the incidence of WT in children >10 years is rare. In children with known predisposing syndromes, WT may be captured during routine screening and often at an earlier age or stage, and these children are even more likely to be asymptomatic than children without predisposing syndromes 93 .
In low income countries (LICs), interactions between multiple factors usually contribute to a delayed diagnosis compared with diagnosis in high income regions 94,95 (bOx 2). These factors include family or rel atives' awareness of a possible cancer, contacting and arrival in primary care, health care staff recognition of cancer and transfer to tertiary care. Furthermore, a much higher number of children in LICs have a dis tended abdomen due to other conditions, for example, malnutrition, parasitic infections and benign blood diseases than in high income regions. Hence, identi fying, differentiating and prioritizing investigations of the relatively few cases of WT is challenging. Moreover, the latency to diagnosis (patient interval and diagnos tic interval) prolongs further, as diagnosis is not only dependent on the recognition by the family, but also by the lack of awareness by the primary care medical per sonnel and poor referral networks 96  ~90% being Wilms tumours (WTs) and ~10% other primary renal tumours. In a paradigm of disrupted organ development eventually leading to tumorigenesis, remains of the multipotent nephrogenic zone of the fetal kidney may persist after birth and appear in up to 1% of routine infant autopsies as nephrogenic rests. The natural history and fate of nephrogenic rests is, however, uncertain. These cells may terminate their differentiation, or eventually regress and become sclerotic and obsolescent, whereas others can progress to form hyperplastic nephrogenic rests, with typical genetic changes. Nephrogenic rests are found in >90% of patients with bilateral WTs and in ~30-40% of patients with unilateral sporadic WT. WTs are then characterized by the acquisition of additional genetic and epigenetic changes, some of them being quite specific for histological subtypes. The percentages indicate the frequency of mutation in sporadic cases. It is unclear whether WTs originate directly from nephrogenic blastema without progression through nephrogenic rest stages. CCSK, clear cell sarcoma of the kidney; CMN, congenital mesoblastic nephroma; LOH, loss of heterozygosity; LOI, loss of imprinting; miRNA, microRNA; RCC, renal cell carcinoma; RTK, rhabdoid tumour of the kidney. a IGF2-H19 LOH/LOI have not been shown in epithelial WTs. metastases in low income regions than in high income regions 14,24 .

Diagnosis, classification and staging
Diagnosis of WT can be made reliably on histol ogy, especially WTs in which all three characteristic components -blastemal, epithelial and stromal -are evident. These components may be mixed in any pro portion, but WTs showing one or two components are not rare. Epithelial and stromal components may show different lines of differentiation and degrees of differen tiation, resulting in a countless number of histological appearances (Fig. 6). WTs composed of only one compo nent may represent a diagnostic challenge and ancillary techniques may be needed to establish the diagnosis 97 . However, no immuno histochemical markers or mole cular biology findings are 100% specific for WT. In addi tion, preoperative chemotherapy, when used, alters the histological appearance of WT, and may result in marked tumour necrosis, or maturation of tumour components. Approximately 7-8% of WTs demonstrate anaplasia, defined as the presence of cells with hyperchromatic, pleomorphic nuclei that are three times larger than adja cent cells and have atypical mitotic figures 98 , and it may occur in any tumour component (blastemal, epithelial or stromal). The definition of anaplasia was further refined to specify whether the anaplasia is diffuse or focal based on the anatomical distribution of anaplastic cells within the tumour 99 . The diagnosis of focal anaplasia is based on the presence of clearly defined one or two foci showing the above mentioned nuclear criteria with sharp demar cation within the primary intrarenal tumour and with out evidence of anaplasia or prominent nuclear atypia (defined as nuclear unrest) in other areas. According to SIOP, up to two foci up to 15 mm in size is allowed for the diagnosis of focal anaplasia 9 whereas according to COG, up to four foci up to 20 mm in size is allowed 99 . Diffuse anaplasia is defined as non localized anaplasia, which may present in any of these situations: focal anaplasia with marked nuclear unrest in the non anaplastic tumour; anaplasia beyond the tumour capsule; anaplastic cells in intrarenal or extrarenal vessels, renal sinus or extra capsular sites, in metastases, or in biopsy. Despite well established criteria, anaplasia represents a diagnostic problem, with ~30-50% discrepancy between institutional pathologists and central pathology review 100,101 . Anaplasia is very rare in the first 2 years of life, and increases after 4 years of age. Anaplasia is usually neither obliterated nor induced by preoperative chemotherapy.
As SIOP and COG have different treatment initial strategies, relevant differences exist in histological clas sifications of WTs between the two groups. The COG classification includes anaplastic (focal and diffuse) and non anaplastic (favourable histology) WTs based on assessment of a chemonaive tumour after up front sur gery. The SIOP classification is based on the assessment of the percentage of preoperative chemotherapy induced changes and viable tumour components, and includes three major WT risk groups: low risk tumours (com pletely necrotic WT), high risk tumours (blastemal type and diffuse anaplasia), and intermediate risk tumours (all other types) (TAble 3). To correctly subclassify the WT, the percentages of chemotherapy induced changes and viable tumour components are assessed and taken into account 9 . COG has reported histology and out comes in patients not eligible for up front surgery using the SIOP post chemotherapy histological classification system but to date has not used this system to guide sub sequent treatment in patients with a unilateral tumour 102 .
The staging criteria between COG and SIOP also differ, making a direct comparison of outcomes stage by stage difficult (Supplementary Table 1).

Diagnostic imaging.
Abdominal ultrasonography is efficient and globally the most available means of inves tigating suspected WT 103 . Ultrasonography provides information about the organ of origin, extension into the renal and inferior cava veins or urinary collecting system, the contralateral kidney and associated urogen ital abnormalities, and may identify liver or lymph node metastases. In resource limited regions, ultrasonography is sufficient for abdominal staging and can be comple mented by chest plain radiography, recognizing that plain radiography may miss smaller pulmonary lesions (typically <1 cm) 95,104 . In better resourced settings, cross sectional imaging is usually undertaken preoperatively with abdominal CT or MRI 105 . The main drawback of CT is radiation exposure, but the procedure is rapid, allows continuous imaging of the chest and abdomen, has moderate specificity for detection of preoperative spill, may help distinguish nephrogenic rests from WT and gives excellent pulmonary detail [106][107][108] . It is noteworthy that COG and SIOP incorporate centrally reviewed CT identification and response to therapy of lung nodules into current risk stratification treatment algorithms 10,109 .
The main hurdle of abdominal MRI is that moderate to deep sedation is often required in young children but it provides excellent organ details in those with bilateral disease or liver metastases. Abdominal MRI is preferen tially recommended for better assessment of potential nephrogenic rests and their distinction from true WT, and by SIOP to attempt to correlate apparent diffusion coefficient mapping with histopathology prediction after preoperative chemotherapy 105,110 . Fluorodeoxyglucose (FDG) PET is not routinely used for imaging WT 105 . Bone scan or cross sectional imaging of other sites is reserved for patients with signs or symp toms suspicious for distant extrapulmonary metastases. Non pulmonary and non hepatic metastatic disease is very rare at primary diagnosis of non anaplastic WT and is more likely observed in anaplastic WT, clear cell sarcoma of the kidney, malignant rhabdoid tumour and renal cell carcinoma, or at WT relapse [111][112][113] .
Laboratory testing. Baseline blood work should be per formed to confirm adequate renal function, support subsequent chemotherapy and rule out acquired von Willebrand's disorder, which although uncommon may be associated with substantial bleeding risks and can be pre emptively managed 114 .
SIOP diagnostic algorithms recommend percutane ous image guided coaxial core needle biopsy through a retroperitoneal approach in patients 7 years of age or older or in patients with imaging findings unusual for WT (psoas muscle infiltration, numerous calci fications, vessel encasement or massive lymphade nopathy) 10,115,116 . The currently used cut off of 7 years to consider a biopsy is under revision, and based on epidemiological data showing the peak incidence of WT versus other non WTs 2 , a new consensus towards raising the age threshold for biopsy providing there are no other atypical presenting features is forming 105,115 . COG recommends that primary nephrectomy should be strongly considered in all patients, but if not feasible, open or Tru Cut needle biopsy should be undertaken with a minimum of 10-12 cores. Notably, needle biopsy cannot reliably distinguish WT from nephrogenic rests, and often misses anaplasia 115 .
Patients with syndromic features should be referred to medical genetics for counselling and possible testing. Circulating blood or urine tumour DNA is being explored for diagnostic and response or relapse assessment but is not yet standard of care 80,117,118 .

Prognosis and prognostic features
It is important to recognize that prognostic markers must be interpreted in the context of the accompany ing treatment regimen. This principle is relevant to WT as COG advocates for immediate nephrectomy in most patients, whereas SIOP advocates for preoperative chemotherapy 119 . Thereafter, prognostic factors used for clinical treatment stratification differ between COG and SIOP 120,121 . According to both groups, tumour histology and stage are key prognostic indicators, although applied differently and together with other factors in clinical practice. Diffuse anaplasia is regarded as a high risk tumour by COG and SIOP, whereas focal anaplasia is regarded as an intermediate risk tumour by SIOP but as a high risk tumour by COG. A blastemal type WT after preoperative chemotherapy is also regarded by SIOP as a high risk tumour and a completely necrotic type WT as a low risk tumour 122 (TAble 3). Similarly, staging cri teria are also different; for example, according to COG, any tumour biopsy results in upstaging to local stage III, whereas according to SIOP, fine needle aspiration and percutaneous core needle biopsy are ignored for stag ing purposes 31 , and according to SIOP, the presence of necrotic tumour or chemotherapy induced changes in

Box 1 | Wilms tumour predisposition and driver genes
Most genes implicated in Wilms tumorigenesis act in gene expression control and growth factor signalling. Approximately 50% of the genes can be present in mutant form in germline or constitutional DNA conferring increased Wilms tumour (WT) risk 51 . Other prognostic factors according to SIOP include tumour histological response to preoperative chemo therapy and tumour volume (>500 ml) after chemotherapy for certain WT types. Additional prognostic factors according to COG include age, tumour weight and biomarkers or tumour biology, that is, loss of heterozy gosity at chromosomes 1p/16q, loss of heterozygosity at chromosome 11p15, and gain at chromosome 1q 121 . For both groups, response of lung metastases to neoadjuvant chemotherapy indicates chemosensitivity and dictates the intensity of subsequent treatment; for example, if lung lesions are not present at 6 weeks after induction chemotherapy, radiotherapy can be omitted in some patients 109,123 .

Kidney development
Although the SIOP and COG strategies differ in their upfront treatment approach, they lead to similar overall survival rates of ~90% 7,39,124 . Patients with stage IV ana plastic WT and/or blastemal type WT have substantially poorer outcomes, with an overall survival rate of <50% despite very intensive therapy 125,126 .
Despite the good prognosis in most children with WT, ~20% of patients will relapse, predominantly within 2 years of diagnosis 113,127,128 . The overall survival rate after relapse is ~50% but varies considerably according to the initial treatment received (which in turn reflects initial tumour stage and histology), time to relapse, site of relapse, and patient age 113,129,130 . Surveillance with abdominal ultrasonography and chest plain radio graphy are offered, and patients with asymptomatic relapse detected by surveillance seem to have better outcomes 113 . Evidence from COG shows a lack of benefit in terms of longer survival after relapse if CT imaging had been used instead of plain radiography and ultrasonography in follow up surveillance 128 . SIOP data also suggest that surveillance beyond 2 years after completion of therapy has a low yield because of the extremely low relapse rate thereafter 113 .

Screening
Genetic testing in children with cancer but also in other (potentially) unhealthy children presenting with certain abnormalities or syndromes is emerging. This testing includes formalized national and regional whole exome or genome sequencing programmes to detect cancer pre disposition in many high income regions. Accordingly, both novel genes and syndromes associated with WT are revealed and additional children with an increased risk of developing WT are identified, expanding the cri teria for screening programmes 131 . Regular screening for early diagnosis in children with a known WT predis position syndrome has been reported to detect smaller and lower stage tumours but robust evidence is lacking regarding the balanced clinical benefits 93 . In addition, the benefits should outweigh the costs and burden. The latter is reflected in the different thresholds for perform ing screening, which typically varies between 1% and 5% for the childhood risk of developing WT 29 .
Screening is typically offered to children with various cancer predisposition syndromes, such as WT1 related syndromes and Beckwith-Wiedemann syndrome or isolated hemihypertrophy (with at least one Beckwith-Wiedemann syndrome feature). Renal ultrasonography is the recommended screening modality, as it avoids radiation and does not require anaesthesia in young children. The screening interval is every 3 months based on the rather rapid growth rate of the tumour and imag ing should be performed by an experienced paediatric ultrasonographer 93 . Screening should start when the WT predisposition is established and should continue, irrespective of the underlying condition, until the child is approximately 7 years old. At this age, the risk of WT development is greatly reduced 29 .

Challenges
• Highly constrained health-care budgets resulting in insufficient paediatric oncologists, surgeons, anaesthetists and pathologists; shortage of chemotherapeutic agents (which leads to incomplete Wilms tumour (WT) treatment); limited or lacking infrastructure and facilities for imaging and radiation therapy. • Lack of high-quality specialized paediatric surgical training to perform complex operations (WT with intracaval extension, nephron-sparing surgery). • Inadequate reporting or data collection within national or hospital registries precludes accurate outcomes assessment. • Inadequate specialist cancer services.
• Late clinical presentation (delay in diagnosis) owing to family's or relatives' reduced awareness about cancer; contacting and arrival to primary care; health-care staff recognition of cancer (a much higher number of children in low-income countries have a distended abdomen than in high-income countries owing to many other non-malignant conditions; thus it is challenging to differentiate and prioritize investigations for the relatively few children with WT). • Many patients are diagnosed with already advanced or metastatic tumours. • Toxicity from surgery and/or chemotherapy can increase mortality and contribute to treatment abandonment. • Malnutrition is a major concern for higher drug toxicity and treatment-related death.
• Patient quality of life largely unrecognized and unprioritized.

Priorities and areas for improvement
• Comprehensive registries are the first step to appropriate resource allocation according to local needs and to the monitoring of improvement. • Earlier diagnosis through increased education among primary health providers concerning WT diagnosis, and parent education on healthy living and concerning symptoms. • Adapted treatment regimens to accommodate frail children, to reduce toxicity, and to face specific (temporary or permanent) drug regimen shortages. • Nutritional programmes, best with locally available calorie-dense foods and fortifiers.
• Implementation of family education programmes may increase compliance with cancer care reducing abandonment. • Twinning programmes (pairing of hospitals in resource-limited countries with hospitals in developed countries) to improve local medical expertise and education. • Clinical trials answering locally relevant questions (such as prognostic factors).
• Prioritizing resources to focus on curable clinical situations.
• Palliative care as the main priority for advanced malignancies.
The purpose of WT screening is to enable early NSS, to give less intensive (that is, less toxic) chemotherapy, and to avoid radiotherapy. Patients with predisposi tion syndromes may develop metachronous WT in the contralateral kidney. Hence, the aim is, on balance, to preserve maximal kidney function and ultimately avoid end stage renal disease whilst still maintaining onco logical control. Genetic testing, screening and NSS in LICs are rarely available and consequently, more chil dren progress and succumb to end stage renal diseases in low income settings with limited options for dialy sis and/or renal transplantation than in high income regions 132 . High income regions are researching the potential for (epi)mutation detection in circulating tumour DNA for early diagnosis; however, this detec tion technique is not yet ready to be used as an alter native to surveillance with ultrasonography in clinical practice 117,118,133 .

Management
Nephrectomy with adequate lymph node sampling is universally the mainstay of treatment for WT. However, the timing of surgery differs between the SIOP and COG recommendations, and underpins the differences in risk stratification 134,135 . The SIOP WT studies have centred around pre nephrectomy therapy since their beginning in 1971. Neoadjuvant chemotherapy allows assessment of in vivo histological response to treatment (that is, completely necrotic tumour indicates high responsive ness whilst a predominance of remaining blastemal cells is a marker of chemotherapy resistance), which may be used to guide therapeutic stratification after nephrec tomy. According to SIOP protocols, patients are divided into low risk, intermediate risk and high risk groups mainly on the basis of the degree of tumour necrosis and the relative proportion of each of the three cell types (epithelial, stromal  viable component of the resected tumour. On the other hand, the COG approach of upfront nephrectomy allows immediate histological diagnosis, molecular analy sis of tumour samples unaltered by chemotherapy, and drug naive local staging assessment (such as the pres ence of tumour spill or lymph node involvement). This knowledge can identify a subset of children with very low risk tumours who may be treated with nephrec tomy alone 136 . Each approach has its pros and cons, yet survival rates are similar with an overall survival rate of >90%. According to both groups, the management of WT incorporates risk based adjuvant chemother apy and radiotherapy informed by multiple prognostic factors 6 (Supplementary Table 2).

COG perspective
COG has a recommended strategy of primary nephrec tomy for unilateral renal masses in patients without WT predisposition (achievable in >90% of patients) or failing feasibility of nephrectomy, core needle or open biopsy to guide subsequent therapy 135 . Exceptions to upfront biopsy is patients with bilateral disease or patients with a bilaterally predisposed syndrome who should receive neoadjuvant chemotherapy (without biopsy) with the aim of preserving renal units, with surgery at 6-12 weeks after initiation of chemotherapy 137,138 . The primary surgery using a transabdominal or thoraco abdominal approach allows accurate prechemotherapy staging including assessment of chemotherapy naive histology and prognostic molecular testing. Essential surgical tasks in completing a tumour related nephrec tomy include avoidance of tumour spill, ipsilateral hilar and regional lymph node sampling, and assessment and control of extrarenal tumour extension including the renal vein and ureter [139][140][141][142] . Less conventional approaches such as laparoscopic nephrectomy, partial nephrectomy and split renal techniques may be carefully considered in patients with selected small tumours and in expert hands but at this point is confined to a small number of patients [143][144][145][146] .
Chemotherapy is a mainstay of adjuvant therapy except in very low risk tumours (defined as stage I, favourable histology WT, <550 g with negative lymph nodes and no syndromic features) where observation alone following nephrectomy may be sufficient, espe cially in the absence of loss of heterozygosity at chromo some 11p15 (reF. 147 ). Based on COG staging, the bulk of patients with favourable histology WT without certain adverse biomarkers receive regimen EE4A (vincristine and actinomycin D for 18 weeks) for stage I and II, or regimen DD4A (vincristine, doxorubicin and actinomy cin D for 24 weeks) for stage III and stage IV favourable histology WTs 102,109,[148][149][150] . COG recommends the use of CT imaging to identify lung metastases, although it is recognized that up to one third of lesions <1 cm in dia meter may be benign nodules. Biopsy of lung nodules is encouraged if there is any doubt about the nature of the lesion. In addition, round, non calcified lung nodules not in a fissure visible on chest CT are considered stage IV, regardless of size, unless histologically proven not to be WT 102 . The COG study AREN0533 demonstrated that ~40% of patients have complete resolution of pulmo nary metastases after 6 weeks of three drug induction therapy (regimen DD4A) and of these patients, those with tumours without 1q gain can safely have radiation omitted 109 . Patients with an incomplete response of lung nodules after 6 weeks of DD4A chemotherapy receive whole lung irradiation and are escalated to chemotherapy regimen M.
In the setting of loss of heterozygosity at 1p/16q, evi dence shows that intensifying therapy to regimen DD4A for stage I and II or to regimen M (DD4A + cyclophos phamide or etoposide) for stage III and IV improves event free survival outcome 39,151 . Patients with diffuse anaplastic tumours seem to benefit from a multiagent regimen UH2 (reF. 125 ). This regimen is associated with considerable toxicity and further modifications are currently being tested in COG protocol AREN1921 (NCT04322318) (Supplementary Table 3). A variety of strategies for salvage treatment in patients with relapse are used based on risk. Patients with low risk relapse are usually managed with stratum B with an expected out come of ~71% event free survival rate 152 and patients with higher risk relapse are typically managed with regimen C 153 or ICE (ifosfamide, carboplatin and etoposide) with an expected outcome of ~42% event free survival rate. Some centres use autologous bone marrow transplanta tion as consolidation therapy in patients with a high risk tumour but this strategy has never been the subject of a randomized trial to confirm efficacy 154 . A detailed sum mary of the impacts of the first generation of COG stud ies on WT was published in 2021 and forms the basis of standard management approaches in the COG for those patients not participating in a research study 155 .
Newer COG research protocols are testing further refined chemotherapy algorithms incorporating stage, lymph node status, additional somatic molecular bio markers, cardioprotection with dexrazoxane and new agents 40, 102,125,151 . WT is highly radiosensitive; 156 radia tion therapy is utilized for the regional management of stage III or IV favourable histology WT, and relapsed and anaplastic WT. COG protocols are incorporating intensity modulated radiation therapy 157 with doses ranging from 10.6 to 30.6 Gy depending upon resid ual tumour and site 156,158 . National Cancer Cooperative Network 159 guidelines provide further detailed manage ment guidelines and recommend that all children with renal tumours participate in a clinical trial.

SIOP perspective
According to the SIOP strategy, all patients >6 months of age with suspected WT receive either 4 weeks of preoper ative chemotherapy with actinomycin D and vincristine (if localized) or 6 weeks of actinomycin D, vincris tine and doxorubicin (if metastatic). The SIOP9 trial showed no advantage on downstaging to more stage I tumours nor on reducing intraoperative tumour rupture by further prolonging the prenephrectomy regimen to 8 weeks 160 . The SIOP approach accounts for the risk of misdiagnosis of WT by recommending upfront nephrec tomy in infants <6 months of age, and percutaneous core needle biopsy in older children (7 years of age and older) or children with uncertain clinical pictures 115 . Radical nephrectomy is regarded as standard in most patients with unilateral WT; however, the systematic use of preoperative chemotherapy may extend NSS opportu nities in selected patients with unilateral non syndromic tumours 134 to maximize preservation of renal function in patients. Following surgery, patients are stratified into three risk groups according to the histopatholog ical features of the tumour 122 (TAble 3); the histological risk group together with tumour stage is used to direct the intensity of adjuvant chemotherapy and the need for radiotherapy ( Supplementary Tables 2 and 4).
The regimen used in the experimental arm of the SIOP2001 trial has been adopted as the new standard regimen for most patients with stage II−III intermediate risk histology WT 7,10 . This regimen consists of 27 weeks of postoperative treatment with vincristine and actino mycin D without doxorubicin. This schedule resulted in a non significant decrease in event free survival and had no effect on overall survival when randomly compared with the historical standard arm of 27 weeks of these two drugs (vincristine and actinomycin D) plus five doses of doxorubicin at 50 mg/m 2 (reF. 7 ). The use of doxorubicin in patients with intermediate risk stage II-III tumours is currently recommended only in those with non stromal or non epithelial large volume tumours (that is, ≥500 ml after prenephrectomy chemotherapy), based on a post hoc analysis of patients in the SIOP 2001 trial 10 .
Radiotherapy to the flank is administered to patients with stage II WT with diffuse anaplasia or stage III WT (intermediate risk and all high risk), and doses range from 14.4 to 25.2 Gy (±10.8 Gy boost only for macro scopic residual disease) 10,161 . To decrease organ toxicity whilst preserving oncological outcome, the conventional approach of flank irradiation is currently being adapted into a guideline for highly conformal image guided flank target volume delineation 162 .
For metastatic disease, CT only nodules are treated as metastases in the current SIOP protocol if they have a transverse diameter of ≥3 mm and imaging appearance suspicious for metastatic nodules after centralized radio logical review 10 . Following a standard 6 week three drug preoperative regimen, 61-67% of patients have complete metastatic response before nephrectomy 123 . Afterward, current SIOP guideline advise stratifying patients to adjuvant regimens consisting of either vincristine plus increasing cumulative doses of doxorubicin (in the range 150-250 mg/m 2 ) or a four drug regimen includ ing etoposide, carboplatin, cyclophosphamide and dox orubicin (cumulative dose 300 mg/m 2 ). In patients with remaining lung nodules, metastasectomy and histologi cal confirmation of metastasis is advised. Stratification is based on the local stage of the primary tumour, histo logy of the primary tumour and the metastatic tumour (if resected), the size of metastases, and their response to preoperative chemotherapy and surgery ( Supplementary  Tables 2 and 4).
Pulmonary radiotherapy is administered for lung metastases lacking complete response by postoperative week 10. Evidence suggests that the majority of patients achieving a complete response after induction chemo therapy with or without surgery have a satisfactory sur vival probability even without radiotherapy to the lungs (5year event free survival 84%, 5 year overall survival 92%) 123 . Patients with viable metastases at surgery or high risk histology of the primary tumour receive radiotherapy to the lungs.
Patients with metastatic and high risk disease are a rare subgroup with a dismal prognosis, justifying test ing of novel and more intensive regimens in first line therapy 126 . Combinations of vincristine, irinotecan, cyclophosphamide, carboplatin, etoposide and doxoru bicin, followed by high dose melphalan and autologous stem cell rescue, are currently being explored by the SIOP RTSG 10,126,130 .
For bilateral tumours, SIOP guidelines aim to limit preoperative chemotherapy to a maximum of 12 weeks, with time intervals for evaluation to 6 weeks, also compar able with the COG approach. In order to maximize the possibility of bilateral NSS, an approach using carboplatinetoposide in patients with an unsatisfactory response to vincristine-actinomycin D is under evaluation 10 . Adjuvant postoperative treatment guidelines generally follow the same principles as for unilateral WT.

Low-income and middle-income regions
Survival in low income and middle income countries (LMICs) is much lower than in high income regions, with overall survival rates ranging from 11% in Sudan to 46% in Malawi [163][164][165][166] . In LMICs factors that reduce the likelihood of a good outcome include delay in diag nosis leading to advanced disease at diagnosis 94,167 , lack of diagnostic services, insufficient trained personnel, chemotherapy and radiation 14,95,168,169 , misdiagnosis 167 and abandonment of therapy 14,169 (bOx 2). Mortality is higher than in high income regions owing to toxic ity from surgery and/or chemotherapy, coupled with malnutrition 167,[170][171][172] . Addressing these psychosocial issues and malnutrition (chronic and acute) may sig nificantly improve outcomes with time 95,170,173 . As expe rience is gained in these countries, there is a need for support for the development of local priorities and for improvement in the availability of curative therapies and palliative care.
Trained physicians, nurses and ancillary personnel are central in providing quality care. The WHO Global Initiative for Childhood Cancer developed the frame work of care CureAll to provide early diagnosis networks for referral to centres of excellence, to introduce child hood cancer to the Universal Health Coverage schemas and Cancer Control Plans, to introduce cancer and sup portive care regimens of care, and to evaluate and mon itor schemas to measure progress. These strategic plans are coupled with enabling actions -advocacy, leveraged financing and linked governance 174 .
The SIOP approach with prenephrectomy chemo therapy provides the optimum and safest strategy in resource limited settings. For large abdominal tumours (>500 ml), upfront surgery is associated with a high risk of surgical complications, tumour rupture and infec tion 175 . Patients with severe malnutrition may have decreased clearance of chemotherapy and dose adjust ment may be necessary with parallel monitoring of liver function and recovery of myelosuppression 172,176 . In a preoperative chemotherapy scenario, close attention to the interpretation of pathology according to the SIOP risk classification is key to correct selection of post operative treatment intensity, but requires specifically trained pathologists 9,10 .
In LMICs, almost no clinical trials are available, with limited data and information on outcomes 177 . Encouraging prospective registration studies and partic ipation in clinical trials has the benefit of building expert clinical capacity, improving facilities and funding treat ment and associated costs with the effect of improved survival 177 , all according to the local sustainability and capacity building 178 .
As an example, the Collaborative WT Africa Project is a multinational prospective clinical study open in seven sub Saharan countries, which have registered prospective outcomes by implementing the SIOP adapted treatment regimen for WT 164,179,180 . A mini mum requirement of ultrasonography of the abdomen was used for diagnosis. The guidelines recommended preoperative chemotherapy followed by surgery and fur ther chemotherapy, coupled with nutritional rescue. The preoperative treatment included either a 4 week two drug (vincristine and actinomycin D) or a 6 week three drug (vincristine, actinomycin D and doxoru bicin) regimen depending on the presence of local or metastatic disease, respectively 104 . Prolongation of pre operative chemotherapy was an option in patients with a large tumour volume. Patients weighing <12 kg or with severe acute malnutrition were given two thirds of the calculated dose of the chemotherapy 164 . The goal was then to achieve safe nephrectomy with lymph node sampling in patients with improved clinical and nutritional conditions, and tumour shrinking, which are all related to a reduced incidence of intraoperative morbidity. The postoperative chemotherapy aimed to follow the standard SIOP guidelines, but with spacing the administration of vincristine every 3 weeks at a dose of 2 mg/m 2 as also used for children >1 year of age in specific phases of some COG regimens (capped at 2 mg absolute dose), with the purpose of reducing the burden of frequent travel to hospital 181 . Acknowledging deficits in radiotherapy provision (lacking across most of the African network), radiation therapy was used only in Kumasi and Accra, Ghana, for metastatic disease and for stage III abdominal tumour 179 .
The lessons learnt from this structured guideline was the need for team members to work according to a shared vision, mission and principles 178 . The impor tance of using local site leaders to set the priorities for a successful clinical trial and keeping processes as simple as possible for data completeness was also appreciated. Good communication, transparency and trust was found to be the cornerstone for successful local implementation of a multinational clinical trial in LMICs.
Just as clinical investigation is the cornerstone for best therapies and practices in high income regions, research is needed to address the best therapy in different settings (better if tuned to prognostic indicators that have been studied and validated in the local context), best prac tices, and quality data to further improve outcomes 164,165 . Multidisciplinary care meetings with mentors improve management and experience of the local and regional groups.

Long-term complications
Despite the greatly improved therapy for WT over time, survivors still report a high frequency (25%) of severe chronic health conditions in adult life 15,16,182,183 . Patients with WT have a higher risk of death than the general population. In an analysis of children enrolled in the NWTS group between 1969 and 1995, the standardized mortality ratio was 24.3 during the first 5 years after diag nosis, but remained increased for >20 years after diagnosis (standardized mortality ratio 4.3) 184 . Although the primary tumour remained the most frequent cause of death >5 years after diagnosis, secondary malignant tumours, cardiac disease and end stage renal disease were also major causes of mortality.
The hazard ratios for hypertension (8.2), conges tive heart failure (23.6) and renal failure (50.7) are all increased among 5 year survivors of WT compared with sibling groups 15,185,186 . The risk of congestive heart fail ure increases with the cumulative dose of doxorubicin administered, with a critical threshold of 240 mg/m 2 . Cardiotoxicity is potentiated by the concurrent use of radiotherapy, with girls and infants more susceptible than boys 15,186 . Similarly, doxorubicin seems to potenti ate the adverse effects related to radiotherapy, probably owing to its radiosensitization of cells. These adverse effects include abnormal tissue growth within the target area and secondary malignancies.
The 20year cumulative incidence of end stage renal disease is reported to be <1% in patients with unilateral WT and ~10% in patients with bilateral disease [187][188][189][190] . The risk factors associated with end stage renal disease owing to chronic renal failure are stromal predominant histol ogy (HR 6.4), intralobar nephrogenic rests (HR 5.9), and age at diagnosis of <24 months or <48 months (HR 1.7 and 2.8, respectively) 191 . Given the increased risk of cardiovascular morbidity with chronic kidney disease, identifying patients with a high risk of progressive renal impairment early is imperative to preserve the quality of life of long term survivors. The wider availability and accuracy of patient genotyping may identify more molecular fingerprints with implications for renal func tion into adulthood, in order to select a subset of patients with WT without clinical renal impairment at pres entation, yet who might benefit from nephron sparing surgical procedures.
Gonadal dysfunction. Gonadal dysfunction is observed in female WT survivors 192,193 , and is strongly associated with exposure of the ovaries to radiotherapy (at any dose) and treatment with alkylating drugs 192,194 . The first line chemotherapy with two drugs used (that is, vincristine and actinomycin D), in general, does not affect either ovarian reserve or male fertility. Whole abdomen radiation usually results in primary ovarian failure or premature menopause. Additionally, exposure of patients with WT to treatment including anthracy clines and lung radiation is associated with cardiovas cular risks that can affect pregnancy outcomes 192 . WT survivors should receive personalized counselling about the type and magnitude of reproductive health risks on the basis of their specific treatment exposure, with older girls with unfavourable histology or high risk WT being at increased risk. Patients at the highest risk should be offered fertility preservation whenever possible, and after accurate counselling 195 . In this view, prior abdom inopelvic surgery (done to perform nephrectomy as part of initial cancer treatment) should not be regarded as a barrier to laparoscopic oophorectomy with tissue cryopreservation for fertility preservation.

Quality of life
Parents, charities and survivors of WT have worked closely with researchers and scientists to ensure that research is focused on what is important to families, and to highlight areas of need (bOx 3). The Wilms Tumour Link Group is an example of a parent led research group in the UK focused on identifying priorities for future research, and uses a social media group of >600 mem bers from around the world to communicate research updates. Parents of children with WT have participated and presented at international scientific meetings and are considered to be partners in the research process, amplifying the patient voice within childhood cancer research 90 .
This level of parent and patient involvement in research provides an important opportunity for physi cians to work collaboratively. This collaboration facil itates the chance to have a greater impact on what is researched and highlights that not only is cure impor tant but also the child's quality of life and happiness and health after cancer. The active involvement of parents and survivors in research helps translate findings in an equitable and accessible way. Findings are all too often kept within scientific journals that do not allow access to non academics, so those that are affected by the disease are less informed. Working collaboratively and honestly with families is the key to patient driven research with real life translatable outcomes.
As a result of their better quality of life and physi cal functioning, children surviving renal tumours can hopefully also enjoy an increased involvement in sports. Patient counselling should include explaining any poten tial contraindication for practising sports carrying a risk of abdominal injuries 196 . Practice guidelines, where avail able, addressing the participation of children and adoles cents with a solitary kidney (like most survivors of WT) in high impact sports do not share a common vision worldwide 197 . In Italy, unlike the rest of Europe and the USA 198 , having only one kidney automatically disqual ifies an individual wishing to participate in any organ ized competitive contact sport, including basketball and soccer, and sometimes, volleyball (bOx 4).
To instigate positive changes in cancer care through exercise, and to endorse change in patients sensitively, patients, families and health care teams must be made more aware of current evidence based information to provide a framework for the harmonization of guide lines for sport participation of renal tumour survivors, to ensure that they can exercise freely yet safely.

Basic research
Increased understanding of the aberrant molecular pathways active in Wilms tumorigenesis has identified many potential targeted therapeutic approaches that could be applied in a clinical setting 36 . These include miRNA therapeutic modulation 43,44 , Wnt signalling 90 and p53 specific biological targeting agents in anaplastic WTs 69, 70 . In addition, retinoic acid, although ineffective as a WT therapy in the all trans form 199 , may impart a differentiation effect on precancerous nephrotic rests in the 13 cis form, potentially mitigating the development of WT in a selected group of patients at risk, particu larly those with hyperplastic nephroblastomatosis 200,201 . Although the options seem intriguing, the main chal lenges are the relatively few patients in each molecu lar subgroup, WT intratumoural heterogeneity 78 , few actionable known targets, selection and conduct of Oophorectomy A surgical procedure to remove one or both ovaries.

Box 3 | Patient experience
The statement provided has not been edited and the patient's emphases remain in place.
Teenage years, the best ones in everyone's life. I was living unforgettable moments, going out and having parties with my friends. And then, after some medical checks, hell overnight. I had cancer. At first, I started imagining what I would have had to go through, how much I would have suffered. I was lost in doubts, fear and contrasting feelings.
To start chemotherapy shocked me. Eight hours, each impressed in my mind, in which milliards of medicine's drops came into my body. I felt exhausted.
As I was left alone for a moment in that hospital room, I abandoned myself to tears at the idea of repeating all of that the next day, and for eight more courses: that thought killed me. I came in that realization in that right moment.
I remember how important it was, for me, to have my friends around and to spend as much time as possible together. I remember they were the only ones who made me laugh, who made me feel normal, like nothing had ever changed. They made me breath, giving me the oxygen I needed. They reminded me how strong I was, when I was totally worn out. They recalled me what it meant to live, as sometimes I forgot how to do it.
Then, the COVID-19 pandemic situation came, and loneliness. My mum and I, stop. Far away from everyone, from everything. Three months of physical pain for chemo that I kept doing, of discouragement and fear. And, if it wasn't enough, there was also the worry of catching the virus.
Finally, after never-ending months, I came back to my lovely Naples. To my friends, to my family. To the sea, as I saw it, I felt free. Everything finally came to an end, and I couldn't believe it.
targeted trials and coordination of timely enrolment in the background of competitive trials 90,202 .

Translational and clinical research
Some novel targets for WT have emerged, mostly based on PDX dependent drug screens, leading to a few phase I and II trials in WT. Examples include phase II study of IGF1 based inhibition (cixutumumab), anti VEGF based therapy (sorafenib, cabozantinib), aurora A kinase inhibition (alisertib), and anti mitotic based therapies either through direct microtubule inhibitory activity (ixabepilone) or via an antibodydrug conjugate linking an antimitotic agent (DM1) to an anti CD56 antibody (lorvotuzumab) (lorvotu zumab mertansine, IMGN901) 202 . With the advance of β catenin targeting, the COG will soon launch the study of tegavivint, a specific β catenin inhibitor, to include a WT cohort 203 .
Advances in refined personalized multilayered bio logically derived WT treatment will emerge, in the shorter term, via expanded creation and use of tumour models, ideally sufficient in number to represent the majority of WT biological subtypes 86,87 . Development of organoids, spheroids and PDX models has progressed from basic investigation to real time patient specific drug screening, and application of these approaches is now feasible, with plans evolving to launch an inter national patient individualized relapse WT protocol harnessing this opportunity 90,204 .
Novel imaging investigations also hold promise for advancing WT treatment. For example, diffusion weighted imaging MRI has been implemented as stand ard for diagnostic and post chemotherapy assessment 105 . Such techniques may non invasively quantify tumours and risk stratify patients with WT prior to surgery, with radiological surrogates (apparent diffusion coeffi cient mapping) for both necrosis (particularly relevant when tumour size remains stable) and blastemal type histology 205 . Radiogenomics holds promise to further expand the utility of imaging in the care of patients with WT, as do artificial learning algorithms, for example, for the detection and quantification of lung nodules (and disease elsewhere), which dictates intensity of treatment. The import of such technology is magnified by the shift from chest plain radiography to adequate CT scans, which provide detailed information on lung lesions, but also reveal that there is considerable inter rater and intra rater variation in interpreting such lesions even by experienced radiologists 206 .
WT may be amenable to advances in liquid biopsy techniques for diagnostics, monitoring of therapy, and detection of minimal residual disease 80,118,207 . As an exam ple, COG study AREN1921 (NCT04322318) is collecting serial blood and urine samples in newly diagnosed ana plastic WT and patients with relapsed favourable histology WT to test the potential benefits of liquid biopsy 208 .

Cooperative group efforts
Both the COG and the SIOP groups have advanced well organized prospective clinical trials and studies that tightly integrate biological aims and clinical insights, both linked to specific clinical protocols (AREN1921; NCT04322318) or via overarching biobanking and risk stratification studies (AREN03B2;NCT00898365, SIOP UMBRELLA) 10 . Cross validation (meta analysis) of data between these groups, especially data from small cohorts of rare patients (such as those with ana plastic tumours, bilateral tumours or relapsing disease) and strategic efforts to synergize intervention trials or observational studies hold promise in continuing to advance diagnostics, risk stratification and therapeu tic options. Such 'harmonization' between cooperative groups has been formally advanced in the form of the Harmonica Initiative that integrates multidisciplinary dialogue, meetings, consensus building, specific research focus and overall strategies on a trans continental, inter cooperative group level 6,90,127,202 . Likewise, dialogue continues regarding potential trans Atlantic collabora tive trials among the Innovative Therapies for Children with Cancer (ITCC), the Pediatric Early Phase Clinical Trials Network (PEP CTN), the Pediatric Preclinical Testing Consortium, and parent representatives 90 .

Global efforts
Childhood cancer therapy in LMIC lags behind in diag nosis, therapy and survival, with minimal clinical or biological research. In 2018, WHO launched the Global Initiative for Childhood Cancer. In 2020, WHO pub lished the CureAll Framework technical package WHO Global Initiative for Childhood Cancer, designed to pro vide guidance to member states for the implementation of childhood cancer services in resource constrained settings 174 . Six tracer cancers, including WT, are targeted to provide guidance for diagnosis, therapy and support ive and survivorship care. With the help of international paediatric cancer partners (academic, regional and global societies and non governmental organizations), the aim is to establish the necessary training and design of basic, and translational and clinical research 174 . Hence, progress in WT survival rates is expected to become more visible in the current decade.
Published online xx xx xxxx Box 4 | International controversies in advice on sporting activities in people with single kidneys • Most children diagnosed with Wilms tumour (WT) become long-term survivors and living with a surgically solitary kidney. • Among injuries occurring during sport exposure, the incidence of injuries to the kidneys is very low (sporting kidney injuries are 0.07-0.5% of all sports-related injuries), less frequent than head injuries, and usually without serious sequelae 235,236 . • The recommendations for the participation of children and adolescents with a solitary kidney in contact or collision sports have changed over time. The last update from the American Academy of Paediatrics (2001) leans towards player participation without restriction in non-contact sports, and with individual assessment for limited-contact, contact, and collision sports to arrive at an unbiased judgement, which is not based only on the fact of having a solitary kidney 198 . • However, national advice towards permission to participate in high-impact sports varies between countries and over time 191 . • Flank protectors have not been rigorously evaluated and an international standard for the protection they may offer is not available 196 . • Individual counselling and decision-making between child, families and oncologist are recommended.