Oligodontia–Colorectal Cancer Syndrome (AXIN2)

Disorders associated with AXIN2*: apparently isolated oligodontia, oligodontia–colorectal cancer predisposition syndrome with attenuated adenomatous polyposis, ectodermal dysplasia, somatic mosaicism and the neurodevelopmental disorder caused by de novo variants in the tankyrase-binding domain; hereditary tooth agenesis as a class and adenomatous polyposes as the differential diagnosis.*

Literature and regulatory information updated: 28 October 2026.

0. Reference Summary Table

1. Introduction

1.1. Historical overview

The link between AXIN2 and colorectal cancer was first somatic. In 2000, Liu and colleagues, at the Mayo Clinic, described AXIN2 mutations in mismatch repair (MMR)-deficient colorectal carcinomas and showed that these alterations activated β-catenin/TCF (T-cell factor) signalling, that is, the same axis that loss of APC deregulates in familial adenomatous polyposis (FAP) [11]. The hereditary dimension emerged four years later. Lammi and colleagues studied a four-generation Finnish family, followed at the Institute of Dentistry of the University of Helsinki, in which severe oligodontia of the permanent teeth and colorectal neoplasia segregated with dominant inheritance; by positional cloning they identified the nonsense variant Arg656Ter (historical notation, = NM_004655.4:c.1966C>T p.(Arg656Ter)) in AXIN2 in all 11 affected individuals and found, in an unrelated boy with severe tooth agenesis, a de novo single-nucleotide insertion in the same segment of the gene [2]. The paper established two ideas that still guide clinical practice: Wnt signalling is required for the normal development of the human dentition, and tooth agenesis may signal a susceptibility to cancer [2].

The following years brought negative results that help to explain the rarity of the syndrome. Peterlongo and colleagues found no germline AXIN2 variants associated with nonsyndromic colorectal cancer [32], and Lejeune and colleagues, in a French series of 39 patients with multiple adenomas or colorectal cancer without an APC variant or MMR deficiency, identified no pathogenic or likely pathogenic AXIN2 variants, whereas MUTYH-associated polyposis (MAP) explained 7 of 22 cases (32%) with multiple adenomatous polyposis [33]. In 2011, Marvin and colleagues described the second family, in Michigan (USA), with the nonsense variant p.(Trp663Ter), oligodontia, colonic polyposis, gastric polyps, mild ectodermal dysplasia, colorectal cancer and early-onset breast cancer, and proposed the name AXIN2-associated autosomal dominant ectodermal dysplasia and neoplasia syndrome [3]. Rivera and colleagues, in Spain, reported in 2014 a family with attenuated adenomatous polyposis and an AXIN2 missense variant without oligodontia or ectodermal dysplasia [8].

From 2019 onwards, the inclusion of AXIN2 in multigene panels added families and cohorts. Beard and colleagues confirmed the phenotype in an Australian family with a frameshift variant [34]; Chan and colleagues described four Australian families with polyposis and colorectal cancer [35]; Jensen and colleagues reported a three-generation Danish family [36]; Leclerc and colleagues assembled the French series of the Groupe Génétique et Cancer, with 11 index cases and comparison with population controls [7]; and Roht and colleagues, within the European Reference Network on Genetic Tumour Risk Syndromes (ERN GENTURIS), gathered 13 carriers from six families and proposed cleft palate as a possible manifestation [15]. The systematic review by Ratanasereeprasert and colleagues (2025) pooled 91 patients and 34 variants and examined the correlations between the affected domain, the pattern of agenesis and cancer [25]. In 2026 two extensions of the spectrum were published: the first description of polyposis caused by somatic AXIN2 mosaicism [9] and a multisystem neurodevelopmental disorder caused by de novo missense variants in the tankyrase-binding domain [10].

1.2. Nomenclature, classifications and reclassifications

OMIM names the phenotype oligodontia–colorectal cancer syndrome (#608615, abbreviation ODCRCS), with the alternative title “tooth agenesis–colorectal cancer syndrome”, and assigns it autosomal dominant inheritance caused by heterozygous AXIN2 variants; in the gene entry (*604025), the same locus is also associated with somatic colorectal cancer (114500) [1, 17]. The colorectal cancer entry (114500) includes oligodontia–colorectal cancer syndrome among the monogenic predisposition syndromes, alongside FAP, Lynch syndrome and recessive MUTYH-associated polyposis, and lists AXIN2 among the genes with somatic alterations in sporadic tumours [18]. The nonsyndromic forms of tooth agenesis form their own phenotypic series in OMIM, “selective tooth agenesis” (STHAG, selective tooth agenesis), numbered in order of discovery: STHAG1 (MSX1), STHAG3 (PAX9), STHAG4 (WNT10A), STHAG7 (LRP6), STHAG8 (WNT10B), STHAG9 (GREM2), STHAG10 (TSPEAR) and STHAGX1 (EDA), plus two loci that are only mapped (STHAG2 at 16q12.1 and STHAG5 at 10q11.2); the former STHAG6 was absorbed into the dental anomalies and short stature syndrome (601216) [19, 37–39]. The AXIN2 syndrome lies outside this numbering and has its own entry in OMIM, because OMIM defines it by the association between agenesis and cancer predisposition [1, 19]. The GeneReviews on tooth agenesis, Lynch syndrome, APC-associated polyposis and juvenile polyposis use variants of this name (“AXIN2-associated polyposis”, “polyposis and oligodontia”) in their differential diagnosis tables [16, 21–23]. Orphanet kept two records for years: ORPHA:300576, “Oligodontia-cancer predisposition syndrome” (synonym “autosomal dominant ectodermal dysplasia-cancer predisposition syndrome”), and ORPHA:401911, “AXIN2-related polyposis”. In the 2 July 2026 version of the Orphanet nomenclature, the former is inactive (“deprecated”) and the latter is the active record, linked to OMIM 608615 and to ICD-10 code D12.6 [4, 5]. The Orphanet application programming interface does not return a formal replacement relationship between the two codes, and papers from 2023 still cite the old code [15]. The GeneGuidance database links OMIM 608615 to ORPHA:401911 and names the entity AXIN2-associated attenuated familial adenomatous polyposis [12]; the same database links OMIM 114500 (somatic colorectal cancer) to ORPHA:873 (desmoid tumour), a mapping without clinical correspondence that is not used. ClinGen classified the relationship between AXIN2 and oligodontia–cancer predisposition syndrome (MONDO:0012075), autosomal dominant inheritance, as definitive [6]. The World Health Organization (WHO) classification of genetic tumour syndromes, presented in 2026, organises the entities by cellular mechanism, pathway, syndrome and gene; the abstract consulted does not indicate the position of AXIN2 in this hierarchy [40].

1.3. Definitions of tooth agenesis

Tooth agenesis is the absence of one or more permanent teeth, excluding the third molars, owing to failure of the early stages of tooth development; it is nonsyndromic when it is the only clinical finding [16]. GeneReviews considers the expression “congenitally missing teeth” inappropriate, because the permanent teeth continue to form after birth and the diagnosis is possible only at around 6 years of age [16]. Hypodontia corresponds to agenesis of up to 5 teeth, oligodontia to agenesis of 6 or more teeth, and anodontia to the absence of all permanent teeth, almost exclusive to the syndromic forms, always excluding the third molars [2, 16]. Some studies of the syndrome use slightly different limits: Roht and colleagues defined hypodontia as 1–6 missing teeth [15], and Broekema and colleagues as 1–5 [14]; the exact count and a statement on inclusion or exclusion of third molars are therefore needed to compare series. OMIM uses the same threshold of six permanent teeth, excluding third molars, to separate oligodontia from hypodontia, considers the expression “partial anodontia” obsolete and notes that inconsistent use of the terms has hampered comparison between studies [19].

The 2021 French PNDS adopts the same three categories: hypodontia of 1 to 5 teeth, oligodontia of six or more teeth not counting the third molars, and anodontia as the absence of all teeth [27]. The summary and the definitions section of the same document, however, state “more than 6 teeth”, which contradicts the definition in the introduction and the upper limit of hypodontia; the threshold consistent with the rest of the protocol and with OMIM is six or more [19, 27]. The protocol calls agenesis of more than two teeth multiple, asks that the number and type of missing teeth be recorded in the Fédération Dentaire Internationale (FDI) notation, points out that agenesis may affect the primary dentition, the permanent dentition or both, unilaterally or bilaterally, and distinguishes agenesis (a tooth germ that never formed) from a tooth absent from the arch because of failed eruption, which should not be classified as agenesis [27]. For the PNDS, anodontia is always associated with a syndrome [27].

1.4. Definitions of adenomatous polyposis

Classic FAP is defined by more than 100 colorectal adenomas; attenuated FAP by fewer than 100, and the term “multiple adenomas” or oligopolyposis applies to lower cumulative counts [29, 41]. Up to 30% of the population develops one or two adenomas by 70 years of age, but only about 1 in 1000 people develops three or more, and multiplicity is empirically associated with a higher risk of colorectal cancer [41, 42]. The AXIN2-associated syndrome usually presents with attenuated polyposis (fewer than 100 adenomas in most patients) of adult onset, with associated serrated and hyperplastic polyps [7].

1.5. Scope of the article

This article covers AXIN2 as the main entity and hereditary tooth agenesis as a syndromic class, with a uniform table of the nonsyndromic tooth agenesis genes (section 2.10) and a differential diagnosis table of the syndromic forms (section 6.2). The adenomatous, hamartomatous and serrated polyposes are included only as differential diagnoses (section 6.1): Lynch syndrome, familial adenomatous polyposis, serrated polyposis syndrome, polymerase proofreading-associated polyposis and juvenile polyposis syndrome. The ectodermal dysplasias, orofacial clefts, anterior segment dysgenesis with Axenfeld-Rieger, Kallmann syndrome and incontinentia pigmenti are likewise included in the differential diagnosis.

2. Molecular Genetics and Pathophysiology

2.1. The gene and the protein

AXIN2 lies at 17q24.1 and has 11 exons; exon 1 is untranslated and the coding region spans 10 exons, encoding a protein of 843 amino acids [14, 15]. The literature before about 2020 numbered the exons from the first coding exon, so the exon designated “exon 7” by Lammi (2004) and by Macklin-Mantia (2020) corresponds to the current exon 8 [2, 14]. The reference transcript is NM_004655.4, which is the MANE Select (ENST00000307078.10) [7, 13]. The human complementary DNA was isolated in 1999 from fetal brain, by homology with rat Axil and mouse conductin; the predicted 843-amino-acid protein shares 89% identity with Axil and contains the RGS, GSK3β-binding, β-catenin-binding and DIX (Dishevelled-binding) domains [17]. The 2001 study of the genomic structure counted 10 exons spanning more than 25 kb, a count consistent with the old numbering from the first coding exon, and showed that the gene is single-copy [17]. Initial radiation hybrid mapping placed it at 17q23–q24, a region with frequent loss of heterozygosity in breast cancer, neuroblastoma and other tumours, and fluorescence in situ hybridisation (FISH) refined the location to 17q24 [17]. On Northern blot, the transcript of about 3.0 kb is ubiquitously expressed, most strongly in thymus, prostate, testis, ovary and small intestine, less in colon and almost undetectable in leucocytes [17]. The protein is a molecular scaffold with distinct domains whose boundaries vary according to the source; the following table presents the two delimitations in use, because the published genotype–phenotype correlations depend on the delimitation chosen.

Sources: [7, 14, 25].

2.2. AXIN2 in the canonical Wnt pathway

In the absence of Wnt ligand, cytoplasmic β-catenin is captured by the destruction complex, formed by APC, AXIN1 or AXIN2, casein kinase 1 and GSK3β; sequential phosphorylation tags it for ubiquitination and proteasomal degradation, which keeps the TCF/LEF-dependent target genes silent [7, 10, 12]. The AXIN proteins are the component whose concentration limits assembly of the complex [7]. AXIN2 is itself a direct transcriptional target of the pathway, so Wnt activation increases AXIN2 and AXIN2 brakes signalling: it is a negative feedback loop [7]. AXIN2 also interacts with SMAD3, facilitates activation by the TGFβ (transforming growth factor β) receptors and is negatively regulated by TGFβ, mediating crosstalk between the two pathways [7]. Aceves-Ewing and colleagues emphasise a dichotomous role: without Wnt, AXIN2 inhibits the pathway in the destruction complex; with active Wnt, it binds the LRP5/6 (low-density lipoprotein receptor-related protein 5/6) co-receptors and helps to maintain their phosphorylation, promoting signalling [10]. AXIN2 stability is controlled by tankyrase, which poly-ADP-ribosylates it and directs it to degradation; the tankyrase-binding motifs at the amino terminus are therefore determinants of the amount of protein available [7, 10].

2.3. Mechanism of tooth agenesis

Axin2 is expressed in the enamel knots and in the developing dental mesenchyme, as Lammi and colleagues showed by in situ hybridisation in the mouse [2]. In animal models, both inhibition of the Wnt pathway (inactivation of Lef1, overexpression of Dickkopf) and its overactivation disturb odontogenesis; in FAP and Gardner syndrome, loss of APC is associated with supernumerary teeth and odontomas [2]. The current interpretation is that the dentition depends on the magnitude and spatiotemporal regulation of the Wnt signal: excess signal in the dental mesenchyme, when the AXIN2 brake is missing, leads to agenesis, whereas other disturbances of the pathway produce supernumerary teeth [15, 25]. Pathogenic AXIN2 variants mainly affect the permanent dentition, particularly the posterior molars, without significant involvement of the primary dentition [2, 15].

Outside the dentition, Axin2 expression serves, in murine lineage-tracing models, as a marker of Wnt-dependent stem cell populations: the stem cells of the interfollicular epidermis, which self-renew through autocrine Wnt signalling; a pericentral diploid population of the hepatic lobule that renews hepatocytes during homeostasis, maintained by Wnt signals from the central vein endothelium; and the cells of the base and lower isthmus of the gastric antral glands, whose expansion depends on stromal R-spondin 3 and increases in Helicobacter pylori infection [17]. These animal model data illustrate the role of AXIN2 as a target and sensor of the Wnt pathway in epithelial renewal niches, without established direct clinical translation.

2.4. Tumour mechanism: haploinsufficiency, stable truncation and second hit

The GeneGuidance database and the ClinGen dosage curation describe the mechanism as haploinsufficiency, with reduced β-catenin degradation and constitutive activation of Wnt transcription, similar to loss of APC [12]. The functional data from the French series qualify this reading. The truncating variants of exons 6 and 8, which account for most polyposis cases, do not appear to undergo nonsense-mediated messenger RNA decay (NMD): in a medulloblastoma with p.(Val506Ter) the truncated protein is expressed, and RT-PCR (reverse transcription followed by polymerase chain reaction) analysis of blood and tumour from a patient with c.2062_2063del showed expression of both alleles [7]. In vitro, the p.(Val506Ter), p.(Arg656Ter) and p.(Leu688Ter) mutants are more stable than the wild-type protein, and their transfection accumulates cytoplasmic and nuclear β-catenin and activates the pathway [7]. Marvin and colleagues had already shown, in HEK293 cells, a truncated protein of about 80 kDa lacking the last three exons, including the DIX domain [3]. These results suggest that the truncated protein, lacking the DIX domain and all or part of the SMAD3-binding domain but with the β-catenin-binding domain preserved, retains some activity and interferes with the pathway, which is compatible with selection of altered-function mutants rather than simple loss [7]. In contrast, the missense variant p.(His660Tyr), described in a family with isolated tooth agenesis, decreased β-catenin and inhibited the pathway, so truncating and missense variants may have opposite effects [7]. Loss of heterozygosity (LOH) of the wild-type allele was documented in two colorectal cancers of the Australian series, as a second hit [35].

Somatic AXIN2 alterations occur in about 20%–25% of MMR-deficient colorectal tumours, almost always single-base-pair insertions or deletions in the mononucleotide repeats of exon 8, with truncation and β-catenin/TCF activation; somatic mutation or promoter hypermethylation has also been described in gastric carcinoma, endometrioid ovarian carcinoma and medulloblastoma [7, 11]. In the original Mayo Clinic study, 11 of 45 MMR-deficient colorectal carcinomas (24%) had AXIN2 alterations; the two somatic variants recorded in OMIM, in historical notation, are the insertion 2084insG, predicted as Glu706Ter and observed in tumours from six patients, and the deletion 2083delG, predicted as Leu688Ter and observed in tumours from three patients [11, 17]. The deletion 2083delG corresponds to NM_004655.4:c.1994del p.(Gly665AlafsTer24) (ClinVar VCV000005880, OMIM allele 604025.0002); the insertion 2084insG is a published (somatic) designation without a confirmed HGVS equivalent, because ClinVar (VCV000005879, OMIM allele 604025.0001) records it as c.2116G>T p.(Glu706Ter), a substitution incompatible with the description of an insertion [43]. The mutant protein proved more stable than the wild type in colon cancer cells, and preliminary data suggested AXIN2 overexpression in these tumours [17]. The germline variant c.1994dup, recurrent in the syndrome, is identical to one of the somatic mutations of these tumours [2]. These somatic alterations do not indicate hereditary predisposition.

2.5. Variant spectrum and normalised nomenclature

The following variants were normalised in VariantValidator on NM_004655.4 and GRCh38, with ClinVar status consulted on 28 October 2026 [13, 43]. The “literature” column retains the historical notation to allow searching in the original sources.

Three nomenclature notes prevent transcription errors. The table of Roht and colleagues presents the Finnish variant as “c.1966C>T p.Arg565*”, a typo: codon 656 is correct [2, 7, 15]. The PMC text of Marvin and colleagues sometimes writes “p.Tyr663X” and “c.1989A>G”, whereas the abstract and the French series confirm c.1989G>A p.(Trp663Ter) [3, 7]. The variant c.254del appears as p.(Leu85TyrfsTer84) in the discussion and as p.(Leu85TyrfsTer24) in the introduction of Leclerc and colleagues and in the table of Broekema and colleagues; the primary source was not consulted to resolve the discrepancy [7, 14]. Exon 8 is a hotspot for germline and somatic variants, with several mononucleotide repeats [14, 15].

By 2021, HGMD (Human Gene Mutation Database) recorded 40 AXIN2 variants of classes 3 to 5 (21 pathogenic or likely pathogenic and 19 of uncertain significance), about half associated with polyposis, colorectal cancer or oligodontia, and ClinVar recorded 44 pathogenic or likely pathogenic variants [15]. The most frequent variants in the 2025 systematic review were c.1966C>T (10 patients), c.1994dup (8) and c.1989G>A (5) [25].

2.6. Genotype–phenotype correlations

In the French series, 8 of the 10 pathogenic or likely pathogenic variants were located in exons 6 (n=2) or 8 (n=6); they were truncating and were predicted to produce a protein lacking the DIX domain and all or part of the SMAD3-binding domain, with the β-catenin-binding domain preserved [7]. Loss-of-function variants in the region encoding the SMAD3-binding domain were present in 7 of 10 patients versus 1 of 8 controls (p=0.025), whereas variants not causing loss of function were not enriched (23/60 versus 136/427; p=0.378) [7]. The two exon 2 variants (c.204_214del and c.-12_8del, which removes the initiation codon, p.0?) were associated with particular phenotypes: the former coexisted with homozygosity for MUTYH c.1145G>A p.(Gly382Asp) and was excluded from the clinical analysis; the latter segregated in a family with oligodontia, ectodermal dysplasia in a child, a single adenoma and a cancer of the ampulla of Vater [7]. The authors found no relationship between tooth agenesis and the nature or location of the variant [7].

The 2025 systematic review, with a broader delimitation of the SMAD3-binding domain (amino acids 476–761), located 20 of 34 variants (59%) and 61 of 91 patients (67%) in that domain [25]. Among the 68 patients with agenesis, 57 (83.8%) had a variant in that domain, and no patient with a variant in the β-catenin-binding domain had agenesis [25]. The mean number of missing teeth, excluding third molars, was 6.93 in 69 patients, with no significant difference by domain, but it differed by variant type: 12.9 with nonsense variants (16 patients, all with agenesis), 4.9 with missense (17) and 4.8 with frameshift (36) [25]. In the SMAD3-binding domain, 40 of 43 patients (93%) had agenesis and cancer or polyps, almost always with frameshift (20/40) or nonsense (19/40) variants [25].

2.7. Penetrance and expressivity

Oligodontia was completely or almost completely penetrant in the two original nonsense families, and colorectal neoplasia occurred in 8 of the 11 affected Finnish individuals and in 3 of the 4 North American carriers [2, 3]. In the North American family, all five affected individuals (the proband, her mother, two maternal aunts and the maternal grandmother) had oligodontia, four had sparse hair and eyebrows and three had colonic polyps or cancer, one of whom also had breast cancer at 44 years of age; the grandmother, who died at 97 years without testing, had no known history of polyps or cancer, and the variant was absent in two unaffected relatives [1, 17]. This course shows that an obligate carrier may reach a very advanced age without known neoplasia, although the absence of a record does not equate to the absence of lesions, given the lack of documented colonoscopy. In other families colorectal penetrance is incomplete and variable: in the Australian series, 5 of 7 carriers of c.1049del, 2 of 6 carriers of c.1994dup and 3 of 3 carriers of c.1972del had colorectal cancer or polyposis [35]. In the systematic review, 68 of 83 patients with dental data (82%) had agenesis, with complete dental penetrance for nonsense variants and incomplete penetrance for missense and frameshift variants [25]. The frequency of c.1994dup in the gnomAD population database (Genome Aggregation Database; allele frequency 0.0006954, 17 heterozygotes, accessed April 2020) is high for a pathogenic variant of a rare syndrome, which is compatible with incomplete penetrance, an attenuated phenotype or common ancestry; the authors who report it did not estimate penetrance from this figure [44]. The available estimates come from families and cohorts selected by phenotype, subject to ascertainment bias (section 3.10); there are no age-specific cumulative risks of colorectal cancer, and the GeneReviews on Lynch syndrome classifies the risk as unknown [21].

2.8. Missense variants, variants of uncertain significance and reclassifications

AXIN2 variants in isolated or syndromic oligodontia without known cancer predisposition are mainly missense: c.923C>T p.(Thr308Met) and c.2490G>C p.(Met830Ile), c.314T>G p.(Val105Gly), c.1978C>T p.(His660Tyr), c.137C>A p.(Thr46Asn), c.2029C>T p.(His677Tyr), c.2522G>A p.(Arg841Gln) and an insertion c.133_134insGCCAGG; the variant c.2285G>A p.(Ser762Asn) was described in an ovarian cancer without agenesis [7, 25]. There are truncating exceptions in young patients with isolated oligodontia: c.1994dup was found in patients aged 13 and 17 years and, later, in a 65-year-old patient with 57 adenomas and in a family with agenesis, polyps and colorectal cancer, which shows that young age and the absence of polyps do not exclude predisposition [7, 44]. The 17-year-old patient belongs to a Swedish series of isolated oligodontia (Bergendal and colleagues, 2011), in which three index cases with mild oligodontia (7 to 10 missing teeth) had one frameshift and two missense AXIN2 variants; the young woman lacked 8 teeth and her mother 9, and none of the three index cases reported ectodermal features or personal or family colorectal cancer, information obtained by report and not by targeted clinical assessment [1, 45]. The series included 93 Swedish probands with isolated nonsyndromic oligodontia, screened for AXIN2, MSX1, PAX9, EDA, EDAR and EDARADD: 10 (10.8%) had variants predicted to be damaging, of which 3 were in AXIN2 (3.2% of probands), 4 in PAX9, 2 in MSX1 and 1 in EDARADD [45]. Oral parameters did not differ between carriers and non-carriers, but a family history of oligodontia was three times more frequent in probands with a variant [45]. The Spanish variant c.1387C>T p.(Arg463Cys) was supported by the authors on the basis of low frequency in controls, loss of heterozygosity, segregation and immunofluorescence [8], but the French series argues that loss, in the polyps, of the mutated allele (and not of the wild-type allele) argues against pathogenicity, and ClinVar classifies it as of uncertain significance [7, 43]. The variant c.1882C>T p.(Arg628Trp), present in 48 of 282,360 gnomAD alleles (0.017%), was classified as of uncertain significance in the Estonian family [15], and c.1822del was reclassified from pathogenic to uncertain significance and now has conflicting interpretations in ClinVar [15, 43]. Three truncating variants upstream of the β-catenin-binding domain (c.254del in a rectal cancer at 79 years without polyposis, c.204_214del with MUTYH homozygosity and c.1049del in an Australian family) have a clinical significance that requires further study [7].

In the ACMG/AMP (American College of Medical Genetics and Genomics/Association for Molecular Pathology) classification, the consensus of the French Groupe Génétique et Cancer adopted for AXIN2 an allele frequency above 1% in controls as a stand-alone benign criterion, instead of the generic 5% [7]. There are no AXIN2-specific classification rules from a ClinGen variant curation expert panel; at the time of the 2024 European guideline, only APC had them among the polyposis genes [26].

2.9. Allelic forms and modifiers

Neurodevelopmental disorder caused by variants in the tankyrase-binding domain. Aceves-Ewing and colleagues, drawing on the Undiagnosed Diseases Network (UDN) and commercial laboratories, described in 2026 five individuals with de novo heterozygous variants in the tankyrase-binding domain: c.196G>A p.(Glu66Lys) in three, absent from gnomAD v4.0.0, c.197A>G p.(Glu66Gly) in one and c.199G>A p.(Gly67Arg) in one [10]. Structural modelling predicts that p.Gly67Arg abolishes and p.Glu66Lys weakens AXIN2–tankyrase binding, reducing poly-ADP-ribosylation and degradation and potentially increasing AXIN2 stability; the equivalent AXIN1 variant, p.Gly74Val, prevents binding [10]. In mice with the variant introduced by prime editing, none of the 30 weaned animals was heterozygous, whereas 14 of 29 embryos at day E18.5 (48%) were (p<0.0001), indicating perinatal lethality, with soft palate cleft, kinked tail and digital and skeletal anomalies [10]. In Drosophila, overexpression in the wing produced more severe phenotypes than the reference, suggesting gain of function or a dominant-negative effect rather than loss of function [10]. This is therefore a mechanism different from that of the truncating variants of the classic syndrome.

Somatic mosaicism. Murakami and colleagues described a Japanese man with more than 200 proximal adenomas, a negative germline test and the same c.623_624del variant in three independent adenomas, with loss of heterozygosity and absence from normal mucosa, the first demonstration of somatic AXIN2 mosaicism [9].

Orofacial clefts. AXIN2 polymorphisms have been associated with tooth agenesis and oral clefts in association studies, and Axin2 and Irf6 are co-expressed in the murine palatal epithelium [3, 15]. The presence of cleft palate in three members of the Estonian family with c.1214_1215dup was proposed as a possible manifestation with incomplete penetrance; none of the 21 pathogenic variants in HGMD was associated with cleft palate at the time [15]. In an association study of 116 Brazilian and 51 Turkish patients with tooth agenesis, Callahan and colleagues analysed three AXIN2 polymorphisms (rs7591, rs11867417 and rs2240308) and found an association of the T-C-A haplotype with agenesis in cases with at least one missing incisor (combined p of 0.02), a result of modest effect, distinct from the Mendelian causality of the truncating variants [17, 46]. The Brazilian sample was from Rio de Janeiro, an admixed population of European (mainly Portuguese) and African origin, with 71 sporadic and 45 familial cases (74 female); the Turkish sample consisted of 51 trios, all sporadic, and in both the most frequently missing teeth were the second premolars, followed by the lateral incisors [46]. The authors chose not to apply the Bonferroni correction; with it, the threshold would fall to 0.0056 (0.05/9) and the association with missing incisors (p = 0.021) would no longer be significant, so the result is exploratory [46]. The 2025 review records divergent association results across populations [25].

Coexistence with other variants. In the French series, one index case was homozygous for MUTYH and another had a heterozygous FLG variant that explained the ichthyosis but not the oligodontia [7]; Macklin-Mantia and colleagues recorded variants of uncertain significance in NF1 and RAD50 in one carrier [24]. These findings reinforce the need for complete panels and joint interpretation.

2.10. The class of hereditary tooth agenesis

About 80% of cases of nonsyndromic tooth agenesis are attributed to variants in genes of craniofacial and dental development, and about 20% to exogenous factors acting during formation of the permanent tooth germs, such as chemotherapy, radiotherapy, maternal rubella, thalidomide and other antineoplastic agents [16]. Many genes have been suggested on the basis of syndromic forms or animal models, and syndromic forms are frequent in orofacial cleft syndromes and in the ectodermal dysplasias [16]. Inheritance is traditionally monogenic, but there is evidence of multilocus or oligogenic inheritance: WNT10A variants cosegregate with variants in GREM2, LAMA3 or BCOR [16]. Inter- and intrafamilial variability and reduced penetrance are common, and genes with an allelic syndromic form appear to form a spectrum between apparently isolated agenesis and the syndrome, so that, after a variant is identified, an active search for syndromic features is recommended [16]. The WNT10A variant c.682T>A p.(Phe228Ile) is the most common pathogenic variant in nonsyndromic agenesis, present in about 3.43% of the general population, with reduced penetrance in heterozygotes [16]. The following table brings together the 11 genes that GeneReviews includes because they have at least two families with nonsyndromic agenesis after careful phenotyping [16]; the OMIM phenotype numbers and the STHAG series names were checked against the OMIM entries, and empty cells correspond to genes without a nonsyndromic agenesis entry in that series [19].

Sources: [16, 19]. AD, autosomal dominant; AR, autosomal recessive; ODCRCS, oligodontia–colorectal cancer syndrome; STHAG, selective tooth agenesis. The column layout of the WNT10B row in the original GeneReviews file is ambiguous; the OMIM entry describes oligodontia, which supports the “+” in that column.

The OMIM series adds two forms to the GeneReviews list: STHAG10, of autosomal recessive inheritance caused by biallelic TSPEAR variants (620173), a gene also responsible for ectodermal dysplasia 14 (section 6.2), and two loci without an identified gene, STHAG2 (602639, 16q12.1, autosomal recessive in a consanguineous Pakistani family) and STHAG5 (610926, 10q11.2, agenesis restricted to the permanent dentition in a Chinese family) [19, 38, 39, 52, 53]. The patterns of agenesis described in the OMIM entries help to guide testing: in MSX1, absence of the upper first premolars (75% of carriers in the series of Kim and colleagues) is the most distinctive feature, together with the second premolars and third molars; in PAX9, absence of the upper and lower second molars exceeds 80%, with the primary dentition sometimes also affected; in WNT10B, the lateral incisors predominate [19, 51, 54]. The GeneReviews cell that assigns MSX1 a pattern of “permanent molars” diverges from these descriptions and from the 2025 systematic review, which also records the predominance of second premolars [16, 25].

For WNT10A, two series summarised in OMIM quantify the weight of the gene and the effect of allelic dosage. In a series of 34 unrelated patients with nonsyndromic agenesis of 6 to 28 teeth, van den Boogaard and colleagues found WNT10A variants in 56%, PAX9 in 9% and MSX1 and AXIN2 in 3% each, and inclusion of WNT10A raised the yield of molecular testing from 15% to 71% [55]; in a population-based Swedish cohort (94 families, 102 individuals, mean of 8.2 missing teeth), Arzoo and colleagues identified WNT10A variants in 27.7% of probands, with more missing teeth in biallelic (mean 11.1) than in monoallelic individuals (6.8) and more relatives with oligodontia (5/11 versus 1/17) [56, 57]. With the seven genes (WNT10A, MSX1, PAX9, AXIN2, EDA, EDAR and EDARADD), 38.3% of probands had a variant, and the authors stress that this population-based cohort had fewer WNT10A alleles and fewer missing teeth than series recruited in specialist centres, which explains part of the difference from the Dutch 56% [57]. In the same series, 11 patients had subtle ectodermal features (sparse hair and eyebrows, short eyelashes, toenail changes) without the signs of odonto-onycho-dermal dysplasia [56]. These data come from selected series and were read in the OMIM synthesis, without consulting the primary articles. In LRP6, the variants of four Chinese patients showed a dominant-negative effect on Wnt/β-catenin signalling, a de novo variant was described and one carrier with oligodontia also had hypohidrosis and sparse hair, absent in his carrier father [49]; the LRP6 missense variants associated with early coronary artery disease with hypertension, dyslipidaemia and type 2 diabetes (ADCAD2) constitute a distinct phenotype, described in an Iranian family and in North American families, and the sources consulted do not describe tooth agenesis in these carriers [50].

GeneReviews is internally inconsistent regarding AXIN2: Table 1 marks it only in the hypodontia column, whereas Table 2 describes oligodontia with agenesis of molars, lower incisors and upper lateral incisors [16]. The primary literature resolves the question in favour of a spectrum ranging from hypodontia (1–5 teeth) to severe oligodontia and, in one case, anodontia: the original families had 8 to 29 missing teeth [2], the French series recorded oligodontia in 8 patients and anodontia in one [7], and the European and North American cohort included carriers with 1 to 4 missing teeth [15]. GeneReviews Table 2 also includes odontomas and osteomas among the features of AXIN2, whereas the French series stresses that, unlike FAP, AXIN2 carriers present only agenesis, without supernumerary teeth, odontomas or impacted teeth [7, 16]; the primary sources consulted do not describe odontomas in AXIN2 carriers.

The 2021 French PNDS reproduces the proportions of 80% genetic and 20% exogenous causes in nonsyndromic oligodontia, adding craniofacial trauma to the exposures, and mentions more than 50 genes responsible for isolated or syndromic forms and more than 60 OMIM syndromes with oligodontia [27]. As the main genes of the nonsyndromic forms it lists EDA, EDAR, EDARADD, MSX1, PAX9, IRF6, GREM2, AXIN2, LRP6, SMOC2, LTBP3, FGFR1, PITX2, WNT10A and WNT10B, considers WNT10A the most frequently implicated gene and stresses that the same genes appear in the syndromic forms, which requires a search for extraoral signs [27]. It also describes topographic signatures: agenesis of incisors and premolars with minor ectodermal signs in WNT10A; agenesis of incisors and premolars in MSX1; agenesis of the posterior segments in PAX9; and, in EDA, a decreasing order of lower incisors, upper lateral incisors, lower second premolars, upper second premolars and second molars in the permanent dentition [27]. In X-linked ectodermal dysplasia, the protocol indicates a mean of 8 missing primary teeth (5.3 in the mandible and 3.5 in the maxilla) and 13.5 missing permanent teeth [27]. The PNDS signature for MSX1 (incisors and premolars) differs from the 2025 systematic review, in which second premolars predominate and canines are spared, and from OMIM, which gives the upper second and first premolars as most affected (section 3.1) [19, 25]. The two French panels annexed to the protocol include AXIN2: the GenoDENT panel, with about 60 genes of agenesis, ectodermal dysplasia and syndromes with dental anomalies, and the Cochin panel, with 31 genes [27].

3. Clinical Spectrum and Manifestations by System

3.1. Dentition

The most constant manifestation is agenesis of permanent teeth. In the Finnish family, 11 members lacked at least 8 permanent teeth, 2 developed only 3 permanent teeth, and the number of missing teeth ranged from 8 to 29; molars and premolars, lower incisors and upper lateral incisors were mainly missing, three individuals had no canines and the upper central incisors were always present [2]. Only one individual had agenesis of four primary incisors, and retention of primary teeth was frequent [2]. In the North American family, the 35-year-old proband had no more than 10 permanent teeth, which were conical and spaced [3]. In the French series, there was multiple tooth agenesis in at least 9 of 13 carriers assessed (69%; the abstract of the paper states “at least 60%”), anodontia in one and oligodontia in eight, with 8 to 23 missing permanent teeth [7]. In the European and North American cohort, agenesis ranged from 2 to 22 teeth, and 4 of 13 carriers (30.8%) lacked 8 or more permanent teeth [15].

The systematic review described the pattern in 39 patients with complete records: the second premolars were the most frequently missing teeth (65%–78%), followed by the lower central incisors (49%) and the upper lateral incisors (46%–49%), with symmetrical involvement; the upper central incisors (8%), the canines (11%–22%) and the lower first premolars (22%–24%) were the least affected, and third molar agenesis reached 80.8%–84.6% in a subgroup of 26 [25]. This pattern differs from that of PAX9, in which second molars predominate (78%) and lower first premolars are spared (4%), and from that of MSX1, in which second premolars predominate (83%–86%) and canines are spared (6%–8%) [25]. In tooth agenesis in general, microdontia (including peg-shaped lateral incisors), malocclusion and retention of primary teeth are common [16], and peg-shaped incisors were recorded in members of the Finnish family [14]. The practical consequences include impaired chewing and feeding in childhood, the need for orthodontics and prosthetic rehabilitation, and psychological and aesthetic impact [16].

The French PNDS describes the anomalies that accompany oligodontia, whatever the gene, and that weigh on the treatment plan: delayed eruption of the permanent teeth (on average by two years), retention of primary teeth with ankylosis and infraocclusion of primary molars, more severe the earlier it occurs, microdontia and conical or rice-shaped teeth, short roots of incisors and premolars, taurodontism, impaction and transposition of upper canines, diastemas, deep overbite and reduced alveolar height and lower facial height [27]. Infraocclusion is classified as mild (occlusal surface about 1 mm below the occlusal plane), moderate (at the level of the contact point of the adjacent tooth) or severe (below the contact point, at gingival level), and may cause tipping of neighbouring teeth, extrusion of antagonists, shortening of the arch, disturbance of posterior occlusion and lateral tongue interposition [27]. At the skeletal level, reduced anterior facial height and a more retrusive maxilla (reduced SNA and ANB angles) predominate, with a tendency towards class III owing to maxillary hypoplasia, a straight or biretrusive profile and eversion of the lower lip, findings that vary with age and ethnic origin [27]. Class III malocclusion predominates in published cases, but in a series of 900 patients with agenesis (De Stefani, 2021, cited by the PNDS) class II division 2 was the most prevalent, particularly with premolar agenesis; the protocol acknowledges a substantial selection bias in published cases [27]. Tongue interposition, disorders of speech and chewing and, in the ectodermal dysplasias, a possible predisposition to obstructive sleep apnoea due to deficient maxillomandibular development (suggested by a study of 9 patients) complete the functional picture; none of these data is specific to AXIN2 [27].

3.2. Skin and appendages

Ectodermal features are frequent in the published families, but not constant. In the Marvin family, the proband had very sparse eyebrows, fine hair, a high palate, hypognathia, malar hypoplasia and a broad nasal bridge, with normal body hair, nails, skin and sweating; the mother and aunts had sparse body and axillary hair, short eyelashes and sparse or absent eyebrows [3]. In the French series, only one 10-year-old child had complete ectodermal dysplasia (oligodontia, ichthyosis, dry and sparse hair, sparse eyebrows, dysplastic nails and recurrent keratitis); the ichthyosis was explained by a coexisting heterozygous FLG variant [7]. Brittle hair and mild ectodermal dysplasia have also been described in the Danish family [14] and sparse eyebrows in a Dutch carrier [15]. In the carrier reported by Macklin-Mantia, eyebrows, hair and nails were normal [24].

3.3. Colon and rectum

The colonic phenotype is an attenuated polyposis of adult onset, with highly variable expression. In the French series, 11 of 15 carriers underwent colonoscopy and 10 of them (91%) had adenomatous polyposis (≥10 adenomas), with a median age at diagnosis of 55 years (36–82); the cumulative number of polyps ranged from more than 20 to several hundred, was below 100 in 7 of 10 patients (70%), the polyps were distributed throughout the colon and sessile serrated adenomas and hyperplastic polyps were frequent [7]. One of the two screened relatives had 40 adenomas [7]. In the Australian series, the mean age at diagnosis of colorectal cancer or polyposis was 52.5 ± 9.2 years, and the polyps were typically pancolonic, numbering between 5 and more than 100 (median 12.5), predominantly adenomatous but also serrated [35]. In the Australian family reported by Beard, the three confirmed carriers had a mean of 49 polyps (5 to more than 100) [34]. The GeneReviews on Lynch syndrome summarises the phenotype as 10 to more than 100 colonic adenomas, with a mean age of 52–62 years at diagnosis of cancer or polyposis [21].

In the Finnish family, the colorectal findings were recorded per individual [2]. The following table reproduces these data because they show, within a single family, the range of expression.

Source: [2]. The summary table of Broekema and colleagues attributes colorectal cancer to II:7, but the original paper describes 68 polyps with severe dysplasia; the primary source prevails [2, 14].

Colorectal cancer occurred in 5 of 15 French carriers (4 index cases and 1 relative; the paper states 30%, whereas 5/15 corresponds to 33%), one with two synchronous carcinomas, at a median age of 62 years (58–82); the two tumours with reported MMR status were proficient [7]. In the North American family, an aunt had metachronous colorectal cancers at 50 and 59 years, and the mother had more than 100 adenomas at 60 years, with sigmoidectomy at 60 and right hemicolectomy at 62 [3]. In the Australian family reported by Beard, the proband's daughter had colorectal cancer at 43 years and died at 46 [24]. In the Spanish family with a missense variant, three relatives had colorectal cancer between 36 and 51 years of age, without hypodontia [8, 14].

3.4. Upper gastrointestinal tract

Lesions of the upper gastrointestinal tract are less frequent than in FAP and MAP, but have been described: five duodenal polyps in one patient, a small-bowel carcinoma at 66 years followed by a caecal carcinoma, and a pancreatobiliary-type carcinoma of the ampulla of Vater with death at 53 years [7]. There was no gastric polyposis or cancer in the French series, but antral intestinal metaplasia not attributable to Helicobacter pylori was recorded in one patient and three fundic gland polyps in another [7]. The Marvin proband had multiple fundic gland polyps at 33 and 35 years, possibly related to proton pump inhibitor use, and a normal capsule endoscopy [3]. Macklin-Mantia and colleagues described a foveolar-type gastric adenoma [24], and Roht and colleagues a gastric adenoma in a female carrier [15]. The GeneReviews on Lynch syndrome includes upper gastrointestinal cancers among the manifestations of AXIN2 [21].

3.5. Extra-intestinal tumours

Breast cancer (at 44 and 53 years), endometrioid ovarian carcinoma, melanoma, unspecified skin cancer, prostate adenocarcinoma, olfactory neuroblastoma, salivary gland cancer and head and neck cancer have been described [3, 7, 15, 24, 44]. In the systematic review, among 63 patients with data, men had colorectal cancer or polyps in 78%, skin cancer in 10% and other tumours in 9% (peritoneal, ampulla of Vater, prostate), and women had colorectal cancer or polyps in 71%, breast cancer in 7%, ovarian cancer or cysts in 7% and skin cancer in 2% [25]. Leclerc and colleagues consider the association with extracolonic cancers (ampulla, ovary, breast, melanoma, neuroblastoma, keratoacanthoma) not established and advise attention to the salivary glands and the head and neck, having observed three relatives with tumours at these sites [7]. The NCI PDQ (National Cancer Institute Physician Data Query) and the NCCN (National Comprehensive Cancer Network) make no extracolonic recommendations owing to the lack of penetrance data [28, 29].

3.6. Orofacial clefts

In the Estonian family with c.1214_1215dup, a 6-year-old girl had Pierre Robin sequence (repaired bilateral cleft palate, micrognathia and retrognathia, microglossia with glossoptosis) and 2 missing teeth, her 4-year-old brother had bilateral cleft palate and their 26-year-old mother had cleft palate and an impacted tooth [15]. The authors propose cleft palate as a possible manifestation, with incomplete penetrance, and recommend taking a family cancer history in patients with unexplained oral clefts [15]. Marvin and colleagues had already noted that families with cleft lip and palate had a higher incidence of breast and colon cancer in a study of 75 families [3].

3.7. Other findings

Two French patients had severe bleeding during polypectomies, one requiring admission to intensive care, and a third had antral angiodysplasia and a tendency to gastrointestinal bleeding; the three carriers of family Li-4 had idiopathic myopathy, and the Marvin proband had fibromyalgia [3, 7]. A causal relationship between these findings and AXIN2 has not been established, but the authors recommend considering them before invasive procedures [7].

3.8. Neurodevelopmental disorder caused by variants in the tankyrase-binding domain

In the five individuals with de novo p.Glu66Lys, p.Glu66Gly or p.Gly67Arg variants, digit or limb anomalies were present in 5/5, global developmental delay in 4/5, microcephaly in 4/5, complex ophthalmological anomalies in 4/5, genitourinary anomalies in 3/5, facial dysmorphism in 4/5, palatal anomalies in 2/5, oligodontia in 4/5, sparse hair in 4/5, sparse eyebrows and eyelashes in 4/5 and nail anomalies in 2/5; Raynaud phenomenon was also described [10]. The only adult, whose variant was found on a colorectal cancer panel, had a milder phenotype (digital and skeletal anomalies, vesicoureteral reflux) and colon cancer with more than 25 polyps, and the authors recommend considering colorectal surveillance in these carriers [10].

3.9. Polyposis caused by somatic mosaicism

The Japanese man described by Murakami and colleagues was 77 years old, asymptomatic and had undergone sigmoid polypectomies at 72; he had more than 200 low-grade tubular adenomas, up to 10 mm, predominantly in the caecum and proximal ascending colon, few and small in the transverse and descending colon and none in the sigmoid and rectum, without serrated lesions, tooth agenesis, ectodermal findings, duodenal adenomas or desmoids [9]. The segmental distribution and the absence of dental manifestations distinguish this form from the germline syndrome.

3.10. Natural history, ascertainment bias and disease under surveillance

The available data come from families and hospital series selected for oligodontia, polyposis or cancer, and almost all carriers with cancer were identified after the cancer diagnosis. The systematic review shows the effect of the route of ascertainment: the mean age was 22.5 ± 19.4 years in the 17 patients with tooth agenesis only, 53.3 ± 16.5 years in the 53 with colorectal cancer or polyps and 61.9 ± 8.4 years in the 10 with other tumours (p<0.05); the authors consider that the proportion of patients with cancer underestimates the age-dependent lifetime penetrance, and that the age recorded is not always the age at diagnosis [25]. Conversely, selection through cancer overestimates the frequency of cancer among identified carriers. The co-occurrence of agenesis and cancer or polyps in 73%, of isolated cancer in 20% and of isolated agenesis in 3% therefore describes the published literature and not the risk of a carrier identified by cascade testing [25]. None of the sources consulted describes a cohort of carriers followed prospectively under surveillance, so it is not possible to separate natural history from treated disease or to estimate the effect of surveillance on cancer incidence.

4. Clinical Recognition and Red Flags

4.1. In dental and paediatric dental clinics

Agenesis of six or more permanent teeth, particularly with simultaneous involvement of the anterior and posterior segments, is the sign that allows the syndrome to be recognised before cancer. Unusual spacing of the dentition in a child should lead the paediatrician and dentist to suspect agenesis, which is confirmed by orthopantomography after 6 years of age [16]. A family history of polyps or colorectal cancer, relatives with multiple dental prostheses from an early age and ectodermal features (sparse hair and eyebrows) increase the probability of AXIN2 [3, 14]. Broekema and colleagues recommend that dentists refer people with oligodontia and families with hypodontia to clinical genetics [14].

The French PNDS identifies four circumstances of discovery: absence of teeth noticed by the parents during eruption, an incidental finding on an orthopantomogram taken for orthodontic assessment, inclusion in an already identified syndromic picture, and a family history of anomalies of tooth number; the most common warning sign is delayed eruption of the permanent teeth between 6 and 12 years of age [27]. Because of the colorectal carcinoma risk described in AXIN2 oligodontia, the protocol requires that the history of any patient with oligodontia include the family history of colorectal cancer, and it places colorectal cancer, polyps and adenomas among the tumour signs associated with AXIN2 in its syndromic correspondence tables; it does not, however, propose any gastrointestinal management for carriers [27].

4.2. In gastroenterology and oncology clinics

Attenuated adenomatous polyposis, with 10 or more adenomas, associated serrated or hyperplastic polyps and adult onset, should prompt questions about the dental history and examination of the mouth and jaws [7, 41]. According to Leclerc and colleagues, the association between polyposis or colorectal cancer and oligodontia is highly predictive of a germline AXIN2 variant [7]. Marvin and colleagues recommend asking about ectodermal phenotypes in families with colorectal neoplasia and studying AXIN2 when neoplasia coexists with oligodontia, sparse hair or eyebrows or oral clefts [3]. Tooth agenesis may go unnoticed in adults, because it is confused with extractions for caries or periodontal disease: in the Beard family, the proband's oligodontia was recognised only after genetic testing [16, 24]. Facial and dental features as clues to cancer predisposition syndromes have been reviewed by Horton and colleagues [58], and a 2025 letter described an African American carrier identified from the physical manifestations [59, 60].

4.3. In paediatrics and developmental genetics

A child with oligodontia, sparse hair and eyebrows, digit or limb anomalies, microcephaly, global developmental delay and ocular or genitourinary anomalies may have a de novo variant in the tankyrase-binding domain [10]. In the 2026 paediatric review of hereditary polyposis syndromes, oligodontia is listed among the extra-intestinal findings that guide the differential diagnosis in children, alongside macrocephaly, mucocutaneous pigmentation, osteomas, desmoids and congenital hypertrophy of the retinal pigment epithelium (CHRPE) [61].

4.4. Tooth agenesis as a population marker

In a Danish population cohort of 2,501,715 singleton live births (1977–2018), followed up to 40 years of age, recorded tooth agenesis (2.8%) was associated with colorectal cancer between 30 and 40 years of age, with a hazard ratio (HR) of 2.81 (95% confidence interval [95% CI] 1.38–5.71), and with several childhood cancers [62]. The systematic review by Hashmi and colleagues (2026) rated the certainty of this evidence as very low using GRADE (Grading of Recommendations Assessment, Development and Evaluation), noted that agenesis of a single tooth occurs in about 3.5%–8% of the population and concluded that tooth agenesis should not be used for universal colorectal cancer screening; as an interim approach, it proposes targeted AXIN2 testing in patients with polyposis and tooth agenesis [63].

4.5. Warning and emergency signs

Rectal bleeding, iron deficiency anaemia, a recent change in bowel habit and bowel obstruction in a carrier or in a person with oligodontia and a suggestive family history are signs that warrant immediate assessment [14, 24]. Obstructive jaundice or persistent upper abdominal pain raise the possibility of ampullary neoplasia, in light of the fatal case in the French series [7]. Unilateral nasal obstruction, hyposmia and epistaxis led to the diagnosis of olfactory neuroblastoma in a female carrier, and the authors propose referral to otorhinolaryngology with imaging for nasal or facial symptoms [24]. Bleeding after polypectomy was severe in two carriers [7]. There is no newborn screening for this condition. The French PNDS does not provide an emergency card for oligodontia, although cards exist for some syndromic forms (ectodermal dysplasias, incontinentia pigmenti), and recommends presenting to emergency services the health record with the diagnosis, associated malformations and ongoing treatments [27].

5. Diagnosis: Criteria, Tests and Algorithms

5.1. Criteria

There are no formal clinical criteria for the syndrome. The diagnosis is molecular: identification of a heterozygous pathogenic or likely pathogenic germline AXIN2 variant in a person with tooth agenesis, polyposis or colorectal cancer, or in an at-risk relative [7, 12]. No biochemical or functional test confirms the disease, and neither polyp histology nor the pattern of agenesis is specific to AXIN2 among the polyposis and agenesis genes [12]. Classification as classic polyposis (more than 100 adenomas), attenuated polyposis (fewer than 100) or multiple adenomas describes the colonic burden and guides testing, but does not identify the gene [41].

5.2. Dental assessment

The diagnosis of agenesis requires clinical and radiographic examination to exclude impacted teeth and losses due to caries or trauma; orthopantomography confirms it after 6 years of age, when all permanent tooth germs except the third molars should be visible [16]. In adults, it is useful to consult old dental records and to assess the edentulous spaces to exclude extractions or periodontal disease; direct examination of relatives by an oral health professional experienced in agenesis may be necessary, because in adults it is difficult to distinguish agenesis from other causes of tooth loss [16]. The assessment includes the exact count and identification of missing teeth, other anomalies of shape and enamel, bitewing, periapical and panoramic radiographs and, in adolescents and adults, lateral cephalometric radiography and cone-beam computed tomography when indicated [12, 16]. The three-generation pedigree should record agenesis, enamel and tooth shape anomalies, hair, nails, skin, sweating and clefts [16, 64].

The 2021 French PNDS details the initial assessment. The dental chart is compared with what is expected for chronological age; agenesis is affirmed only when it is shown that the tooth never existed, excluding traumatic or infectious losses and early extractions, and the history seeks exposure to toxins, drugs or teratogenic agents and other affected family members [27]. The diagnosis of agenesis of permanent teeth is possible with relative certainty between 6 and 8 years of age [27]. For imaging, the protocol applies the ALADAIP principle, which replaces ALARA and seeks the lowest diagnostically acceptable exposure: isolated intraoral periapical radiographs as first line; orthopantomography as first line when the anomaly of number is generalised, with sequences limited to the arches and, if absence of permanent tooth germs is suspected, from 4–6 years of age; avoidance of the complete 21-film periapical survey; follow-up with isolated intraoral radiographs; lateral cephalometric radiograph with cephalometric analysis for dentofacial deformities; and cone-beam computed tomography only in the pre-implant phase, close to surgery, with a small or medium field of view and low-dose protocols [27]. The assessment is completed by study models (conventional or optical impression) mounted on an articulator, intraoral and extraoral photographs, orthodontic assessment (skeleton, soft tissues, orofacial functions) and extraoral examination of the skin appendages and sweating, particularly on the palms and soles [27]. When the end of growth is uncertain before implants, skeletal maturity is assessed by radiography of the left hand and wrist according to the Greulich and Pyle atlas; fusion of the epiphyseal growth plates occurs at about 14–15 years in girls and 17–18 years in boys, with wide individual variation [27]. GeneReviews places radiographic confirmation after 6 years of age, and the PNDS accepts the first orthopantomogram from 4 years of age when absence of tooth germs is suspected (section 7.12) [16, 27].

5.3. Gastrointestinal assessment

Colonoscopy with histology documents the number, distribution and type of polyps (adenomatous, serrated, hyperplastic) [12]. In the general approach to polyposis, monocryptal microadenomas and multiple fundic gland polyps are characteristic of FAP and attenuated FAP, and a large pigmented retinal lesion with a pale halo (CHRPE) is diagnostic of FAP, although its absence does not exclude it [41]. Upper gastrointestinal endoscopy assesses the duodenum, ampulla and stomach, in light of the duodenal polyps, ampullary neoplasia and gastric adenomas described [7, 24].

5.4. Whom to test

The 2024 joint European guideline of the European Hereditary Tumour Group (EHTG) and the European Society of Coloproctology (ESCP) recommends considering a multigene panel in the presence of more than 20 cumulative colorectal adenomas (recommendation OAPS.1A; low evidence; 97% agreement), lowering the threshold to 10 or more adenomas if the diagnosis occurs before 60 years of age, if there is a family history of polyposis or colorectal cancer or if there are compatible extracolonic manifestations (OAPS.1B; low evidence; 93%), testing gastrointestinal cancers before 50 years of age with a panel (OAPS.1C; low evidence; 94%) and studying APC mosaicism in tissue when unexplained polyposis meets 1A or 1B (OAPS.1D; low evidence; 94%) [26]. Additional indications include the extracolonic manifestations of FAP, the somatic KRAS c.34G>T transversion and a first-degree relative with more than 10 adenomas; above 70 years of age, with fewer than 20 adenomas, the yield of testing is about 1% [26]. The Danish guideline for non-APC polyposes includes AXIN2 in the national panel [65]. The 2026 American College of Gastroenterology (ACG) guideline on adenomatous polyposis syndromes, developed with GRADE methodology, was consulted only in abstract form, and its specific content on AXIN2 could not be verified [66].

On the dental side, GeneReviews recommends establishing the genetic cause of agenesis to guide prognosis and counselling, based on the medical and dental history, physical examination, radiographs, family history and genetic testing [16]. Hashmi and colleagues propose targeted AXIN2 testing when polyposis and tooth agenesis coexist, rather than genetic screening of everyone with agenesis [63]. For variants in the tankyrase-binding domain, the context is that of a child with a syndromic picture investigated by exome or genome sequencing [10].

5.5. Which test

The GeneReviews on tooth agenesis proposes sequential single-gene testing when the clinical picture points to a gene: in posterior agenesis, MSX1 and/or PAX9 first; in hypodontia or oligodontia with anterior and posterior involvement, AXIN2, IRF6 and/or WNT10A first; in microdontia with peg-shaped incisors or abnormally shaped teeth, WNT10A and/or EDA first [16]. In practice, a multigene panel with the genes of GeneReviews Tables 1 and 2 is the option most likely to identify the cause and to limit variants of uncertain significance and non-explanatory findings; the genes and sensitivity vary between laboratories, some panels do not include all the rare genes, and the exome is the most widely used alternative [16].

In polyposis, the European guideline recommends that the panel include APC, MUTYH, MLH1, MSH2, MSH6, PMS2, MSH3, MLH3, POLE and POLD1 (exonuclease domain), NTHL1, MBD4 and AXIN2, plus STK11, BMPR1A, SMAD4, PTEN and RNF43 because of phenotypic overlap [26]. The panels published in the AXIN2 series followed similar compositions: the French panel included APC, MUTYH, the MMR genes, POLE, POLD1, SMAD4, BMPR1A, STK11, PTEN and, depending on the laboratory, NTHL1, GREM1/SCG5 and MSH3 [7]; Broekema and colleagues used a 23-gene panel [14]. One French case was diagnosed by a genodermatosis panel [7], and in the Roht cohort carriers were identified by research genome sequencing, clinical exome, a 34-gene ectodermal dysplasia and hypodontia panel, commercial cancer panels of 70 and 84 genes and a polyposis panel, and relatives by targeted Sanger sequencing [15].

In France, the 2021 PNDS requires genetic testing to be prescribed in hospital by a clinical geneticist or by an organ specialist from a reference or competence centre, who explains its scope to the patient and family, on blood or saliva [27]. The first line is a next-generation sequencing panel (for example, GenoDENT); exome or genome sequencing is reserved for second line, after a negative panel, or for first line when the clinical picture is broad and does not point to a diagnosis, and targeted sequencing of a single gene is justified only by a suggestive dental signature [27]. The protocol on one occasion describes the GenoDENT panel as “DNA chip analysis”, in contradiction with its description as a sequencing panel; the latter prevails. Genome sequencing is decided with the family, the O-Rares network and the local genetics clinic, and is accessible under the France Genomic Medicine Plan 2025 through the pre-indication “syndromic forms of rare diseases with oral and dental expression” or through the pre-indications of the associated syndromes [27]. Analysis is usually performed as a trio, and variants of uncertain significance are reported as such and may be reinterpreted as techniques and bioinformatics evolve [27].

5.6. Diagnostic yield

The yield for AXIN2 is low in all the populations studied, and what each figure measures depends on the population tested.

The discrepancy between the OR in the abstract (11.89) and that in the text and Table 2 (11.39) of Leclerc and colleagues is internal to the paper; the confidence interval is identical in both versions. Direct calculation from the published counts (8/3322 versus 12/59,095) gives an odds ratio of about 11.9, closer to the value in the abstract [7]. The frequency of 0.25% mentioned in the discussion of the same paper rounds the value of 8/3322 [7].

5.7. Limitations of testing

A negative panel result excludes neither the syndrome nor the other polyposes. Copy number analysis may not cover non-coding regions: in the Rowell series, one of the APC variants identified on repeat testing was located in the 5′UTR, not covered by the earlier deletion and duplication analysis [69]. Variants that alter splicing may be misinterpreted by protein-level prediction, as shown by c.2405G>C, predicted as missense p.(Arg802Thr) but which creates an alternative transcript with a frameshift [14, 35]. Somatic mosaicism confined to the colon is not detectable in blood: the European guideline estimates APC mosaicism in 20%–50% of unexplained polyposes with more than 20 adenomas, often detectable only in DNA from more than two adenomas [26], and the Murakami case showed the same phenomenon for AXIN2 [9]. Missense variants and some truncating variants upstream of the β-catenin-binding domain remain of uncertain or conflicting significance, and family segregation and functional and RNA studies are the tools for reclassifying them [7, 15]. Interpretation of population frequency has particular features in AXIN2: c.1994dup is present in gnomAD at a higher frequency than expected for a variant of a rare syndrome [44].

5.8. Report content

The report should state the gene, the transcript with version (NM_004655.4), the variant in HGVS (Human Genome Variation Society) notation at the coding DNA and protein levels, the genomic coordinate with the assembly, the zygosity, the ACMG/AMP class with the criteria applied, the coverage of exon 8 and the regions analysed and the method for detecting copy number variants; for variants of uncertain significance, it should propose segregation and RNA studies [7, 13]. When the report numbering uses the earlier transcript NM_004655.3, as in several publications [14, 24, 44], conversion to the MANE transcript should be performed with a validation tool.

5.9. Practical algorithm

Prenatal diagnosis and preimplantation genetic testing are covered in section 8.

6. Differential Diagnosis

6.1. Adenomatous polyposes and hereditary colorectal cancer

The colonic phenotype of AXIN2 (attenuated polyposis of adult onset, with associated serrated and hyperplastic polyps) overlaps with that of MAP and attenuated FAP, and only a multigene panel reliably distinguishes the entities [7, 26]. The table brings together the differential diagnoses with data from the reference sources.

AD, autosomal dominant; AR, autosomal recessive.

The most useful clinical distinction between AXIN2 and APC lies in the dentition: FAP is associated with supernumerary teeth, odontomas and impacted teeth, and the AXIN2 syndrome with agenesis [7, 22]. Tooth agenesis does not, however, exclude other polyposes: a patient with MAP, multiple adenomas and severe tooth agenesis was described in France, and hypodontia is frequent in the general population [16, 33]. In the French series, the AXIN2 phenotype was considered very close to that of MAP, which is why the authors propose similar surveillance protocols [7]. In an estimate prior to the inclusion of AXIN2 in panels, an attenuated FAP phenotype with demonstrated dominant inheritance was explained by APC in about 10% of families and remained unexplained in about 90%; without demonstrated dominant inheritance, MUTYH explained about 25% [41]. Defined Mendelian syndromes explain less than 4% of colorectal cancers, and about 20% occur in people with multifactorial susceptibility without evident familial clustering, according to the OMIM synthesis [18]. The table in the ESMO (European Society for Medical Oncology) handbook of cancer genomics on polyposes and hereditary colorectal cancer does not include AXIN2 [73].

6.2. Syndromic tooth agenesis and ectodermal dysplasias

GeneReviews organises the differential diagnosis of apparently isolated agenesis into allelic disorders of genes also involved in the nonsyndromic form, genes with preliminary data on isolated agenesis, and syndromes with ectodermal findings without characteristic dysmorphism or cognitive impairment [16]. The table reproduces this organisation and adds the tankyrase-binding domain disorder.

AD, autosomal dominant; AR, autosomal recessive.

X-linked hypohidrotic ectodermal dysplasia is the most common ectodermal dysplasia: affected males have reduced or absent sweating, dry skin, sparse hair and missing and conical teeth, with a risk of hyperpyrexia in infancy if the condition is not recognised [64]. Agenesis ranges from hypodontia to anodontia, with a mean of nine permanent teeth present, and 60%–80% of female carriers are identifiable by conical teeth; EDA accounts for about 70% of cases, EDAR 10%–15%, EDARADD 1%–2% and WNT10A 5%–6% [88]. Witkop syndrome has variable hypodontia, spaced and conical teeth and small, dysplastic nails, with normal sweating [64]. In Axenfeld-Rieger syndrome, PITX2 accounts for 40%–55% of cases and almost 80% when there are dental or umbilical defects, and FOXC1 for 8%–25%, with earlier glaucoma in FOXC1 variants [85]. In the AXIN2 syndrome, ectodermal features are usually limited to sparse hair and eyebrows, without abnormal sweating, and the family cancer history is the distinguishing clue [3, 7].

The OMIM entries add discriminating elements to this differential. In Witkop syndrome, the nail changes are more evident in childhood and may go unnoticed in adults, so the history of the nails in childhood is useful; the syndrome has been described frequently among Dutch Mennonites in Canada [74]. In odonto-onycho-dermal dysplasia, the most specific criterion is severe oligodontia of the permanent dentition with a normal or less affected primary dentition, and the porocarcinoma described in a patient with Schöpf-Schulz-Passarge syndrome led the authors to accept an increased risk of skin tumours in the biallelic WNT10A forms [75, 76]. In the LTBP3 dental anomalies and short stature syndrome, agenesis is accompanied by amelogenesis imperfecta and, in some families, by aortic aneurysms and dissection, which makes cardiovascular assessment part of the diagnosis and not only of follow-up [37]. In cleft lip and palate with ectodermal dysplasia caused by NECTIN1, the Margarita Island (Venezuela) form shows a founder effect, with an estimated heterozygote frequency of about 1 in 22 in the native population, and heterozygosity for the founder variant has been associated with nonsyndromic cleft lip with or without cleft palate in northern Venezuela (orofacial cleft 7) [87]. In GAPO syndrome, optic atrophy is not constant, and glaucoma and keratopathy may replace it [77]. In all these entities, the absence of a cancer history and the presence of extradental signs point away from AXIN2.

The tankyrase-binding domain disorder is distinguished from the classic syndrome by developmental delay, microcephaly and malformations of the limbs, eyes and genitourinary tract, and from the ectodermal dysplasias by these same malformations [10]. In an oral cleft with a family history of cancer, CDH1, responsible for hereditary diffuse gastric cancer, should be considered in addition to AXIN2 [15]. The patterns of agenesis described in section 3.1 (second premolars and lower incisors in AXIN2, second molars in PAX9, sparing of canines in MSX1) guide the choice of test, but the authors of the systematic review consider that their usefulness for risk stratification requires prospective validation [25].

The French PNDS adds to the differential the exclusion of non-genetic causes (traumatic or infectious tooth loss, multiple extractions, antineoplastic treatments and other exposures during odontogenesis) and a list of syndromes with oligodontia not included in the table: trisomy 21, blepharo-cheilo-dontic syndrome, Johanson-Blizzard syndrome, ulnar-mammary syndrome, Wolf-Hirschhorn syndrome, Williams-Beuren syndrome, Floating-Harbor syndrome, craniosynostoses, oral-facial-digital syndromes and chromosomal anomalies [27]. It groups the ectodermal dysplasias, more than 170 entities with a combined frequency of about 7 per 10,000 births, by their extra-ectodermal signs: major skeletal involvement (TP63 syndromes, cranioectodermal dysplasia, tricho-dento-osseous syndrome, ectodermal dysplasia with clefting), corneal involvement and deafness, endocrinopathies or retinal degeneration, atopic background and immunodeficiency [27]. The ectodermal dysplasias with immunodeficiency, caused by hypomorphic IKBKG (NEMO) variants or hypermorphic variants in the IκBα gene (NFKBIA), may be associated with osteopetrosis and lymphoedema, hypogammaglobulinaemia, hyper-IgM syndrome, NK cytotoxicity deficiency or autoimmunity, and may present with severe respiratory, opportunistic or mycobacterial infections; their recognition changes the dental plan, because it may contraindicate implants and bone grafts and requires antibiotic prophylaxis (section 7.7) [27]. Among the features associated with genes of the class, the protocol's tables include, for GREM2, hypohidrosis, periorbital hyperpigmentation, mitral valve prolapse and short stature, and, for LTBP3, increased bone density, platyspondyly, short stature and amelogenesis imperfecta [27].

7. Treatment and Risk Management Strategies

7.1. Principles

There is no targeted pharmacological therapy for the syndrome and no interventional trial in carriers of AXIN2 variants [12]. Management rests on three pillars: endoscopic surveillance with removal of polyps, colonic surgery when the polyp burden can no longer be controlled endoscopically, and oral rehabilitation of tooth agenesis [16, 26, 28]. Because AXIN2-specific data are scarce, the 2024 European guideline recommends applying the MAP recommendations to the rare polyposes, including AXIN2 polyposis, and the French series proposes the same because of the phenotypic similarity [7, 26]. The following recommendations for colonic management are therefore extrapolations from MAP and attenuated FAP, and should be presented to the patient as such.

7.2. Endoscopic management

Systematic polypectomy during surveillance colonoscopy is the basis of colorectal cancer prevention. In MAP, with a phenotype similar to attenuated FAP, the European guideline prefers endoscopic resection to surgery (recommendation MAP.LGM.4A; 86% agreement) and adapts the colonoscopy interval, between 1 and 2 years, to the polyp burden (MAP.LGM.3; 97%) [26]. The NCCN, according to the two secondary sources consulted, recommends colonoscopy every 1–2 years in AXIN2 carriers when polyps are present and individualised surgical assessment when the polyp burden cannot be controlled [28, 29]. In the duodenum and stomach, the European guideline recommends, in MAP, removal of polyps regardless of size (MAP.UGM.3; 100%) [26]. In polyposis due to segmental somatic mosaicism, endoscopic submucosal dissection (ESD) of the caecum, repeated at 3 months, followed by cold snare polypectomy every 6 months, allowed the colon to be preserved, with no recurrence at 3 years [9]; the authors suggest this colon-preserving strategy only for selected patients with segmental polyposis without advanced neoplasia.

7.3. Surgery

The surgical decision is individual and should be discussed in a multidisciplinary team, taking into account the polyp burden in the colon and rectum, age, comorbidities, patient preferences and adherence to surveillance (MAP.LGM.4B) [26]. In MAP, the European guideline recommends colectomy with ileorectal anastomosis as the first option, reserves proctocolectomy with ileal pouch–anal anastomosis (IPAA) for dense rectal polyposis that cannot be controlled (MAP.LGM.4C), advises against prophylactic surgery in carriers without polyps or cancer (MAP.LGM.4D), considers that total colectomy may be preferable to segmental colectomy when colorectal cancer is present (MAP.LGM.5) and recommends pouch surveillance after IPAA (MAP.LGM.6B) [26]. In FAP, prophylactic colectomy is typically performed between 16 and 20 years of age, with ileorectal anastomosis in most patients and restorative proctocolectomy in cases with dense polyposis [41]; this scheme does not apply to AXIN2, in which polyposis appears in adulthood.

The surgical procedures reported in carriers illustrate the variability of decisions. In the Marvin family, the mother underwent sigmoidectomy at 60 years and right hemicolectomy at 62 for more than 100 adenomas [3]. In the Roht cohort, a Norwegian carrier underwent subtotal colectomy at 70 years for polyposis and was still undergoing rectoscopy at 89, and a Dutch carrier underwent subtotal colectomy at 62 years for caecal adenocarcinoma [15]. For a 65-year-old patient with 57 adenomas removed over four colonoscopies, Macklin-Mantia and colleagues presented two options, laparoscopic subtotal colectomy with flexible rectosigmoidoscopy every 6–12 months or colonoscopy with polypectomy every 6–12 months, and the patient chose the latter [44]. These approaches are case reports and do not replace guideline recommendations.

7.4. Precautions for invasive procedures

The French series recorded severe bleeding after polypectomy in two carriers, one of whom was admitted to intensive care, antral angiodysplasia with a tendency to gastrointestinal bleeding in a third, and idiopathic myopathy in three members of one family [7]. There is no established causal relationship, but the authors recommend taking these findings into account before invasive procedures [7]. In practice, this means asking about a bleeding history before extensive polypectomies and planning haemostasis; the sources consulted do not propose a specific protocol.

7.5. Chemoprevention

None of the sources consulted recommends chemoprevention in AXIN2 carriers, and there are no trials in this population. The 2026 ACG guideline includes a section on the role of chemoprevention in adenomatous polyposis syndromes, but the full text was not accessible, so it is not known whether it addresses AXIN2 [66]. Trial NCT02394769 (ASPIRED), with acetylsalicylic acid after colorectal cancer in the general population, was the only colon-related ClinicalTrials.gov record returned by the search for “AXIN2” and is not aimed at carriers (accessed 28 October 2026).

7.6. Treatment of cancers

Colorectal and extracolonic cancers in carriers are treated according to the standards for the tumour type and stage; the sources consulted describe no differences in response associated with AXIN2. The two colorectal carcinomas in the French series with reported MMR status were proficient [7]. In the Roht cohort, a carrier with metastases from caecal adenocarcinoma received pembrolizumab, with no indication of the MMR or microsatellite instability status of the tumour [15]; MMR testing of the tumour is necessary to decide on immunotherapy and to exclude coexisting Lynch syndrome. The carrier with olfactory neuroblastoma underwent endoscopic resection and adjuvant radiotherapy of 60 Gy in 30 fractions (section 10) [24]. In choosing the extent of colectomy for colorectal cancer, recommendation MAP.LGM.5 of the European guideline applies [26].

7.7. Oral rehabilitation and caries prevention

GeneReviews considers that there is no standard of care or clinical practice guideline for nonsyndromic tooth agenesis and proposes a manifestation-based approach, which applies equally to the AXIN2 syndrome [16]. In France, the November 2021 HAS PNDS on multiple tooth agenesis (oligodontia and anodontia, isolated or syndromic) is a national protocol based mainly on expert consensus of the O-Rares network, because there are few high-level evidence studies, and it expressly includes nonsyndromic AXIN2 oligodontia among the indications for early implants [27].

Source: [16]. In the ectodermal dysplasias, the Oxford Desk Reference mentions removable prostheses from 2 years of age, with several replacements during growth and later implants, and specialist orthodontic follow-up [64]. Assessment of feeding and chewing is recommended in children with oligodontia, with referral to a feeding specialist according to severity [16].

GeneReviews does not state fluoride concentrations or doses. The 2013 recommendations of the American Dental Association (ADA) Council on Scientific Affairs for people at risk of caries are the reference used, and their application to tooth agenesis is an extrapolation, because there are no data specific to this population [89].

Sources: [16, 89]. Before 6 years of age, the ADA panel considered that the risk of swallowing the gel and mouthrinse could outweigh the benefit [89]. The sources consulted do not state the trade names or the authorisation status of these products in Portugal or Brazil.

Rehabilitation by age according to the French PNDS. The aims are to preserve the teeth present and the bone stock, restore chewing, speech and aesthetics, accompany maxillomandibular growth and limit the psychosocial impact; treatment is long, coordinated by a reference or competence centre, and may be discussed at a multidisciplinary team meeting [27]. The preventive measures of the protocol coincide with those of GeneReviews: dietary advice, fluoride varnish, pit and fissure sealants and six-monthly reviews during the transitional rehabilitation phase [16, 27].

Source: [27].

Early implants. The PNDS draws on the HAS assessments of 2004 and 2006, which accepted, in oligodontia due to a rare disease, 2 implants (maximum 4) only in the anterior region of the mandible, between 6 years of age and the end of growth, after failure of or intolerance to a conventional prosthesis [27]. The criteria are: severe oligodontia, defined as at least 6 missing permanent mandibular teeth (not counting the third molars), or mandibular anodontia; minimum age of 6 years, for anatomical reasons and neuropsychological maturity; failure of or intolerance to a conventional resin prosthesis; and ectodermal dysplasia or nonsyndromic oligodontia due to genes such as MSX1, PAX9 or AXIN2, without a molecular diagnosis being a mandatory condition [27]. Two symphyseal implants are sufficient for an implant-stabilised complete denture; four are justified only for a fixed bridge, a solution that the protocol considers outside the HAS recommendations [27]. The recommended technique uses lengths of 10–13 mm (minimum of 8 mm in a symphysis with little bone) and diameters of 3.0–3.75 mm, avoiding the more fragile 2.5–2.75 mm implants and preferring conical connections that can be used in adulthood; planning is by cone-beam tomography with a surgical guide, and surgery is under general anaesthesia in most children, with slow drilling and irrigation, a two-stage surgical protocol, 4 months of osseointegration, micromovement below 100 µm and single ball attachments, which tolerate the changes of growth [27]. The most frequent complication is relative submergence of the implant, because the ankylosed implant does not follow vertical alveolar growth; it is resolved by a minor procedure or by adapting the abutments, and reintervention is not considered a complication; failure of osseointegration leads to removal of the implant without bone resection [27]. According to the literature reviewed by the protocol, about 20% of treated patients had received implants before 18 years of age [27].

Adults and anodontia. In adults, the first-line solution is implant-prosthetic (fixed prosthesis or implant-supported complete denture), often with bone or mucosal grafts and other pre-implant techniques (distraction, Le Fort I osteotomy, nerve lateralisation) [27]. In mandibular anodontia good results have been described with 4, 6 or 8 implants; in total anodontia in adults, the treatment of choice is a fixed prosthesis on 6–8 maxillary and 4–6 mandibular implants, with bone augmentation of both arches, and alternatives include bars on 8 maxillary and 6 mandibular implants (provided there is no neuromuscular deficit preventing handling of the prosthesis), zygomatic implants or, as a last resort, mucosa-supported complete dentures [27].

Orthodontics. Orthodontics is involved at every stage: interception and early orthopaedics for transverse or sagittal skeletal discrepancy; expansion and bone anchorage plates in severe forms in adolescence; preservation of primary molars without a successor and without severe infraocclusion (especially the mandibular second primary molar); planning of extraction of ankylosed primary teeth; skeletal anchorage with mini-screws when anchorage teeth are missing; and preparation of a mesiodistal space of 6 mm, coronal and radicular, for an upper lateral incisor implant [27]. Retention maintains the spaces until the age for implants; in adults, orthognathic surgery is preceded by provisional rehabilitation that stabilises the arches [27]. Re-education of orofacial dysfunctions (tongue interposition, swallowing) should precede rehabilitation, and frenulum surgery may be necessary [27].

Antibiotic prophylaxis in implant surgery. The 2021 PNDS considers antibiotic prophylaxis mandatory for early implant placement in children, without specifying drug or dose [27]. The more recent HAS recommendation of July 2026 on antibiotic prescribing in dentistry does not recommend systematic antibiotic prophylaxis for implant placement in the general population (grade B), accepts it according to the patient profile (history of osseointegration failure, poorly vascularised bone) and the complexity of the procedure (more than 3 implants, extraction with immediate implantation, duration over 1 h 30) and recommends it for pre- or peri-implant bone grafts (grade C), sinus lift and patients at increased infectious risk, including congenital immune deficiencies (expert consensus) [90]. For all indications of a single preoperative dose, the recommended dose, outside the marketing authorisation, is oral amoxicillin 30–60 minutes before the procedure: 2 g in adults and 50 mg/kg in children, not exceeding 2 g [90]. In proven penicillin allergy, contraindication or stock shortage, the alternatives in order of preference are clarithromycin 500 mg (children 15 mg/kg, maximum 500 mg), azithromycin 500 mg (children 20 mg/kg, maximum 500 mg) or pristinamycin 1 g (children 25 mg/kg, maximum 1 g) [90]. By the parenteral route, between 60 minutes before and the incision, amoxicillin 2 g IV (children 50 mg/kg, maximum 2 g) is used, with cefazolin 1 g IV/IM (children 50 mg/kg, maximum 1 g), ceftriaxone 1 g IV/IM (children 50 mg/kg, maximum 1 g) or clindamycin 900 mg IV (children 20 mg/kg, maximum 900 mg) as alternatives; if the dose is forgotten or the procedure is urgent, the antibiotic may be given up to 2 hours afterwards [90]. Application of these regimens to early implantology in oligodontia is an extrapolation, because the 2026 recommendation does not address this population specifically. Where the two documents diverge (systematic prophylaxis in the PNDS, individualised in the 2026 recommendation), the more recent recommendation, specific to antibiotic use, is the reference; in ectodermal dysplasias with immunodeficiency, prophylaxis is indicated in both [27, 90].

7.8. Regulatory status

The absence of records in the search performed does not demonstrate the absence of processes in other databases.

7.9. Clinical trials

There are no interventional trials, so it is not possible to present efficacy estimates with confidence intervals for any management measure. The surveillance and surgical recommendations result from expert consensus and extrapolation from MAP and attenuated FAP, with evidence rated as low by the European guideline, and the case series do not allow strategies to be compared [15, 26].

7.10. Developments in the last 12 months (November 2025 to October 2026)

No therapies in development, regulatory applications or authorisations related to the syndrome were identified.

7.11. Special populations

Pregnancy and fertility. The sources consulted do not describe pregnancies in female AXIN2 carriers. By extrapolation from FAP, the European guideline found no convincing evidence of differences in fertility between IPAA and ileorectal anastomosis in women with FAP (9/49 and 9/51 with fertility problems in the largest retrospective study), although a historical meta-analysis, mainly of patients with ulcerative colitis, estimated a relative risk of infertility of 3.91 (95% CI 2.06–7.44) after IPAA [26]. This information is relevant when the type of surgery is discussed with a young woman.

Perioperative care. The history of bleeding at polypectomy and the myopathy described in the French series should be taken into account in anaesthetic and surgical planning [7]. According to GeneReviews, dental implants and orthognathic surgery are performed only after the end of growth [16]; the French PNDS accepts mandibular symphyseal implants from 6 years of age under strict criteria (section 7.7) and indicates, for implant surgery in children, general anaesthesia as first choice depending on age and cooperation, so that an anaesthetic contraindication calls early implantology into question [27]. In the ectodermal dysplasias with immunodeficiency, implants and grafts may be contraindicated [27].

Children and transition. Children with oligodontia need paediatric dentistry, assessment of chewing and orthodontic and prosthetic planning; transition to an adult dentist takes place at about 14 years of age according to GeneReviews and between 15 and 20 years according to the French PNDS [16, 27]. Transition to colorectal surveillance occurs between 18 and 30 years of age, depending on the scheme adopted (section 11). Children with the tankyrase-binding domain disorder need multidisciplinary developmental, ophthalmological and urological follow-up and, in adulthood, colorectal surveillance [10].

Agents to avoid. The sources consulted identify no drugs specifically contraindicated in the syndrome. The Marvin proband developed fundic gland polyps during treatment with a proton pump inhibitor, which should be taken into account when interpreting upper endoscopy [3].

Psychological support and therapeutic education. The French PNDS recommends assessing oral health-related quality of life with subjective data in addition to dental indices and offering psychological and medico-social support to the patient, parents and siblings from the disclosure of the diagnosis and throughout follow-up; it adopts the American Academy of Pediatric Dentistry recommendation of a psychological assessment at each stage of treatment [27]. Therapeutic education begins with an individual educational diagnosis and continues in individual or group sessions, open to parents, on the disease, adaptation of oral hygiene and diet, the care pathway, risk situations (gingival inflammation, fracture or debonding of a prosthesis or a bracket) and psychological aspects, with a final assessment and booster sessions; in France there is a specific programme for rare dental diseases, DentO-RarEduc, coordinated by the Strasbourg centre [27].

7.12. Guideline recommendations and divergences

The main divergence is the age at which colonoscopy starts: 25–30 years in the NCCN, according to the secondary sources, and 18 years in the European guideline by extrapolation from MAP [26, 29]. The second divergence is upper endoscopy, absent from the NCCN and recommended by the European guideline (by extrapolation) and by the ERN GENTURIS experts [15, 26]. The original NCCN text and that of the 2026 ACG guideline were not read, so the most recent NCCN version may have changed these values.

On the dental side, the French PNDS diverges from GeneReviews on five points: the age for implants (GeneReviews: only after the end of skeletal growth; PNDS: mandibular symphyseal implants from 6 years of age under strict criteria); the removable prosthesis (GeneReviews: 6–18 years; PNDS: from 2–3 years, depending on cooperation); the first orthopantomogram (GeneReviews: confirmation after 6 years of age; PNDS: from 4–6 years if absence of tooth germs is suspected); transition to an adult dentist (GeneReviews: about 14 years; PNDS: between 15 and 20 years); and the frequency of visits (GeneReviews: twice a year; PNDS: six-monthly before 6 years of age and during the phases of dentition change, at least annually after 6 years of age and in adults) [16, 27]. The PNDS considers antibiotic prophylaxis mandatory in early implantology, whereas the 2026 HAS recommendation on antibiotics in dentistry individualises it for implant placement and reserves it for grafts, sinus lift and patients at increased infectious risk (section 7.7) [27, 90]. None of the dental documents addresses colorectal surveillance, which follows the gastrointestinal guidelines.

8. Reproductive Options and Family Counselling

8.1. Inheritance and recurrence risks

The AXIN2 syndrome is transmitted in an autosomal dominant manner: each child of a carrier has a 50% chance of inheriting the variant, regardless of sex [16, 28]. Variable expressivity and incomplete penetrance mean that a carrier child may have only hypodontia, isolated oligodontia, polyposis without agenesis or the complete picture, and that a carrier parent may appear unaffected [15, 16, 35]. An apparently negative family history may result from unrecognised agenesis in adults, reduced penetrance or variable expressivity, and only clinical and dental assessment of the parents or their molecular testing clarifies it [16].

In an isolated case with a molecular diagnosis, the parents are tested. If the variant is absent in both, it arose de novo (once parentage has been confirmed) or there is parental germline mosaicism; germline mosaicism is theoretically possible, and leucocyte DNA may not detect it [16]. De novo variants have been described: c.1994dup in a 13-year-old boy [2] and the tankyrase-binding domain variants in the five individuals reported in 2026 [10]. For the siblings of a proband, the risk is 50% when a parent is a carrier; when the variant is not detected in the parents' leucocytes, the risk is about 1% owing to theoretical germline mosaicism; when the parents are clinically unaffected and their genetic status is unknown, the risk is low, but higher than that of the population [16]. Somatic mosaicism confined to the colon, as described by Murakami and colleagues, does not involve the germline, so it carries no risk for offspring; relatives do not need testing targeted at that variant [9].

For the other forms of the tooth agenesis class, the risks depend on the gene. In autosomal recessive WNT10A inheritance, the parents are presumed heterozygous, siblings have a 25% chance of being biallelic (often with oligodontia), 50% of being heterozygous (asymptomatic or with hypodontia) and 25% of not being carriers, and all offspring of a biallelic patient are obligate heterozygotes [16]. In X-linked EDA inheritance, the mother of an affected male is an obligate carrier if there is more than one affected individual in the family; a carrier mother transmits the variant in 50% of pregnancies (affected sons and heterozygous daughters with variable manifestations), and an affected male transmits it to all his daughters and none of his sons [16]. In female carriers of X-linked hypohidrotic ectodermal dysplasia, the starch–iodine sweat test reveals mosaicism along the lines of Blaschko [64].

8.2. Cascade testing

The European guideline recommends offering testing to all first-degree relatives in autosomal dominant polyposis syndromes, in cascade (OAPS.2B; low evidence; 97% agreement) [26]. Testing of relatives has a dual purpose: it identifies those who need colorectal surveillance and allows dental care to start early in the youngest, with intraoral and extraoral examination and radiographs [16, 26]. In the French series, only 10 relatives were tested, which the authors attribute to the absence of guidelines [7]. Non-carriers in a family with a known variant follow general population recommendations, and carriers enter the surveillance scheme of section 11.

The French PNDS recommends, after the molecular diagnosis in the index case, study of the parents to confirm segregation, usually as a trio, and then testing of symptomatic relatives and at-risk adult relatives, in accredited medical genetics laboratories, with the support of a genetic counsellor at disclosure and counselling [27].

8.3. Testing of minors

Colorectal surveillance starts at 18 years of age (European guideline, by extrapolation) or at 25–30 years (NCCN, according to the secondary sources), so predictive testing of asymptomatic minors does not change colonic management in childhood [26, 29]. It may be justified earlier when there is a clinical indication: oligodontia requiring orthodontic and prosthetic planning, cleft palate or a syndromic picture [15, 16]. In the families studied by Roht and colleagues, the children were tested on clinical indication and surveillance was planned from 18 years of age [15]. Counselling of affected or at-risk young adults should include the surveillance plan and reproductive options [16].

8.4. Prenatal and preimplantation diagnosis

Once the familial variant has been identified, prenatal testing and preimplantation genetic testing are technically possible; the risk should be discussed before pregnancy, and views on the use of prenatal testing vary among professionals and families, particularly for a condition of adult onset, with incomplete penetrance and with surveillance available [16]. Prenatal diagnosis of the tankyrase-binding domain disorder is a different matter: the variants are de novo, the congenital phenotype is severe, and the recurrence risk for the parents of an affected child is low, but the authors do not quantify it and germline mosaicism cannot be excluded [10, 16].

The French PNDS does not address prenatal diagnosis or preimplantation genetic testing; at the transition to adulthood, it recommends explaining the mode of transmission again and assessing the risk of recurrence in a pregnancy [27].

8.5. Content of counselling

Counselling should explain that tooth agenesis and predisposition to colorectal cancer are part of the same condition, that dental severity does not predict colonic risk, that the cancer risk has not been quantified and that the extracolonic cancers described have no estimated penetrance [7, 21, 29]. It should also state the need to share the result with relatives and the availability of patient associations and European Reference Networks [28, 30]. In people with an unexplained oral cleft, taking a family cancer history is recommended [15], and in families with colorectal neoplasia, asking about the dentition and ectodermal features [3].

9. Special Considerations: Population Diversity and Epidemiology

9.1. Frequency of the syndrome and what each figure measures

There is no validated population prevalence estimate for the AXIN2 syndrome. Orphanet assigns ORPHA:401911 to the class <1/1,000,000, based on a small number of published families; it is an upper-bound class and not a measurement [4, 12]. The NCI PDQ, in its revision of 21 March 2025, considers the prevalence of pathogenic variants and their contribution to polyposis and colorectal cancer uncertain [29]. The same PDQ mentions eight families from Australia and New Zealand, attributed to the study by Chan and colleagues, whose abstract describes four families with pathogenic variants; the difference cannot be resolved with the sources consulted [29, 35].

The proportion of 0.020% in gnomAD controls measures loss-of-function variants, not patients, and is not a prevalence of the syndrome; the proportion of 0.24% applies only to patients already selected for suspected predisposition [7].

9.2. Epidemiology of tooth agenesis

The prevalence of nonsyndromic tooth agenesis ranges from 1.6% to 36.5% depending on the population and the inclusion of third molars; in most studies, hypodontia affects 3%–10% and oligodontia 0.1%–0.5% of the population, excluding third molars [16]. More than 20% of people lack one or more third molars, and more than 5% lack other permanent teeth, most often second premolars or upper lateral incisors [2]. The review by Al-Ani and colleagues gives a prevalence of hypodontia of 1.6%–6.9%, excluding third molars, and cites the Polder meta-analysis (33 studies) with 5.5% in Europe and 6.3% in Australia, with lower values in North America; in about 80% of people with hypodontia only one or two teeth are missing, and primary agenesis occurs in less than 1% of Caucasians [92]. GeneReviews gives 1% for North America, a value that the Al-Ani review does not confirm [16, 92]. In the white population, hypodontia has been estimated at 2.2%–10.1% and oligodontia at about 0.14% [14]. In the Danish registry of 2.5 million live births, 2.8% had recorded tooth agenesis, with a mean age at diagnosis of 13.2 ± 4.1 years [62]. Agenesis is more frequent in women, with a ratio of 3:2, and in Asians, followed by people of northern European origin and African Americans; agenesis of the lower incisors is more frequent in Asians, and in a Japanese population of orthodontic patients the prevalence was 8.5% (selected population) [16]. The OMIM synthesis estimates the frequency of some form of agenesis in the population at about 20%, counting third molars, whose absence affects a third or more of people in many populations; outside the third molars, the most affected teeth are the upper lateral incisors and the lower second premolars, and agenesis of the first and second molars is rare [19]. Nieminen and colleagues placed hypodontia of incisors and premolars at 5%–10% in European and Asian populations, and Witkop estimated the frequency in the USA of the trait of small, peg-shaped or missing upper lateral incisors at about 1.5% [56]. These values, from different periods and methods, measure distinct dental traits and are not comparable with one another.

The 2021 French PNDS presents values compiled from the literature, without its own data from France: tooth agenesis in 1.6%–9.6% depending on the study and 4.34% as a pooled value, with no difference between the sexes; hypodontia in about 5% in the permanent dentition (Polder meta-analysis) and 0.4%–0.9% in the primary dentition; agenesis of third molars in about 20%, of second premolars in 2.91%–3.22% and of upper lateral incisors in 1.55%–1.78%; oligodontia in 0.14% [27]. The absence of a sex difference does not match the Polder meta-analysis that the protocol itself cites: in the Caucasian populations analysed, the prevalence of agenesis was 1.37 times higher in women, with 4.6% in men and 6.3% in women in Europe, 5.5% and 7.6% in Australia and 3.2% and 4.6% in North American Caucasians, which agrees with the 3:2 ratio in favour of women given by GeneReviews [16, 27, 93]. The protocol considers that oligodontia, although it may be nonsyndromic, is more often associated with a syndrome, even one of mild expression, and that anodontia is always syndromic; the most frequently associated syndromes are the ectodermal dysplasias, incontinentia pigmenti, Axenfeld-Rieger syndrome and colorectal cancer predispositions [27]. In the summary for the general practitioner, the percentages of missing teeth (premolars 15%–32%, upper lateral incisors 27%, third molars 25%, canines and first and second molars 1%, upper central incisors 0.05%) do not state the denominator, so they are not comparable with the population prevalences above [27].

9.3. Geographical distribution and founder populations

The published families come from Finland (founder variant c.1966C>T), the USA, Spain, France, Australia, Denmark, Sweden, Norway, Estonia, the Netherlands and Japan [2, 3, 7–9, 14, 15, 34–36]. The variant c.1994dup is recurrent: it has been described de novo in Finland, in Swedish and Danish families and in a patient from Georgia (USA) with no known link to Scandinavia, and the authors hypothesise common European ancestry of the North American families [36, 44]. The Roht cohort included only Caucasians (11 Europeans and 2 North Americans of probable European ancestry) [15], and most published patients are of European ancestry; an African American carrier has been published [59], and the Dutch patient had mixed European and Chinese ancestry [14]. The sources consulted contain no data from Portugal or Brazil on the syndrome; from Brazil there is only the cohort of 116 patients from Rio de Janeiro with tooth agenesis included in the association study of common AXIN2 polymorphisms (section 2.9), which does not describe carriers of pathogenic variants [17, 46]; the lead author of the GeneReviews on tooth agenesis is Brazilian, but the text presents no Brazilian data [16]. This absence in the sources consulted does not demonstrate that there are no cases in these countries. In the differential diagnosis of the polyposes, the APC p.Ile1307Lys variant (8% of Ashkenazi Jews) and the duplication upstream of GREM1 are examples of a founder effect in Ashkenazi Jews [41, 42].

9.4. Special populations

In children, the syndrome manifests as tooth agenesis and, rarely, as cleft palate, with no polyps described in childhood in the series consulted [7, 15]. In older people, polyposis may appear late, as shown by the diagnoses at 78 and 82 years in the French series and the mosaic polyposis at 77 years [7, 9]. In women, breast and ovarian cancers have been described, without quantified risk [25].

9.5. Reference networks and patient associations

ERN GENTURIS (https://www.genturis.eu) covers the hereditary polyposis syndromes, and the multicentre study by Roht and colleagues was carried out within it [15, 30]. ERN CRANIO (https://www.ern-cranio.eu) covers craniofacial anomalies, including dental anomalies [31]. Patient associations include the National Foundation for Ectodermal Dysplasias (NFED, USA; https://www.nfed.org), the Ectodermal Dysplasia Society (United Kingdom), the Colorectal Cancer Alliance (USA; https://www.ccalliance.org) and, among the GeneReviews resources, NORD (National Organization for Rare Disorders, “Tooth Agenesis” page) and the Parent to Parent network [16, 64, 94, 95]. The Mayo Clinic authors mention building a registry of patients with AXIN2 variants [14, 24].

Organisation in France. The 2021 PNDS was coordinated by the reference centre for rare oral and dental diseases O-Rares (coordinating site at the Strasbourg University Hospitals; constituent site at the Rothschild Hospital, Assistance Publique–Hôpitaux de Paris), supported by 16 competence centres (Angoulême, Besançon, Bordeaux, Dijon, Lyon, Marseille, Montpellier, Nancy, Nantes, Créteil, Pitié-Salpêtrière, Reims, Rennes, Rouen, Toulouse and Tours) and by the TETECOU health filière (malformations of the head, neck and teeth); in the syndromic forms, the FIMARAD (rare dermatological diseases), AnDDI-Rares (developmental anomalies) and OSCAR (rare diseases of bone, calcium and cartilage) filières are involved [27]. Isolated or syndromic oligodontia may be recognised as an off-list long-term condition (ALD 31); the care protocol is requested by a hospital doctor from a reference or competence centre, or by a dentist or stomatologist with the signature of the general practitioner, and includes a recent orthopantomogram and, in boys aged at least 17 and girls aged at least 14, a hand and wrist radiograph for bone age [27]. Health insurance has reimbursed implant-prosthetic procedures for multiple agenesis due to a rare disease since 28 June 2007 in children over 6 years of age until the end of growth (2 to 4 implants in the anterior region of the mandible, with at least 6 missing permanent mandibular teeth and after failure of or intolerance to the conventional prosthesis) and since 9 January 2012 in adults (up to 10 implants, with at least 6 missing permanent teeth, one of which belongs to a defined list of teeth) [27]. Orthodontics started after 16 years of age is not reimbursed, so the request for orthodontic treatment should be made before that age; the protocol describes a high residual cost and forgoing of care for financial reasons, and provides travel support when the centre is more than 150 km away or 4 journeys of at least 50 km are needed within two months [27]. The French associations cited are the Association Française des Dysplasies Ectodermiques (AFDE, https://afde.net), linked to the O-Rares network and TETECOU, and the Association pour la Reconnaissance de l'Agénésie Dentaire (ARAD) [27].

10. Illustrative Clinical Case

Macklin-Mantia and colleagues, from the Mayo Clinic, published in 2020 the case of a 49-year-old woman referred to the otorhinolaryngology clinic for a right nasal mass [24]. For three months she had had progressive nasal congestion, reduced smell and taste, epistaxis, right retro-orbital pain with mild eyelid oedema, occipital pain and pain in the upper gum. Physical examination was normal, except for unilateral cervical lymphadenopathy, and nasal endoscopy showed a grade 3 polyp in the middle meatus. Biopsy suggested a neuroendocrine neoplasm, probably olfactory neuroblastoma (aesthesioneuroblastoma), and magnetic resonance imaging showed tumour in the right nasal cavity, with thickening of the frontal, ethmoid and sphenoid sinuses and contact with the dura mater, the medial orbital wall and the maxillary sinus (figure 1a, https://pmc.ncbi.nlm.nih.gov/articles/PMC7433097/figure/Fig1/) [24].

About three months after the first visit, and six months after symptom onset, she underwent exclusively endoscopic resection (bilateral maxillary antrostomies, total ethmoidectomies, frontal Draf III procedure and dural biopsy, with skull base reconstruction using a nasoseptal flap), lasting 300 minutes, with discharge on the second day without complications [24]. Histology confirmed Hyams grade 3, Kadish stage C olfactory neuroblastoma, with the characteristic “salt and pepper” appearance and strong immunoreactivity for synaptophysin (figure 2, https://pmc.ncbi.nlm.nih.gov/articles/PMC7433097/figure/Fig2/); MRI the following day showed macroscopically complete resection (figure 1b) [24]. She began radiotherapy 30 days after surgery, with 60 Gy in 30 fractions, and imaging 363 days after surgery showed no recurrence [24].

Germline testing, performed within a universal testing programme for cancer patients with an 83-gene panel, identified AXIN2 c.1822del p.(Leu608PhefsTer81) (NM_004655.3, classified at the time as pathogenic) and variants of uncertain significance in NF1 and RAD50 [24]. Only after the result, at the genetics clinic, did the patient report hypodontia: she lacked the upper lateral incisors, the lower second premolars and three third molars, out of a total of 24 teeth, and the upper canines had been veneered to resemble lateral incisors (orthopantomogram, figure 3, https://pmc.ncbi.nlm.nih.gov/articles/PMC7433097/figure/Fig3/) [24]. Eyebrows, hair and nails were normal, and she had no gastrointestinal symptoms or previous endoscopies. Upper endoscopy showed a 4 mm polyp in the gastric body, a foveolar-type gastric adenoma (figure 4, https://pmc.ncbi.nlm.nih.gov/articles/PMC7433097/figure/Fig4/), and colonoscopy a 5 mm hyperplastic polyp in the ascending colon [24]. In the pedigree (figure 5, https://pmc.ncbi.nlm.nih.gov/articles/PMC7433097/figure/Fig5/), her father, not available for testing, lacked several lower teeth and had a malignant cerebellar tumour; a paternal aunt had breast cancer; one of her three daughters had peg-shaped upper lateral incisors [24].

The authors proposed that the gastric adenoma and the olfactory neuroblastoma were part of the phenotype, arguing that gastric adenomas occur in about 10% of patients with FAP and in less than 0.01% of the general population, and suggested interim guidance with upper endoscopy and colonoscopy from 18 years of age, repeated at least every 3 years, annual history and physical examination and referral to otorhinolaryngology with imaging for nasal or facial symptoms [24]. Three aspects limit the generalisation of the case. First, the variant was later reclassified as of uncertain significance and now has conflicting interpretations in ClinVar, so the causal attribution of the olfactory neuroblastoma and gastric adenoma to AXIN2 is uncertain [15, 43]. Second, the authors' interim guidance differs from the guidelines: the NCCN, according to the secondary sources, starts colonoscopy at 25–30 years and does not recommend upper endoscopy, and the European guideline starts colonoscopy at 18 years and upper endoscopy at 35 [26, 29]. Third, the case shows how mild hypodontia in an adult, disguised by dental rehabilitation, was recognised only after genetic testing, which reinforces the value of asking cancer patients and patients with polyposis about their dental history [24, 63].

11. Prognosis and Long-Term Follow-up

11.1. Prognosis

Prognosis depends mainly on the prevention and early detection of colorectal cancer. There are no age-specific cumulative risk estimates for colorectal cancer: the GeneReviews on Lynch syndrome classifies the risk as unknown, with a mean age of 52–62 years at diagnosis of cancer or polyposis [21]. In the available series, colorectal cancer occurred in 5 of 15 French carriers, at a median age of 62 years (58–82), and 40% of the French carriers developed colorectal cancer or another gastrointestinal cancer [7]; in the Finnish family, colorectal neoplasia occurred in 8 of the 11 carriers with oligodontia, including a metastatic adenocarcinoma at 54 years [2]. These values come from selected families and series and do not estimate the risk of a carrier identified by cascade testing (section 3.10). Cancer deaths have been described: a carcinoma of the ampulla of Vater at 53 years and a colorectal cancer diagnosed at 43 years, with death at 46 [7, 24]. The patients under surveillance in the Roht cohort were mostly young or middle-aged and without cancer, and a 95-year-old carrier with subtotal colectomy at 70 years was still undergoing rectoscopy at 89 [15]. In the tankyrase-binding domain disorder, the developmental prognosis has not been described beyond the initial series of five individuals, and the only adult had colon cancer [10]. The dental prognosis of agenesis is outside the scope of GeneReviews; prosthetic and implant rehabilitation restores function in most cases after the end of growth [16].

11.2. Surveillance table

The NCCN colorectal surveillance recommendations for AXIN2 were not confirmed in the original NCCN document: they are known through the NCI PDQ (update of 21 March 2025, which cites version 3.2024) and a laboratory leaflet citing version 1.2025, with the caveat that the supporting data are evolving [28, 29]. The same recommendation, with the same starting age and intervals, had already been cited by Roht and colleagues from an NCCN consultation in March 2021 [15].

11.3. Other published proposals

For comparison, the 2010 British scheme reproduced in the Oxford Desk Reference recommended colonoscopy from 25 years of age every 18–24 months in carriers of MMR variants, and in MAP colonoscopy from 25 years and upper endoscopy from 30 [96]. In the same book, the authors consider the annual sigmoidoscopy scheme of FAP inappropriate in families with attenuated FAP, because polyps appear later, spare the rectosigmoid and vary widely in number, and prefer biennial colonoscopy [41].

11.4. Particular follow-up situations

With a high polyp burden, the options include colonoscopy with polypectomy every 6–12 months or subtotal colectomy with rectal surveillance every 6–12 months [44]. After IPAA, the pouch should be monitored [26]. In segmental mosaic polyposis, endoscopic control every 6 months until eradication allowed the colon to be preserved [9]. In reported practice, carrier children do not undergo gastrointestinal surveillance and start it at 18 years of age [15].

11.5. Transition and coordination

Follow-up involves paediatric dentistry, orthodontics and prosthodontics in childhood and adolescence, transition to an adult dentist at about 14 years of age and entry into the colorectal surveillance programme in adulthood, coordinated by a familial gastrointestinal cancer risk clinic [16, 26]. ERN GENTURIS and ERN CRANIO offer access to centres with experience in both areas [30, 31].

The French PNDS places transition between 15 and 20 years of age, adapted to each patient, with detailed reports between specialists and transfer of images to avoid repeated examinations; at the first adult visit, the history is reviewed, the examination is repeated, the need for follow-up and the reasons for urgent consultation are reinforced, information on social rights is provided, understanding of the disease is checked, transmission and the risk of recurrence are explained again and the treatment plan is adapted [27]. In AXIN2 carriers, this visit coincides with entry into the colorectal surveillance programme (section 11.2).

12. Summary and Concluding Remarks

AXIN2 encodes a scaffold of the β-catenin destruction complex and a feedback brake on the canonical Wnt pathway. Heterozygous germline variants, mainly truncating variants of exon 8, cause an autosomal dominant condition that combines agenesis of permanent teeth, often oligodontia with loss of the second premolars, lower incisors and upper lateral incisors, with attenuated adenomatous polyposis of adult onset and colorectal cancer. ClinGen classifies the gene–disease relationship as definitive, but the cumulative cancer risk remains unknown, because the data come from families and series selected by phenotype, and the extracolonic cancers described (duodenal and ampullary, breast, ovary, melanoma, prostate, olfactory neuroblastoma) have no estimated penetrance. Expression ranges from isolated hypodontia to anodontia and from no polyps to hundreds of adenomas, and there are families with polyposis without agenesis. The spectrum widened in 2026 with polyposis due to somatic mosaicism confined to the colon and with a multisystem neurodevelopmental disorder caused by de novo missense variants in the tankyrase-binding domain, probably with a mechanism different from loss of function.

Tooth agenesis is common in the population (hypodontia in 3%–10%) and is not useful for universal colorectal cancer screening; what has clinical value is the association of polyposis or colorectal cancer with oligodontia, which should lead to a multigene panel including AXIN2, and oligodontia with a family history of colorectal neoplasia, which should lead to clinical genetics. The diagnosis is molecular, on transcript NM_004655.4, and requires attention to the historical exon nomenclature, missense variants of uncertain significance and mosaicism. There is no pharmacological therapy and there are no trials; management is based on endoscopic surveillance with polypectomy, surgery when polyposis can no longer be controlled and oral rehabilitation, which the 2021 French PNDS organises by age, with mandibular symphyseal implants accepted from 6 years of age under strict criteria, including AXIN2 oligodontia. The guidelines diverge: the NCCN, known through secondary sources, proposes colonoscopy from 25–30 years of age, every 2–3 years or every 1–2 years with polyps, without upper surveillance; the 2024 European guideline, for lack of data, applies the MAP scheme, with colonoscopy from 18 years and upper endoscopy from 35. Cascade testing of first-degree relatives, with a 50% risk of transmission, allows dental care to start early and colorectal surveillance to be adjusted. The research priorities are cohorts of carriers identified by cascade testing and followed prospectively, which would allow penetrance to be estimated and the starting age and intervals of surveillance to be grounded.

13. References

1. OMIM® Online Mendelian Inheritance in Man. #608615 OLIGODONTIA-COLORECTAL CANCER SYNDROME; ODCRCS. Johns Hopkins University; last edited 12/12/2024. https://omim.org/entry/608615. Accessed 28 October 2026.
2. Lammi L, Arte S, Somer M, Jarvinen H, Lahermo P, Thesleff I, et al. Mutations in AXIN2 cause familial tooth agenesis and predispose to colorectal cancer. Am J Hum Genet. 2004;74(5):1043-1050. doi:10.1086/386293. PMID: 15042511; PMC1181967.
3. Marvin ML, Mazzoni SM, Herron CM, Edwards S, Gruber SB, Petty EM. AXIN2-associated autosomal dominant ectodermal dysplasia and neoplastic syndrome. Am J Med Genet A. 2011;155A(4):898-902. doi:10.1002/ajmg.a.33927. PMID: 21416598; PMC3094478.
4. Orphanet. AXIN2-related polyposis (ORPHA:401911). Available from: https://www.orpha.net/en/disease/detail/401911
5. Orphanet. Oligodontia-cancer predisposition syndrome (ORPHA:300576). Available from: https://www.orpha.net/en/disease/detail/300576
6. ClinGen Hereditary Cancer Gene Curation Expert Panel. Gene-disease validity: AXIN2 – oligodontia-cancer predisposition syndrome (MONDO:0012075), autosomal dominant. Classification: Definitive (SOP9). 2023 Dec 20. Available from: https://search.clinicalgenome.org/kb/gene-validity/CGGV:assertion_cb1ae839-6186-44f7-85fe-8642b6c180ba-2023-12-20T180000.000Z
7. Leclerc J, Beaumont M, Vibert R, Pinson S, Vermaut C, Flament C, et al. AXIN2 germline testing in a French cohort validates pathogenic variants as a rare cause of predisposition to colorectal polyposis and cancer. Genes Chromosomes Cancer. 2023;62(4):210-222. doi:10.1002/gcc.23112. PMID: 36502525; PMC10107344.
8. Rivera B, Perea J, Sánchez E, Villapún M, Sánchez-Tomé E, Mercadillo F, et al. A novel AXIN2 germline variant associated with attenuated FAP without signs of oligondontia or ectodermal dysplasia. Eur J Hum Genet. 2014;22(3):423-426. doi:10.1038/ejhg.2013.146. PMID: 23838596; PMC3925274.
9. Murakami T, Arai M, Eguchi H, Hirasawa Y, Hata W, Otsuki Y, et al. Somatic Mosaic AXIN2-associated Colonic Polyposis Predominantly Involving the Proximal Colon: A Case Report. DEN Open. 2027;7(1):e70409. doi:10.1002/deo2.70409. PMID: 42643670; PMC13504014.
10. Aceves-Ewing NM, Lanza DG, Marcogliese PC, Lu D, Hsu CW, Hashimoto H, et al. Uncovering phenotypic expansion in AXIN2-related disorders through precision animal modeling. Genet Med. 2026;28(9):102648. doi:10.1016/j.gim.2026.102648. PMID: 42429102; PMC13678866.
11. Liu W, Dong X, Mai M, Seelan RS, Taniguchi K, Krishnadath KK, et al. Mutations in AXIN2 cause colorectal cancer with defective mismatch repair by activating beta-catenin/TCF signalling. Nat Genet. 2000;26(2):146-147. doi:10.1038/79859. PMID: 11017067.
12. GeneGuidance. AXIN2 — gene record and clinical curation (version of 2026-09-26).
13. Freeman PJ, Hart RK, Gretton LJ, Brookes AJ, Dalgleish R. VariantValidator: Accurate validation, mapping, and formatting of sequence variation descriptions. Hum Mutat. 2018;39(1):61-68. PMID: 28967166; PMC5765404. Tool: https://rest.variantvalidator.org (accessed 2026-10-28).
14. Broekema MF, Redeker EJW, Uiterwaal MT, van Hest LP. A novel pathogenic frameshift variant in AXIN2 in a man with polyposis and hypodontia. Hered Cancer Clin Pract. 2023;21(1):16. doi:10.1186/s13053-023-00260-6. PMID: 37626374; PMC10464116.
15. Roht L, Hyldebrandt HK, Stormorken AT, Nordgarden H, Sijmons RH, Bos DK, et al. AXIN2-related oligodontia-colorectal cancer syndrome with cleft palate as a possible new feature. Mol Genet Genomic Med. 2023;11(6):e2157. doi:10.1002/mgg3.2157. PMID: 36860143; PMC10265056.
16. Letra A, Chiquet B, Hansen-Kiss E, Menezes S, Hunter E. Nonsyndromic Tooth Agenesis Overview. 2021 Jul 22. In: Adam MP, Feldman J, Mirzaa GM, et al., editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993–. Available from: https://www.ncbi.nlm.nih.gov/books/NBK572295/
17. OMIM® Online Mendelian Inheritance in Man. *604025 AXIS INHIBITOR 2; AXIN2. Johns Hopkins University; last edited 28/11/2017. https://omim.org/entry/604025. Accessed 28 October 2026.
18. OMIM® Online Mendelian Inheritance in Man. #114500 COLORECTAL CANCER; CRC. Johns Hopkins University; last edited 29/06/2026. https://omim.org/entry/114500. Accessed 28 October 2026.
19. OMIM® Online Mendelian Inheritance in Man. #106600 TOOTH AGENESIS, SELECTIVE, 1; STHAG1. Johns Hopkins University; last edited 11/06/2024. https://omim.org/entry/106600. Accessed 28 October 2026.
20. Orphanet. Oligodontia (ORPHA:99798). Available from: https://www.orpha.net/en/disease/detail/99798
21. Idos G, Hampel H, Valle L. Lynch Syndrome. 2004 Feb 5 [updated 2026 Sep 17]. In: Adam MP, Feldman J, Mirzaa GM, et al., editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993–. Available from: https://www.ncbi.nlm.nih.gov/books/NBK1211/
22. Yen T, Stanich PP, Axell L, Patel SG. APC-Associated Polyposis Conditions. 1998 Dec 18 [updated 2022 May 12]. In: Adam MP, Feldman J, Mirzaa GM, et al., editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993–. Available from: https://www.ncbi.nlm.nih.gov/books/NBK1345/
23. Larsen Haidle J, MacFarland SP, Howe JR. Juvenile Polyposis Syndrome. 2003 May 13 [updated 2022 Feb 3]. In: Adam MP, Feldman J, Mirzaa GM, et al., editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993–. Available from: https://www.ncbi.nlm.nih.gov/books/NBK1469/
24. Macklin-Mantia SK, Hines SL, Chaichana KL, Donaldson AM, Ko SL, Zhai Q, et al. Case report expanding the germline AXIN2- related phenotype to include olfactory neuroblastoma and gastric adenoma. BMC Med Genet. 2020;21(1):161. doi:10.1186/s12881-020-01103-0. PMID: 32807118; PMC7433097.
25. Ratanasereeprasert N, Intarak N, Chaweewannakorn C, Abid M, Marya A, Cho SD, et al. AXIN2 variants, tooth agenesis, and cancer risk: a systematic review. BMC Oral Health. 2025;25(1):1738. doi:10.1186/s12903-025-07022-x. PMID: 41188821; PMC12584280.
26. Zaffaroni G, Mannucci A, Koskenvuo L, de Lacy B, Maffioli A, Bisseling T, et al. Updated European guidelines for clinical management of familial adenomatous polyposis (FAP), MUTYH-associated polyposis (MAP), gastric adenocarcinoma, proximal polyposis of the stomach (GAPPS) and other rare adenomatous polyposis syndromes: a joint EHTG-ESCP revision. Br J Surg. 2024;111(5):znae070. doi:10.1093/bjs/znae070. PMID: 38722804; PMC11081080.
27. Haute Autorité de Santé. Protocole National de Diagnostic et de Soins (PNDS): Agénésies dentaires multiples : oligodontie et anodontie. Saint-Denis La Plaine: HAS; November 2021. Available from: https://www.has-sante.fr/jcms/p_3301900
28. Ambry Genetics. Clinician Management Resource for AXIN2 / Understanding Your Positive AXIN2 Genetic Test Result (MKT-ONCO-FLYR-10041-EN v3). Aliso Viejo (CA): Ambry Genetics; 2025 Jul 15. Available from: https://www.ambrygen.com/file/material/view/2512/CMR_UYR_Combo_AXIN2_0725_FINAL.pdf
29. PDQ Cancer Genetics Editorial Board. Genetics of Colorectal Cancer (PDQ®)–Health Professional Version. Bethesda (MD): National Cancer Institute; updated 2025 Mar 21. Available from: https://www.cancer.gov/types/colorectal/hp/colorectal-genetics-pdq. PMID: 26389505.
30. European Reference Network on Genetic Tumour Risk Syndromes (ERN GENTURIS). Available from: https://www.genturis.eu/l=eng/index.html
31. European Reference Network for rare and/or complex craniofacial anomalies and ENT disorders (ERN CRANIO). Available from: https://www.ern-cranio.eu
32. Peterlongo P, Howe LR, Radice P, Sala P, Hong YJ, Hong SI, et al. Germline mutations of AXIN2 are not associated with nonsyndromic colorectal cancer. Hum Mutat. 2005;25(5):498-500. doi:10.1002/humu.20189. PMID: 15841489.
33. Lejeune S, Guillemot F, Triboulet JP, Cattan S, Mouton C, PAFNORD Group, et al. Low frequency of AXIN2 mutations and high frequency of MUTYH mutations in patients with multiple polyposis. Hum Mutat. 2006;27(10):1064. doi:10.1002/humu.9460. PMID: 16941501.
34. Beard C, Purvis R, Winship IM, Macrae FA, Buchanan DD. Phenotypic confirmation of oligodontia, colorectal polyposis and cancer in a family carrying an exon 7 nonsense variant in the AXIN2 gene. Fam Cancer. 2019;18(3):311-315. doi:10.1007/s10689-019-00120-0. PMID: 30671715.
35. Chan JM, Clendenning M, Joseland S, Georgeson P, Mahmood K, Walker R, et al. Rare germline variants in the AXIN2 gene in families with colonic polyposis and colorectal cancer. Fam Cancer. 2022;21(4):399-413. doi:10.1007/s10689-021-00283-9. PMID: 34817745.
36. Jensen JM, Skakkebæk A, Gaustadness M, Sommerlund M, Gjørup H, Ljungmann K, et al. Familial colorectal cancer and tooth agenesis caused by an AXIN2 variant: how do we detect families with rare cancer predisposition syndromes?. Fam Cancer. 2022;21(3):325-332. doi:10.1007/s10689-021-00280-y. PMID: 34637023.
37. OMIM® Online Mendelian Inheritance in Man. #601216 DENTAL ANOMALIES AND SHORT STATURE; DASS. Johns Hopkins University; last edited 05/06/2018. https://omim.org/entry/601216. Accessed 28 October 2026.
38. OMIM® Online Mendelian Inheritance in Man. %602639 TOOTH AGENESIS, SELECTIVE, 2; STHAG2. Johns Hopkins University; last edited 21/11/2017. https://omim.org/entry/602639. Accessed 30 October 2026.
39. OMIM® Online Mendelian Inheritance in Man. %610926 TOOTH AGENESIS, SELECTIVE, 5; STHAG5. Johns Hopkins University; last edited 22/07/2009. https://omim.org/entry/610926. Accessed 30 October 2026.
40. Cree IA, Arends MJ, Khoury JD, Denton ERE, Gill AJ, Lazar AJ, et al. The WHO Classification of Genetic Tumour Syndromes: Considerations for Genetics. Clin Genet. 2026;110(3):389-401. doi:10.1111/cge.70195. PMID: 42304730; PMC13431718.
41. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. p. 568–573.
42. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. p. 556–561.
43. National Center for Biotechnology Information. ClinVar — records VCV000005881, VCV000040260, VCV000545741, VCV000582788, VCV001303554, VCV000408788, VCV000182008, VCV001415792 (AXIN2). Available from: https://www.ncbi.nlm.nih.gov/clinvar/ (E-utilities, accessed 2026-10-28)
44. Macklin-Mantia SK, Riegert-Johnson DL. An American patient with polyposis carrying a Scandinavian AXIN2 pathogenic variant. Hered Cancer Clin Pract. 2020;18:14. doi:10.1186/s13053-020-00149-8. PMID: 32754288; PMC7391027.
45. Bergendal B, Klar J, Stecksén-Blicks C, Norderyd J, Dahl N. Isolated oligodontia associated with mutations in EDARADD, AXIN2, MSX1, and PAX9 genes. Am J Med Genet A 2011;155A(7):1616-1622. doi:10.1002/ajmg.a.34045. PMID: 21626677.
46. Callahan N, Modesto A, Meira R, Seymen F, Patir A, Vieira AR. Axis inhibition protein 2 (AXIN2) polymorphisms and tooth agenesis. Arch Oral Biol 2009;54(1):45-49. doi:10.1016/j.archoralbio.2008.08.002. PMID: 18790474. PMCID: PMC2643013.
47. OMIM® Online Mendelian Inheritance in Man. #313500 TOOTH AGENESIS, SELECTIVE, X-LINKED, 1; STHAGX1. Johns Hopkins University; last edited 16/08/2016. https://omim.org/entry/313500. Accessed 30 October 2026.
48. OMIM® Online Mendelian Inheritance in Man. #617275 TOOTH AGENESIS, SELECTIVE, 9; STHAG9. Johns Hopkins University; last edited 11/01/2024. https://omim.org/entry/617275. Accessed 28 October 2026.
49. OMIM® Online Mendelian Inheritance in Man. #616724 TOOTH AGENESIS, SELECTIVE, 7; STHAG7. Johns Hopkins University; last edited 11/12/2025. https://omim.org/entry/616724. Accessed 28 October 2026.
50. OMIM® Online Mendelian Inheritance in Man. #610947 CORONARY ARTERY DISEASE, AUTOSOMAL DOMINANT 2; ADCAD2. Johns Hopkins University; last edited 28/03/2014. https://omim.org/entry/610947. Accessed 28 October 2026.
51. OMIM® Online Mendelian Inheritance in Man. #617073 TOOTH AGENESIS, SELECTIVE, 8; STHAG8. Johns Hopkins University; last edited 16/08/2016. https://omim.org/entry/617073. Accessed 28 October 2026.
52. OMIM® Online Mendelian Inheritance in Man. #618180 ECTODERMAL DYSPLASIA 14, HYPOHIDROTIC/HAIR/TOOTH/NAIL TYPE; ECTD14. Johns Hopkins University; last edited 10/04/2025. https://omim.org/entry/618180. Accessed 28 October 2026.
53. OMIM® Online Mendelian Inheritance in Man. #620173 TOOTH AGENESIS, SELECTIVE, 10; STHAG10. Johns Hopkins University; last edited 20/12/2022. https://omim.org/entry/620173. Accessed 30 October 2026.
54. OMIM® Online Mendelian Inheritance in Man. #604625 TOOTH AGENESIS, SELECTIVE, 3; STHAG3. Johns Hopkins University; last edited 13/03/2025. https://omim.org/entry/604625. Accessed 28 October 2026.
55. van den Boogaard MJ, Créton M, Bronkhorst Y, van der Hout A, Hennekam E, Lindhout D, et al. Mutations in WNT10A are present in more than half of isolated hypodontia cases. J Med Genet 2012;49(5):327-331. doi:10.1136/jmedgenet-2012-100750. PMID: 22581971.
56. OMIM® Online Mendelian Inheritance in Man. #150400 TOOTH AGENESIS, SELECTIVE, 4; STHAG4. Johns Hopkins University; last edited 06/03/2017. https://omim.org/entry/150400. Accessed 28 October 2026.
57. Arzoo PS, Klar J, Bergendal B, Norderyd J, Dahl N. WNT10A mutations account for ¼ of population-based isolated oligodontia and show phenotypic correlations. Am J Med Genet A 2014;164A(2):353-359. doi:10.1002/ajmg.a.36243. PMID: 24449199.
58. Horton A, Fostier W, Winship I, Rajan N. Facial Features of Hereditary Cancer Predisposition. JCO Oncol Pract. 2024;20(9):1182-1197. doi:10.1200/op.23.00610. PMID: 38713892.
59. Adegun AH, Schalet R, Berezowski I, Borum ML. Importance of Identifying Physical Manifestations That Are Associated With Hereditary Cancer Predisposition: AXIN2 Mutation in an African American Patient. JCO Oncol Pract. 2025;21(3):440-441. doi:10.1200/op-24-00609. PMID: 39413343.
60. Fostier W, Horton A, Winship I, Rajan N. Reply to Importance of Identifying Physical Manifestations That Are Associated With Hereditary Cancer Predisposition: AXIN2 Mutation in an African-American Patient. JCO Oncol Pract. 2025;21(3):441. doi:10.1200/op-24-00713. PMID: 39413338.
61. Khalifeh H, Zhang X, Hoskins BJ. Pediatric Hereditary Polyposis Syndromes: Diagnosis, Surveillance, and Management Across Evolving Guidelines. Dig Dis Sci. 2026;71(11):5994-6008. doi:10.1007/s10620-026-10108-3. PMID: 42423930; PMC13585806.
62. Eiset SE, Schraw J, Sørensen GV, Gregersen PA, Rasmussen SA, Ramlau-Hansen CH, et al. Congenital Tooth Agenesis and Risk of Early-Onset Cancer. JAMA Netw Open. 2024;7(3):e240365. doi:10.1001/jamanetworkopen.2024.0365. PMID: 38488796; PMC10943407.
63. Hashmi AS, Ranganathan R, Vats K, Baburaj D, Sampaio P, Chauhan D, et al. Tooth agenesis as a potential clinical indicator of colorectal cancer susceptibility: a systematic review. Front Oral Health. 2026;7:1865287. doi:10.3389/froh.2026.1865287. PMID: 42609260; PMC13477944.
64. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. p. 330–332.
65. Jelsig AM, Karstensen JG, Jespersen N, Ketabi Z, Lautrup C, Rønlund K, et al. Danish guidelines for management of non-APC-associated hereditary polyposis syndromes. Hered Cancer Clin Pract. 2021;19(1):41. doi:10.1186/s13053-021-00197-8. PMID: 34620187; PMC8499431.
66. Mankaney G, Idos GE, Stoffel EM, Weiss JM, Shah S, Qayed E, et al. ACG Clinical Guideline: Diagnosis and Management of Adenomatous Colorectal Polyposis Syndromes. Am J Gastroenterol. 2026;121(9):2086-2120. doi:10.14309/ajg.0000000000004105. PMID: 42683623.
67. Lorca V, Rueda D, Martín-Morales L, Fernández-Aceñero MJ, Grolleman J, Poves C, et al. Contribution of New Adenomatous Polyposis Predisposition Genes in an Unexplained Attenuated Spanish Cohort by Multigene Panel Testing. Sci Rep. 2019;9(1):9814. doi:10.1038/s41598-019-46403-5. PMID: 31285513; PMC6614360.
68. Frostberg E, Petersen AH, Bojesen A, Rahr HB, Lindebjerg J, Rønlund K. The Prevalence of Pathogenic or Likely Pathogenic Germline Variants in a Nationwide Cohort of Young Colorectal Cancer Patients Using a Panel of 18 Genes Associated with Colorectal Cancer. Cancers (Basel). 2021;13(20):5094. doi:10.3390/cancers13205094. PMID: 34680242; PMC8534092.
69. Rowell B, Roberts ME, Brock PL, Cooper J, Pearlman R, Kalady MF, et al. Updated genetic testing in individuals with unexplained adenomatous polyposis and the diagnostic yield. Fam Cancer. 2025;24(4):90. doi:10.1007/s10689-025-00513-4. PMID: 41296145.
70. OMIM® Online Mendelian Inheritance in Man. #617100 FAMILIAL ADENOMATOUS POLYPOSIS 4; FAP4. Johns Hopkins University; last edited 29/07/2025. https://omim.org/entry/617100. Accessed 28 October 2026.
71. OMIM® Online Mendelian Inheritance in Man. #610069 POLYPOSIS SYNDROME, HEREDITARY MIXED, 2; HMPS2. Johns Hopkins University; last edited 20/08/2021. https://omim.org/entry/610069. Accessed 30 October 2026.
72. OMIM® Online Mendelian Inheritance in Man. #601228 POLYPOSIS SYNDROME, HEREDITARY MIXED, 1; HMPS1. Johns Hopkins University; last edited 16/10/2013. https://omim.org/entry/601228. Accessed 28 October 2026.
73. Harrold E, Stadler Z. Cancer Predisposition — Familial Adenomatous Polyposis. In: ESMO Handbook — Cancer Genomics. Lugano: ESMO Press. p. 27–28.
74. OMIM® Online Mendelian Inheritance in Man. #189500 WITKOP SYNDROME. Johns Hopkins University; last edited 12/05/2016. https://omim.org/entry/189500. Accessed 28 October 2026.
75. OMIM® Online Mendelian Inheritance in Man. #257980 ODONTOONYCHODERMAL DYSPLASIA; OODD. Johns Hopkins University; last edited 18/09/2025. https://omim.org/entry/257980. Accessed 28 October 2026.
76. OMIM® Online Mendelian Inheritance in Man. #224750 SCHOPF-SCHULZ-PASSARGE SYNDROME; SSPS. Johns Hopkins University; last edited 17/11/2022. https://omim.org/entry/224750. Accessed 28 October 2026.
77. OMIM® Online Mendelian Inheritance in Man. #230740 GAPO SYNDROME; GAPOS. Johns Hopkins University; last edited 12/03/2019. https://omim.org/entry/230740. Accessed 28 October 2026.
78. OMIM® Online Mendelian Inheritance in Man. #193530 WEYERS ACROFACIAL DYSOSTOSIS; WAD. Johns Hopkins University; last edited 04/01/2023. https://omim.org/entry/193530. Accessed 28 October 2026.
79. OMIM® Online Mendelian Inheritance in Man. #620193 LACRIMOAURICULODENTODIGITAL SYNDROME 3; LADD3. Johns Hopkins University; last edited 01/02/2024. https://omim.org/entry/620193. Accessed 30 October 2026.
80. OMIM® Online Mendelian Inheritance in Man. #620192 LACRIMOAURICULODENTODIGITAL SYNDROME 2; LADD2. Johns Hopkins University; last edited 09/01/2023. https://omim.org/entry/620192. Accessed 30 October 2026.
81. OMIM® Online Mendelian Inheritance in Man. #149730 LACRIMOAURICULODENTODIGITAL SYNDROME 1; LADD1. Johns Hopkins University; last edited 03/04/2024. https://omim.org/entry/149730. Accessed 28 October 2026.
82. OMIM® Online Mendelian Inheritance in Man. #616029 ECTODERMAL DYSPLASIA/SHORT STATURE SYNDROME; ECTDS. Johns Hopkins University; last edited 26/09/2014. https://omim.org/entry/616029. Accessed 28 October 2026.
83. OMIM® Online Mendelian Inheritance in Man. #617392 ECTODERMAL DYSPLASIA 13, HAIR/TOOTH TYPE; ECTD13. Johns Hopkins University; last edited 29/02/2024. https://omim.org/entry/617392. Accessed 28 October 2026.
84. OMIM® Online Mendelian Inheritance in Man. #602482 AXENFELD-RIEGER SYNDROME, TYPE 3; RIEG3. Johns Hopkins University; last edited 23/04/2025. https://omim.org/entry/602482. Accessed 30 October 2026.
85. Jones KL, Jones MC, del Campo M. Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. p. 834–837.
86. OMIM® Online Mendelian Inheritance in Man. #180500 AXENFELD-RIEGER SYNDROME, TYPE 1; RIEG1. Johns Hopkins University; last edited 25/02/2026. https://omim.org/entry/180500. Accessed 28 October 2026.
87. OMIM® Online Mendelian Inheritance in Man. #225060 CLEFT LIP/PALATE-ECTODERMAL DYSPLASIA SYNDROME; CLPED1. Johns Hopkins University; last edited 17/12/2024. https://omim.org/entry/225060. Accessed 28 October 2026.
88. Jones KL, Jones MC, del Campo M. Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. p. 760–763.
89. Weyant RJ, Tracy SL, Anselmo TT, Beltrán-Aguilar ED, Donly KJ, Frese WA, et al. Topical fluoride for caries prevention: executive summary of the updated clinical recommendations and supporting systematic review. J Am Dent Assoc. 2013;144(11):1279-1291. doi:10.14219/jada.archive.2013.0057. PMID: 24177407; PMC4581720.
90. Haute Autorité de Santé. Prescription des antibiotiques en pratique bucco-dentaire. Recommandation de bonne pratique. Saint-Denis La Plaine: HAS; July 2026. Available from: https://www.has-sante.fr/jcms/p_3525810
91. Guilmot A, Belpaire M, Olinger E, De Leener A, Punie K, Duhoux FP. Inherited cancer syndromes in adult oncology: a referral guide based on malignant tumor and polyp features. Clin Transl Oncol. 2026;28(8):3488-3507. doi:10.1007/s12094-025-04134-2. PMID: 41670817; PMC13675360.
92. Al-Ani AH, Antoun JS, Thomson WM, Merriman TR, Farella M. Hypodontia: An Update on Its Etiology, Classification, and Clinical Management. Biomed Res Int. 2017;2017:9378325. doi:10.1155/2017/9378325. PMID: 28401166; PMC5376450.
93. Polder BJ, Van't Hof MA, Van der Linden FP, Kuijpers-Jagtman AM. A meta-analysis of the prevalence of dental agenesis of permanent teeth. Community Dent Oral Epidemiol. 2004;32(3):217-226. doi:10.1111/j.1600-0528.2004.00158.x. PMID: 15151692.
94. National Foundation for Ectodermal Dysplasias (NFED). Available from: https://www.nfed.org
95. Colorectal Cancer Alliance. Available from: https://www.ccalliance.org
96. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. p. 857–862.